0163-769X/92/1302-0146$03.00/0 Endocrine Reviews Vol. 13, No. 2 Copyright 0 1992 by The Endocrine Society Printed in U.S.A.

The Molecular Biology of RU486. Is There a Role for Antiprogestins in the Treatment of Breast Cancer?*

KATHRYN B. HORWITZ DeDartments of Medicine and Patholot!v. Universitv of Colorado Health Sciences Center, Division of Endocrinology,. Denuer, Colorado 8026~ ’

I. Introduction spread use of tamoxifen reflects its efficacy and low II. Progesterone and the Normal Breast toxicity, and the fact that it makes good physiological III. Progesterone and Breast Cancer sense to block the local proliferative effects of estrogens A. Progestin agonists and tumor induction directly at the breast. But are estrogens the only hor- B. Progestin agonists and growth of established tumors mones with a proliferative impact on the breast and on IV. Molecular Mechanisms of Progesterone Antagonists A. Progesterone receptors breast cancers? This review focuses on evidence that B. Molecular mechanisms of progesterone antagonists progesterone also has proliferative actions in the breast; 1. Affinity for inactive PR - on the role of synthetic progestins in breast cancer 2. PR activation treatment; on the molecular biology of progesterone an- 3. DNA binding tagonists; and on the preliminary data showing that 4. Dimerization progesterone antagonists may be powerful new tools for 5. Transcription the management of metastatic breast cancer, because 6. PR “processing” or down-regulation they block the local effects of endogenous progesterone V. Progesterone Antagonists and the Treatment of Breast on breast cell proliferation. The reader is also referred Cancer to the excellent general review on progestin regulation A. Human breast cancer cell lines B. Animal models of mammary cancer of cell proliferation by Clarke and Sutherland (1). The C. Human clinical trials structures of the agonists and antagonists discussed in VI. Progestin Resistance this review are shown in Fig. 1. VII. Summary and Future Prospects I. Introduction II. Progesterone and the Normal Breast NDOCRINE therapy used either prophylactically Conventional wisdom holds that the mechanisms by E or therapeutically for the treatment of locally ad- which estradiol and progesterone regulate the prolifera- vanced or metastatic breast cancers offers many advan- tion and differentiation of uterine epithelial cells apply tages to patients whose tumors contain functional estro- equally to the breast. This is probably inaccurate (2-4). gen and progesterone receptors (ER and PR). The range In the uterus, estrogens are clearly mitogenic, and addi- of treatments defined as endocrine include surgical abla- tion of progesterone to the estrogenized endometrium tion of endocrine glands, administration of pharmacolog- leads to the appearance of a secretory pattern character- ical doses of steroid hormones, chemical blockade of ized by cells engaged in protein synthesis rather than steroid hormone biosynthesis, and inhibition of endoge- cell division (5). That is, in the uterus, estradiol is a nous steroid hormone action at the tumor with synthetic proliferative hormone; progesterone is a differentiating antagonists. The last of these approaches is the most hormone. For this reason the unopposed actions of estra- widely used, making the antiestrogen tamoxifen the pre- diol are considered to be tumorigenic in the uterus, while ferred first-line therapeutic agent for treatment of hor- the risk of endometrial hyperplasia and cancer is lowered mone-dependent metastatic breast cancer. The wide- when estrogens are combined with progestins. In fact, the combined regimen may even be protective since a Address requests for reprints to: Kathryn B. Horwitz, Ph.D., Divi- decrease in endometrial cancers has been reported in sion of Endocrinology, Box B151, University of Colorado Health Sci- women prescribed combined estrogens and progestins, ences Center, 4200 East Ninth Avenue, Denver, Colorado 80262. *Supported in part by grants from the National Cancer Institute compared to women receiving no treatment (Ref. 6 and and the American Cancer Society. references therein). 146

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER

FIG. 1. The structure of progesterone, the synthetic progestin agonist R5020, and four progesterone antagonists dis- cussed in this review.

However, considerable evidence has now accrued to progestin components of oral contraceptives increase the suggest that in the epithelium of the breast, progesterone thymidine labeling index with progestin-only formula- has a different influence. That, like estradiol, progester- tions exhibiting high activity (2-4). The investigators one in the breast has a strong proliferative effect. Studies conclude that it is difficult to sustain the idea that in support of this come both from experimental models progestins are protective in the breast (3). It would seem and from normal cycling women. Both the proliferation that more work must be done to understand the actions of normal mammary epithelium in virgin mice and the of progestins in the normal breast, but that clinical lobular-alveolar development of mammary tissues in decisions based on an inappropriate uterine model sys- pregnant mice require progesterone (7,8). A fundamental tem are unjustified (14). difference in the actions of estradiol and progesterone in the breast is that the latter stimulates DNA synthesis, III. Progesterone and Breast Cancer not only in the epithelium of the terminal bud, but also A discussion of the role of progestins in breast cancer in the ductal epithelium (9). The stimulating effects of must distinguish between their effects on carcinogenesis progesterone on the development of mammary gland and their role in regulating proliferation of established buds can be inhibited by progesterone antagonists (10). cancers. Data from normal human mammary cells have been more difficult to obtain and are often equivocal. Com- A. Progestin agonists and tumor induction pared to the increase caused by estradiol treatment (11.3- fold), progesterone treatment only marginally (2.0-fold) Progestin agonists have been shown to be carcinogenic increases the mitotic index of normal human breast or to increase the incidence of spontaneous mammary ductal epithelium maintained in intact athymic nude tumors in dogs and mice (E-19). In mice, results vary mice (11). In fact, Mauvais-Jarvis and colleagues (12, 13) with the strain tested, suggesting the contribution of a concluded, using primary cultures of epithelial cells from genetic component; however, tumorigenic effects of pro- normal human mammary glands, that while estradiol gestins have been observed whether or not the strain treatment stimulates growth, progestins inhibit growth. harbors the mouse mammary tumor virus (MMTV). The Their data are difficult to interpret, however, since the importance of progesterone in carcinogen-induced rat experiments using estradiol were done with cells growing mammary cancers is documented by the early reports of in minimally supplemented medium, while the progestin Huggins et al. (20-22), who showed that pregnancy pro- treatment studies were done with cells in optimally sup- motes the growth of dimethylbenzanthracene (DMBA)- plemented medium, and any progestin growth-stimula- induced mammary tumors, and that administration of tory effect might have been masked. In contrast to the progesterone together with the carcinogen to intact rats sparse and conflicting in. vitro data are studies of the accelerates the appearance of tumors, increases the num- mitotic rate in breast epithelial cells during the normal ber of tumors, and augments the growth rate of estab- menstrual cycle and in women taking oral contraceptives. lished tumors. The relationship between progestins and These data show that the highest thymidine labeling carcinogenesis is temporally complex. In general, proges- indices occur during the progestin-dominated, secretory terone administered simultaneously with, or after, the phase of the menstrual cycle. Both the estrogen and the carcinogen enhances tumorigenesis, while progesterone

Downloaded from edrv.endojournals.org on February 27, 2005 148 HORWITZ Vol. 13. No. 2 given before the carcinogen inhibits tumorigenesis (Ref. that progestin-sensitive cells can generate resistant sub- 23 and references therein). Thus, the high progesterone populations (48); and it is possible that no physiological levels associated with pregnancy can be protective if the consensus is likely to be forthcoming using these in vitro hormone precedes administration of the carcinogen (24). models. These models remain invaluable, however, for Extrapolation of these experimental models to human the analysis of molecular mechanisms of progestin ac- disease is unclear since the only data available for the tions, and to underscore the complexities inherent in latter are epidemiological in nature and relate hormone tumor cell biology. use, particularly oral contraceptive use, to the risk of Where does this leave us on the critical issue of the breast cancer. The trend toward increased risk with use of progestin agonists in breast cancer treatment? increased duration of hormone use appears repeatedly Interestingly, here there is more agreement, but the data (25, 26), and a possible adverse effect of progestins ap- contradict the conclusion that physiological levels of pears to be likely (27, 28). This is discouraging when progestins are growth stimulatory. Especially at high taken together with the likelihood that in the breast, doses, progestins appear to be antiproliferative in breast unlike the uterus, progestins enhance proliferation dur- cancers. A comprehensive review of the clinical literature ing the menstrual cycle. (Ref. 49 and references therein and Refs. 50-55) shows that synthetic progestins, used at pharmacological doses B. Progestin agonists and growth of established tumors for first- or second-line therapy, are as effective as ta- moxifen in the treatment of advanced breast cancer. Carcinogen-induced rat mammary tumors are a major That is, approximately 30% of patients whose tumors model for in uiuo studies of progestin-regulated growth are not screened for steroid receptors have an objective, (23). After ovariectomy, progesterone alone is usually positive response. Since in addition, progestins are well- unsuccessful in preventing regression of established tu- tolerated and have a relatively low toxicity (56), their mors. The rapid decrease of PR levels due to estrogen use in the treatment of advanced breast cancer is expe- withdrawal is probably a critical factor (29). In intact riencing a resurgence (57,58). However, the mechanisms animals, which more closely mimic the clinical situation, underlying the actions of intermediate and high doses of progestin agonists at moderate doses have been reported progestin agonists in breast cancer regression remain to promote tumor growth and to reverse the antitumor unclear, when compared to their proliferative actions at effects of tamoxifen (30). Thus, it is possible that endog- physiological doses. While some studies suggest that PR- enous circulating progesterone may enhance breast can- negative tumors respond just as well as do PR-positive cer growth. Enigmatically, progestins at higher pharma- tumors (implying that PR are not involved), others sug- cological doses appear to be growth inhibitory (31). The gest that methodological problems produce false PR- molecular mechanisms responsible for the opposing ac- negative values in responders (58), and that PR are tions of physiological and high-dose progestins remain indeed required to obtain a response to progestins. An unclear and are discussed later in this review. interesting study, in which tamoxifen therapy was com- In vitro cell culture models designed to assess the role pared to therapy in which tamoxifen was alternated with of progestin agonists in tumor cell proliferation have medroxyprogesterone acetate (MPA) in ER-positive pa- generated contradictory results. Experiments can be tients, showed a 40% response to tamoxifen alone us. a cited in support of any argument, that progestins stim- 62% response with the alternating treatment (53). It is ulate (32-34), inhibit (35-39), or have no effect (40) on postulated that when tamoxifen is cycled, its agonist growth. Explanations for the lack of a consensus are as properties predominate, which increases PR levels varied as the results. Responses of cells in culture are thereby enhancing the efficacy of MPA. The same ar- critically dependent on the conditions in which they are gument is made for enhancing the therapeutic efficacy grown. In a rich medium, where growth is optimized, of antiprogestins (see below). In general, tamoxifen in- further growth enhancement is difficult to demonstrate, ducibility of PR is considered to be a good indicator for while inhibitory stimuli may be exaggerated. In a de- a positive response to hormone therapy (51). Thus, while prived medium the reverse is true, although here, key definitive data are still lacking, it is likely that positive cofactors may be lacking. There is no simple solution to responses to therapy with progestins in breast cancer are these inherent problems. Couple this generic uncertainty mediated by the PR in the tumors. with other variables including the use of different cell In order to consider mechanisms by which progestin lines (35, 37, 40); heterogeneity and genetic instability agonists and antagonists regulate tumor cell growth, and even within the same cell lines (41,42); a burgeoning list how the hormone dose influences cellular responses, the of factors other than estradiol and progestins that di- following section will review recent advances in the mo- rectly or indirectly modulate progestin sensitivity lecular biology of PR especially as they apply to proges- through regulation of PR levels (43-47); the possibility terone antagonists.

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER 149

IV. Molecular Mechanisms of Progesterone 94 kDa A receptors (74-76, 80). While A receptors were Antagonists originally thought to be produced by a proteolytic artifact Like all steroid hormones, progesterone and the syn- (81), it is now clear that the amino-truncated receptors, thetic progestins function by binding to intranuclear at least in chickens and humans, are a naturally synthe- proteins, the PR, the presence of which specifies a pro- sized form (76,82-85). In human endometrial carcinoma gesterone target tissue and target cell. PR, when bound (86) and breast cancer cell lines (74,75), the two receptor to ligand, control the transcription of genes, which in the isoforms are expressed in approximately equimolar breast include, among others, the genes for ER (59), for amounts. It is not known whether this quantitative re- insulin receptors (37, 60), and for epidermal growth lationship between the two isoforms is maintained in all factor (EGF) receptors (61), as well as the genes for EGF human target tissues and tumors, and the mechanisms and tranforming growth factor-a (62, 63). Thus, the for their differential regulation are not known, but at proliferative effects of progesterone may in part be a least two of the nine mRNA species lack the translation reflection of its ability to modulate the levels of growth initiation site for B receptors and can therefore encode factors and their cognate receptors. The mechanisms by only A receptors (77, 84). These messages arise by tran- which this regulation is achieved and the structure of the scription from an internal promoter in the human PR PR proteins are complex and are described here with gene (84). Five other message species can potentially specific reference to human PR. For additional infor- encode both receptor isoforms, synthesized by alternate mation the reader is referred to several recent reviews of translation initiation from two in-frame AUG codons. In the steroid receptor family of proteins (64-67). theory, use of the upstream codon generates the B recep- tors and use of the downstream codon generates the A A. Progesterone receptors receptors, but it is not known whether initiation at the downstream site actually occurs in intact cells (85). In- Complementary DNAs for the chicken PR were cloned terestingly, two additional message species have been independently by Jeltsch et al. (68) and Conneely et al. mapped that lack both the A and B translation start (69) in 1986, and for the human PR by Misrahi et al. sites and would encode neither of the two receptor iso- (70) in 1987. Structure/function analyses of the encoded forms. These messages do, however, contain downstream proteins followed quickly. PR belong to the steroid/ translation start sites that if utilized would generate a thyroid hormone receptor superfamily of ligand-acti- protein with a partial DNA-binding domain, and com- vated DNA binding proteins which are composed of a plete hormone binding domain and dimerization se- number of independently functioning domains required for nuclear localization, ligand binding, DNA binding, quences (77). This “C-receptor,” when reconstructed by site-directed mutagenesis and expressed in PR-negative dimerization, and transcriptional activation. Additional sites on the receptors are responsible for their interaction cells, binds the synthetic progestin R5020 with the same with one or more accessory proteins including heat shock affinity as the A and B receptors (K. B. Horwitz, unpub- proteins, and several of the serine residues are covalently lished). Expression of such an N-terminally truncated , modified by phosphorylation. Not all of these domains protein in breast cancer cells would lead to a functionally have been definitively mapped, and the functional re- impaired but PR-positive phenotype. The existence of gions exhibit some overlap, but in general the N terminus such variant forms of receptors, not only for progester- is involved in transactivation and contains several phos- one, but also for estrogen, may in part explain the recur- phorylation sites; the DNA-binding domain is centrally rent descriptions of receptor-positive, but hormone-re- positioned and may have a weak dimerization function; sistant breast cancers. and ligand-binding activity is restricted to the C termi- PR are unique among steroid receptors in having two nus, which also contains structural features required for naturally occurring hormone binding forms, and this protein-protein interactions plus a second transcrip- structural feature may have important functional impli- tional activation domain (64-67, 71). cations with respect to antagonist action (discussed be- The single-copy human PR gene encodes at least nine low). Originally, the unliganded A and B isoforms were messenger RNA species ranging in size from 2.5-11.4 thought to be subunits of a larger heteromeric 8S receptor kilobases (72,73). The nine messages direct the synthesis (87). This is not the case; instead, each receptor species of at least two (74-76), and possibly three (77), structur- forms an independent 8s heteromeric complex on su- ally related receptor proteins. The two major protein crose gradients (88-90) by binding to as many as five species, the B- and A receptors, were originally described non-hormone-binding proteins including two molecules by O’Malley and co-workers (78,79) in the chick oviduct. of the 90 kDa heat shock protein (hsp) and one molecule Subsequent studies using the breast cancer cell line T47D of hsp 70 (91-94). In this oligomeric state, the receptor showed that human PR also exist as two isoforms, 116 proteins are basally phosphorylated and unable to bind kilodalton (kDa) B receptors and N-terminally truncated DNA, but they are maintained in a stable conformation

Downloaded from edrv.endojournals.org on February 27, 2005 150 HORWITZ Vol. 13, No. 2 that allows them to bind to progestins. Hormone binding molar amounts of A and B receptors have been demon- to the SS receptor form activates a rapid series of strated in breast cancer cell lines and endometrial can- changes, the exact sequential order of which is still cers, it is not known whether this stoichiometry applies unknown. Within minutes, if not seconds after hormone to normal target tissues, and to PR-positive breast can- binding, the receptors become hyperphosphorylated (95- cers. A scenario can be envisioned in which a tumor is 97), and several of the associated proteins, but not hsp PR-positive by ligand or immunological assays, but con- 70, dissociate, leading to generation of a faster sediment- tains only B receptors that either cannot activate a key ing 4s receptor form (93, 94, 98). Two 4S receptor proliferation regulatory gene, or that confer agonist prop- molecules dimerize (98) and then bind to DNA at specific erties to an antagonist, providing molecular explanations transcription enhancer sites called progesterone response for a PR-positive but hormone-resistant state. elements (PREs) (Refs. 64-67 and references therein). The biological significance of two or more PR isoforms Since both homo- and heterodimers can form between extends beyond breast cancer; there is evidence that, in the A and B isoforms, three possible classesof receptor the chicken, the two receptor isoforms are differentially dimers (A:A, A:B, B:B) can bind to a PRE, each having expressed developmentally and differentially regulated a potentially different transcription regulatory capacity by hormone treatment (100, 101). The two homodimers (71, 98). This diversity in the repertoire of responses to and the heterodimer may therefore differentially regulate one hormone sets progesterone apart from the regulatory transcription depending on the gene, the t,arget cell, the information carried by other sex steroid hormones stud- hormonal milieu, or the developmental program, leading ied to date. However, no formal proof exists that both to a complex combinatorial response that is unique to isoforms are in the same cell. the progestins. That this molecular heterogeneity is indeed translated into functional heterogeneity was first demonstrated by B. Molecular mechanisms of progesterone antagonists the study of Tora et al. (99) that assessedthe cell-specific transcriptional activation of two different target genes Synthetic hormone antagonists are exceptionally use- by the chicken A and B receptors. When a vector con- ful as both clinical and experimental tools. Their efficacy taining the PRE from the MMTV promoter linked to a and low toxicity have made antiestrogens the drugs of reporter gene was transfected into HeLa cells, transcrip- choice for the treatment of advanced hormone-depend- tional activation by form B was twice as efficient as that ent breast cancer. Experimentally, these compounds by form A; both receptor isoforms were similarly effective have been crucial for unraveling the mechanisms of ER at stimulating the MMTV promoter when t,hey were action. Antiprogestins have even greater potential, not transfected into chick embryo fibroblast cells. In con- only as experimental tools and as anticancer agents, but trast, when the PRE from the ovalbumin promoter was as midcycle contraceptives and implantation inhibitors tested, neither receptor form was functional in HeLa and as abortifacients. By the end of the 1970s more than cells, and only the A receptor activated transcription in a decade of computer-assisted design and testing of mol- chick embryo fibroblast cells. Transcription from the ecules likely to have antiprogestational activity had not ovalbumin promoter can also be stimulated through an led to any clinically or experimentally useful products. estrogen response element activated by estradiol-bound In 1981, however, Philibert et al. (102) announced the ER. When ER and one of the two PRs were cotrans- synthesis of RU38486 [17/Y-hydroxy-lip-(4dimethylam- fected, transcription from the ovalbumin promoter was inophenyl)-17cY-(l-propynyl)estra-4,9-diene-3-one]. additive with ER and the A receptor, but the B receptor RU38486 was one of a large number of ll/%substituted negated the transcriptional activation by ER. It seems steroid derivatives that were synthesized by scientists at that the N-terminal region of PR, present in B but Roussel-Uclaf exploiting their extensive experience in missing in A receptors, plays a specific role in transcrip- analyzing the structure/activity relationships of com- tion that differs depending both on the gene being mod- pounds that bind both glucocorticoid receptors (GR) and ulated and the cell being analyzed. Thus, the A receptor PR (103). The antiglucocorticoid activity of RU38486 can be stimulatory in a setting where the B receptor is was demonstrated first (102), and its antiprogestin and inactive or is inhibitory. The impact of this heterogeneity abortifacient properties in experimental animals were on antagonist action is discussed below, but it suggests reported soon thereafter (104). Testing in humans fol- that the N terminus of the B receptor confers specific lowed rapidly, and it was shown that oral administration transcription regulatory functions to PR. The data also of RU38486 to women in preliminary clinical trials in- suggest that as we learn more about the tissue-specific duced interruption of the luteal phase of the menstrual effects of the two PR isoforms, it may be important not cycle and termination of early pregnancy (105). The only to measure PR levels but also to identify the partic- mechanisms of action of RU38486 (renamed RU486 or ular PR isoform present in a breast tumor. While equi- mifepristone) have since been the subject of intense

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER 151 investigation (106). Using RU486 as a starting-point, appears to be related to rapid dissociation of the multiple scientists at Schering (107) reported the synthesis of a accessory proteins, including hsp 90, from PR (92-94, second generation of antagonists with potent antiproges- 98), and to the equally rapid hormone-dependent phos- tational activity but weaker antiglucocorticoid activity phorylation of PR (95-97). The two reactions occur in a than RU486 (108). These compounds hold much promise still undetermined order. The dissociation of accessory for long-term clinical use, as preliminary studies suggest. proteins is readily measured on sucrose density gradients In theory, an antagonist can act by blocking agonist as a decrease in the sedimentation coefficient. Mullick activity at any of the multiple stages between its entry and Katzenellenbogen (115) first showed that while ac- into the target cell and first interaction with the recep- tivated R5020-receptor complexes sediment at approxi- tors, through its role in terminating the transcriptional mately 4S, the RU486-bound receptors sediment as a signal. The effects of antiprogestins at each of these prominent 6s species as well as a 4s species; t.he heavier stages are reviewed below. species is likely to be due to persistent receptor oligo- 1. Affinity for inactive PR. RU486 binds to the PR of merization with other proteins (116,117). The suggestion rodents and humans with a dissociation constant (Kd) that RU486 traps hetero-oligomeric 8s PR in an inactive of approximately 2 x lo-’ M, roughly equivalent to the state, thereby preventing their interaction with DNA affinity of the synthetic progestin R5020 for PR (109, (112), is not supported by in viuo data (71). llO), but tighter than the affinity of the natural hormone, In addition to the loss of accessory proteins, PR acti- progesterone. Interestingly, unlike progesterone, RU486 vation may be regulated by phosphorylation (95-97). does not bind to chicken or hamster PR (106), which Human PR are phosphorylated at serine residues at the suggests that there are subtle differences in the contact time of their synthesis, and in the absence of progestins sites of agonists and antagonists at the hormone binding this basal phosphorylation continues for several hours. domain of the receptors. As in R5020-bound PR, unac- However, progestin binding to the receptors rapidly pro- tivated RU486-bound PR complexes sediment at 8s on motes additional serine phosphorylation concomitant sucrose density gradients (111, 112). RU486 absorbs UV with receptor activation and is accompanied by struc- at approximately 300 nm, and this wavelength can be tural changes that alter the mobility of human PR on used to covalently cross-link the drug to PR in intact electrophoretic gels (118, 119). The function of this hor- cells. This method was used to show that, like R5020, mone-dependent PR phosphorylation is still unknown, low concentrations of [3H]RU486 bind to both the A but it may play a role in the release of hsp proteins, in receptors and the B receptors of human PR (110). In receptor dimerization, in DNA binding, and in regulating intact breast cancer cells, R5020 and RU486 have iden- the functional state of PR while they are bound to the tical affinity for PR, and low concentrations (6 to 8 IIM) transcription complex. Since a subset of the hormone- of RU486 rapidly deplete more than 95% of hormone- dependent phosphorylation precedes DNA binding, and unoccupied PR and activate them to a state in which the remainder is DNA-dependent (K. B. Horwitz, un- they are tightly bound to DNA and require high concen- published), it is likely that phosphorylation of human trations of salt for their extraction (110). Thus, it was PR subserves more than one function. Of interest is the clear soon after its development that RU486 promotes fact that compared to agonists like R5020, binding of the binding of PR to DNA in intact cells. Like other syn- antagonist RU486 results in hyperphosphorylation of the thetic progestins, RU486 is not metabolized in intact PR (118, 119), so that this stage of receptor activation breast cancer cells (113), allowing its activity to be pro- is, if anything, enhanced by the antagonist. It remains longed. Consequently, it chronically suppresses receptor to be determined whether appropriate or inappropriate replenishment and lowers the levels of inactive 8s PR serine residues are the targets of this intensive RU486- available for ligand binding (but see Section IV.B.6 be- dependent phosphorylation. low). In this fashion RU486 resembles long-acting syn- 3. DNA binding. Whether RU486-bound PR complexes thetic agonists, so that this property is unlikely to be have altered DNA binding capacity when compared to involved in its antagonist actions. agonist-bound PR complexes remains a matter of con- 2. PR actiuation. Initial studies by Rauch et al. (114) tention, partially due to the lack of consensus for any suggested that while RU486-bound PR complexes were hormone requirement! This issue has been addressed by activation-deficient and had lower affinity for DNA than in vivo studies, by in vitro DNA binding studies, and agonist-bound complexes, these differences were modest most recently by cell-free transcription studies, with and could not fully explain its antagonist properties. often contradictory conclusions. The following example These investigators suggested that the major defect was illustrates the methodological conundrum for glucocor- at the transcriptional step. The receptor activation step ticoids: In consecutive articles published in Nature, that follows hormone binding is unclearly defined; it Becker et al. (120) concluded that, based on in uiuo

Downloaded from edrv.endojournals.org on February 27, 2005 152 HORWITZ Vol. 13, No. 2 genomic footprinting analysis, GR interactions with a lost components are restored. This is clearly evident in glucocorticoid response element (GRE) require the pres- more recent studies using in vitro transcription methods. ence of hormone, and Willmann and Beato (121) con- O’Malley and his colleagues have developed such an cluded that based on in vitro DNase I footprinting, assay using purified chick PR or crude human PR. The hormone-free GR bind specifically to a GRE. purified chick PR (124) lacks a hormonal requirement There is little question that progesterone is required for transcription that the cruder human PR (125) retains. to activate PR in intact cells, and that this is accom- In addition to issues of receptor purification, accurate panied by increased affinity of receptors for chromatin/ kinetic data obtained from in vitro DNA binding studies DNA as assessed by the concentration of salt required require that the reactants be at equilibrium conditions, to extract the receptors from cells or nuclei [for a caveat and that many factors including pH, incubation times, see recent studies by O’Malley and collaborators (122)]. salt concentrations, DNA and receptor concentrations, Hormone-free receptors can be extracted with little or etc., be controlled. Differences among these parameters, no salt in buffers; after hormone treatment at least 0.3 plus differences among the hormones and DNAs tested, M salt is required. When PR are bound by RU486 in may account for the lack of consensus about hormone viuo, the receptors also require high salt concentrations requirements, or about the kinetics of reactions. The for their extraction, suggesting that the antagonist in- greatest kinetic difference between agonist and antago- deed promotes PR binding to DNA (110). That both B nist regulation of PR binding to DNA was reported by and A receptors can bind DNA when occupied by RU486 Schauer et al. (126) who showed that in the presence of was first demonstrated by their high salt requirement for agonist and under equilibrium conditions, the on-rate of extraction after in situ photoaffinity labeling (110). receptor binding to DNA was accelerated 2- to 5-fold, These in vivo assays cannot, however, distinguish be- and the off-rate was accelerated lo- to 20-fold compared tween PR that are bound to nonspecific DNA sequences to ligand-free receptors. It is thus surprising that the and PR bound to specific PREs. Becker et al. (120) used actual DNA occupancy time of receptors was shortened in vivo protein DNA footprinting to show that GR bind- by the agonist. The investigators postulated that this ing to the GRE of the tyrosine amino transferase gene reduced occupancy time accelerated the DNA scanning could be detected only after glucocorticoid treatment of potential of PR, but it was apparent that these results hepatoma cells. Similar studies have not been done to were unexpected. In the same study, the kinetic values analyze PR binding. Instead, in vitro kinetic methods observed with RU486 occupancy were closer to those have been used in an attempt to measure the affinity of observed with ligand-free receptors. That is, both hor- receptors for DNA under varying conditions of ligand mone-free receptors and RU486-occupied receptors bind occupancy. Using crude receptors, Milgrom and his col- tighter to DNA than do agonist-bound receptors. This laborators (114) showed that RU486-PR complexes bind apparently tighter DNA binding of RU486-bound com- to nonspecific DNA with a slightly lower affinity than plexes compared to agonist-bound complexes (due to the do R5020-PR complexes, but that a ligand was required slowed off-rate with RU486) had previously been ob- for any PR/DNA interactions. The affinity differences served by Bagchi et al. (127) using gel mobility shift between R5020- and RU486-bound PR were insufficient assays. In contrast, ligand-free receptors fail to bind to explain antagonist action, but it appeared that hor- PREs when gel mobility shift methods are used (71,127), mone occupancy can indeed modulate binding, even to results that differ from those observed using the kinetic nonspecific DNA. It was therefore surprising that in a equilibrium assays. Finally, while confirming that ligand later study (123), this same group found that PR, whether is required for PR binding to DNA by gel mobility shift hormone-free or bound to an agonist or antagonist, in- assay, Meyer et al. (71) found no differences with respect teracted with a similar affinity and protected the same to specificity, affinity, or stability when comparing R5020 DNA regions on three fragments of the uteroglobin gene. to RU486. A major difference between the two studies was that Newly developed cell-free assays reaffirm the hormone instead of the crude PR used in the earlier study, the and DNA binding requirements for receptor-mediated later study used immunopurified PR, which may have transcription (124, 125, 128). Of interest is the report by lost a component necessary to demonstrate the hormone Bagchi et al. (125) that nuclear PR, salt-extracted from effect. It has been suggested that this repressor is hsp T47D breast cancer cells grown in the absence of hor- 90. Alternatively, it is possible that extensive purification mone, are in a 4s form and dissociated from hsp 90. irreversibly modifies the structure of receptors in a man- While these criteria characterize “activated” receptors, ner that mimics the structural changes associated with these 4s PR complexes nevertheless require hormone for hormone binding. Thus, paradoxically, excessive purifi- DNA binding and transcriptional activation, suggesting cation may hinder in vitro study of the. relationship’ that hsp 90 dissociation is necessary but not sufficient between hormone occupancy and DNA binding,-unless to generate receptors capable of binding DNA. Other

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER 153

nuclear factors or receptor modifications are likely to wild-type RU486-occupied PR (133). As with R5020, regulate DNA binding of the 4s complexes, with phos- RU486 and the Schering antiprogestins, ZK98434 and phorylation being a leading candidate. ZK112993, promote both homo- and heterodimerization Finally, it is important to recall the danger of deriving between forms A and B, excluding the possibility that generalized conclusions about the effects of antagonists antiprogestins induce only one dimeric form of PR as a on PR/DNA binding when only one antagonist is tested. mechanism for their action (129). Complexes formed Recently, Klein-Hitpass et al. (129) compared RU486 between the PRE and the PR dimers migrate faster on and three newer Schering antiprogestins with regard to nondenaturing gels when the receptors are occupied by the DNA binding properties of PR by gel mobility shift RU486 (71, 117); these differences are likely to be due to assay. They found that while three of the antiprogestins altered conformation of the receptors at the hormone promoted DNA binding, one, ZK98299, did not induce binding domain generated by the antagonist, or to differ- binding of PR to PREs, and in competition assays, ences in DNA bending produced by occupancy of the blocked the DNA binding of RU486-bound PR com- receptors. How this structural change is translated into plexes. The investigators postulate that there are at least an antagonist effect is still unknown. This finding, how- two types of antiprogestins: type I that block PR binding ever, raises an interesting question: can dimerization to PREs; and type II that promote PR binding to PREs occur if one PR monomer is bound by agonist and but impede transcription, echoing a general mechanism another by antagonist? Since the receptor dimerization of antagonist action proposed by Guiochon-Mantel et al. site is probably located in the hormone-binding domain, (130) based on studies of antiglucocorticoids and anti- conformational differences at the dimerization domain estrogens. interface of the two monomers may be incompatible. To summarize, in vitro analysis of PR binding to DNA Indeed, Meyer et al. (71) found that R5020 and RU486- for the purposes of assessing the function of agonists and liganded receptors are unable to bind to PREs as hetero- antagonists remains complex, and the data obtained may dimers, thereby introducing a novel mechanism for an- depend on variables in assay conditions that are difficult tagonist action. This is confirmed by in vitro kinetic to control. Critical among them is the structural status studies showing that the slope of the competition curve of the PR. Warming and salt exposure are among the for RU486 is different than the slopes for progesterone factors known to artifactually activate receptors in uitro; and R5020 at receptor concentrations that favor dimer- the former is often included in binding reactions, and ization. This is consistent with a preference of the recep- the latter is commonly used to extract or elute receptors. tor to have RU486 occupy both sites of a dimer, rather When artifactual activation is minimized, it appears than having progesterone and RU486 each occupy one certain that hormone is required to generate the DNA of the two monomers (134). On the other hand, failure binding form of PR. It is also clear that the most widely to promote formation of stable receptor dimers is the tested antagonist, RU486, also promotes PR binding to postulated mechanism for antagonism of the non-DNA DNA, but that not all antiprogestins do so. However, binding antiprogestin ZK98299 (129). because equally convincing studies have yielded contra- 5. Transcription. The ultimate effect of antiprogestins is dictory conclusions, it is still not clear whether antago- to modify transcription of genes regulated by progester- nist-bound receptors bind to DNA with kinetics that one through PR. While it is usually assumed that antag- differ from those of agonist-bound receptors. onists inhibit the effects of agonists, this is not always 4. Dimerization. The palindromic structure of PREs, the the case since antagonists often have agonist actions as dyad symmetry shown by receptor DNA contact sites, well. The paradoxical effects of steroid hormone antag- the failure of PR to bind to half-palindromic PREs in onists are well-known (see below), but the mechanisms uitro, and the leucine zipper/helix-loop-helix motif found that underlie them are poorly understood. That antago- in PR all suggest that receptors bind to DNA as dimers nists regulate the transcription of specific genes, rather (64-67,71,131-133). Since human PR exist naturally as than having global, nonspecific effects, is certain (114), two size classes, the A and B receptors, homo- and and it is also likely that whether agonist or antagonist heterodimers formed between them can be distinguished effects predominate depends on complex interacting fac- by differences in their mobility on nondenaturing elec- tors that include the antagonist concentration, the spe- trophoretic gels. This interaction between monomers is cific target gene, other transcriptional regulators of that activated by ligand binding and takes place in solution gene, and the cellular environment in which that gene is independent of DNA binding (98, 133). Similar to the active. Studies to unravel this complexity are still in the agonists, RU486 promotes dimerization as shown by its preliminary stages. While it is clear that progestins and ability to induce nuclear transfer of a nuclear localiza- antiprogestins regulate the expression of a variety of tion-deficient mutant, when the mutant dimerizes to proteins in intact breast cells, transcriptional regulation

Downloaded from edrv.endojournals.org on February 27, 2005 154 HORWITZ Vol. 13, No. 2 of PRE-containing genes by PR has not been analyzed RU486 were not tested in these studies, but Strahle et in situ. Instead, two structurally different model systems al. (135) showed that RU486 was able to inhibit an R5020 have been used: the palindromic PRE of the tyrosine stimulatory effect using the same PRE-promoter con- amino transferase (TAT) promoter (137) and the com- structs in transient transfection experiments. Unfortu- plex promoter of the MMTV long terminal repeat which nately, in this study the agonist effects of RU486 alone has, in addition to one distal palindromic PRE, a cluster were not tested. Clearly, even in this simple system, a of three tandemly arrayed (138-141) proximal PRE half- considerable amount of work remains to be done in order sites. to understand the different agonist and antagonist ac- The first model, the PRE of the TAT gene, consists of tions of RU486. Another observation that requires expla- two symmetrical half-sites, each of which may be occu- nation is that the agonist actions of RU486 using the pied by one molecule of the PR dimer (131, 135). This TAT PRE, can be blocked by the type I antagonist type of dimeric binding has been characterized as coop- ZK98299 (129). erative for GR, i.e. that (using a DNA-binding domain The second model used for transcription studies is the fragment) occupancy of one half-site by one monomer MMTV-long terminal repeat which may behave primar- enhances binding to the second half-site by another ily as a PRE in mammary cells (143). By analogy with monomer (131). This DNA binding cooperativity is dif- calculations of GR binding to the MMTV promoter ficult to reconcile with the fact that receptor monomers (141), it is likely that one PR dimer interacts with the dimerize in solution. Assuming that only the preformed distal palindromic PRE. Additionally, four to six PR dimer binds DNA, one would have to postulate that when monomers bind the three proximal half-sites so that at one monomer of the pair encounters the first PRE half- full occupancy six to eight PR molecules are associated site, the other monomer is in close physical proximity with the MMTV promoter. The distal and proximal and binds more efficiently to the second half-site. On the PREs act synergistically to stimulate transcription (144, other hand, since dimers and monomers are likely to be 145), and mutations in any of the four PREs reduce in equilibrium in solution (141)) it may be the monomeric significantly the overall hormone inducibility of the form that accounts for the binding cooperativity seen in MMTV promoter (144, 145). The complexity is obvious. the in vitro assays. The receptor concentration used in How many RU486 bound PR molecules would it take to these assays controls the state of the reactants at equi- modify progesterone-stimulated transcription from this librium with more dilute receptor concentrations favor- promoter? What is the functional difference between PR ing a monomeric state (134); these concentrations are monomer and dimer binding sites? rarely carefully controlled. When two PREs (PREP) are Several studies using both stable and transient trans- tandemly linked, occupancy of one PRE by receptor fection models show that, when acting at the MMTV dimers increases the binding affinity of additional recep- promoter, RU486 is a pure antagonist with no apparent tor dimers to the second PRE by as much as loo-fold agonist activity. In kinetic analyses Guiochon-Mantel et depending on the distance between the PREs (136,137). al. (130) calculate that when 6% of PR molecules are This probably involves protein-protein contacts between occupied by RU486, they can inhibit 50% of stimulatory dimers, with looping of the intervening DNA (142). Co- PR activity (in this case the constitutive activity from a operative binding of PR to PREs leads to synergistic mutant PR). They propose that since all four PREs of induction of a coupled reporter gene (136, 137). the MMTV promoter are necessary for optimal tran- Transcriptional studies with the TAT PRE use either scriptional activity, partial occupancy of one PRE by a single (71, 135) or tandem copies (124). Assuming that PR-RU486 complex is inhibitory, even if the other sites R5020/RU486 PR heterodimers do not form (71), a single are occupied by stimulatory PR complexes. These con- PRE should give clearly interpretable results when clusions are consistent with other studies using MMTV RU486 function is analyzed since both monomers would (117, 143), which show that lower concentrations of be occupied by the same hormone. But what happens in RU486 than R5020 are needed to shut down the agonist the tandem PRE, system when one of the dyads is effects of R5020, despite the fact that both hormones occupied by a progestin-bound homodimer, and the other bind to human PR with similar affinities. dyad is bound by an RU486-bound homodimer? The Finally, by comparing both the TAT PREz and the answer to this is unknown. Two studies using the TAT MMTV model systems in one study, Meyer et al. (71) PREP linked to the thymidine kinase and/or ovalbumin lend support to the findings that RU486 can be a partial promoter, and using human PR in cell-free transcription agonist or a pure antagonist, depending on the context assays, suggest that like the agonist R5020, the antago- of the promoter and the simplicity or complexity of the nist RU486 stimulates transcription (124, 125); that is, PREs. They confirmed the agonist activity of RU486- RU486 acts like a weak agonist having 25-30% the occupied human PR when they are bound to PRE, in activity of R5020. The effects of combining R5020 and the context of a thymidine kinase promoter and their

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER 155 lack of agonist activity when bound to the MMTV pro- known about the role of PR phosphorylation in initiating moter. In addition, these investigators tested the func- or sustaining the transcriptional signal, or in regulating tional capacity of the B and A isoforms of human PR receptor down-regulation. It is possible that PR struc- separately with totally unexpected findings: only the B tural changes associated with hyperphosphorylation pre- form of the receptors exhibited the agonist properties vent the down-regulatory machinery from being acti- when bound by RU486. Since the A and B isoforms also vated or from functioning. had quite different inductive capacities when bound to In the case of ER, it is known that this structural R5020 (the A receptors had only 10% the agonist activity block to down-regulation imparted by antagonist occu- of the B receptors on the MMTV promoter but 60% of pancy of the receptors is rapidly reversible. In intact B on the TAT PREJ, the combinatorial possibilities for breast cancer cells, ER occupied by the antiestrogen modulating transcriptional responsiveness (by type of nafoxidine bind persistently to DNA and are not down- ligand, by receptor isoform, by promoter, or by cell type) regulated for at least 24 h. If, at 24 h, cells are switched are staggering. Tora et al. (99) showed that the cell type to medium containing estradiol, the excess agonist ex- into which receptors are transfected and tested strongly changes with the antagonist on the DNA-bound recep- influences transcriptional capacity. Note also that the tors, and down-regulation is initiated and proceeds with transcriptional actions of an AB heterodimer, likely to a normal time course. The estradiol-substituted ER, be the most abundant naturally occurring human PR which were transcriptionally silent when bound to an- form, remain unknown. tagonist, are rapidly reactivated; in fact, a superinduction of the transcriptional signal is seen (154). This is an 6. PR “processing” or down-regulation. A persistent tran- interesting observation for which no molecular explana- scriptional signal by hormone-receptor complexes cannot tions are presently forthcoming. These studies also ad- be tolerated by the cell, so mechanisms must exist to dress persistent questions regarding the length of time terminate it. Little is known about this final stage in that a ligand must occupy receptors after they bind to receptor action. Receptor recycling and covalent modi- DNA. Is the function of the ligand then terminated and fications are among the mechanisms that have been is it dispensible? The antagonist to agonist switching proposed, but down-regulation may be the most common study suggests either that ligand occupancy persists or feature (146-153) associated with receptor inactivation. that conformational changes are permanently imprinted After agonist treatment, both DNA-bound ER and PR on the receptor by ligand binding and are preserved when are “processed” or down-regulated, a stage that begins the ligand dissociates. The imprints imparted by agonists approximately 1 h after DNA binding and is complete 6- and antagonists differ but are exchangeable. 12 h later. Since this loss of DNA-bound receptors is not immediately accompanied by replenishment of unacti- V. Progesterone Antagonists and the Treatment vated receptors, total cellular receptor levels fall by 80- of Breast Cancer 90%. The down-regulated receptors are restored only many hours (>24) later, presumably after synthesis of Because progestin antagonists are relatively new com- new receptors. The receptor replenishment time differs pounds, and because of the political controversy that depending on the persistence of agonists within the cell surrounds them, their promise and use in the treatment (147-153). of breast cancer are just beginning to be evaluated. What While down-regulation has been documented for ER is the rationale for their use? What is the explanation and PR bound to agonists, the scenario is entirely differ- for the paradox that both antiprogestins and high-dose ent for receptors bound to antagonists. Antagonist- progestin agonists inhibit growth of breast cancers? bound receptor/DNA complexes down-regulate only par- tially or not at all; instead nuclear receptor levels remain A. Human breast cancer cell lines elevated for hours, even days, after hormone treatment. Two PR-positive human breast cancer cell lines that In the specific case of RU486 in breast cancer cells, PR are phenotypically different have served as the major remain complexed to DNA for at least 48 h (72,97,118). models for studies of growth regulation by antiprogestins. The sustained DNA occupancy of RU486-bound PR may, The MCF-7 cell line is classically estrogen-responsive at least in part, explain its antagonist actions. Is it cells; the cells are ER-positive but have only low PR possible that agonist effects of RU486 result from the levels unless these are induced by estradiol (153). The initial contact between RU486-bound PR and DNA, and T47D cell line (155) is more complex; it is genetically that antagonist effects follow from the subsequent pro- unstable (41, 42) and differs phenotypically among (34, longed DNA occupancy? Recall also that RU486-PR 35, 37) and within laboratories, which leads to reported complexes bound to DNA are hyperphosphorylated com- differences in response to hormone treatment. One major pared to their R5020 counterparts (97, 118). Nothing is T47D subline, clone 11, is ER-positive and PR-positive,

Downloaded from edrv.endojournals.org on February 27, 2005 156 HORWITZ Vol. 13, No. 2 and the cells respond to estradiol treatment by prolifer- (nonprogestin) growth factors including estradiol. 3) ating, a response that is tamoxifen-inhibitable (35). An- “Nonspecific cytotoxic” effects are not receptor-me- other subline, T47D,,, is estrogen-resistant, and cell diated and are seen with most steroid hormones at high growth is neither accelerated by estradiol nor inhibited doses. The molecular explanations that underlie these by tamoxifen (147). The response to antiprogestins is three mechanisms are unknown, but they serve as an generally similar among these cell lines, but interpreta- important departure for further research. For example, tion of the results differs. In general, RU486 inhibits the the prolonged DNA occupancy time of RU486-bound PR growth of both T47D cells and MCF-7 cells (110, 156- compared to R5020-bound PR may account for cytotoxic 161). The antiproliferative effects are evident at low vs. cytostatic effects (97). As described above, variables doses, and their magnitude correlates loosely with PR like the gene or cell being tested (99) and regulation by levels; T47D > estrogen primed MCF-7 > unprimed either the A or B receptor (71) may dictate an agonist or MCF-7 (156). That the antiproliferative effects are me- antagonist response. Alternatively, it has been suggested diated by PR is also shown by the fact that in T47D cells that R5020 is inhibitory because it is “antiestrogenic” growth inhibition is confined to progestins; other steroid (35) while RU486 is inhibitory through a direct antipro- hormones are ineffective (156, 158). Moreover, the fact liferative effect involving PR (156). While these expla- that RU486 is not antiproliferative in PR-negative breast nations may begin to address the paradox that allows cancer cell lines also argues for a receptor-mediated both R5020 and RU486 to be growth inhibitory in the mechanism of action (156). appropriate physiological setting, it is clear that a con- The antiproliferative actions of RU486 in these models siderable amount of research remains to be done. Prelim- of breast cancer would support a logical treatment strat- inary analyses of cell cycle parameters show no definitive egy if it were not for one disturbing fact: progestin differences between agonists and antagonists. Both ap- agonists, including R5020, also inhibit their growth. This pear to inhibit growth by significantly decreasing the effect of R5020 is seen even at low doses in the T47D,, proportion of cells in the S-phase of the cell cycle; cells cells (110) but requires higher doses in the clone 11 cells, accumulate in Go/G1 possibly due to an increase in the in which, interestingly, low doses actually protect the G1 transit time (158, 162). cells from the antiproliferative effects of RU486 (156). While the majority of studies using cell culture models Furthermore, in the estrogen-responsive clone 11 cells, ascribe growth- inhibitory properties to progestins and R5020 inhibits only the estradiol-stimulated growth frac- to antiprogestins through direct effects involving PR, tion where it is cytostatic, while RU486 reduces cell contradictory results have also been reported. Given the numbers below the estrogen-untreated baselines, sug- fact that, at physiological levels, progesterone is believed gesting that the latter has a more profound cytotoxic to be mitogenic in the normal breast (see above), it is effect (160, 161). Thus the antiproliferative mechanisms not entirely surprising that Hissom and Moore (33) for progestin agonists like R5020 and antagonists like consistently report proliferative effects of R5020 at all RU486 may be fundamentally different. R5020 appears doses in T47D cells. It is surprising, however, that RU486 to have dual proliferative/antiproliferative effects de- also stimulates growth in these cells (163). The expla- pending on the dose that is tested, and the antiprolifer- nation for this discrepancy is unknown. However, it is ative doses produce growth stasis. By contrast, RU486 is now clear that T47D cells are exceptionally unstable; more purely antiproliferative even at low doses, and it during prolonged time in culture, subpopulations can can produce growth regression. Indeed, ultrastructural develop that are phenotypically different from the paren- studies show effects of RU486 on cell and chromatin tal stocks, and some of these subpopulations may respond condensation and pyknosis consistent with the induction to hormones in a manner that differs markedly from the of cell death by apoptosis (161). This cytotoxic effect is expected response. For example, some sublines or sub- prevented by low doses of R5020. To explain these re- populations of T47D cells respond to high doses of ta- sults, Bardon et al. (161) propose that there are three moxifen by growth stimulation (48, 164). In the case of different mechanisms by which progestins inhibit growth tamoxifen, these aberrant responses may be mediated by of breast tumor cells: 1) “PR-mediated cytotoxic” mech- the presence of mutant or variant ER (164). By analogy, anisms that apply to antagonists like RU486. These are it is possible that the T47D cells of Moore and his observed only in PR-positive cells using “physiological” colleagues have arisen from a subpopulation harboring a hormone doses and are preventable by receptor occu- mutant PR. This scenario would be very interesting, but pancy with an agonist. This cytotoxicity is characterized the PR of these cells have not been analyzed in detail. by ultrastructural evidence of cell death. 2) “PR-me- Less interesting trivial explanations for discrepancies diated cytostatic” effects are produced by physiological among laboratories studying growth regulation by pro- doses of antagonists or agonists and are characterized by gestins in cell culture were discussed above (see also Ref. inhibition of the growth-stimulatory actions of unrelated 1).

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER 157

B. Animal models of mammary cancer having no agonist activity. Second, among the physiolog- ical effects seen in RU486-treated intact female rats are The antiproliferative properties of progesterone antag- increased plasma levels of LH, PRL, estradiol, and pro- onists are well documented in animal models of hormone- gesterone, as well as the persistence of numerous and dependent mammary cancer. These include rats bearing actively secretory corpora lutea associated with hyper- DMBA-induced or nitrosomethylurea (NMU)-induced trophic pituitaries (165, 166, 168, 169). It has therefore tumors, and mice bearing the transplantable MXT tumor been proposed that the efficacy of simultaneous tamox- line. Growth of these tumors is inhibited by ovariectomy ifen results from its ability to counteract the proliferative and maintained by physiological doses of estrogens (165 effects of the high estrogen levels. 170). While treatment of rats with progestins at the time Several newer antiprogestins, ORG31710 and of DMBA administration accelerates tumor formation ORG31806 (166), and ZK98299 and ZK112993, have (21-23), prophylactic treatment of rats with RU486 at equal or greater antiproliferative actions than RU486 the time of DMBA administration delays the initial (166,168). In the hormone-dependent MXT-transplant- appearance of tumors from an average of 39 days, to an able tumor model, treatment with ZK98299 or RU486 average of 81 days (157). The reversal by progesterone starting 1 day after transplantation led to an almost of the inhibition of tumor induction produced by tamox- complete inhibition of tumor growth. Their effect on ifen can in turn be blocked by RU486 (30). These impli- established tumors was equivalent to that of ovariectomy cate a PR-mediated mechanism. Treatment of estab- (168, 169). In this model, the potent antiproliferative lished tumors with RU486 for 3 weeks prevents their actions of the antiprogestins completely counteracted further enlargement, but tumor remission is not observed the growth-stimulatory actions of estradiol, or of approx- (165). In contrast to the stasis seen with RU486, ovar- imately equimolar doses of MPA, but at higher MPA iectomy leads to a decrease in tumor size (169), which in doses, the agonist actions of the progestin prevailed MXT tumors is accompanied by necrosis and cytolysis (170). It appears that antiprogestins inhibit growth by of the tumor cells (168). While ovariectomy-induced direct antagonism of progesterone action at the tumor, tumor regression is known to be accompanied by exten- probably mediated by PR. This conclusion is bolstered sive loss of PR (29), loss of tumor PR was also seen with by the fact that the hormone-independent MXT tumor RU486 treatment (165). Since in the latter, the PR assay is antiprogestin-resistant (168). was not performed under exchange conditions or by In DMBA-induced tumors, ZK98299 was more potent immunological methods, the validity of this decrease than an equal concentration of RU486 (168). It produced requires reexamination in light of other studies that show tumor regression analogous to that of ovariectomy, persistently high levels of PR in RU486-treated human rather than the tumor stasis observed with RU486. A breast cancer cells (97). Since adrenal weights are un- similar trend was observed with NMU-induced rat mam- changed by RU486, participation of the antiglucocorti- mary tumors (168-170). However, lack of comparative coid effects in the antitumor activity is considered to be metabolic and pharmacokinetic data on the two antipro- unlikely (169). This is supported by studies in human gestins in rats and mice makes these quantitative differ- breast cancer cell lines, where the inhibitory effects of ences uninterpretable at present. Also of interest is the RU486 cannot be rescued by dexamethasone (156). An- finding that strong antitumor activity was noted at 20% tigonadotrophic effects have also been excluded as a of the doses needed to obtain abortifacient actions in mechanism (169). these rodent systems. This is important because by use In established DMBA tumors inhibition of growth by of lower doses of antiprogestins, their antiglucocorticoid tamoxifen resembles the inhibition seen with RU486. effects may be mitigated. After treatment with the anti- The two hormones have equal growth-inhibitory effects progestins, the morphology of the hormone-dependent when each is used alone (165). When the two drugs are MXT and DMBA tumors showed signs of differentiation combined, however, the inhibitory effects are additive, of the mitotically active polygonal epithelial tumor cells and tumor remission similar to that induced by ovariec- toward the nonproliferating glandular secretory pattern tomy is observed (166). This effect of combined treat- with formation of acini and evidence of secretory activity ment with an antiprogestin and antiestrogen is extremely (168). Based on this it is suggested that the antiprolif- exciting and has considerable therapeutic promise. The erative efficacy of the antiprogestins is related to their mechanisms underlying these effects remain unclear, but ability to induce terminal differentiation. Recall that several proposals have surfaced. First, tamoxifen can they block tumor cells in Go/G1 (158, 162). Note that an have agonist actions among which is the induction of PR antiproliferative mechanism based on induction of ter- (153). A tumor with increased, or restored, PR may have minal differentiation is fundamentally different from a greater or more sustained sensitivity to RU486. This mechanism involving tumor cell death. Tumor cell deg- hypothesis could be tested by the use of an antiestrogen radation and cytolysis were features of ovariectomy-

Downloaded from edrv.endojournals.org on February 27, 2005 158 HORWITZ Vol. 13. No. 2

induced regression (168). The mechanisms underlying VI. Progestin Resistance the antitumor effects of antiprogestational agents require The emergence of hormone-resistant cells eventually further study, especially in human tissues and cells. reduces the effectiveness of all therapies in advanced breast cancer, and progestin agonists or antagonists are C. Human clinical trials unlikely to be exceptions. This is essentially an unex- plored field. Unlike the case for other members of the The enormous promise of progestin antagonists in the steroid receptor family, no examples of natural PR mu- treatment of breast cancer remains largely unexplored in tants have yet been reported. It is possible that unlike clinical practice. Only two small clinical trials using mutations in androgen receptors, systemic mutations in RU486 have been reported, both from European labora- PR are incompatible with life. It is likely, however, that tories. The first involved a series from France (171) of acquired mutations can develop in tumors as one mech- 22 oophorectomized or postmenopausal patients in whom anism for the development of resistance, and that a chemotherapy, radiotherapy, or tamoxifen and other hor- systematic search would demonstrate them. In a similar monal therapy had already been used. RU486 at 200 mg/ manner, resistance to tamoxifen appears to be due in day led to partial regression or stabilization of lesions in part to the development of mutant ER (164). To address 12 of 22 (53%) women after 4 to 6 weeks of treatment. other possible mechanisms of progestin resistance Mur- The response rate at 3 months had dropped to 18%. It is phy et al. (173) generated a subline of T47D cells that important to note that for ethical reasons, this untried are resistant to the growth-inhibitory effects of proges- therapy was used only in patients with advanced breast tins. This was done by sequential selection in medium cancers in whom other treatment modalities had already containing 1 pM MPA. The cells remained PR-positive, failed. PR levels were not measured in all patients, but but receptor levels were halved. Transforming growth of the responders, four of four were PR-positive, while factor-a and EGF receptor mRNA levels were both in- of the nonresponders zero of four were PR-positive. In creased. The investigators suggest that increased growth general, RU486 was well-tolerated in long-term treat- factor expression and action and decreased PR levels ment with few symptoms of adrenal dysfunction, but may be involved in the development of progestin resist- plasma cortisol levels were elevated. Of interest was the ance. Finally, it is likely that extensive heterogeneity in fact that a strong analgesic effect was observed in most PR content exists within cell subpopulations of tumors of the patients with bone metastases. that are PR-positive based on analyses of solid tumors The second trial, from the Netherlands (166, 170), (174) and of human breast cancer cell lines (48). Factors involved 11 postmenopausal patients with metastatic or treatments that lead to the selection and expansion breast cancer who were treated with 200-400 mg RU486 of PR-poor or PR-negative populations would, in the for 3-34 weeks as second-line therapy after first-line long run, produce progestin resistance. treatment with tamoxifen, irrespective of the response to tamoxifen. Six of 11 patients had a short-term (3 to 8 VII. Summary and Future Prospects months) stabilization of disease, and one had an objective response lasting 5 months after RU486 treatment. Again, The foregoing suggests that progesterone antagonists response was associated with presence of PR in the could have an important place in the routine manage- tumors. In this study, which involved prolonged use of ment of hormone-dependent breast cancers. Our knowl- RU486, two patients had undesirable side-effects asso- edge of the actions of these compounds is rudimentary, ciated with the antiglucocorticoid actions of the drug. however. The following points provide an outline for Three days of treatment with dexamethasone reversed future directions: these symptoms after RU486 was stopped. As in the 1. If endogenous progesterone has the mitogenic ac- animal studies, plasma estradiol levels increased despite tions in normal breast epithelia that the current data the fact that these women were postmenopausal, and it would indicate, then it is likely that physiological pro- is suggested that the simultaneous administration of gesterone is also a mitogen in breast cancers. Blockade tamoxifen might be beneficial because of its ability to of endogenous progesterone with antiprogestins, espe- blockade tumor ER. Alternatively, symptoms of adrenal cially in premenopausal women, would seem to be an hypersecretion and elevated plasma estradiol levels important therapeutic goal. However, nothing is known might be reduced with concurrent aminoglutethimide or about the pattern of mitosis in breast cancer cells during aromatase inhibitors. By modification of the dose and the normal menstrual cycle. Obtaining such data is im- time of RU486 administration, its antiglucocorticoid ef- portant, and it should be possible to analyze the mitotic fects might be further minimized (172), although main- patterns of tumors taken from cycling patients. The tenance of high sustained blood levels of the drug is likely problem, of course, is the difference in mitotic indices to be important. among tumors. Ideally, each tumor should serve as its

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER 159 own control, with multiple samples analyzed for prolif- are the genes in breast cells that are regulated by proges- erative activity at different times of the cycle. This tins, whose inhibition is associated with lowered cell approach is fraught with ethical problems. However, it proliferation? What is the underlying cause of the tissue- might be possible to obtain a fine needle aspirate of a specific differences in progestin action at the breast and tumor for initial mitotic analysis, then to acutely treat uterus? What is the molecular explanation for the dual the patient with RU486 before the tumor is removed 24 agonist/antagonist effects seen on only some promoters, h later for reanalysis. Additionally, it is important to with only one or the other PR isoform, and in only some know whether RU486 is cytostatic or cytolytic in human cells when using RU486? Can pure progesterone antag- breast cancers. Electron microscopy of tumors taken onists devoid of antiglucocorticoid activity be synthe- from patients entered into trials may provide answers to sized? Are there mutant PR in breast cancers? this. However, while it is always preferable to ask biolog- Where do we begin? Ensuring that scientists and cli- ical questions using clinical tissues, much of the work on nicians have access to antiprogestins, unencumbered by the mechanisms of the mitogenic actions of progestins the Byzantine bureaucratic obstacles and the “antago- and progestin antagonists will require well-controlled nistic” political climate currently encountered in the studies using organ-cultured human breast tumors, hu- United States, is a good place to start. man breast cancer cell lines, and human tumors im- planted into nude mice. Acknowledgments 2. There are sufficient theoretical and preclinical data Dedicated to the memory of William L. McGuire. to justify large-scale clinical trials. Trials must include I thank the past and present members of my laboratory who gener- accurate measurements of ER and PR in tumors, and ated the data cited here, Linda Trefry for her expert secretarial work, analysis of pre- and posttreatment levels of endogenous and my colleague, William Wood, for reviewing the manuscript. hormones to monitor the status of the pituitary-adrenal- ovarian axis. Progestin antagonists may be useful both References for adjuvant endocrine therapy when used either alone 1. Clarke CL, Sutherland RL 1990 Progestin regulation of cellular or in combination with tamoxifen, and for therapy of proliferation. Endocr Rev 11:266 2. Going JJ, Anderson TJ, Battersby S, MacIntyre CCA 1988 Pro- locally advanced or metastatic cancers again used either liferative and secretory activity in human breast during natural alone or in combination with tamoxifen. The usual issues and artificial menstrual cycles. Am J Path01 130:193 of drug doses and of metabolism and other pharmacoki- 3. Anderson TJ, Battersby S, King RJB, McPherson K, Going JJ 1989 Oral contraceptive use influences resting breast prolifera- netic parameters must be addressed. Schedules in which tion. Hum Path01 20:1139 the antiprogestins and antiestrogens are combined or 4. Anderson TJ, Howell A, King RJB 1987 Comment on progester- alternated must be tested. Is an antiestrogen with agonist one effects in breast tissue. Breast Cancer Res Treat 10:65 5. Berliner JA, Gerschenson LE 1976 Sex steroid induced morpho- properties preferable because it induces PR or is a pure logical changes in primary uterine cell cultures. J Steroid Biochem antiestrogen preferable because it blocks the actions of 7:153 elevated circulating estrogens? Is the answer different in 6. Henderson BE, Ross R, Bernstein L 1988 Estrogens as a cause of human cancer. The Richard and Hinda Rosenthal Foundation premenopausal vs. postmenopausal women? Award Lecture. Cancer Res 48246 3. The antiglucocorticoid side effects of RU486 remain 7. Haslam SZ 1988 Progesterone effects on deoxyribonucleic acid a major impediment to its long-term use. However, as synthesis in normal mouse mammary glands. Endocrinology 122:464 the newer Schering antiprogestins show, it appears to be 8. Imagawa W, Tomooka Y, Hamamoto S, Nandi S 1985 Stimulation possible to design molecules with maximal antiprogestin of mammary epithelial cell growth in uitro: interaction of epider- activity and minimal antiglucocorticoid activity. Further, ma1 growth factor and mammogenic hormones. Endocrinology 116:1514 even RU486 may be used for long-term treatment if it is 9. Bresciani F 1971 Ovarian steroid control of cell proliferation in combined with drugs that block adrenal steroidogenesis the mammary gland and cancer. In: Basic Actions of Sex Steroids or prevent peripheral aromatization of adrenal steroids on Target Organs. Karger Publishing Company, Basel, p 130 10. Michna H, Nishino Y, Schneider MR, Louton T, El Etreby MF to estrogens. Additionally, since the antiglucocorticoid 1991 A bioassay for the evaluation of antiproliferative potencies effects of RU486 are apparently tolerable in the short- of progesterone antagonists. J Steroid Biochem Mol Biol38:359 term (i.e. 2 to 3 months), perhaps it is reasonable to ask 11. McManus MJ, Welsch CW 1984 The effect of estrogen, proges- terone, thyroxine, and human placental lactogen on DNA synthe- whether alternating RU486 with tamoxifen maximizes sis of human breast ductal epithelium maintained in athymic its antiprogestin effects while it minimizes its antiglu- nude mice. Cancer 54:1920 cocorticoid ones. 12. Mauvais-Jarvis P, Kuttenn F, Gompel A 1986 Antiestrogen action of progesterone in breast tissue. Breast Cancer Res Treat 8179 4. Basic tumor biological and molecular research must 13. Gompel A, Malet C, Spritzer P, Lalardrie J-P, Kuttenn F, Mau- continue in order for us to understand the precise molec- vais-Jarvis P 1986 Progestin effect on cell proliferation and 17/3- ular targets and mechanisms of antagonist action. If hydroxysteroid dehydrogenase activity in normal human breast cells in culture. J Clin Endocrinol Metab 63:1174 different antagonists target different molecular sites, 14. McCarty Jr KS 1989 Proliferative stimuli in the normal breast: which antagonist would be best for clinical use? What estrogens or progestins? Human Path01 20:1137

Downloaded from edrv.endojournals.org on February 27, 2005 160 HORWITZ Vol. 13, No. 2

15. Frank DW, Kirton KT, Murchism TE, Quinlan WJ, Coleman TJ, of insulin receptors in antiestrogen-resistant T47D,, human Gilbertson TJ, Feenstra ES, Kimball FA 1979 Mammary tumors breast cancer cells by progestins: implications for endocrine ther- and serum hormones in the bitch treated with medroxy- apies. Cancer Res 45:167 progesterone acetate and progesterone for four years. Fertil Steril 38. Purohit A, Lai LC, Singh A, Coldham NG, Ghilchik MW, Reed 31:340 MJ, James VHT 1989 The effect of medroxyprogesterone acetate 16. Lanari C, Molinolo AA, Dosne PC 1986 Induction of mammary on aromatase and DNA polymerase (Y activities in breast tumours. adenocarcinomas by medroxyprogesterone acetate in BALB/c J Steroid Biochem 34:443 female mice. Cancer Lett 33:215 39. Poulin R, Baker D, Poirier D, Labrie F 1990 Multiple actions of 17. Nagasawa H, Aoki M, Sakagami N, Ishida M 1988 Medroxy- synthetic ‘progestins’ on the growth of ZR-75-1 human breast progesterone acetate enhances spontaneous mammary tumorigen- cancer cells: an in vitro model for the simultaneous assay of esis and uterine adenomyosis in mice. Breast Cancer Res Treat androgen, progestin, estrogen, and glucocorticoid agonistic and 12:59 antagonistic activities of steroids. Breast Cancer Res Treat 17:197 18. Lanari C, Kordon E, Molinolo A, Dosne PC, Charreau EH 1989 40. Lippman M, Bolan G, Huff K 1976 The effects of glucocorticoids Mammary adenocarcinomas induced by medroxyprogesterone and progesterone on hormone-responsive human breast cancer in acetate: hormone dependence and EGF receptors of BALB/c in long-term tissue culture. Cancer Res 364602 uiuo sublines. Int J Cancer 43:345 41. Reddel RR, Alexander IE, Koga M, Shine J, Sutherland RL 1988 19. Kordon E, Lanari C, Meiss R, Elizalde P, Charreau E, Pasqualini Genetic instability and the development of steroid hormone in- CD 1990 Hormone dependence of a mouse mammary tumor line sensitivity in cultured T47D human breast cancer cells. Cancer induced in uiuo by medroxyprogesterone acetate. Breast Cancer Res 48:4340 Res Treat 17:33 42. Graham II ML, Dalquist KE, Horwitz KB 1989 Simultaneous 20. Huggins C, Moon RC, Morii S 1962 Extinction of experimental measurement of progesterone receptors and DNA indices bv flow mammary cancer. I. Estradiol-17P and progesterone. Proc Nat1 cytometry: analysis of breast cancer cell mixtures and genetic Acad Sci USA 48:379 instabilitv of the T47D line. Cancer Res 49:3943 21. Huggins C, Yang NC 1962 Induction and extinction of mammary 43. Sarup JC; Rao KVS, Fox CF 1988 Decreased progesterone binding cancer. Science 137:257 and attenuated progesterone action in cultured human breast 22. Huggins C 1965 Two principles in endocrine therapy of cancers: carcinoma cells treated with epidermal growth factor. Cancer Res hormone deprival and-hormone interference. Cancer-Res 25:1163 485071 23. Welsch CW 1985 Host factors affecting the growth of carcinoeen- 44. Katzenellenbogen BS, Norman MJ 1990 Multihormonal regula- induced rat mammary carcinomas: a revview and tribute to Charles tion of the progesterone receptor in MCF-7 human breast cancer Brenton Huggins. Cancer Res 45:3415 cells: interrelationships among insulin, insulin-like growth factor- 24. Russo IH, Gimotty P, Dupuis M, Russo J 1989 Effect of medrox- I, serum, and estrogen. Endocrinology 126:891 yprogesterone acetate on the response of the rat mammary gland 45. Clarke CL, Graham J, Roman SD, Sutherland RL 1991 Direct to carcinogenesis. Br J Cancer 59:210 transcriptional regulation of the progesterone receptor by retinoic 25. Meirik 0, Lund E, Hans-Olov A, Bergstrom R, Christoffersen T, acid diminishes progestin responsiveness in the breast cancer cell Bergsjo P 1986 Oral contraceptive use and breast cancer in young line T-47D. J Biol Chem 266:18969 women. Lancet 2:650 46. Aronica SM, Katzenellenbogen BS 1991 Progesterone receptor 26. Hulka BS 1990 Hormone-replacement therapy and the risk of regulation in uterine cells: stimulation by estrogen, cyclic adeno- breast cancer. CA 40:289 sine 3’,5’-monophosphate, and insulin-like growth factor I and 27. Bergkvist L, Adami HO, Persson I, Hoover R, Schairer C 1989 suppression by antiestrogens and protein kinase inhibitors. En- The risk of breast cancer after estrogen and estrogen-progestin docrinology 128:2045 replacement. N Engl J Med 321:293 47. Eckert RL, Katzenellenbogen BS 1983 Modulation of progestin 28. Ewertz M 1988 Influence of non-contraceptive exogenous and binding activity in cultured human breast carcinoma cells: the endogenous sex hormones on breast cancer risk in Denmark. Int effect of serum type and concentration. J Recept Res 3599 J Cancer 42:832 48. Graham II ML, Smith JA, Jewett PB, Horwitz KB 1992 Hetero- 29. Horwitz KB, McGuire WL 1977 Progesterone and progesterone geneity of progesterone receptor content and remodeling by ta- receptors in experimental breast cancer. Cancer Res 37:1733 moxifen characterize subpopulations of cultured human breast 30. Robinson SP, Jordan VC 1987 Reversal of the antitumor effects cancer cells: analysis by quantitative dual parameter flow cytom- of tamoxifen by progesterone in the 7,12-dimethylbenzanthra- etry. Cancer Res, 52:593 cene-induced rat mammary carcinoma model. Cancer Res 47:5386 49. Sedlacek SM, Horwitz KB 1984 The role of progestins and pro- 31. Danguy A, Legros N, Develeeschouwer N, Heuson-Stennon JA, gesterone receptors in the treatment of breast cancer. Steroids Heuson JC 1980 Effects of medroxyprogesterone acetate (MPA) 44:467 on growth of DMBA-induced rat mammary tumors: histopatho- 50. Canobbio L, Galligioni E, Gasparini G, Fassio T, Crivellari D, logical and endocrine studies. In: Iacobelli S, Di Marco A (eds) Villalta D, Santini G, Monfardini S, Boccardo F 1987 Alternatine Role of Medroxyprogesterone in Endocrine-Related Tumors. tamoxifen and medroxyprogesterone acetate in postmenopausal Raven Press, New York, p 21 advanced breast cancer patients short and long term endocrine 32. Dao TL, Sinha DK, Nemoto T, Pate1 J 1982 Effect of estrogen effects. Breast Cancer Res Treat lo:201 and progesterone on cellular replication of human breast tumors. 51. Howell A, Harland RNL, Barnes DM, Baildam AD, Wilkinson Cancer Res 42:359 MJS, Hayward E, Swindell R, Sellwood RA 1987 Endocrine 33. Hissom JR, Moore MR 1987 Progestin effects on growth in the therapy for advanced carcinoma of the breast: relationship be- human breast cancer cell line T47D-possible therapeutic impli- tween the effect of tamoxifen upon concentrations of progesterone cations. Biochem Biophys Res Commun 145:706 receptor and subsequent response to treatment. Cancer Res 47:300 34. Hissom JR, Bowden RT, Moore MR 1989 Effects of progestins, 52. Lundgren S, Gundersen S, Klepp R, Lonning PE, Lund E, estrogens, and antihormones on growth and lactate dehydroge- Kvinnsland S 1989 Megestrol acetate uersus aminoelutethimide nase in the human breast cancer cell line T47D. Endocrinology for metastatic breast cancer. Breast Cancer Res Treat 14:201 125:418 53. Gundersen S, Kvinnsland S, Lundgren S, Klepp 0, Lund E, 35. Vignon F, Bardon S, Chalbos D, Rochefort H 1983 Antiestrogenic Bormer 0, Host H 1990 Cyclical use of tamoxifen and high-dose effect of R5020, a synthetic progestin in human breast cancer medroxyprogesterone acetate in advanced estrogen receptor pos- cells in culture. J Clin Endocrinol Metab 56:1124 itive breast cancer. Breast Cancer Res Treat 17:45 36. Chalbos D, Rochefort H 1984 Dual effects of the progestin R5020 54. Parnes HL. Abrams JS. Tchekmedvian NS. Tait N. Aisner J 1991 on proteins released by the T47D human breast cancer cells. J A phase I/II study of high-dose megestrol acetate in the treatment Biol Chem 259:1231 of metastatic breast cancer. Breast Cancer Res Treat 18:171 37. Horwitz KB, Freidenberg GR 1985 Growth inhibition and increase 55. Pronzato P, Brema F, Amoroso D, Bertelli G, Conte PF, Martini

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER 161

MC, Pastorino G, Rosso R 1990 Megestrol acetate: phase II study synthesized intracellularly and are biologically functional. of a single daily administration in advanced breast cancer. Breast Biochem Bionhvs Res Commun 157:278 Cancer Res Treat 17:51 77. Wei LL, Gonzalez-Aller C, Wood WM, Miller LA, Horwitz KB 56. Henderson IC, Harris JR, Kinne DW, Hellman S 1989 Cancer of 1990 5’-heterogeneity in human progesterone receptor transcripts the breast. In: DeVita VT, Hellman S, Rosenberg SA (eds) Cancer: predicts a new amino-terminal truncated “C”-receptor and unique Principles & Practice of Oncology. JB Lippincott Company, Phil- A-receptor messages. Mol Endocrinol4:1833 adelphia, p 1197 78. Sherman MR, Corvol PL, O’Malley BW 1970 Progesterone-bind- 57. McGuire WL, Cavalli F, Bonomi P, Alexieva-Figusch J 1985 ing components of chick oviduct. J Biol Chem 245:6085 Pregestin therapy for breast cancer. Breast Cancer Res Treat 79. Schrader WT, O’Malley BW 1972 Progesterone-binding compo- 6:213 nents of chick oviduct. Characterization of purified subunits. J 58. McGuire WL, Johnson PA, Muss HB, Osborne CK 1989 Meges- Biol Chem 247:51 trol acetate in breast cancer-a panel discussion. Breast Cancer 80. Horwitz KB, Francis MD, Wei LL 1985 Hormone-dependent Res Treat 14:33 covalent modification and processing of human progesterone re- 59. Alexander IE, Shine J, Sutherland RL 1990 Progestin regulation ceptors in the nucleus. DNA 4:451 of estrogen receptor messenger RNA in human breast cancer cells. 81. Logeat F, Pamphile R, Loosfelt H, Jolivet A, Fournier A, Milgrom Mol Endocrinol 4:821 E 1985 One-step immunoaffinity purification of active progester- 60. Papa V, Reese CC, Brunetti A, Vigneri R, Siiteri PK, Goldfine one receptor: further evidence in favor of the existence of a single ID 1990 Progestins increase insulin receptor content and insulin steroid binding subunit. Biochemistry 24:1029 stimulation of growth in human breast carcinoma cells. Cancer 82. Conneely OM, Maxwell BL, Toft DO, Schrader WT, O’Malley Res 50:7858 BW 1987 The A and B forms of the chicken progesterone receptor 61. Murphy LC, Murphy LJ, Shiu RPC 1988 Progestin regulation of arise by alternate initiation of translation of a unique mRNA. EGF-receptor mRNA accumulation in T-47D human breast can- Biochem Biophys Res Commun 149:493 cer cells. Biochem Biophys Res Commun 150:192 83. Conneely OM, Kettelberger DM, Tsai M-J, Schrader WT, O’Mal- 62. Murphy LC, Dotzlaw H 1989 Endogenous growth factor expres- ley BW 1989 The chicken progesterone receptor A and B isoforms sion in T47D human breast cancer cells associated with reduced are products of an alternate translation initiation event. J Biol sensitivity to antiproliferative effects of progestins and antiestro- Chem 264:14062 gens. Cancer Res 49:599 84. Kastner P, Krust A, Turcotte B, Stropp U, Tora L, Gonemeyer 63. Murphy LC, Murphy LJ, Dubik D, Bell GI, Shiu RPC 1988 H, Chambon P 1990 Two distinct estrogen-regulated promoters Epidermal growth factor gene expression in human breast cancer generate transcripts encoding the two functionally different hu- cells: regulation of expression by progestins. Cancer Res 484555 man progesterone receptor forms A and B. EMBO J 9:1603 64. Green S, Chambon P 1986 A superfamily of potentially oncogenic 85. Kastner P, Bocquel M-T, Turcotte B, Garnier J-M, Horwitz KB, hormone receptors. Nature 324:615 Chambon P, Gronemeyer H 1990 Transient expression of human 65. Evans RM 1988 The steroid and thyroid hormone receptor super- and chicken progesterone receptors does not support alternative family. Science 240889 translational initiation from a single mRNA as the mechanism 66. Beato M 1989 Gene regulation by steroid hormones. Cell 56335 generating two receptor isoforms. J Biol Chem 265:12163 67. Carson-Jurica MA, Schrader WT, O’Malley BW 1990 Steroid 86. Feil PD, Clarke CL, Satyaswaroop PG 1988 Progesterone receptor receptor family: structure and functions. Endocr Rev 11:201 structure and protease activity in primary human endometrial 68. Jeltsch JM, Krozowski Z, Quirin-Stricker C, Gronemeyer H, carcinoma. Cancer Res 481143 Simpson RJ, Garnier JM, Krust A, Jacob F, Chambon P 1986 87. Birnbaumer ME, Schrader WL, O’Malley BW 1979 Chemical Cloning of the chicken progesterone receptor. Proc Nat1 Acad Sci cross-iinking of chick oviduct progesterone-receptor subunits by USA 83:5424 using a reversible bifunctional cross-linking agent. Biochem J 69. Conneely OM, Sullivan WP, Toft DO, Birnbaumer M, Cook RG, 181:201 Maxwell BL. Zarucki-Schulz T. Greene GL. Schrader WT. O’Mal- 88. Dougherty JJ, Toft DO 1982 Characterization of two 8 S forms of chick oviduct progesterone receptor. J Biol Chem 257:3113 ley BW 1986 Molecular cloning of the’chicken progesterone 89. Puri RK, Grandics P, Dougherty JJ, Toft DO 1982 Purification receptor. Science 233:767 of “nontransformed” avian progesterone receptor and preliminary 70. Misrahi M, Atger M, d’Aurio1 L, Loosfelt H, Meriel C, Fridlansky characterization. J Biol Chem 257:10831 F, Guiochon-Mantel A, Galibert F, Milgrom E 1987 Complete 90. Wei LL, Sheridan PL, Krett NL, Francis MD, Toft DO, Edwards amino acid sequence of the human progesterone receptor deduced DP, Horwitz KB 1987 Immunologic analysis of human breast from cloned cDNA. Biochem Biophys Res Commun 143:740 cancer progesterone receptors. 2. Structure, phosphorylation and 71. Meyer M-E, Pornon A, Ji J, Bocquel M-T, Chambon P, Grone- processing. Biochemistry 26:6262 meyer H 1990 Agonistic and antagonistic activities of RU486 on 91. Schuh S, Yonemoto W, Brugge J, Bauer VJ, Riehl RM, Sullivan the functions of the human progesterone receptor. EMBO J 9:3923 WP, Toft DO 1985 A 90,000-dalton binding protein common to 72. Wei LL, Krett NL, Francis MD, Gordon DF, Wood WM, O’Mal- both steroid receptors and the Rous Sarcoma virus transforming ley BW, Horwitz KB 1988 Multiple human progesterone receptor nrotein. DD~O’.““. J Biol Chem 260:14292 messenger ribonucleic acids and their autoregulation by progestin 92. Catelli ‘MG, Binart N, Jung-Testas I, Renoir JM, Baulieu EE, agonists and antagonists in breast cancer cells. Mol Endocrinol Feramisco JR, Welch WJ 1985 The common 90-kd protein com- 2:62 ponent of non-transformed ‘8s’ steroid receptors is a heat-shock 73. Read LD, Snider CE, Miller JS, Greene GL, Katzenellenbogen protein. EMBO J 4:3131 BS 1988 Ligand-modulated regulation of progesterone receptor 93. Pratt WB 1990 Interaction of hsp90 with steroid receptors: orga- messenger ribonucleic acid and protein in human breast cancer nizing some diverse observations and presenting the newest con- cell lines. Mol Endocrinol 2:263 cepts. Mol Cell Endocrinol 74:C69 74. Horwitz KB, Alexander PS 1983 In situ photolinked nuclear 94. Smith DF, Faber LE, Toft DO 1990 Purification of unactivated progesterone receptors of human breast cancer cells: subunit progesterone receptor and identification of novel receptor-asso- molecular weights after transformation and translocation. Endo- ciated proteins. J Biol Chem 265:3996 crinology 113:2195 95. Loaeat F. LeCunff M. Pamnhile R. Milerom E 1985 The nuclear 75. Lessey BA, Alexander PS, Horwitz KB 1983 The subunit struc- bo;nd form of the progesterone receptor is generated through a ture of human breast cancer progesterone receptors: characteriza- hormone-dependent phosphorylation. Biochem Biophys Res tion by chromatography and photoaffinity labeling. Endocrinol- Commun 131:421 ogy 112:1267 96. Dougherty JJ, Puri RK, Toft DO 1982 Phosphorylation in uiuo of 76. Krett NL, Wei LL, Francis MD, Nordeen SK, Gordon DF, Wood chicken oviduct nroaesterone recentor. J Biol Chem 257:14226 . WM, Horwitz KB 1988 Human progesterone A-receptors can be 97. Sheridan PL, Krett-NL, Gordon-JA, Horwitz KB 1988 Human

Downloaded from edrv.endojournals.org on February 27, 2005 162 HORWITZ Vol. 13, No. 2

progesterone receptor transformation and nuclear down-regula- 118. Sheridan PL, Evans RM, Horwitz KB 1989 Phosphotryptic pep- tionare independent of phosphorylation. Mol Endocrinol 2:i329 tide analysis of human progesterone receptors. New phosphory- 98. DeMarzo AM. Beck CA. Onate SA. Edwards DP 1991 Dimeriza- lated sites formed in nuclei after hormone treatment. J Biol Chem tion of mammalian progesterone receptors occurs in the absence 264:6520 of DNA and is related to the release of the 90-kDa heat shock 119. Sheridan PL, Francis MD, Horwitz KB 1989 Synthesis of human protein. Proc Nat1 Acad Sci USA 88:72 progesterone receptors in T47D cells. Nascent A- and B-receptors 99. Tora L, Gronemeyer H, Turcotte B, Gaub M-P, Chambon P 1988 are active without a phosphorylation dependent posttranslational The N-terminal region of the chicken progesterone receptor_ spec-_ maturation step. J Biol Chem 264:7054 ifies target gene activation. Nature 333:185 120. Becker PB, Gloss B, Schmid W, Strahle U, Schutz G 1986 In uiuo 100. Soelsbere TC. Halbere F 1980 Circannual rhvthms in steroid protein-DNA interactions in a glucocorticoid response element receptor concentration and nuclear binding in the chick oviduct. require the presence of the hormone. Nature 324:686 Endocrinology 107:1234 121. Willmann T, Beato M 1986 Steroid-free glucocorticoid receptor 101. Boyd-Leinen PA, Fournier D, Spelsberg TC 1982 Nonfunctioning binds specifically to mouse mammary tumour virus DNA. Nature progesterone receptors in the developed oviducts from estrogen- 324:688 withdrawn immature chicks and in aged nonlaying hens. Endo- 122. Denner LA, Weigel NL, Maxwell BL, Schrader WT, O’Malley crinology 111:30 BW 1990 Regulation of progesterone receptor-mediated transcrip- 102. Philibert D, Deraedt R, Teutsch G, RU38486, a potent antiglu- tion by phosphorylation. Science 250:1740 cocorticoid in uiuo. In Abstracts, VIIIth International Congress 123. Bailly A, Le Page C, Rauch M, Milgrom E 1986 Sequence-specific of Pharmacology, Tokyo, Japan, 1981 (No. 1463) DNA binding of the progesterone receptor to the uteroglobin gene: 103. Ravnaud J-P. Oiasoo T 1986 The design and use of sex-steroid effects of hormone, antihormone and receptor phosphorylation. antagonists. J Steroid Biochem 25:811 - EMBO J 5:3235. 104. Philibert D, Deraedt R, Teutsch G, Tournemine C, Sakiz E, RU 124. Klein-Hitpass L, Tsai SY, Weigel NL, Allan GF, Riley D, Rodri- 38486 -a new lead for steroidal anti-hormones. Program of the guez R, Schrader WT, Tsai M-J, O’Malley BW 1990 The proges- 64th Annual Meeting of The Endocrine Society, San Francisco, terone receptor stimulates cell-free transcription by enhancing California, 1982, p 246 (Abstract 668) the formation of a stable preinitiation complex. Cell 60~247 105. Hermann DW, Wyss R, Riondel A, Philibert D, Teutsch G, Sakiz 125. Bagchi MK, Tsai SY, Tsai M-J, O’Malley BW 1990 Identification E, Baulieu E-E 1982 The effects of an antiprogesterone steroid in of a functional intermediate in receptor activation in progester- women: interruption of the menstrual cycle and of early preg- one-dependent cell-free transcription. Nature 345:547 nancy. C R Acad Sci Paris 294:933 126. Schauer M, Chalepakis G, Willmann T, Beato M 1989 Binding of 106. Baulieu E-E 1989 Contragestion and other clinical applications hormone accelerates the kinetics of glucocorticoid and progesier- of RU 486, an antiprogesterone at the receptor. Science 245:1351 one recentor binding to DNA. Proc Nat1 Acad Sci USA 86:1123 107. Neef G, Beier S, Elger W, Henderson D, Wiechert R 1984 New 127. Bagchi MK, Elliston JF, Tsai SY, Edwards DP, Tsai M-J, O’Mal- steroids with antiprogestational and antiglucocorticoid activities. ley BW 1988 Steroid hormone-dependent interaction of human Steroids 44:349 progesterone receptor with its target enhancer element. Mol En- 108. Elger W, Beier S, Chwalisz K, Fahnrich M, Hasan SH, Henderson docrinol2:1221 D, Neef G, Rohde R 1986 Studies on the mechanisms of action of 128. Kalff M, Gross B, Beato M 1990 Progesterone receptor stimulates progesterone antagonists. J Steroid Biochem 25:835 transcription of mouse mammary tumour virus in a cell-free 109. Ojasoo T, Raynaud J-P, Mornon J-P 1989 Steroid hormone system. Nature 344:360 receptors. In: Sammes PG, Taylor JB (eds) Comprehensive Me- 129. Klein-Hitpass L, Cato ACB, Henderson D, Ryffel GU 1991 Two dicinal Chemistry, Volume 3: Membranes and Receptors. Perga- types of antiprogestins identified by their differential action in mon Press, New York, p 1175 transcriptionally active extracts from T47D cells. Nucleic Acids 110. Horwitz KB 1985 The antiprogestin RU38 486: receptor-mediated Res 19:1227 progestin uersu.s antiprogestin actions screened in estrogen-insen- 130. Guiochon-Mantel A, Loosfelt H, Ragot T, Bailly A, Atger M, sitive T47D,, human breast cancer cells. Endocrinology 116:2236 Misrahi M. Perricaudet M. MileromI E 1988 Recenters bound to 111. Renoir J-M, Radanyi C, Baulieu E-E 1989 The antiprogesterone antiprogestin form abortive complexes with hormone response RU486 stabilizes the heterooligomeric, non-DNA-binding, 8S- elements. Nature 336:695 form of the rabbit uterus cytosol progesterone receptor. Steroids 131. Tsai SY, Carlstedt-Duke J, Weigel NL, Dahlman K, Gustafsson 53:l J-A, Tsai M-J, O’Malley BW 1988 Molecular interactions of 112. Renoir J-M, Radanyi C, Jung-Testas I, Faber LE, Baulieu E-E steroid hormone receptor with its enhancer element: evidence for 1990 The nonactivated progesterone receptor is a nuclear heter- receptor dimer formation. Cell 55:361 ooligomer. J Biol Chem 265:14402 132. Fawell SE, Lees JA, White R, Parker MG 1990 Characterization 113. Horwitz KB, Pike AW, Gonzalez-Aller C, Fennessey PV 1986 and colocalization of steroid and dimerization activities in the Progesterone metabolism in T47D,, human breast cancer cells. mouse estrogen receptor. Cell 60:953 II. Intracellular metabolic path of progesterone and synthetic 133. Guiochon-Mantel A, Loosfelt H, Lescop P, Sar S, Atger M, progestins. J Steroid Biochem 25:911 Perrot-Applanat M, Milgrom E 1989 Mechanisms of nuclear 114. Rauch M, Loosfelt H, Philibert D, Milgrom E 1985 Mechanism localization of the progesterone receptor: evidence for interaction of action of an antiprogesterone, RU486, in the rabbit endome- between monomers. Cell 57:1147 trium. Effects of RU486 on the progesterone receptor and on the 134. Skafar D 1991 Differences in the binding mechanism of RU486 expression of the uteroglobin gene. Eur J Biochem 148:213 and progesterone to the progesterone receptors. Biochemistry 115. Mullick A, Katzenellenbogen BS 1986 Antiprogestin-receptor 30:10829 complexes: differences in the interaction of the antiprogestin 135. Strahle U, Klock G, Schutz G 1987 A DNA sequence of 15 base RU38,486 and the progestin R5020 with the progesterone receptor pairs is sufficient to mediate both glucocorticoid and progesterone of human breast cancer cells. Biochem Biophys Res Commun induction of gene expression. Proc Nat1 Acad Sci USA 84:7871 135:90 136. Jantzen H-M. Strahle U. Gloss B. Stewart F. Schmid W. Boshart 116. Groyer A, Schweizer-Groyer G, Cadepond F, Mariller M, Baulieu M, Miksicek ‘R, Schutz’G 1987 ‘Cooperativity of glucdcorticoid E-E 1987 Antiglucocorticosteroid effects suggest why steroid hor- response elements located far upstream of the tyrosine amino mone is required for receptors to bind DNA in uiuo but not in transferase gene. Cell 49:29 uitro. Nature 328624 137. Tsai SY, Tsai M-J, O’Malley BW 1989 Cooperative binding of 117. El-Ashry D, Oxiate SA, Nordeen SK, Edwards DP 1989 Human steroid hormone receptors contributes to transcriptional syner- progesterone receptor complexed with the antagonist RU486 gism at target enhancer elements. Cell 57:443 binds to hormone response elements in a structurally altered 138. Majors J, Varmus HE 1983 A small region of the mouse mammary form. Mol Endocrinol 3:1545 tumor virus long terminal repeat confers glucocorticoid hormone

Downloaded from edrv.endojournals.org on February 27, 2005 May, 1992 RU486 AND BREAST CANCER 163

regulation of a linked heterologous gene. Proc Nat1 Acad Sci USA 158. Sutherland RL, Hall RE, Pang GYN, Musgrove EA, Clarke CL 80:5866 1988 Effect of medroxyprogesterone acetate on proliferation and 139. Buetti E, Diggelmann H 1983 Glucocorticoid regulation of mouse cell cycle kinetics of human mammary carcinoma cells. Cancer mammary tumour virus: identification of a short essential DNA Res 485084 region. EMBO J 2:1423 159. Rochefort H, Chalbos 1985 RU 486: A full progestin antagonist 140. Nordeen SK, Suh BJ, Kuhnel B, Hutchinson CA 1990 Structural in human breast cancer cell lines. In: Baulieu E-E, Segal SJ (eds) determinants of a glucocorticoid receptor recognition element. The Antiprogestin Steroid RU 486 and Human Fertility Control. Mol Endocrinol 4:1866 Plenum Publishing Corporation, New York, p 307 141. Perlmann T, Eriksson P, Wrange 0 1990 Quantitative analysis 160. Gill PG, Vignon F, Bardon S, Derocq D, Rochefort H 1987 of the glucocorticoid receptor-DNA interaction at the mouse Difference between R5020 and the antiprogestin RU486 in anti- mammary tumor virus glucocorticoid response element. J Biol proliferative effects on human breast cancer cells. Breast Cancer Chem 265:17222 Res Treat l&37 142. Theveny B, Bailly A, Rauch C, Rauch M, Delain E, Milgrom E 161. Bardon S, Vignon F, Montcourrier P, Rochefort H 1987 Steroid 1987 Association of DNA-bound receptors. Nature 329:79 receptor-mediated cytotoxicity of an antiestrogen and an antipro- 143. Cato ACB, Miksicek R, Schutz G, Arnemann J, Beato M 1986 gestin in breast cancer cells. Cancer Res 47:1441 The hormone regulatory element of mouse mammary tumour 162. Michna H, Schneider M, Nishino Y, El Etreby MF, McGuire WL virus mediates progesterone induction. EMBO J 5:2237 1990 Progesterone antagonists block the growth of experimental 144. Cato ACB, Skroch P, Weinmann J, Butkeraitis P, Ponta H 1988 mammary tumors in Go/G1. Breast Cancer Res Treat i7:155 DNA sequences outside the receptor-binding sites differentially 163. Bowden RT. Hissom JR. Moore MR 1989 Growth stimulation of modulate the responsiveness of the mouse mammary tumour virus T47D human breast cancer cells by the anti-progestin_ - RU486. promoter to various steroid hormones. EMBO J 7:1403 Endocrinology 124:2642 145. Buetti E, Kuhnel B 1986 Distinct sequence elements involved in 164. Graham ML. Krett NL. Miller LA. Leslie KK. Gordon DF. Wood the glucocorticoid regulation of the mouse mammary tumor virus WM, Wei LL, Horwitz KB 1990 T47D,, cells, genetically unstable promoter identified by linker scanning mutagenesis. J Mol Biol and containing estrogen receptor mutations, are a model for the 190:379 progression of breast cancers to hormone resistance. Cancer Res 146. Burnstein KL, Bellingham DL, Jewel1 CM, Powell-Oliver FE, 50:6208 Cidlowski JA 1991 Autoregulation of glucocorticoid receptor gene 165. Bakker GH, Setyono-Han B, Portengen H, De Jong FH, Foekens expression. Steroids 56:52 JA, Klijn JGM 1989 Endocrine and antitumor effects of combined 147. Horwitz KB, Mockus MB, Lessey BA 1982 Variant T47D human treatment with an antiprogestin and antiestrogen or luteinizing breast cancer cells with high progesterone-receptor levels despite hormone-releasing hormone agonist in female rats bearing mam- estrogen and antiestrogen resistance. Cell 28633 mary tumors. Endocrinology 125:1593 148 Mockus MB, Horwitz KB 1983 Progesterone receptors in human 166. Bakker GH, Setyono-Han B, Portengen H, De Jong FH, Foekens breast cancer. Stoichiometric translocation and nuclear process- JA, Klijn JGM 1990 Treatment of breast cancer with different ing. J Biol Chem 258:4778 antiprogestins: preclinical and clinical studies. J Steroid Biochem 149. Horwitz KB. Mockus MB. Pike AW. Fennessev PV. Sheridan RL 37:789 1983 Progesterone receptor replenishment in T47Dhuman breast 167. Klijn JGM, De Jong FH, Bakker GH, Lamberts SWJ, Rodenburg cancer cells. Roles of protein synthesis and hormone metabolism. CJ, Alexieva-Figusch J 1989 Antiprogestins, a new form of endo- J Biol Chem 2587603 crine therapy for human breast cancer. Cancer Res 49:2851 150. Mullick A, Katzenellenbogen BS 1986 Progesterone receptor syn- 168. Michna H, Schneider MR, Nishino Y, El Etreby FM 1989 Anti- thesis and degradation in MCF-7 human breast cancer cells as tumor activity of the antiprogestins ZK98.299 and RU38.486 in studied by dense amino acid incorporation. Evidence for a non- hormone dependent rat and mouse mammary tumors: mechanistic hormone binding receptor precursor. J Biol Chem 261:13236 studies. Breast Cancer Res Treat 14:275 151. Nardulli AM, Katzenellenbogen BS 1988 Progesterone receptor 169. Schneider MR, Michna H, Nishino Y, El Etreby FM 1989 Anti- regulation in T47D human breast cancer cells: analysis by density tumor activity of the progesterone antagonists ZK 98.299 and RU labeling of progesterone receptor synthesis and degradation and 38.486 in the hormone-dependent MXT mammary tumor model their modulation by progestin. Endocrinology 122:1532 of the mouse and the DMBA- and the MNU-induced mammary 152. Horwitz KB, McGuire WL 1978 Nuclear mechanisms of estrogen tumor models of the rat. Eur J Cancer Clin Oncol25:691 action. Effects of estradiol and antiestrogens on estrogen receptors 170. Michna H. Schneider MR. Nishino Y. El Etrebv MF 1989 The and nuclear receptor processing. J Biol Chem 253:8185 antitumor mechanism of progesterone antagonists is a receptor 153. Horwitz KB, McGuire WL 1978 Estrogen control of progesterone mediated antiproliferative effect by induction of terminal cell receptor in human breast cancer. Correlation with nuclear proc- death. J Steroid Biochem 34:447 essinn of estrogen recenters. J Biol Chem 253:2223 171. Maudelonde T, Romieu G, Ulmann A, Pujol H, Grenier J, Khalaf 154. Hor&z KB, &ginger P, Kuttenn F, McGuire WL 1981 Nuclear S, Cavalie G, Rochefort H 1987 First clinical trial on the use of estrogen receptor release from antiestrogen suppression: amplified the antiprogestin RU486 in advanced breast cancer. In: Klijn induction of progesterone receptor in MCF-7 human breast cancer JGM. Paridaens R. Foekens JA (eds) Hormonal Maninulation of cells. Endocrinology 108:1703 Cancer: Peptides, Growth Factors’ and New (AntL)Steroidal 155. Keydar I, Chen L, Karby S, Weiss FR, Delarea J, Radu M, Agents. Raven Press, New York, p 55 Chaitcik S, Brenner HJ 1979 Establishment and characterization 172. Gaillard RC, Riondel A, Muller AF, Herrmann W, Baulieu E-E of a cell line of human carcinoma origin. Eur J Cancer 15:659 1984 RU486: a steroid with antiglucocorticosteroid activity that 156. Bardon S, Vignon F, Chalbos D, Rochefort H 1985 RU486, a only disinhibits the human pituitary-adrenal system at a specific progestin and glucocorticoid antagonist, inhibits the growth of time of day. Proc Nat1 Acad Sci USA 81:3879 breast cancer cells via the progesterone receptor. J Clin Endocri- 173. Murphy LC, Dotzlaw H, Johnson Wong MS, Miller T, Murphy no1 Metab 60:692 LJ 1991 Mechanisms involved in the evolution of progestin re- 157. Bakker GH, Setyono-Han B, Henkelman MS, De Jong FH, Lam- sistance in human breast cancer cells. Cancer Res 51:2051 berts SWJ. van der Schoot P. Kliin JGM 1987 Comnarison of the 174. Howell A, Harland RNL, Barnes DM, Hayward E, Redford J, actions of the antiprogestin mifepristone (RU486), the progestin Swindell R, Sellwood RA 1987 Endocrine therapy for advanced megestrol acetate, the LHRH analog buserelin, and ovariectomy carcinoma of the breast: effect of tumor heterogeneity and site of in treatment of rat mammary tumors. Cancer Treatment Rep biopsy on the predictive value of progesterone receptor estima- 71:1021 tions. Cancer Res 47:296

Downloaded from edrv.endojournals.org on February 27, 2005