ELSEVIER Molecular and Cellular Endocrinology 102 (1994) 45-52

Synthetic progestins induce proliferation of breast tumor cell lines via the progesterone or estrogen receptor

Eric Kalkhoven *, Linda Kwakkenbos-Isbriicker, Siegfried W. de Laat, Paul T. van der Saag, Bart van der Burg

Hubrecht Laboratory, Netherlands Institute for Developmental Biology Uppsalalaan 8,3584 CT Utrecht, Netherlands

(Received 20 December 1993; accepted 1 February 1994)

Abstract

Proliferation of the human breast tumor cell lines T47D and MCF7 was stimulated by high concentrations (10e6 M) of the synthetic progestins gestodene and 3-ketodesogestrel, but not by Org2058, comparable to the stimulation by low dosages of estradiol (lo-” M). At physiological concentrations of the progestins (lo- lo M) only T47D cells responded. Using specific antihormones it was shown that the effect at pharmacological dosages is mediated by a crossreaction of these compounds with the estrogen receptor (ER), while the stimulation of T47D cells at physiological concentrations seems progesterone receptor (PR) mediated. This was further substantiated using transient transfection assays with ER- and PR-inducible reporter constructs and mRNA induction of the ER- and PR-target genes pS2 and fatty acid synthetase, respectively. Using a whole cell ligand binding assay, 20-fold higher amounts of PR were measured in T47D compared to MCF7 cells. This was in line with a much higher PR-dependent transactivation in T47D cells and suggests that the level of transcriptionally active PR is a major determinant for the response to physiological concentrations of progestins in human breast cancer cells.

Key words: Synthetic progestin; Proliferation; Breast cancer cell; Estrogen receptor; Progesterone receptor

1. Introduction

Sex steroids are important regulators of growth and development of the human mammary gland. A stimula-

tory role of estrogens is well established, but progestins may be important in regulating breast cell proliferation as well (King, 1991). In normal breast tissue the prolif- eration rate of epithelial cells coincides with high serum levels of progesterone during the menstrual cycle (Longacre and Bartow, 1986). Although experimental evidence is lacking momentarily, progestins may there- fore stimulate cell proliferation in this situation. Part of the human breast neoplasms are also steroid hor- mone-dependent for their proliferation. In vitro studies have firmly established that estrogens stimulate prolif- eration (Van der Burg et al., 1989 and references therein), but progestins can either stimulate (Horwitz and Freidenberg, 1985; Braunsberg et al., 1987; Hissom

* Corresponding author. Tel.: 030-510211; Fax: 030-516464.

and Moore, 1987; Papa et al., 19901, inhibit (Brauns- berg et al., 1987; Vignon et al., 1983; Sutherland et al., 1988; Poulin et al., 1990, Colletta et al., 1991; Gill et al., 1991) or have no effect (Schatz et al., 1985; Van der Burg et al., 1992) on the proliferation of hormone-de- pendent breast cancer cells.

Interestingly, it has been reported that synthetic progestins used in oral contraceptives (OCs) have very different effects on the proliferation of human breast tumor cells in vitro (Sutherland et al., 1988; Poulin et al., 1990; Colletta et al., 1991). We have previously compared the effect of four different synthetic pro- gestins (Org30659, gestodene, 3-ketodesogestrel and levonorgestrel) on the proliferation of the hormone-de- pendent breast tumor cell line MCM (Van der Burg et al., 1992). For this purpose the cells were grown under steroid- and growth factor-defined culture conditions. Under these conditions, MCF7 cells were found to be highly responsive to 17/3-estradiol (E2), especially in the additional presence of low concentrations of in- sulin (Van der Burg et al., 1988). When physiological

0303-7207/94/$07.00 0 1994 Elsevier Science Ireland Ltd. All rights reserved SSDI0303-7207(94)00037-A

46 E. Kalkhoven et al. /Molecular and Cellular Endocrinology 102 (1994) 45-52

dosages (1O-8-1O-9 M) of synthetic progestins were used, little evidence for a specific progesterone recep- tor (PR) mediated growth response was observed in this cell line, although it expressed functional PR. Interestingly, all progestins mentioned above were able to stimulate proliferation synergistically with insulin, but only at very high concentrations (low6 M), while E2 induces a similar response at much lower concen- tration (lO-i” M). Since anti-estrogens were able to block the progestin-mediated stimulation, it appeared to be mediated by a crossreaction with the estrogen receptor (ER) of these compounds (Van der Burg et al., 1992).

2.2. Cell culture and proliferation assay

MCF7 cells were kindly provided by Dr. C. Quirin- Stricker (Strasbourg, France) and T47D cells by Dr. R.L. Sutherland (Sydney, Australia). Cells were cul- tured in bicarbonate-buffered DF medium containing phenol red, supplemented with 7.5% FCS in a 7.5% CO, humidi~ed atmosphere. Cells were passaged twice a week using trypsin and EDTA.

Other studies have provided evidence that specific PR-mediated stimulation of proliferation can occur in certain situations (Braunsberg et al., 1987; Hissom and Moore, 1987; Papa et al., 1990). To assure that the lack of specific PR-mediated effects in MCF7 cells that we observed Wan der Burg et al., 1992) is not caused by the relatively low PR levels, similar experiments have now been carried out with the T47D cell line, which expresses higher PR levels than MCF7 (Sutherland et al., 1988 and this study). The suggestion that growth stimulation by high pha~acological dosages of syn- thetic progestins can occur through the ER was sub- stantiated by studying ER-mediated transcription and mRNA expression of EZregulated genes in both cell lines. In addition, at physiological concentrations, a specific PR-mediated stimulation was observed in the T47D cell line.

The effect of mitogens on DNA synthesis was tested as described (Van der Burg et al., 1988). In short, cells were plated at a density of 1.0 x 104/cm2 in DF con- taining 30 nm selenite, 10 pg/ml transferr~ and 0.2% BSA (referred to as DF+), supplemented with 5% DCC-FCS. After 24 h the medium was refreshed and the compounds to be tested were added in 2-[bis(2-hy- droxy-ethyljamino] ethanesulfonic acid (50 mM; pH 6.8) buffered DF”, After 4 days the total amount of DNA per well was determined by ~uore~ent staining with Hoechst 33248. Analysis in variance was used to test for differences in the means of data points (Snedecor and Cochran, 1967). Differences were con- sidered significant when P values were 0.05 or less.

2.3. Progesterone receptor (PRf assay

2. Materials and methods

Cells were grown in DF containing 5% DCC-FCS for 2 days, while control cells were grown in DF supplemented with 7.5% FCS. A whole cell PR binding assay was performed as described, for 1 h at 37°C (Van der Burg et al., 1992). Data were analyzed according to the method of Scatchard (19491, using LIGAND soft- ware.

2.1. Materials 2.4. RNA isolation and Northern blotting

A phenol red-free 1: 1 mixture of Dulbecco’s modi- fied Eagle’s medium and Ham’s F12 medium (DF) was obtained from Gibco (Grand Island, NY). Fetal calf serum (FCS) was purchased from Integro (Linz, Aus- tria), bovine insulin, E2, human transferrin and bovine serum albumin (BSA) were from Sigma (St Louis, MO). Trypsin and EDTA used for cell culture were obtained from Flow Laboratories (Irvine, UK). The progestins Org2058, gestodene and 3-ketodesogestrel were provided by Organon International (Oss, Nether- lands). RU486 was obtained from Roussel-U&f (Ro- mainville, France) and 4’-hydro~amo~fen from ICI Pharmaceuticals (Macclesfield, UK). DCC-FCS was prepared by treatment of FCS with dextran-coated charcoal (DC0 to remove steroids, as described (Van der Burg et al., 1988). DCC-SH-FCS was prepared by a treatment with dithiothreitol to inactivate ~l~eptide growth factors (Van Zoelen et al., 19851, and subse- quently treated with DCC, as described previously (Van der Burg et al., 1988).

Total RNA was isolated by the acid-phenol method of Chomczynski and Sacchi (1987). Northern blotting and (pre-jhybridization was carried out as described (Van der Burg et al., 1989). Probes were labeled with [Lu-32P]dCTP using the Multiprime DNA labeling sys- tem (~ersh~). Final, most stringent washing was done in 0.1 x SSC, 0.1% SDS at 68°C. Subsequently filters were exposed for autoradiography. The probes for the progestin-inducible fatty acid synthetase (FAS) mRNA (Chalbos et al., 1986) and the estrogen-induci- ble pS2 mRNA (Brown et al., 1984) were kind gifts of Dr. D. Chalbos and Dr. P. Chambon, respectively.

2.5. Plasmid constructs

The vitERE-tk-CAT construct was obtained by an- nealing the two complemental strands 5’-TCGACAG- GTCACAGTGACCTA-3’ and 5’-AGCTTAGGT- CACTGTGACCI’G-3’, and ligating it into a Hi&III- Sal1 digested pBLCAT2 vector (Luckow and Schiitz,

E. &&oven et ai /hfolecuiar and Cedar Endocrinology 102 (1994) 45-52 47

nated pG29G-tk-CAT, contains two PREs and was a kind gift of Dr. M. Muller (Schiile et al., 1988). Plas- mids were checked by dideoxy sequencing. The human PR-B cDNA (Misrahi et al., 1987) cloned into PGEM 32 using a PCR created EcoRI site on the 5’ end and a DruI site on the 3’ end of the cDNA. The cDNA was sub~quently cloned into a ~coRI-BumHI digested eukaryotic expression vector pSG5 (Green et al., 19881, using the EcoRI site of the cDNA and the BarnHI site of PGEM 32.

2.6. Transfection and CAT assay

The ceils were cultured in 6 well tissue culture plates in DF”, supplemented with 5% DCC-FCS. They were transfected by calcium phosphate co-precipitation using 5 kg CAT reporter, 2 pg PDM-LacZ plasmid (Boer et al., 1990) and in some experiments 1 pg of eukaryotic expression plasmid (pSG5; Green et al.,

T47D

90 .

0 11 10 9 6 7 6 -Log IHonnonol

6 80

a. 70

5 60

E so

3 ;I

20

1988) ~ntaining the human PR cDNA or the empty

expression plasmid. pUC18 plasmid was added to ob- tain a total amount of 10 pg DNA per well. After 6 h the medium was refreshed and (anti-) hormones were added. Cells were harvested 24 h later and assayed for CAT activity as described (German et al., 19821. The reaction products separated by thin-layer chromatogra- phy were quantized using a Molecular Dynamics Phos- pho Imager with Image Quant software. Values were corrected for transfection efficiency by measuring /3- galactosidase activity (Pfahl et al., 1990).

3. Results

3.1. Effects of ~~f~~~c pro~~r~~ on the proliferu~~n of breast tumor cells

In medium containing steroid-depleted serum @CC-FCS), the proliferation of T47D cells was stimu-

T47D

I I

0 11 10 6 6 7 6

-Log [Hormone1

MCF

0 11 10 6 6 7 6 0 11 10 9 8 7 6

-Lo6 [Hormonal -Lo6 [Hormone1

Fii. 1. The effect of progestins on the p~l~erat~n of T47D (A and 8) and MCF’? (C and D) cells.. Various ~n~ntrat~ns of Org2058 (A ; panel A and 0, gestodene (q panel B and Df, 3-ketodesogestrel ( r ; panel B and D) or E2 (m; panel A and C) were added to the ceils. After 4 days the total amount of DNA/well was determined as a measure for proliferation, relative to the stimulation by 10% FCS. Each paint represents the combined data of two indenendent ernerimenta with ckterminatinnc in trinlirat~ + %FM

lated by E2 in a dose-dependent manner (Fig. 1A): The synthetic progestins Org2058, gestodene and 3-ketode- sogestrel showed a stimulatory effect (40-45% of FCS values) at low dosages (10-i’ M), while a significantly stronger proliferative effect (60% of FCS values) was observed at high ~on~ntrations floe6 M) of gestodene and 3-ketodesogestrel, comparable to the stimulation caused by 10-i’ M E2 (Fig. 1). This latter effect did not occur with the specific progestin Org2058, indicat- ing that it is not PR mediated. When similar experi- ments were performed on T47D cells from a different source (American Type Culture Collection), similar results were obtained (data not shown).

Under the same conditions MCF7 cells were strongly stimulated by E2, while low dosages of the synthetic progestins did not affect proliferation (Fig. lC,D). However, high concentrations (10-7-10-6 M) of gesto- dene or 3-ketodesogestrel (Fig. lD), but not Org2058 (Fig. lC), strongly stimulated proliferation, confirming previous experiments using growth factor-defined con- ditions (Van der Burg et al., 1992). Interestingly, MCF7 cells respond to lower concentrations of E2 and the

T47D

stimulatory progestins when compared to T47D cells (Fig. 1).

Since progestins have been described to raise the insulin receptor (Horwitz and Freidenberg, 1985; Papa et al., 1990) and EGF receptor (Murphy et al., 1986) levels, the response of cells to progestins could be affected by the presence of the growth factors. There- fore, we also performed experiments in medium with steroid- and growth factor-depleted serum (Van der Burg et al., 1988). Under these ~nditions no effect of the additional presence of insulin or EGF was seen on the response to progestins (data not shown). We there- fore performed all subsequent assays in medium con- taining DCC-FCS.

To investigate which receptor is mediating growth stimulation in T47D cells by synthetic progestins, anti- hormones were used. Both antihormones used, the anti-progestin RU486 (Fig. 2A) and in particular the anti-estrogen 4’-hydroxytamoxifen (Fig. 2B) had signifi- cant agonistic effects. The stimulation at low dosages by the progestins (lO_” M) could be inhibited to the level in cells receiving antihormone only by RU486

T471

C 010 06 010 G6 KlO K6 C 010 06 610 G6 KlO K6

MCF7 MlX

Ru466 _DHl

C 010 06 010 06 K10 K6 C 010 06 GlO Gb KlO I(6

Fig. 2. Inhibition of progestin-induced proliferation by antihormones in T47D (A and B) and MC%7 (C and D) cells. Org2058 (01, gestodene (G) and 3-ketodesogestrel (K) were tested at 10F6 M (6) or at lo-” M (101, either alone (black bars), or in the additional presence of RU486 (panel A and C) or 4’-hydroxytamoxifen (OHT, panel B and D), both at low7 M (striped bars). After 4 days the total amount of DNA/well was determined as a measure for proliferation. Each point represents the combined data of two independent experiments with determinations in _ .~._.&. . net,

E. Kalkhoven et al. /Molecular and Cellular Endocrindoey 102 (1994) 45-52 49

(lo-’ M; Fig. 2A) and not by 4’-hydroxytamoxifen (Fig. 2B), suggesting that it is PR mediated.

However, the stimulation by gestodene or 3-ketode- sogestrel at high dosages apparently is ER mediated since it could be blocked by addition of 4’-hydroxy- tamoxifen (Fig. 2B), and not by addition of RU486 (Fig. 2A). In MCF7 ce 11 s, which lack a response to low dosages of the synthetic progestins (Fig. 11, the same result was found (Fig. 2C,D).

Taken together, these findings suggest that in hor- mone-dependent breast cancer cells, growth stimula- tion by low dosages of progestins takes place via the PR and depends on the cell type used, while stimula- tion by high dosages is mediated by ER and can be observed in both MCF7 and T47D cells.

3.2. Induction of ER- and PR-target gene expression

To further substantiate the role of the ER and PR in T47D and MCF7 cells by the synthetic progestins, we investigated their ability to induce ER- and PR-reg- ulated genes. We studied the expression of two genes which are transcriptionally up-regulated by their lig- ands: the pS2 gene (Brown et al., 1984) and the gene

A

B

%Ilc@=q* FAS

FAS

PS2

GAPDH

GAPDH

Fig. 3. Northern blot analysis of pS2 and FAS mFWA expression in MCF7 (A) and T47D cells (B). Total RNA was isolated from cells treated for 16 h with vehicle (-1, or steroids as indicated. The blot was hybridized with pS2 and FAS cDNA and subsequently with GAPDH EDNA as a control for equal loading (10 pg/lane).

encoding fatty acid synthetase (FAS; Chalbos et al., 19861, which are E2- and progesterone-inducible, re- spectively.

In MCF7 cells, pS2 mRNA expression was found to be elevated by low concentrations of E2 (10e9 M), while Org2058 and low concentrations of gestodene or 3-ketodesogestrel did not induce expression but re- duced basal expression to some extent. High concen- trations of gestodene and 3-ketodesogestrel did not reduce pS2 mRNA expression, consistent with their mixed estrogenic/progestagenic nature (Fig. 3A). In T47D cells the pS2 mRNA was not detectable on Northern blots, as has been reported before (Brown et al., 1984).

The ability of all progestins to induce PR-mediated gene regulation was confirmed by studying expression of the progesterone-inducible FAS gene. All progestins induced FAS mRNA to a similar level in T47D cells, whereas E2 had no effect (Fig. 3B). In MCF7 cells all progestins also appeared to be equally potent in induc- ing FAS expression (Fig. 3A), but here the level of induction is much lower, as can be expected from the lack of proliferative response of this cell line to pro- gestins (Fig. 1).

3.3. ER- and PR-mediated transcription by synthetic pr0gestin.s

We next studied ER-mediated transactivation in transient transfection assays using a reporter construct containing the estrogen response element (vitERE) present in the Xenopus vitellogenin promoter in front of the tk promoter coupled to CAT @k-CAT). We found that both gestodene and 3-ketodesogestrel stim- ulated transactivation of this construct at 10e6 M, while Org2058 had no effect in MCF7 (Fig. 4A) and T47D cells (Fig. 4B). This effect was clearly ER medi- ated and could be blocked by the anti-estrogen 4’-hy- dro~amo~fen, but not by the ~ti-progestin RU486 (Fig. 4). When dd d 1 a e a one, both 4‘-hydroxytamoxifen and to a lesser extent RU486 stimulated estrogen-de- pendent transactivation (respectively 2-5-fold and 1.5 fold). This agonistic activity is probably responsible for the stimulator effect of these compounds on the pro- liferation of T47D and MCF7 cells (Fig. 2).

To study PR-mediated transactivation we transiently transfected the cells with a reporter construct which contains two progesterone responsive elements (PRE) in front of tk-CAT. Fig. 5 shows that both MCF7 and T47D cells clearly expressed unctions .PRs and that transactivation occurred at low concentrations of the synthetic progestin Org2058 (10m9 M). Gestodene and 3-ketodesogestrel induced PR-mediated transactivation to a similar level (data not shown). Transactivation by Org2058 (Fig. 51, gestodene or 3-ketodesogestrel (data not shown) could be blocked by RU486.

50 E. Kalkhoven et al. /Molecular and Cellular Endocrinology 102 (1994) 45-52

A

4oljl ----- + + + + + -----

1.9 67.4 3.1 38.4 37.2 10.3 18.9 8.2 9.3 10.6 2.6 67.3 2.9 36.2 36.6

B

40H7 - - - - - + + + * + - - - - -

6.3 52.4 9.0 18.6 21.0 10.5 11.6 11.4 9.5 9.6 8.6 6l.l 9.1 17.1 t6.4

Fig. 4. Effect of hormones on ER-mediated transactivation. MCF7 cells (A) and T47D cells (B) were transiently transfected with a vitERE-tk-CAT construct for 6 h. The cells were then incubated with hormones for another 24 h. CAT activity is expressed as percentage conversion. A representative experiment is shown.

Whole cell l&and binding using cells grown in medium supplemented with 5% DCC-FCS, as used in our proliferation studies, showed that T47D cells ex- press approximately 820 000 progesterone receptors per cell (K, 0.5 nM), while MCF7 cells expressed 40-fold lower levels (22000 receptors/cell; K, 0.3 nM), con- firming the earlier reported difference between these cell lines (Sutherland et al., 1988). This could be the cause of the observed low PR-mediated transactivation in MCF7 cells compared to T47D cells (Fig. 5). Fig. 5 shows that indeed by ~-transfection with an eukary- otic expression vector containing the human PR-B cDNA Org2058-induced transactivation levels can be raised 6-fold in this cell line, and only 2-fold in T47D

cells. Therefore, the lack of specific PR-mediated ef- fects on the proliferation of MCF7 cells may be due to expression of a limiting amount of PR.

4. Discussion

Previously, we have shown that proliferation of MCF7 cells can be stimulated by high dosages of synthetic progestins (Van der Burg et al., 1992). In the present study, we demonstrate that this stimulation occurs via cross-reaction of these progestins with the ER. We also found that this phenomenon is not re- stricted to MCM cells but is also evident in T47D cells.

E. Kalkhoven et al. /Molecular and Cellular Endocrinology 102 (1994) 45-52 51

30

6 ‘-

3 20 ._

0 9

10

0

0

EZ +- +- +- +- -+ +-

R” -- ++ -- -- _- ++ -- --

-- --

PR-6 pSQ5 PR-5 pSG5

Fig.‘5 Effect of hormones on PR-mediated transactivation. MCF7 (A) and T47D cells (B) were transiently transfected with a PRE,-tk- CAT construct and with an expression vector containing the human PR-B cDNA or the empty expression vector (pSGS), as indicated, for 6 h. The cells were then incubated with E2 (E2, lo-’ MI, Org2058 (0, lo-’ M) and RU486 (RU, lo-’ M), as indicated, for another 24 h. CAT activity is expressed as fold induction over non-treated cells. Each point represents the combined data of two independent experi- ments with determinations in duplicate f SEM.

Furthermore, T47D cells also have a significant prolif- erative response to physiological concentrations (10-‘“-10-9 M) of progestins. As the PR is clearly activated at these concentrations, and since this effect is inhibited by antiprogestins, it seems to be specifically PR mediated. Therefore, the proliferative effects of 3-ketodesogestrel and gestodene are the result of a combined action of the ER or the PR, depending on the dosage used. Other studies in which the effects of progestins on cell proliferation were described (Horwitz and Freidenberg, 1985; Braunsberg et al., 1987; Hissom and Moore, 1987; Papa et al., 19901, confirm the possi- bility of a specific stimulation by progestins via the PR, since low concentrations of ligands were used (lo-*- 10m9 M), concentrations at which we found that spe- cific PR-mediated effects are predominant. However, our study clearly shows that the effect of progestins on proliferation is minor compared to that of estrogens, and is only observed in T47D cells in which the PR is overexpressed through amplification (Savouret et al., 1991). No effect on proliferation was observed in MCF7 cells despite expression of significant amounts of PR that are functional with respect to activation a PRE,- tk-CAT construct and activation of the endogenous FAS gene. Therefore, our study does not support a role for progestins in breast tumor proliferation except those in which PR overexpression occurs.

Besides growth stimulation by progestins, other studies have shown growth inhibition by this type of steroids (Horwitz and Freidenberg, 1985; Braunsberg et al., 1987; Vignon et al., 1983; Sutherland et al., 1988;

Poulin et al., 1990; Colletta et al., 1991; Gill et al., 1990, in some cases depending on the presence of E2 (Vignon et al., 1983; Hissom and Moore, 1987). We assume that the fact that our experiments were per- formed in the absence of E2 and phenol red, which has a weak estrogenic potential (Berthois et al., 1986), is one of the reasons why we did not observe growth inhibition by synthetic progestins. In the presence of E2 the T47D cell line which we use is inhibited by all progestins (unpublished data). Surprisingly, however, we did not find a specific inhibitory effects of gesto- dene, which has been reported to occur in the absence of E2 (Colletta et al., 1991). The reason for this dis- crepancy is unclear.

Other synthetic progestins used in OCs, like norethindrone, norethynodrel and norgestrel, have re- cently also been shown to stimulate proliferation of breast tumor cells when added in high dosages (Poulin et al., 1990; Jeng et al., 1992). Like we have shown for the progestins used in this study, action via the ER has been suggested. Interestingly, all these progestins share with E2 the 17/3-OH group in their structure, which has been implicated in E2 functioning. The absence of this hydroxyl-group on the strong progestin Org2058 could be the reason for its lack of estrogenic activity.

Studies in which estrogenic activity of progestins has been described involve hormone incubations of at least 24 h (Poulin et al., 1990; Jeng et al., 1992). In this period of time metabolization of progestins could oc- cur, so that the cause of the effects observed could not only be direct binding of the progestins but also bind- ing of their metabolization products to the ER. Al- though we cannot totally exclude metabolism, the fact that gestodene and 3-ketodesogestrel, like E2 (Van der Burg et al., 19891, induced c-fos mRNA expression within 30 min after their addition (unpublished data), makes this less probable. Moreover, this possibility was further made unlikely by whole cell ligand binding assays, involving hormone incubations of 1 h. At 10m6 M all estrogenic progestins were able to compete with E2 for ER binding in this assay (unpublished data).

Presently there is no evidence that the use of oral contraceptives by older women has a great impact on breast tumor development (reviewed by Thomas, 1991). This is in line with our data showing progestins to have no effect on progressed breast tumor cells expressing functional PR. Some epidemiological studies, however, have shown that certain subgroups may be at risk, especially younger women using OCs. This suggests that if there is any effect of OC components on breast tumor development it is likely to occur in an early stage of tumor development, e.g. on premalignant tissue. Since it has been found that proliferation of normal breast epithelium coincides with elevated progesterone levels (Longacre and Bartow, 19861, progestins may fulfill this role.

Estrogens play an important role in both the devel- opment and the progression of breast tumors (Hender- son et al., 1988). Although pharmacological dosages of the synthetic progestins tested in this study show estro- genie activity, we do not value the importance of this effect very high for the in vivo situation, since the endogenous levels of estrogens and levels included in OCs wil1 exceed the contribution made by these pro- gestins by far.

Acknowledgements

We thank Drs. P. Chambon and D. Chalbos for their kind gift of cDNA probes of pS2 and FAS, respectively. Drs. M. Misrahi and M.Muller are thanked for their respective gifts of the human PR-B cDNA and the PRE,-tk-CAT construct. Furthermore, we thank J. Heinen and F. Vervoordeldonk for their excel- lent photographic reproductions.

References

Berthois, Y., Katzenellenbogen, J.A. and Katzenellenbogen, B.S. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 2496-2500.

Boer, P.H., Potten, H., Adra, C.N., Jardine, K., Mulhofer, G. and McBumey, M.W. (1990) B&hem. Genet. 28,299-308.

Braunsberg, H., Coldham, N.G. Ieake, R.E., Cowan, SK. and Won& W. (1987) Eur. J. Cancer Clin. Oncol. 23, 563-571.

Brown, A.M.C., Jeltsch, J.-M., Peters, M. and Chambon, P. (19841 Proc. Natl. Acad. Sci. USA 81, 6344-6348.

Chalbos, D., Westley, B., May, F., Alibert, C., and Rochefort, H. (1986) Nucleic Acids Res. 14, 965-982.

Colletta, A.A., Wakefield, L.M., Howell, F.V., Danielpour, D., Baum M. and Sporn, M.B. (1991) J. Clin. Invest. 87, 277-283.

Chomczynski, P. and Sacchi, N. (1987) Anal. Biochem. 162,156-159. Gorman, C.M., Moffat, L.F. and Howard, B.H. (1982) Mol. Cell.

Biol. 2, 1044-1051. Gill, P.G., Tilley, W.D., De Young, N.J., Lensink, IL., Dixon, P.D.

and Horsfall, D.J. (19911 Breast Cancer Res. Treat. 20, 53-62. Green, S., Issemann, I. and Sheer, E. (1988) Nucleic Acids Res. 16,

369.

Henderson, B.E., Ross, R.K. and Bernstein, L. (19881 Cancer Res. 48,246-253.

Hissom, J.R. and Moore, M.R. (1987) Biochem. Biophys. Res. Com- mun. 14.5,706-711.

Hors&z, K.B. and Freidenberg, G.R. (1985) Cancer Res. 45,167-173. Jeng, M.-H., Parker, C.J. and Jordan, V.C. (1992) Cancer Res. 52,

6539-6546. King, R.B.J. (1991) J. Steroid B&hem. Mol. Biol. 39, 811-818. Longacre, T.A. and Bartow, S.A. (1986) Am. J. Surg. Pathol. 10,

382-393. Luckow, B. and Schiltz, G. (1987) Nucleic Acids Res. 15, 5490. Misrahi, M., Atger, M., d’Aurio1, L., Loosfelt, H., Meriel, C., Frid-

iansky, F., Guiochon-Mantel, A., Galibert, F. and Milgrom, E. (1987) B&hem. Biophys. Res. Commun. 143,740-748.

Murphy, L.J., Sutherland, R.L., Stead, B., Murphy, L.C. and Lazarus, L. (1986) Cancer Res. 46, 728-734.

Papa, V., Reese, C.C., Brunetti, A., Vigneri, R., Siiteri, PK. and Goldfine I.D. (1990) Cancer Res. 50,7858-7862,

Pfahl, M., Tzuke~n, M., Zhang, X-K, Lehmann, J.M., Hemamr, T., Wills, K.N. and Graupner, G. (19901 Methods Enzymol. 189, 256-270.

Poulin, R., Baker, D., Poirier, D. and Labrie, F. (19%) Breast Cancer Res. Treat. 17, 197-210.

Savouret, J.-F., Fridlanski, F., Atger, M., Misrahi, M., Berger, R. and Milgrom, E. (1991) Mol. Cell. Endocrinol. 75, 157-162.

Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51, 660-672. Schatz, R.W., Soto, A.M. and Sonnenschein, C. (1985) J. Cell.

Physiol. 124, 386-390. Schiile, R., Muller, M., Kaltschmidt, C. and Renkawitz, R. (19881

Science 242,1418-1420. Snedecor, G.W. and Cochran, W.G. (1967) Statistical Methods, Iowa

State University Press, Ames, L4. Sutherland, R.L., Hall., R.E., Pang, G.Y.N., Musgrove, EA. and

Clarke, CL. (1988) Cancer Res. 48,5084-5091. Thomas, D.B. (1991) In Oral Contraceptives and Breast Cancer

(Committee on the relationship between oral contraceptives and breast cancer, Institute of Medicine), 102-142, National Academy Press, Washington.

Van der Burg, B., Rutteman, G.R., Bla~enstein, MA. and de Laat, S.W. (1988) J. Cell. Physiol. 123, 101-108.

Van der Burg, B., van Selm-Miltenburg, A.J.P., de Laat, S.W. and van Zoelen, E.J.J. (1989) Mol. Cell. Endocrinol. 64, 223-228.

Van der Burg, B., Kalkhoven, E., Isbrilcker, L. and de Laat, S,W. (1992) J. Steroid Biochem. Mol. Biol. 42, 457-465.

Van Z&en, E.J.J., van Oostwaard, Th.M.J., Van der Saag, P.T. and de Laat, SW. (1985) J. Cell. Physiol. 123, 151-160.

Vignon, F., Bardon, !I., Chalbos, D. and Rochefort, H. (1983) J. Clin. Endocr. Metab. 56, 1124-1130.