Division of Tumour Biology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands

Correspondence to: R Michalides, E-mail: r.michalides@nki.nl

^aCurrent address: Inserm U482, Centre de Recherche St Antoine, 75571 Paris, Cedex 12, France

^bThese authors contributed equally to this work

Estrogen receptor-mediated transcription is enhanced by overexpression of G1/S cyclins D1, E or A in the presence as well in the absence of estradiol. Excess of G1/S cyclins also prevents the inhibition of transactivation of estrogen receptor (ER) by the pure antiestrogen ICI 182780. Cyclin D1 mediates this transactivation independent of complex formation to its CDK4/6 partner. This raises the possibility that overexpression of G1/S cyclins renders growth of ER-positive breast cancer hormone-independent and resistant to treatment with antiestrogens. Transient transfection of ER-positive breast cancer cell lines T47D and MCF7 with G1/S cyclins could overcome the growth arrest induced by ICI 182780 treatment. The ability of various cyclin D1 mutants to overcome the ICI 182780 mediated growth arrest corresponded with their ability to stimulate cyclin A- and E2F- promoter based reporter activities in the presence of ICI 182780. Transfection of a mutant cyclin D1 (cyclin D1-KE) that was unable to bind CDK4 and was reported to transactivate ER in the presence of ICI 182780, could not stimulate proliferation in ICI 182780 treated cells. On the other hand, cyclin D1-LALA, which is unable to stimulate ERE transactivation, could overcome the ICI 182780 cell cycle arrest. Furthermore, transient transfection of T47D cells using cyclin D1 together with a catalytic inactive mutant of CDK4 (CDK4-DN) indicated that the observed effect is due to binding to CDK inhibitors. However, a moderate, sixfold overexpression of cyclin D1 in stably transfected MCF7 cells did not overcome the ICI 182780 mediated growth arrest. These results indicate that CDK-independent transactivation of the estrogen receptor by cyclin D1 is by itself, not sufficient to result in estradiol-independent growth of breast cancer cells, whereas a vast overexpression of G1/S cyclins is able to do so, most likely by capturing of CDK inhibitors.

Oncogene (2002) 21, 8158-8165. doi:10.1038/sj.onc.1206012

antiestrogens; G1/S cyclins; breast cancer; cell cycle; estrogen receptor

Introduction

Normal breast epithelial cells and estrogen receptor (ER) expressing breast cancer cells require estrogens to stimulate proliferation. By acting through the ER, estrogen can regulate the transcription of its responsive genes, which in turn direct cellular proliferation. Therefore, patients with ER-positive breast tumours often undergo adjuvant treatment with antiestrogen tamoxifen. However, in one-third of these patients resistance to treatment with antiestrogens still emerges (Favoni and de Cupis, 1998). Several breast cancer cell lines have served as a model for estrogen responsive, antiestrogen sensitive breast tumours (Lippman et al., 1976; Musgrove et al., 1993). These studies reported that antiestrogen treatment or withdrawal of estrogens results in an arrest in the G1 phase of the cell cycle (Sutherland et al., 1983), implying a central role for cell cycle regulation in estrogen dependent proliferation. In this, transition of the G1 phase of the cell is controlled by the activation of CDK4/6 and CDK2 (Sherr, 1996). Binding to cyclins activates these CDKs. Active cyclin D-cdk4/6 complexes drive cells through the early G1 phase of the cell cycle, whereas the later G1 phase requires cyclin E-CDK2 activity. CDK-activity is additionally controlled by interaction with two families of CDK inhibitors (CDKi): the INK4 family of proteins (p15/p16/p18/p19) and the KIP family (p21/p27/p57) (Sherr and Roberts, 1999). The G1/S-cyclin/cdk activity promotes cell cycle progression by phosphorylation of key substrates including the retinoblastoma protein (pRb). Hyperphosphorylated pRb can no longer bind to, and thereby suppress the activity of E2F/DP family members that mediate transcription of G1/S phase genes.

Frequent overexpression and amplification of G1/S cyclin genes have been observed in a number of primary breast cancers and in tumour derived cell lines. Cyclin D1 is upregulated in up to 50% of breast cancers (Gillett et al., 1996; Michalides et al., 1996). Also cyclin E is upregulated in a significant proportion of breast cancers and this is associated with poor prognosis (Keyomarsi et al., 1994; Nielsen et al., 1999). The relevance of G1/S cyclin overexpression in breast cancer is further strengthened by the finding that tissue-specific transgenic expression of cyclin D1 or cyclin E in mice results in mammary hyperplasia and adenocarcinoma (Wang et al., 1994; Bortner and Rosenberg, 1997). Consistent with a pivotal role of cyclin D1 in control of growth of breast epithelium, the mammary glands of cyclin D1 knockout mice fail to undergo full development during pregnancy (Fantl et al., 1995; Sicinski et al., 1995). The above reports illustrate the significance of G1/S cyclin action as transducers for steroid hormone-induced breast cell proliferation. However, recent investigations have shown a potential deviant role for G1/S cyclins in regulating breast cancer cell proliferation. We and others have shown that ER-signalling is not only induced by estrogens, but ER is also activated in a hormone-independent manner by either phosphorylation of serines 104/106 of ER by cyclin A-cdk2 activity (Rogatsky et al., 1999) or by direct physical binding of cyclin D1 with the ER (Zwijsen et al., 1997; Neuman et al., 1997). Both modulations are thought to promote the interaction between ER and coactivators necessary for ER-dependent transcriptional activation (Rogatsky et al., 1999; Zwijsen et al., 1998), omitting the necessity for estrogen binding to the ER. This raises the possibility that overexpression of G1/S cyclins renders growth of ER-positive breast cancer hormone-independent and resistant to treatment with antiestrogens. Therefore, we investigated whether estrogen independent transactivation of the ER via overexpression of G1/S cyclins could overcome an ICI 182780 induced cell cycle arrest.

Results

Activation of ER in T47D cells

We examined at first the use of cell line T47D as a genuine model for estrogen responsive, antiestrogen sensitive breast tumours by measuring BrdU incorporation in asynchronous growing T47D cells that were treated with estradiol or with the antiestrogens tamoxifen or ICI 182780. As shown in Figure 1a, antiestrogens inhibit DNA-synthesis in T47D cells. The dependence of breast cancer cell proliferation on exogenous estradiol in the cell culture medium is not absolute, demonstrated by the effect of 5% CTS (charcoal treated serum) on T47D cell proliferation, as was also described earlier (Lippman et al., 1976). ICI 182780 is a steroidal antiestrogen with no agonist activities (Wakeling et al., 1991), and it decreased incorporation of 5-Bromo-2'deoxyuridine (BrdU) to approximately 2%. The non-steroidal antiestrogen tamoxifen exhibits both agonist and antagonist activities (Favoni and de Cupis, 1998). Although it decreased proliferation below that of 5% CTS-treated cells, it was much less potent than ICI 182780. For this reason, ICI 182780 was used as antiestrogen treatment in the remainder of our experiments. Treatment of T47D cells with ICI 182780 reduced cyclin A and cyclin D1 protein levels, reduced the activation of CDK2, as indicated by the Thr-160 band of CDK2, and resulted in a decrease of hyperphosphorylated pRB (Figure 1b), which is in line with previous reports (Foster and Wimalasena, 1996; Planas-Silva and Weinberg, 1997; Carroll et al., 2000). The diminished activities of cdk4 and cdk2, as judged by the phospho-specific antibodies pRB249/252 (CDK2) and pRB821 (CDK4), contributed both to this reduction of hyperphosphorylated pRb. These results show that T47D cells are clearly responsive to antiestrogen treatment.

Cyclin D1 and ER mediated proliferation

To determine the effect of cyclin overexpression on antiestrogen sensitivity, plasmids encoding G1/S cyclins were introduced in T47D cells together with a plasmid encoding Green Fluorescence Protein (GFP)-coupled Histone 2B (H2B-GFP) in a ratio of 10 : 1. This enabled us to test various combinations of input DNA using GFP expression as a marker of transfected cells (Strobeck et al., 2000). The ability of these transfected cells to incorporate BrdU was visualized and counted in GFP expressing cells (Figure 2c-f). The percentage of mock-transfected cells that incorporated BrdU was similar to non-transfected, ICI 182780 treated cells (Figure 2a,b), indicating that the transfection protocol did not cause proliferation. In contrast, overexpression of cyclin D1 resulted in a marked increase in the fraction of cells incorporating BrdU (Figure 2a) to levels comparable to that of estradiol treated cells (Figure 1a). Also overexpression of cyclin E and, to a lesser extent, of cyclin A could overcome ICI-mediated cell cycle arrest in this assay (Figure 2b).

Cyclin D1 protein is normally very unstable with an estimated half-life of <30 min. Cyclin D1 turnover is regulated by phosphorylation on threonine-286 followed by ubiquitination and proteosomal degradation (Diehl et al., 1997). Mutation of threonine-286 to an alanine residue (cD1-T286A) results in a markedly stabilized protein, which remains nuclear throughout the cell cycle. In sharp contrast to wild type cyclin D1, this stabilized cD1-T286A protein is able to induce cellular transformation and promote tumour growth in immune compromised mice (Alt et al., 2000). We investigated whether this cD1-T286A is also more potent in inducing ER-transactivation and DNA synthesis in ICI-treated T47D cells. We found that overexpression of this mutant protein, cD1-T286A, had an even higher ability to stimulate ER-activation (7.8-fold) than wild type cyclin D1 (5.6-fold) (Figure 2a). Accordingly, transient overexpression of cD1-T286A in ICI 182780-treated T47D cells resulted in the highest fraction of BrdU incorporating cells, which was comparable to estrogen treated control cells (Figure 2a). Representative photomicrographs of cells cotransfected with cD1-T286A and H2B-GFP, which incorporated BrdU, are shown in Figure 2c-f.

Cyclin D1 stimulates ER-activation independent of cdk4/6 activity (Zwijsen et al., 1997; Neuman et al., 1997), whereas also overexpression of cyclin E and cyclin A could stimulate estrogen-independent transactivation of ER (Rogatsky et al., 1999). We confirmed these previous findings using the several cD1-mutants in our present transfection protocol (Figure 2a,b). Strikingly, we found that transfection of a mutant cyclin D1 that was unable to bind CDK4 (cyclin D1-KE) but which transactivates ER in the presence of ICI 182780, could not stimulate proliferation in ICI 182780 treated cells. On the other hand, cyclin D1-LALA, which is unable to stimulate ER transactivation (Zwijsen et al., 1998), could overcome the ICI 182780 induced cell cycle arrest (Figure 2a). These findings indicate that CDK-independent transactivation of the estrogen receptor by cyclin D1 is not responsible for the observed resistance to antiestrogen treatment in breast cancer. Therefore, an alternative mechanism must be responsible for the ability of cyclin D1 overexpressing breast cancer cells to overcome an ICI 182780 mediated growth arrest. As a likely possibility, we examined whether excess cyclin D1/CDK complexes could act by titrating CDK-inhibitors away from endogenous CDKs resulting in their activation. Such a mechanism was strongly suggested from experiments showing that even kinase-inactive mutant forms are able to bind CDK inhibitors and thereby re-activating endogenous CDKs in synergism with exogenous cyclin D1 (Latham et al., 1996; Haas et al., 1997; Latella et al., 2001). To test this hypothesis, T47D cells were transfected with cyclin D1 and kinase inactive CDK4 mutant, CDK4-DN (van den Heuvel and Harlow, 1993) (Figure 2a). This combination was ineffective in estradiol-independent transactivation of the ER receptor (Zwijsen et al., 1997; Figure 2a). However, transient expression of cyclin D1 and CDK4-DN induces ICI 182780 resistance in T47D cells (Figure 2a), indicating that the most likely mechanism by which cyclin D1 induces antiestrogen resistance is by sequestrating CDK-inhibitors. This finding extends other reports, which show that antiestrogenic treatment leads to a reduced cyclin D1 expression and a consequent shift of CDKi p21 and to a lesser extent, p27 from the cyclin D1-cdk4/6 complexes to cyclin E-cdk2 complexes (Planas-Silva and Weinberg, 1997; Carroll et al., 2000). This redistribution of CDKi could be counteracted by antisense p21 and p27 treatment that also abrogates antiestrogen-mediated cell cycle arrest (Carroll et al., 2000; Cariou et al., 2000).

Functioning of promoter cyclin A and antiestrogen sensitivity

Since the protein level of cyclin A was reduced upon ICI 182780 treatment of T47D cells (Figure 1b), we examined whether transient overexpression of cyclin D1 could overcome this ICI 182780 mediated downregulation of cyclin A using a -7300/+11 nt cyclin A promoter-luciferase construct (Schulze et al., 1995) as a reporter of cyclin A transcription. ICI 182780 treatment resulted in a strong reduction of cyclin A promoter activity in control T47D cells (Figure 3a). Transient overexpression of wild type cyclin D1 could abolish the ICI 182780 mediated reduction of cyclin A promoter activity (Figure 3b). Furthermore, a positive correlation was observed between the ability to stimulate DNA synthesis and activation of the cyclin A promoter by cyclin D1 mutants (Figure 2a and 3b). These results indicate that transient overexpression of cyclin D1 is able to retain normal cyclin A transcription in ICI 182780 treated T47D cells, whereas cyclin A promoter activity is downregulated in T47D control cells.

Transient and stably transfected cyclin D1 and ER-mediated proliferation in MCF7 cells

Our results clearly showed that G1/S cyclins could confer antiestrogen resistance to T47D cells in a short-term assay (<48 h). In these short-term assays, supraphysiological levels of G1/S cyclins are reached. We therefore also examined whether G1/S cyclin overexpression under more physiological conditions was sufficient for antiestrogen insensitivity in breast cancer cells. For that purpose, we used MCF7/3 cells, which have a sixfold overexpression of cyclin D1 when cultured in the absence of tetracycline (-Tet) and express only endogenous cyclin D1 in the presence of tetracycline (+Tet) (Zwijsen et al., 1996). MCF7/3 cells overexpressing cyclin D1 exhibit an accelerated entry into S phase, a reduced serum-dependence (Zwijsen et al., 1996), a higher incorporation of (³H)-thymidine (see Figure 4a,b), an elevated level of cyclin A protein (compare zero day time points in Figure 4c), and an elevated phosphorylation of pRb and in vitro cdk2 kinase activity (data not shown) as compared to MCF cells with a normal level of cyclin D1. However, ICI 182780 treatment inhibited cellular proliferation of MCF7/3 cells with and without overexpression of cyclin D1 (Figure 4a,b), as we also reported previously (Pacilio et al., 1998). Also stable transfectants of MCF7 cells with an ectopic overexpression of cyclin E under a CMV-promoter could not overcome antiestrogen treatment in long-term assays (data not shown). The observed cell cycle arrest in MCF7/3 cells with a stable overexpression of cyclin D1 was not due to a downregulation of the ectopically expressed cyclin D1 (see Figure 4c). However, ICI 182780 treatment of MCF7/3 cells markedly reduced cyclin A protein levels irrespective of cyclin D1 levels (Figure 4c). This suggested that stable, sixfold overexpression of cyclin D1 was not sufficient to overcome the reduction of cyclin A protein levels after antiestrogen treatment. Transient transfection of cyclin D1 in the parental MCF7 cells yielded the same effects on BrdU incorporation and activation of a cyclin A promoter (-89/+11nt) based reporter activity in the presence of ICI182780 (Figure 5a,b) as was observed in T47D cells (Figures 2 and 3). The ability of cyclin D1 mutants to stimulate cyclin A promoter activity in the presence of ICI 182780 was reduced when the cyclin A promoter lacked the E2F binding site in the promoter-reporter construct (Figure 5b). This indicated that cyclin D1 stimulated cyclin A promoter activity, at least in part, via E2F.

Discussion

The results of this study indicate that transient overexpression of G1/S cyclins can overcome an ICI 182780 mediated cell cycle arrest. In case of cyclin D1, the observed antiestrogen insensitivity was not caused by CDK-independent activation of ER, but was likely due to sequestration of CDKi's by the vast excess of the exogenous cyclin D1. Why hormone-independent activation of the ER by cyclin D1 does not automatically result in hormone-independent growth is puzzling. We argue that the 5-6-fold activation of ER by cyclin D1 (Figure 2a) is either not enough as compared with the 25-30-fold activation of ER by estradiol (Figure 3a) to give a high enough expression of S-phase genes, or that only a subset of ER-responsive genes is activated by the ER/cyclin D1 interaction. This hypothesis is fortified by the observation that in lactating breast cells cyclin D1 needs co-stimulation with cAMP for ER-transactivation (Lamb et al., 2000). Furthermore, the present and our previous results (Pacilio et al., 1998) suggest that a stable, sixfold overexpression of cyclin D1 was too low to retain the resistance to antiestrogens. Alternatively, recent research has shown that besides cyclin D1, the CDK2 activating kinase CDC25A is also a potential target of antiestrogens (Cangi et al., 2000; Foster et al., 2001), suggesting that ectopic expression of cyclin D1 alone might not be sufficient to induce antiestrogen resistance. However, also ectopic expression of a constitutively active cdk2 mutant, CDK2-AF (Gu et al., 1992), or of CDC25A did not yield antiestrogen-insensitive proliferation in T47D cells, nor did they act in synergy with cyclin D1 (data not shown). It was furthermore shown that cyclin E expressed under the control of the natural cyclin D1 promoter (cyclin ED1 knock-in) can rescue the defects in the mammary gland development that are associated with cyclin D1 nullizygosity (Geng et al., 1999). This casted further doubt on the pivotal role of cyclin D1 in mammary gland development and suggested that cyclin E is a key regulator in mammary gland development and possibly also contributes to antiestrogen resistance in ER-positive breast cancer cells. But also here we found, however, that cyclin E overexpression could overcome an ICI 182780 induced cell cycle arrest only in transiently transfected cells.

Our present results clearly show that a vast overexpression of G1/S cyclins contributes to antiestrogen resistance. The mechanism behind this is a sustained E2F activity leading to transition through the ICI 182780 mediated growth arrest (see also Figure 5b). Essential to this is overexpression of G1/S cyclins, as is demonstrated in this study in which cell lines with a wild type pRb have been used. A similar effect has been reported for introduction of antisense p21 or p27 (Carroll et al., 2000; Cariou et al., 2000), and of a SV40 large T mutant that results in inactivation of pRb, leading to sustained E2F activity (Varma and Conrad, 2000). Translating these findings to the clinical situation, it is remarkable that overexpression of cylin A and absence of Rb expression in the primary breast cancer have been associated with failure to tamoxifen treatment in large series of breast cancer cases (Michalides et al., 2002; Anderson et al., 1996). In the former study cyclin D1 overexpression, however, was not found to be associated with tamoxifen resistance. This may be due to either a non-sufficient level of overexpression of cyclin D1, induction of apoptosis by overexpression of cyclin D1 (Sofer-Levy and Resnitzky, 1996), or due to cyclin D1 mediated induction of p21 that counterbalances the effect of overexpression of cyclin D1 (Hiyama et al., 1997; de Jong et al., 1999).

The effect of antiestrogens is influenced by the various ways of ER activation; estrogen independent, but still ERE dependent activation of ER by Protein kinase C and MAP kinase overcomes in part the effect of antiestrogens (Ali and Coombes, 2002). Furthermore, ER dependent stimulation and antiestrogen mediated inhibition of DNA synthesis also occurs independent of ERE when ER directs association between Ras-Raf1 and results in MAP kinase signaling. This mechanism is, however, still depending on estrogen (Castoria et al., 1999). The present study indicates deregulated control over G1/S cyclin activity as another putative cause of antiestrogen resistant proliferation in breast cancer, in addition to the various defaults in ER activation (Ali and Coombes, 2002).

Materials and methods

Proliferation assay and transfection of T47D and MCF7 cells

For transient transfections, 4´10⁵ cells were plated in 6-well culture dishes containing coverslips and incubated overnight. The expression vectors used in the transient transfection studies were pCMV-cyclin D1, cyclin A, cyclin E, cyclin D1-KE and cdk4 described by Hinds et al. (1994), pCMV-cyclin D1-Lala (Zwijsen et al., 1998), pCMV-cyclin D1-T286 (Diehl et al., 1997), pCMV-cdk4DN (Gu et al., 1992), and ERE-tk luciferase (Zwijsen et al., 1997). One g of plasmid DNA was added to the cells using Lipofectamine Plus reagents (Invitrogen) following the protocol of the manufacturer. In the BrdU experiments, co-transfections were done using 4.5 g input DNA plus 0.5 g H2B-GFP DNA (Kanda et al., 1998). After 3 h, the transfection medium was replaced by phenol red free DMEM containing 5% Charcoal Treated Serum (CTS) and 100 nM ICI 182780 (Tocris, UK). Next, cells were cultured for an additional 45 h. The last 4 h, cells were pulse labelled with 35 g/ml 5-Bromo-2'deoxyuridine (BrdU) (Roche) for 4 h. Cells were fixed and BrdU-positive nuclei were visualised as described by Strobeck et al. (2000). Proliferative activity was determined by counting BrdU-positive nuclei at a total of >100 nuclei in random areas in non-transfected cells and of at least 100 transfected cells (GFP-positive).

For ³[H]-Thymidine incorporation, cells were seeded in 24-well plates (10 000 cells/well), 24 h later the medium was replaced with fresh ones with or without ICI 182780 and cultured for the indicated period of time. ³[H]-Thymidine (2 Ci/ml) incorporation was evaluated by 1 h pulses. Thereafter, 400 l of 1M ascorbic acid was added to the medium, the cells were kept at 4°C for 4 h, were then incubated with trichloroacetic acid, TCA, (5%) at 4°C for 15 min, and washed with ice-cold TCA. Cells were then lysed with 300 l of 0.1 N NaOH for 30 min at room temperature, and radioactivity of a 250 l aliquot was measured with a liquid scintillation counter.

Antibodies used

Expression of proteins in cells, control and ICI 182780 treated was examined by SDS-PAGE (Laemmli, 1970), (7.5 or 12.5% acrylamide) followed by Western blot analysis with antibodies against cyclin A, cyclin E, CDK2, (Santa Cruz), cyclin D1 (Progen), pRb (Pharmingen), pRb 249/252, pRb 821 (Biosource) and -tubulin (Sigma). Immunoreactive proteins were detected using an ECL detection kit (Amersham).

Luciferase reporter assays

A Firefly luciferase construct carrying the region -7300 to +11 of the cyclin A promoter region (Schulze et al., 1995) was used to measure the ability of cells to transcribe cyclin A. We also used the shorter cyclin A promoter-reporter construct, containing the -89/+11nt fragment of the cyclin A promoter and the same fragment lacking the E2F binding site in that region (Zerfass-Thome et al., 1997). For their ability to transcribe ERE containing genes, we used the ERE-tk-luciferase construct as described by Zwijsen et al. (1997). T47D cells were transiently co-transfected with 4 g cyclin plasmid DNA, 1 g promoter cyclin A-luc or 1 g ERE-tk-luc and 5 ng SV40 Renilla luciferase using Lipofectamine Plus (Invitrogen). After 3 h, cells were washed and treated with phenol red free DMEM containing 5% CTS, 10 nM estradiol and either or not 100 nM ICI. After 48 h, cells were washed with PBS and subsequently lysed in 500 l Passive Lysis Buffer. Cellular debris was removed by spinning at 14 000 r.p.m. in a microcentrifuge for 2 min. The supernatant was stored at 80°C until use. The Firefly luciferase reporter and Renilla luciferase (internal control) activities of 15 l of lysate were measured employing the Dual Luciferase Reagent Assay Kit (Promega) using 75 l of LARII and 75 l of Stop and Glo buffers. Luminescence was determined in a luminometer (Berthold). Results are represented as mean±s.d. from a triplicate experiment.

Acknowledgements

We thank Drs Agami, Bernards, Chambon, Henglein and Morgan for providing various plasmid DNA's, and Dr Agami also for critical reading of the manuscript. This work was supported by grant 97-1431 of the Dutch Cancer Society to E Bindels and a grant from l'Arc to F Lallemand.

Ali S, Coombes RC. (2002). Nat. Rev. Cancer, 2: 101-112.

Alt JR, Cleveland JL, Hannink M, Diehl JA. (2000). Genes Dev., 14: 3102-3114. Article MEDLINE

Anderson JJ, Tiniakos DG, McIntosh GG, Autzen P, Henry JA, Thomas MD, Reed J, Horne GM, Lennard TW, Angus B, Horne CH. (1996). J. Pathol., 180: 65-70. MEDLINE

Bortner DM, Rosenberg MP. (1997). Mol. Cell Biol., 17: 453-459. MEDLINE

Cangi MG, Cukor B, Soung P, Signoretti S, Moreira G, Ranashinge M, Cady B, Pagano M, Loda M. (2000). J. Clin. Invest., 106: 753-761. MEDLINE

Cariou S, Donovan JCH, Flanagan WM, Milic A, Bhattacharya N. (2000). Proc. Natl. Acad. Sci. USA, 97: 9042-9046. MEDLINE

Carroll JS, Prall OWJ, Musgrove EA, Sutherland RL. (2000). J. Biol. Chem., 275: 38221-38229.

Castoria G, Barone MV, Di Domenico M, Bilancio A, Ametrano D, Migliacco A, Auricchio F. (1999). EMBO, 9: 2500-2510.

De Jong JS, van Diest PJ, Michalides RJAM, Baak JPA. (1999). J. Clin. Path. Mol. Pathol., 52: 78-83.

Diehl JA, Zindy F, Sherr CJ. (1997). Genes Dev., 11: 957-972. MEDLINE

Fantl V, Stamp G, Andrews A, Rosewell I, Dickson C. (1995). Genes Dev., 9: 2364-2372. MEDLINE

Favoni RE, De Cupis A. (1998). TIPS, 19: 406-415.

Foster JS, Wimalasena J. (1996). Mol. Endocrinol., 10: 488-498. MEDLINE

Foster JS, Henley DC, Bukovsky A, Seth P, Wimalasena J. (2001). Mol. Cell Biol., 21: 794-810. MEDLINE

Geng Y, Whoriskey W, Park MY, Bronson RT, Medema RH, Li T, Weinberg RA, Sicinski P. (1999). Cell, 97: 767-777. MEDLINE

Gilett C, Smith P, Gregory W, Richards M, Millis R, Peters G, Barnes D. (1996). Int. J. Cancer, 69: 92-99. Article MEDLINE

Gu Y, Rosenblatt J, Morgan DO. (1992). Embo J., 11: 3995-4005. MEDLINE

Haas K, Staller P, Geisen C, Bartek J, Eilers M, Möröy T. (1997). Oncogene, 15: 179-192. Article MEDLINE

Hinds PW, Dowdy SF, Eaton EN, Arnold A, Weinberg RA. (1994). Proc. Natl. Acad. Sci. USA, 91: 709-713. MEDLINE

Hiyama M, Iavarone A, Labaer J, Reeves SA. (1997). Oncogene, 14: 2533-2542. Article MEDLINE

Kanda T, Sullivan KF, Wahl GM. (1998). Curr. Biol., 8: 377-385. MEDLINE

Keyomarsi K, O'Leary N, Molnar G, Lees E, Fingert HJ, Pardee AB. (1994). Cancer Res., 54: 380-385. MEDLINE

Laemmli UK. (1970). Nature, 227: 680-685.

Lamb J, Ladha MH, McMahon C, Sutherland RL, Ewen ME. (2000). Mol. Cell. Biol., 20: 8667-8675. MEDLINE

Latella L, Sacco A, Pajalunga D, Tiainen M, Macera D, d'Angelo M, Felici A, Sacchi A, Crescenzi M. (2001). Mol. Cell. Biol., 21: 5631-5643. MEDLINE

Latham KM, Eastman SC, Wong A, Hinds PW. (1996). Mol. Cell. Biol., 16: 4445-4455. MEDLINE

Lippman M, Bolan G, Huff K. (1976). Cancer Res., 36: 4595-4601. MEDLINE

Michalides R, Hageman Ph, van Tinteren H, Houben L, Wientjens E, Klompmaker R, Peterse H. (1996). Br. J. Cancer,, 73: 728-734. MEDLINE

Michalides R, van Tinteren H, Balkenende A, Vermorken J, Benraadt J, Huldij J, van Diest P. (2002). Br. J. Cancer, 86: 402-408.

Musgrove EA, Hamilton JA, Lee CSL, Sweeney KJE, Watts CKW, Sutherland RL. (1993). Mol. Cell. Biol., 13: 3577-3587. MEDLINE

Nielsen NH, Loden M, Cajander J, Emdin SO, Landberg G. (1999). Breast Cancer Res. Treat., 74: 874-880.

Neuman EMH, Ladha N, Lin TM, Upton SJ, Miller SJ, Direnzo J, Pestell RG, Hinds PW, Dowdy SF, Brown M, Ewen ME. (1997). Mol. Cell. Biol., 17: 5338-5347. MEDLINE

Pacilio C, Germano D, Addeoa R, Altucci L, Petrizzi VB, Cancemi M, Cicatiello L, Salzano S, Lallemand F, Michalides RJAM, Bresciani F, Weisz A. (1998). Cancer Res., 58: 871-876. MEDLINE

Planas-Silva MD, Weinberg RA. (1997). Mol. Cell Biol., 17: 4059-4069. MEDLINE

Rogatsky I, Trowbridge JM, Garabedian MJ. (1999). J. Biol. Chem., 274: 22296-22302. MEDLINE

Schulze A, Zerfass K, Spitkovsky D, Middendorp S, Berges J, Helin K, Jansen-Durr P, Henglein B. (1995). Proc. Natl. Acad. Sci. USA, 92: 11264-11268. MEDLINE

Sherr CJ. (1996). Science, 274: 1672-1677. Article MEDLINE

Sherr CJ, Roberts JM. (1999). Genes Dev., 13: 1501-1512. MEDLINE

Sicinski P, Donaher JL, Parker SB, Li T, Fazeli A, Gardner H, Haslam SZ, Bronson RT, Elledge SJ, Weinberg RA. (1995). Cell, 82: 621-630. MEDLINE

Sofer-Levy Y, Resnitzky D. (1996). Oncogene, 13: 2431-2437. MEDLINE

Strobeck MW, Fribourg AF, Puga A, Knudsen ES. (2000). Oncogene, 19: 1857-1867.

Sutherland RL, Reddel RR, Green MD. (1983). Eur. J. Cancer Clin. Oncol., 19: 307-319.

Van den Heuvel S, Harlow E. (1993). Science, 262: 2050-2054.

Varma H, Conrad SE. (2000). Oncogene, 19: 4746-4753.

Wakeling AE, Dukes M, Bowler J. (1991). Cancer Res., 51: 3867-3873. MEDLINE

Wang TC, Cardiff RD, Zukerberg L, Lees E, Arnold A, Schmidt EV. (1994). Nature, 369: 669-671. MEDLINE

Zerfass-Thome K, Schulze A, Zwercke W, Vogt B, Hekin K, Bartek J, Henglein B, Jansen-Durr J. (1997). Mol. Cell. Biol., 17: 407-415. MEDLINE

Zwijsen RML, Klompmaker R, Wientjens EB, Kristel PM, Van der Burg B, Michalides RJ. (1996). Mol. Cell. Biol., 16: 2554-2560. MEDLINE

Zwijsen RML, Wientjens E, Klompmaker R, van der Sman J, Bernards R, Michalides RJAM. (1997). Cell, 88: 405-415. MEDLINE

Zwijsen RML, Buckle RS, Hijmans EM, Loomans CJM, Bernards R. (1998). Genes Dev., 12: 3488-3498. MEDLINE

Figure 1 Antiestrogens block proliferation of T47D breast cancer cells. (a) Asynchronously growing T47D cells were cultured in phenol red free DMEM containing 5% charcoal treated serum (CTS; Hyclone) and additions of either 10 nM estradiol (E2), 100 nM ICI 182780 (ICI) or 100 nM 4-hydroxytamoxifen (TAM). Cells were incubated for 44 h and then assayed for BrdU incorporation as described in Materials and methods. (b) Effect of ICI 182780 on the expression of cell cycle proteins in T47D cells. Cells were cultured for 48 h in the presence of 5% CTS +/- 100 nM ICI and analysed for the presence of indicated proteins by Western blotting as described in Materials and methods. An arrowhead depicts CDK2 activated by phosphorylation of Thr-160 and the asterisks represent the hyperphosphorylated forms of retinoblastoma protein (pRb)

Figure 2 Overexpression of G1/S cyclins results in antiestrogen resistance in T47D cells. T47D cells were transfected with the various cyclins and H2B-GFP and subsequently cultured in the presence of 5% CTS and ICI 182780. The ability of H2B-GFP expressing cells to replicate DNA was then assayed by BrdU incorporation using immunofluorescent visualisation as described in Materials and methods. (a). Effect of cyclin D1 and cyclin D1 mutants on proliferation of ICI 182780 treated T47D cells (black bars) and on proliferation in 5% CTS (white bars), (b) Effect of cyclin E or A on the proliferation of ICI 182780 treated T47D cells (black bars). As a comparison, control T47D cells in 5% CTS (white bars). The results represent the mean±s.e. of three independent experiments. In parallel to the proliferation assays, we tested the ability of G1/S cyclins to stimulate ER transactivation (the fold transactivation in relation to the mock-transfected cells is indicated below). (c-f) Immunofluorescent visualization of T47D cells transiently co-transfected with cD1-T286A and H2B-GFP and labelled with BrdU. (e) BrdU-incorporation was detected using a rhodamine-conjugated secondary antibody (red). (c) Total nuclei were visualized with Hoechst staining. (f) Overlay images were made to show nuclei that were positive for H2B-GFP (d) and BrdU-incorporation (e)

Figure 3 Effect of transient overexpression of cyclin D1 and cyclin D1 mutants on ICI 182780 mediated downregulation of cyclin A transcription. Transfected T47D cells were cultured for 48 h in the presence of 5% CTS and ICI 182780 and cell lysates were then tested for the ability to activate an ERE or a -7300/+ 11nt cyclin A promoter-reporter construct as described in Materials and methods. (a) ICI 182780 treatment reduces the expression of a cyclin A promoter- and ERE-reporter construct in T47D cells. (b) Overexpression of cyclin D1, cyclin D1-LALA or cyclin D1-T286A can overcome an ICI 182780 mediated downregulation of a cyclin A promoter-reporter construct

Figure 4 Effect of exogenous cyclin D1 expression on long-term ICI sensitivity in stably transfected MCF7/3 cells. MCF-7/3 cells were constructed expressing exogenous cyclin D1 under the control of a tetracycline (Tet) sensitive promoter (Zwijsen et al., 1996). (a,b) Quadruplicate cultures of MCF 7/3 cells were cultured in the presence (a) or absence (b) of 10 g/ml tetracycline and either normal medium (control) or medium containing 10 nM ICI 182780 (ICI). At the indicated times incorporation of [³H]-thymidine was measured by 1 h pulses. (c) Effect of ICI 182780 on the expression of cell cycle proteins in MCF7/3 cells. At the times indicated cells were isolated for Western blot analysis with antibodies against cyclin D1 or cyclin A

Figure 5 Effect of overexpression of cyclin D1 mutants on antiestrogen resistance in MCF7 cells. (a) Effect on proliferation. MCF7 cells were transfected with the various cyclin D1 mutants and H2B-GFP and subsequently cultured in the presence of 5% CTS containing ICI 182780. The ability of H2B-GFP expressing cells to replicate DNA was then assayed by BrdU incorporation using immunofluorescent visualization as described in Materials and methods. Cyclin D1 and cyclin D1-T286, but not cyclin D1-KE can overcome an ICI 182780 mediated growth arrest. (b) Effect on 182780 mediated downregulation of cyclin A transcription. Transfected MCF7 cells were cultured for 48 h in the presence of 5% CTS and ICI 182780 and cell lysates were then tested for the ability to activate cyclin A-promoter constructs containing either the -89/+11 nt region of the cyclin A promoter (cA-luc) or the -89/+11 nt region lacking the E2F binding sites (cA dE2F-luc) (Zerfass-Thome et al., 1997), as described in Materials and methods. Cyclin D1 and cyclin D1-LALA can overcome the ICI 182780 mediated downregulation of a cyclin A promoter-reporter construct, which is, in part, sensitive to E2F binding