Main

Breast cancer is the most prominent malignant neoplasm in women, at least in economically developed Western countries. Tamoxifen stands as the standard first-line endocrine therapy for female patients with breast carcinoma expressing oestrogen receptor alpha (ER), who represent the majority of breast cancer cases. Tamoxifen inhibits the growth of breast tumors by competitively antagonising oestrogen effect on cognate receptors (Bardon et al, 1984). It also has partial agonist (i.e. oestrogenic) activity which can be beneficial since it prevents bone loss in postmenopausal women (Love et al, 1994; Powles et al, 1996).

A novel approach for the hormonal therapy of breast carcinoma relies on the use of aromatase inhibitors (AIs) in place of tamoxifen. AIs markedly suppress plasma oestrogen level in postmenopausal women by inhibiting or inactivating aromatase enzymes responsible for the synthesis of oestrogen from androgenic substrates (Brodie and Njar, 2000). They also markedly inhibit intratumoral aromatase activity present in the majority of breast cancers (Harper-Wynne and Dowsett, 2001; Miller et al, 2001).

Last generation nonsteroidal AIs, such as letrozole (Femara®) and anastrozole (Arimidex®), are widely recommended as second-line hormonal agents for postmenopausal breast cancer patients (Goss and Strasser, 2001) and they have recently proven to be superior to tamoxifen as first-line treatment for advanced disease (Bonneterre et al, 2000; Nabholtz et al, 2000; Mouridsen et al, 2001) and, even more importantly, in the adjuvant setting as well (Baum et al, 2002). Thus, AIs are bound to progressively replace tamoxifen for the first-line treatment of postmenopausal patients with ER-positive cancers (Smith and Dowsett, 2003).

Nevertheless, the risk of osteoporotic fractures is likely to increase with the use of AIs since the maintenance of bone density partly depends on oestrogen, even in postmenopausal women. On the other hand, tamoxifen reduces bone mineral loss through its weak agonist effect, at least in postmenopausal women (Powles et al, 1996). Short-term use of letrozole has been shown to be associated with an increase in bone resorption markers in plasma and urine (Harper-Wynne et al, 2002; Heshmati et al, 2002) and adjuvant therapy with anastrozole is associated with a higher incidence of fractures than adjuvant therapy with tamoxifen (Boccardo et al, 2001; Baum et al, 2002).

It should be possible to prevent bone demineralization associated with oestrogen deprivation therapy by combining the latter with the use of bisphosphonates (Cummings et al, 1998). Bisphosphonates are indeed potent inhibitors of osteoclasts and represent the reference therapy for the management of postmenopausal osteoporosis (Fleisch, 2003). Moreover, many studies indicate that they prevent the development of metastatic bone diseases in animals (Yoneda et al, 2000). A few human trials have already been performed to evaluate oral clodronate in this setting. Although the clinical usefulness of this compound in the adjuvant treatment of breast cancer remains somewhat controversial (Diel et al, 1998; Saarto et al, 2001; Coleman, 2003), the largest trial so far has shown that clodronate can decrease the incidence of bone metastasis and prolong survival (Powles et al, 2002). Several trials are currently in progress or planned to disclose possible combined effect of AIs and various bisphosphonates in the treatment of breast cancers, with the prospect of preventing bone mineral loss and reducing the incidence of bone metastasis.

Bisphosphonates are also able to inhibit breast cancer cell growth in vitro (for a review, see Neville-Webbe et al, 2002). However, all studies reported so far were performed in steroid-containing medium. The present work has investigated the effects of clodronate on the growth of the ER-positive MCF-7 breast cancer cell line cultured in steroid-free medium (SFM) in order to mimic the effect of AIs in vivo. We have also analysed the effect of clodronate in combination with ER agonist and antagonists, with particular emphasis on ER expression and activity.

Material and methods

Cell culture conditions

The ER-positive MCF-7 breast cancer cell line (ATCC HTB-22) was initially obtained in 1977 from the Michigan Cancer Foundation (Detroit, MI, USA). IBEP-2 cell line was previously established in our laboratory from a pleural effusion due to metastatic breast carcinoma (Siwek et al, 1998) and also expresses functional ER. MDA-MB-231 breast carcinoma cells (ATCC HTB-26) lack ER expression.

All plastic flasks, dishes and multiwell plates for cell culture were obtained from Nunc™ (Naperville, IL, USA). Cells were cultured at 37°C in a humidified 95% air and 5% CO2 atmosphere.

For routine maintenance, cells were cultured in 75-cm2 flasks containing Eagle's minimum essential medium (MEM) with Phenol Red, supplemented with 10% heat-inactivated fetal calf serum (FCS), and containing standard concentrations of L-glutamine, penicillin and streptomycin (Gibco BRL, Life Technologies, Merelbeke, Belgium). Cells were harvested by trypsinisation (0.1% trypsin – 0.02% EDTA) and subcultured twice a week.

For all experiments, cells were plated in SFM (day 0) consisting of RPMI Medium 1640 without Phenol Red (Gibco BRL, Life Technologies, Merelbeke, Belgium), supplemented with 10% charcoal-stripped FCS as previously described (Devleeschouwer et al, 1987). At day 1, the seeding medium was replaced by fresh SFM containing clodronate (gift from Roche Diagnostics GmbH, Mannheim, Germany), 17β-oestradiol (E2; Sigma, St Louis, MO, USA), 4-hydroxy-tamoxifen (4-OH-TAM; Sigma, St Louis, MO, USA) fulvestrant (ICI 182,780; Tocris, Bristol, UK) at concentrations specified in ‘Results’. In control cultures, drug solution was replaced by an equivalent amount of vehicle. Treatment lasted from 1 to 3 days.

MVLN cells (generously provided by Dr M Pons, INSERM U58, Monpellier, France) are MCF-7 cells stably transfected with the oestrogen-response element cloned upstream the luciferase reporter gene (Demirpence et al, 1993). They were cultured in the same conditions as that used for MCF-7 cells.

Crystal violet staining

Cell number was assessed indirectly by staining with crystal violet dye, since the total amount of cell-bound dye is proportional to cell number (Lee et al, 2001). Cancer cells were seeded in 96-well plates (density 5000 cells well−1) in SFM or complete medium, and cultured for 24 h. Cells were then exposed to clodronate, E2, antioestrogens or vehicle as described in ‘Results’. After incubation, the medium was removed and cells were gently rinsed with PBS. Cells were fixed with 1% glutaraldehyde/PBS for 15 min and stained with crystal violet (0.1% w/v in dd H2O). Cell surfaces were destained under slightly floating tap water for 15 min and subsequently lysed with 0.2% Triton X-100 (v/v in dd H2O). The absorbance was measured at 550 nm using a Microplate Autoreader EL309 (BIO-TEK Instruments). Blanks (wells containing medium alone) were processed in parallel, and their values were subtracted from those of the sample plates.

Western blot analysis (ER and PgR)

Oestrogen receptor (ER) and progesterone receptor (PgR) amounts per milligram proteins were determined by Western blotting. Cells were plated in 60-cm2 Petri dishes (density 10 000 cells cm−2) in SFM and cultured for 24 h. After incubation with clodronate, E2 or vehicle as specified in ‘Results’, cell cultures were rinsed twice with tris-buffered saline (TBS, 50 mM Tris-HCl pH 7.5 and 150 mM NaCl) and harvested using 500 μl lysis buffer (TBS, 0.5% sodium deoxycholate, 1% Nonidet P-40, 0.1% SDS, 50 mM NaF, 0.1 mM Na3VO4 and 5 mM EDTA) with freshly added proteolysis inhibitors. Total cell lysates were mixed with standard electrophoresis sample buffer. Denatured samples (20 and 40 μg of cellular proteins for ER and PgR assays, respectively) were subjected to SDS–PAGE under reducing conditions. Proteins were subsequently electrotransferred onto a nitrocellulose membrane (Amersham Pharmacia Biotech, Roosendaal, The Netherlands). The membrane was then incubated for 3 h at room temperature in a blot solution (10 mM Tris-HCl pH 8, 150 mM NaCl, 0.05% Tween 20 and 7% skimmed milk) to block nonspecific binding, then incubated overnight at 4°C in a fresh blot solution containing polyclonal rabbit anti-human ERα antibody (HC-20, Santa Cruz Biotechnology, Santa Cruz, CA, USA) diluted 1 : 5000 or monoclonal mouse anti-human PgR (A/B isoforms) antibody (NCL-PGR-AB, Novocastra Laboratories, Newcastle upon Tyne, UK) diluted 1 : 500. Immunoblots were then incubated for 2 h at room temperature in a blot solution containing peroxidase-labelled donkey anti-rabbit IgG antibody (1 : 5000) or peroxidase-labelled sheep anti-mouse IgG antibody (1 : 5000) (Amersham Pharmacia Biotech, Roosendaal, The Netherlands), depending on primary used antibody. The bound peroxidase activity was revealed using the Lumi-Light Western Blotting Substrate (Roche Diagnostics GmbH, Mannheim, Germany). The immunoreactive band intensity was estimated using a computer-assisted gel scanning densitometer (GS-710 Callibrated Imaging Densitometer) and Quantity One software, both from Bio-Rad (Hercules, CA, USA).

Luciferase induction assay

MVLN cells were used to study the transcriptional activity of ER by determining ER-induced luciferase activity (Demirpence et al, 1993) using the Luciferase Assay System from Promega (Madison, WI, USA). Cells were plated in six-well plates (density 100 000 cells well−1) in SFM and were cultured for 72 h. At the end of treatment with clodronate, E2, antioestrogens or vehicle as described in ‘Results’, the medium was removed and the cell monolayers rinsed twice with PBS. A minimal volume (250 μl) of a five-fold diluted lysis solution (Promega E153A) was added and the culture was maintained under mild agitation for 20 min to extract luciferase. Detergent-lysed cells were scraped and suspensions were clarified by centrifugation for 5 min at 10 000 g. Finally, 20 μl of extracts were mixed at room temperature with 100 μl of a luciferase reactant medium (Promega E151A/E152A) prepared according to the manufacturer's protocol. Luminescence was measured in a Berthold luminometer (Lumat LB 9507). Luciferase induction was expressed in arbitary units (relative luciferase units, RLU) calculated per mg of protein and data are given in percent of mean value obtained from untreated cells.

Protein determination

Protein concentrations in total cell lysates obtained by detergent extraction were determined by the BCA Protein Assay (Pierce, Rockford, IL, USA) using bovine serum albumin (BSA) as standard.

Statistical analysis

Data are reported as means±s.d. Statistical analysis was performed using analysis of variance (ANOVA) with a statistical significance level arbitrarily set at 0.05 (Fisher's PLSD) (StatView version 4.02).

Results

Clodronate-induced stimulation of MCF-7 cell growth in SFM

MCF-7 cells were cultured in SFM for 24 h before additional exposure to 10−4 M clodronate or 10−9 M 17β-oestradiol (E2) during 72 h in order to assess cell growth. In these steroid-free culture conditions, clodronate stimulated cell proliferation as evidenced by a marked increase in cell number by 51±21% (mean±s.d., P<0.05) compared with controls (Figure 1). As expected, E2 also exerted a growth stimulatory effect (60±20%, P<0.05) under similar conditions. By contrast, neither clodronate nor E2 stimulated MCF-7 cell growth in medium supplemented with normal, steroid-containing serum (complete medium) (Figure 1). These results suggest that clodronate exerts a mitogenic activity on ER-positive breast cancer cells when the latter are cultured in absence of oestrogens.

Figure 1
figure 1

Clodronate stimulated MCF-7 cell proliferation in SFM but not in complete medium. Cancer cells were incubated for 72 h with 10−4 M clodronate (Clod), 10−9 M E2 or vehicle (control) in complete medium or SFM. Cell proliferation was determined by crystal violet staining assay (optical density for control value was 2.129 in complete medium, and 1.197 in SFM). Data are presented as percentages of control values (mean±s.d.). Mean of results is pooled from four experiments (n=24). *ANOVA, P<0.05 vs control.

Full size image

Time-course experiments performed over 72 h indicated that clodronate increased cell proliferation in a time-dependent manner similar to what is seen with E2 (Figure 2). Enhancement of cell growth was significant (P<0.05) after 48 h of exposure to either compound.

Figure 2
figure 2

Time-course experiments. Cells were cultured for up to 3 days in SFM containing 10−4 M clodronate (circle), 10−9 M E2 (square) or vehicle (control, triangle). Measurement was performed as described in Figure 1. *ANOVA, P<0.05 vs control.

Full size image

The dose–response relationships characterising the mitogenic effect of E2 or clodronate on hormone-deprived MCF-7 cells after 72 h of culture are illustrated in Figure 3 (upper panel). Clodronate (10−7–10−3 M) induced a dose-dependent response with a maximal stimulation of about two-fold at 10−3 M. 17β-oestradiol tested at concentrations ranging from 10−12 to 10−8 M produced a typical sigmoid dose–response curve with a maximal response similar to the one observed for clodronate but occurring at a much lower concentrations. The EC50 values (concentrations leading to half maximal stimulation) were around 6 × 10−5 and 2 × 10−12 M for clodronate and E2, respectively. Of note, because of the higher doses that were required, no true plateau phase was reached for clodronate. However, by analogy with the dose–response curve obtained with E2, the two-fold stimulation of growth induced by 10−3 M clodronate in MCF-7 cells was considered as the maximum effect.

Figure 3
figure 3

Dose-dependent effects of 17β-oestradiol and clodronate on the proliferation of MCF-7, IBEP-2 and MDA-MB-231 cells. Breast cancer cells were treated for 72 h with increasing concentrations of E2 (10−12–10−8 M), clodronate (Clod) (10−7–10−3 M) or vehicle (control) in SFM. The rest is the same as in Figure 1. *ANOVA, P<0.05 vs control.

Full size image

Moreover, we obtained closely similar findings using IBEP-2 cell line, which has recently been reported as an oestrogen-responsive cell line closely resembling the MCF-7 cell line (Journe et al, 2004a). We indeed showed that clodronate as well as E2 stimulated the proliferation of IBEP-2 cells cultured in SFM (Figure 3, middle panel), indicating that the mitogenic stimulation induced by the bisphosphonate was not restricted to the MCF-7 cells. By contrast, clodronate weakly inhibited the proliferation of MDA-MB-231 cells which are ER-negative breast cancer cells, thus nonsensitive to E2 stimulation (Figure 3, bottom panel). Altogether, these data suggest that ER status may be important in the mitogenic effects of clodronate.

Suppression of clodronate-induced MCF-7 cells growth by antioestrogens

As suggested above (see Figures 1 and 3), the ER pathway was somehow involved in the mitogenic effect of clodronate. We tested this hypothesis by using two ER antagonists: the partial antioestrogen 4-OH-TAM and the pure antioestrogen fulvestrant (ICI 182,780). IN SFM, exposure of MCF-7 cells to 10−7 M antioestrogen for 72 h reduced cell populations by 20±9% (mean±s.d., P<0.05) (4-OH-TAM) and by 40±6% (P<0.05) (fulvestrant) (Figure 4). This ability of antioestrogens to impede the growth of ER-positive breast cancer cells (either by interfering with cell proliferation and/or by inducing cell apoptosis) even in the absence of oestrogenic stimulation has also been found by other investigators (Jensen et al, 2003). Simultaneous addition of partial or pure antioestrogen with clodronate completely abolished the mitogenic effects of the bisphosphonate and decreased cell growth to values similar to those observed in the presence of antioestrogen alone. As expected, the proliferative response to 10−10 M E2 was similarly suppressed by concomitant addition of 4-OH-TAM or fulvestrant (Figure 4). Interestingly, previous exposure of MCF-7 cells to antioestrogen for 8 h prior to treatment with clodronate was as efficient as the addition of antioestrogen in combination with clodronate in suppressing the mitogenic effect of the latter (data not shown). These data suggest that preliminary blockade of ER-mediated signalling was sufficient to suppress clodronate-induced cell proliferation, and indicate that clodronate may indeed act through the ER pathway.

Figure 4
figure 4

Clodronate-induced cell growth was suppressed by antioestrogens. MCF-7 cells were exposed for 3 days to 10−4 M clodronate (Clod), 10−10 M E2, 10−7 M 4-OH-TAM (TAM), 10−7 M fulvestrant (ICI) or vehicle (control) in SFM. The rest is the same as in Figure 1. *ANOVA, P<0.05 vs control.

Full size image

Oestrogenic effects of clodronate in SFM: downregulation of oestrogen receptor, induction of progesterone receptor and transactivation of an oestrogen-responsive reporter gene

As suggested by the use of antioestrogens (see Figure 4), the mitogenic action of clodronate might be mediated by ER. This led us to investigate ER expression and activity under clodronate exposure.

We determined the effect of clodronate on ER expression by Western blot analysis using an antibody raised against the F domain of the receptor (Figure 5). The amount of immunoreactive ER was determined in MCF-7 cells after 24 h of treatment with clodronate and/or E2 in SFM. Densitometric analysis allowed semiquantitative evaluation of ER protein changes. We found that 10−4 M clodronate decreased ER amount by about 50%. As expected, 10−9 M E2 led to a more drastic downregulation of the ER protein. Interestingly, the combination of clodronate and E2 induced a complete disappearance of the receptor, suggesting that both compounds may produce additive effects on ER downregulation (Figure 5), and that the pathway for ER downregulation differs according to the drug.

Figure 5
figure 5

Clodronate downregulated oestrogen receptor expression as assessed by Western blot. MCF-7 cells were incubated for 24 h with 10−4 M clodronate (Clod), 10−9 M E2, Clod+E2 or vehicle (control) in SFM. Equal quantities of proteins (20 μg) were subjected to 10% SDS–PAGE and electrotransferred onto nitrocellulose membranes. Immunodetection was performed with anti-human ER antibody raised against its F domain. Blots show representative experiments performed twice. Quantitative data were obtained from densitometric analyses and are presented as percentages of control values (mean). ND=not detectable.

Full size image

We then investigated if clodronate could induce ER-mediated gene transactivation. The effect of clodronate on the oestrogen-inducible PgR gene was assessed by Western blot analysis using an antibody raised against the A/B isoforms of human PgR. Exposure of MCF-7 cells for 72 h to 10−4 M clodronate in SFM doubled the expression of PgR-B isoform (114 kDa, the isoform known to function as a transcriptional activator) (Figure 6). This induction was much less than the one observed with 10−9 M E2, which increased the PgR level by more than six-fold. The combination of clodronate and E2 did not further increase the amount of PgR (possible upper limit with this method beyond which no further increase in protein level can be documented). In control conditions, the A isoform of PgR (94 kDa, the isoform known to act as a transcriptional inhibitor) was not detectable in MCF-7 cells. Small amounts of the A isoform were observed after clodronate stimulation. Here also, E2 exerted a more pronounced effect on PgR-A expression. The B/A ratios determined for clodronate- and E2-induced stimulation were, however, similar (averaging 15), suggesting no difference in ER-mediated gene transactivation.

Figure 6
figure 6

Clodronate induced progesterone receptor expression as determined using Western blot. MCF-7 cells were incubated for 72 h in SFM containing 10−4 M clodronate (Clod), 10−9 M E2, Clod+E2 or vehicle (control). Equal amounts of proteins (40 μg) were subjected to 8% SDS–PAGE and electrotransferred onto nitrocellulose membranes. Immunodetection was assessed with anti-human PgR antibody raised against A/B isoforms of the receptor. Blots show representative experiments performed twice. Quantitative data were obtained from densitometric analyses of the B-isoform and are presented as percentages of control values (mean).

Full size image

Finally, the transcriptional activity of ER was investigated in MVLN cells obtained by stable transfection of MCF-7 cells with an oestrogen-responsive luciferase reporter gene. Culture for 24 h in SFM containing 10−4 M clodronate resulted in a significant increase in ER-induced luciferase expression by 106±56% (mean±s.d., P<0.05) (Figure 7). Reporter gene transactivation caused by 10−9 M E2 was stronger (279±64%, P<0.05). The activation of the reporter gene was further increased when clodronate and E2 were combined (431±106%, P<0.05). The stimulation of luciferase activity by clodronate and E2 were both completely abrogated by simultaneous incubation with 4-OH-TAM or fulvestrant, again suggesting that Clod may specifically induce ER activation.

Figure 7
figure 7

Clodronate stimulated ER transcriptional activity. MVLN cells were incubated in SFM for 24 h with 10−4 M clodronate (Clod), 10−10 M E2, 10−7 M 4-OH-TAM (TAM), 10−7 M fulvestrant (ICI), combination of Clod or E2 with antioestrogens, Clod+E2 or vehicle (control) in SFM. Luciferase activities were normalised with protein levels. Data are given as percentages of control values (mean±s.d.). Experiments were performed four times in replicate (n=8). *ANOVA, P<0.05 vs control.

Full size image

Discussion

Bisphosphonates, including clodronate, are widely used for the treatment of metastatic bone disease (Body et al, 1998; Lipton, 2003). In particular, they are administered to alleviate skeletal complications resulting from osteoclast-mediated tumour bone disease associated with breast and prostate cancers, and multiple myeloma. Although human data are still scarce, a double-blind trial involving more than 1000 breast cancer cases after surgery indicates that a 2-year treatment with clodronate can reduce the incidence of bone metastases by about one-half, and prolong patient survival (Powles et al, 2002).

Bisphosphonates are also currently evaluated in combination with third-generation AIs (Coleman, 2003), which appear to be more efficient than tamoxifen as a first-line treatment of advanced breast cancer (Bonneterre et al, 2000; Nabholtz et al, 2000; Mouridsen et al, 2001), as well as in the adjuvant setting (Baum et al, 2002). Thus, the administration of bisphosphonates, notably clodronate, in combination with AIs seems to be an attractive choice in order to prevent the risk of bone loss associated with complete oestrogen deprivation therapy, with the additional possibility of simultaneously decreasing the incidence of bone metastases (Cummings et al, 1998).

In our study, we investigated the effects of clodronate on the proliferation of MCF-7 cells cultured in SFM, an in vitro model that mimics breast tumours exposed to treatment combining the bisphosphonate clodronate and AI-induced oestrogen deprivation. Quite surprisingly, we found that clodronate stimulated the proliferation of MCF-7 cells in these particular conditions. Moreover, in the same conditions, clodronate exhibited oestrogen-like activity, inducing the downregulation of ER, the expression of progesterone receptor and the transactivation of a reporter gene. Hence, clodronate might increase MCF-7 cell proliferation through specific ER activation, as suggested by a complete suppression of its mitogenic effects by partial or pure antioestrogens.

The concentration of clodronate (10−4 M) that was needed to stimulate the growth of MCF-7 cells in SFM appears to be at least 10-fold higher than the circulatory levels in patients treated with bisphosphonates (Osterman and Lauren, 1991; Villikka et al, 2002). Nevertheless, it has been shown that bisphosphonates accumulate in bone (Lin et al, 1992), and that their effective local concentrations at sites of active bone resorption are much higher than serum levels, and may reach up to 10−3 M in the resorption lacunae (Sato et al, 1991). It is unclear whether the bioavailability of bone bisphosphonates for cancer cells would result in effective concentration as high as 10−4 M. However, we can speculate that the close association of breast cancer cells in chronically treated patients with actively resorbing osteoclasts within lytic lesions will result in exposure of the tumor cells to high levels of bisphosphonates. Therefore, it is likely that metastatic tumor cells in patients could be exposed to high doses of clodronate for long periods of time.

This is the first time that clodronate is reported to exert a significant proliferative action on MCF-7 breast cancer cells cultured in SFM for 2 or 3 days. Previous studies examining the influence of clodronate on MCF-7 cell growth have shown either no effect, a modest inhibition of cell growth after short exposure in serum-free medium or a moderate one after high-dose exposure or prolonged treatment (more than 6 days) in serum-containing medium (Busch et al, 1998; Fromigue et al, 2000; Senaratne et al, 2000). The weak proapoptotic effects of clodronate that has been occasionally observed could be attributed to the accumulation of toxic analogues of ATP generated by bisphosphonate metabolism (Frith et al, 1997). By contrast, in our study, clodronate showed no effect on cells grown in medium containing complete serum, suggesting that serum growth factors may prevent clodronate-induced apoptosis, and that serum steroids may obscure clodronate-induced cell growth.

Importantly, the oestrogenic stimulation induced by clodronate was not cell line specific. In SFM, we indeed obtained similar results using the IBEP-2 cell line which was previously established in our laboratory from breast carcinoma (Siwek et al, 1998). IBEP-2 cells have recently been characterised as an oestrogen-responsive cell line closely resembling the MCF-7 cells with regard to ER expression, cell mitogenic response to oestrogenic stimulation and sensitivity to antioestrogens (Journe et al, 2004a). We thus observed that clodronate stimulated IBEP-2 cell proliferation, and also provoked ER downregulation and induced PgR expression (two last data not shown). By contrast, the MDA-MD-231 cells, which do not express functional ER, exhibited no oestrogenic response to the bisphosphonate in SFM. Furthermore, at the highest concentration, clodronate inhibited MDA-MB-231 cell growth, probably because of the accumulation of nonhydrolysable analogues of ATP. Consequently, it appears that the mitogenic actions of clodronate can only be detected in oestrogen-responsive breast cancer cells cultured in SFM, the sole conditions appropriate for the observation of E2-induced stimulation.

Clodronate belongs to the group of non-nitrogen bisphosphonates closely resembling pyrophosphate. As mentioned above, it can be metabolised to nonhydrolysable analogues of ATP affecting numerous biosynthetic, metabolic or signal transduction pathways involving the hydrolysis or synthesis of ATP, including DNA synthesis, glycolysis and protein phosphorylation (Frith et al, 1997). Accumulation of toxic ATP analogues could account for the weak toxicity of clodronate reported in previous studies (Fromigue et al, 2000; Senaratne et al, 2000). However, in SFM, it is likely that the weak toxicity of nonhydrolysable ATP analogues may be masked by the oestrogenic effect of clodronate, as we describe in oestrogen-sensitive cell lines. The molecular mechanisms by which clodronate stimulates MCF-7 cell growth through ER activation in SFM remain a matter of speculation. Direct interaction of clodronate with ER, leading to the activation of the receptor, is conceivable. Indeed, pyrophosphate has been reported to enhance E2-mediated binding of rat uterine ER to oligodeoxythymidylate cellulose (Myatt et al, 1985). On the other hand, because of its close structural similarity with pyrophosphate, clodronate could alter intracellular processes involving protein phosphorylation/dephosphorylation. An attractive possibility would be that clodronate interferes with the activity of phosphatases similar to alendronate or tiludronate in osteoclasts (Schmidt et al, 1996; Murakami et al, 1997), and causes subsequent accumulation of phosphorylated key proteins implicated in mitogenic signaling pathways such as ERK1/2. Supporting this view, pervanadate, a potent inhibitor of tyrosine phosphatases, was reported to activate ERK1/2 (Zhao et al, 1996), and etidronate, another pyrophosphate-like bisphosphonate, has been reported to produce a rapid and transient increase in the phosphorylation of ERK1/2 (Plotkin et al, 1999). Hence, clodronate could act as a phosphatase inhibitor promoting ERK1/2 phosphorylation, which in turn could activate ER and stimulate its binding to DNA (at ERE and/or AP-1 sites), inducing gene transcription and further cell proliferation (Kushner et al, 2000). Of note, several lines of evidence indicate that E2 can also increase ERK1/2 phosphorylation in MCF-7 cells, thereby promoting cell growth through the activation of the MAPK pathway (Improta-Brears et al, 1999; Driggers and Segars, 2002). Further experiments are warranted to confirm this hypothesis.

By contrast, nitrogen-containing bisphosphonates (e.g. pamidronate, ibandronate and zoledronic acid) exhibit cytotoxic effects by interfering with the mevalonate pathway, leading to inhibition of the post-translational prenylation of small GTP-binding proteins such as Ras, that are necessary for cell function and survival (Luckman et al, 1998; Russell and Rogers, 1999). The antitumoral activity of nitrogen-containing bisphosphonates on breast cancer cells is now largely documented and has been associated with induction of apoptosis (Fromigue et al, 2000; Senaratne et al, 2000; Jagdev et al, 2001). Differences between clodronate and nitrogen-containing bisphosphonate actions were also observed in steroid-free culture conditions since 10−4 M ibandronate strongly inhibited MCF-7 cell growth (IC50=10−4 M) and did not affect ER expression or activity (Journe et al, 2004b).

In conclusion, we report a previously unknown mitogenic effect of the bisphosphonate clodronate on MCF-7 breast cancer cells cultured in SFM, an environment mimicking that created by the administration of AIs in vivo. Even though obtained by in vitro studies, our observations raise the possibility of growth stimulation of ER-positive breast cancer in patients receiving AIs and clodronate, a combined therapy that will probably be used extensively in future. Our data thus suggest that caution has to be exercised before extensive combined clinical use of AIs with clodronate. Further experimental work must, however, be undertaken to establish the relevance of our in vitro observations.