These two authors contributed equally to this work.
Effects of raloxifene on breast cancer cell migration and invasion through the actin cytoskeleton
Article first published online: 16 SEP 2008
© 2008 The Authors Journal compilation © 2009 Foundation for Cellular and Molecular Medicine/Blackwell Publishing Ltd
Journal of Cellular and Molecular Medicine
Volume 13, Issue 8b, pages 2396–2407, August 2009
How to Cite
Flamini, M. I., Fu, X.-D., Sanchez, A. M., Giretti, M. S., Garibaldi, S., Goglia, L., Pisaneschi, S., Tosi, V., Genazzani, A. R. and Simoncini, T. (2009), Effects of raloxifene on breast cancer cell migration and invasion through the actin cytoskeleton. Journal of Cellular and Molecular Medicine, 13: 2396–2407. doi: 10.1111/j.1582-4934.2008.00505.x
- Issue published online: 26 NOV 2009
- Article first published online: 16 SEP 2008
- Received: July 9, 2008; Accepted: September 3, 2008
- breast cancer;
Raloxifene (RAL) is a selective oestrogen receptor modulator (SERM) approved for the prevention and treatment of osteoporosis and for the prevention of breast cancer in postmenopausal women. However, little is known on the effects of this SERM on breast cancer cell metastasis, which is the main cause of morbidity and death. Cell movement is critical for local progression and distant metastasis of cancer cells. These processes rely on the dynamic control of the actin cytoskeleton and of cell membrane morphology. The aim of the present study was to characterize the effects of RAL or of 17β-estradiol (E2) plus RAL on oestrogen receptor (ER) positive T47-D breast cancer cell cytoskeletal remodelling, migration and invasion. Our findings show that, when given alone, RAL induces a weak actin cytoskeleton remodelling in breast cancer cells, with the formation of specialized cell membrane structures implicated in cell motility. However, in the presence of physiological amounts of estradiol, which potently drives breast cancer cell cytoskeletal remodelling and motility, RAL displays a powerful inhibitory effect on oestrogen-promoted cell migration and invasion. These actions are plaid through an interference of RAL with an extra-nuclear signalling cascade involving G proteins and the RhoA-associated kinase, ROCK-2, linked to the recruitment of the cytoskeletal controller, moesin. Hence, in the presence of estradiol, RAL acts as an ER antagonist. These results highlight a novel mechanism of action of the SERM raloxifene that might be important for the interference of breast cancer progression or metastasis induced by oestrogens in postmenopausal women.
Hormone replacement therapy (HRT) alleviates menopausal symptoms and protects postmenopausal women from osteoporosis. However, HRT is associated with a slightly increased risk of breast cancer. This adverse effect limits to some extent the clinical use of HRT for long periods for the prevention of osteoporosis .
Selective oestrogen receptor modulators (SERMs) are a family of compounds that interact with oestrogen receptors (ER) in target tissues inducing functional effects that are distinct from those of oestrogens . Raloxifene belongs to this family and is approved for the prevention and treatment of postmenopausal osteoporosis. This compound acts as an oestrogen agonist in the skeleton and on the cardiovascular system, whereas it counteracts the effects of oestrogen in the breast and endometrium . Raloxifene reduces the risk of breast cancer and of cardiovascular events in older women with osteoporosis . Recently, the NSABP Study of Tamoxifen and Raloxifene (STAR) trial also demonstrated that raloxifene is as effective as tamoxifen in reducing the risk of invasive breast cancer in postmenopausal women . Moreover, compared with tamoxifen, raloxifene does not increase the risk of endometrial cancer . Therefore, this compound represents a potentially good tool to maintain and improve bone quality in postmenopausal women while reducing breast cancer risk. Recently, raloxifene has received from the FDA the indication for its use as primary chemoprevention of breast cancer in postmenopausal women with osteoporosis or at high risk for breast cancer.
The clinical efficacy of adjuvant therapy with tamoxifen on oestrogen receptor positive (ER+) breast cancers is well-established . However, little is known on the possible effects of raloxifene on breast cancer progression and metastasis. The ability of cancer cells to migrate to distant organs, which is essential for cancer spread, invasion and metastasis, is achieved through a complex array of functional modifications. One of the initial steps is the organization of dynamic rearrangements of the actin cytoskeleton leading to changes in cell morphology and to the development of membrane structures bridging the cytoskeleton to the extracellular matrix . Sex steroids, including oestrogen and progestins, promote T47-D breast cancer cell migration and invasion through rapid signalling to the actin cytoskeleton . These actions are mediated by the rapid activation of the actin-binding protein moesin, which belongs to the ezrin/radixin/moesin (ERM) family. Activated moesin triggers the de-polymerization of actin fibres and the re-assembly toward the cell membrane edge, leading to the formation of cortical actin complexes and specialized cell membrane structures implicated in the generation of the cellular locomotive force . Interestingly, ERM proteins have been recently reported as key regulators of metastasis in aggressive cancers .
The aim of the present study was to investigate the effects of raloxifene on human breast cancer cell migration and invasion and the underlying mechanisms involved. To this aim, we studied the actions of raloxifene alone or combined with estradiol on the activation of moesin and on actin remodelling in ER+ breast cancer cells.
Materials and methods
Cell cultures and treatments
The human breast carcinoma cell line T47-D was obtained from the American Type Culture Collection. T47-D cells were grown in RPMI 1640 supplemented with L-glutamine (2 mM), 10% foetal bovine serum. Before treatments, T47-D cells were kept 24 hrs in medium containing steroid-deprived FBS. Before experiments investigating non-transcriptional effects, cancer cells were kept in culture medium containing no FBS for 8 hrs. Whenever an inhibitor was used, the compound was added 30 min. before starting the treatments. Raloxifene was obtained from Eli Lilly Research Laboratories, 17β-estradiol, PTX, Y-27632, PD98059 were from Sigma-Aldrich (Saint-Louis, MO, USA), ICI 182,780 was obtained by Tocris Cookson (Avonmouth, UK). Raloxifene and PD98059 were dissolved in DMSO, 17β-estradiol, and ICI 182,780 were dissolved in 100% ethanol. PTX and Y-27632 were dissolved in PBS. The final concentration of the solvents was 1 μl of solvent per 1 ml of medium.
Cell lysates were separated by SDS-PAGE. Antibodies used were: moesin (clone 38, Transduction Laboratories, Lexington, KY), Thr558-P-moesin (sc-12895, Santa Cruz Biotechnology, Santa Cruz, CA, USA). Primary and secondary Abs were incubated with the membranes with standard technique. Immunodetection was accomplished using enhanced chemiluminescence.
T47-D breast cancer cells were grown on coverslips. Cells were fixed with 4% paraformaldehyde for 30 min. and permeabilized with 0.1% Triton for 5 min. Blocking was performed with PBS containing 3% BSA for 30 min. Then cells were incubated with Texas Red-phalloidin (Sigma) for 30 min. After washing, the nuclei were counterstained with or 4′-6-diamidino-2-phenylindole (DAPI) (Sigma) and mounted with Vectashield mounting medium (Vector Laboratories, Burlingame, CA, USA). Immunofluorescence was visualized using an Olympus BX41 microscope and recorded with a high-resolution DP70 Olympus digital camera (Olympus, Milan, Italy).
Cell migration assays
Cell migration was assayed with razor scrape assays as previously described . Briefly, a razor blade was pressed through the confluent T47-D breast cancer cell monolayer into the plastic plate. T47-D cells were swept away on one side of that line. Cells were washed, and 2.0 ml of RPMI containing steroid-deprived FBS were added. Cytosine β-D-arabinofuranoside hydrochloride (Sigma) (10 μM), a selective inhibitor of DNA synthesis, which does not inhibit RNA synthesis, was used 1 hr before the test substance was added. Migration was monitored for 48 hrs. Cells were digitally imaged and migration distance was measured by using phase-contrast microscopy.
Cell invasion assay
Cell invasion were assayed with the BD BioCoat™ Growth Factor Reduced (GR) Matrigel™ Invasion Chamber (BD Bioscience, Milan, Italy). In brief, after rehydrating the GFR Matrigel inserts, test substance were added to the wells of the BD Falcon™ TC Companion Inserts (no GRF Matrigel coating). Cytosine β-D-arabinofuranoside hydrochloride (Sigma) (10 μM) was used 1 hr before the test substance was added. An equal number of BD BioCoat™ Control Inserts (no GFR Matrigel coating) were likewise prepared as control. Then 0.5 ml of T47-D cell suspension (density, 2.5 × 104 cells) were added to the inside of the inserts. The chambers were put in incubator at 37°C, 5% CO2 atmosphere for 24 hrs. After incubation, the cells on the lower surface of the membrane were stained with Diff-Quick stain. The invading cells were counted in the central field of triplicate membranes. The invasion index is calculated as follows: % invasion test cell/% invasion control cell.
All values are expressed as mean ± S.D. Statistical differences between mean values were determined by anova, followed by the Fisher’s protected least significance difference (PLSD). All statistical analyses were performed with the version 12 of the SPSS software package (SPSS Inc, IL, USA). A value of P < 0.05 was considered significant.
RAL induces a rapid rearrangement of the actin cytoskeleton in OR+ T47-D breast cancer cells
We previously demonstrated that 17β-estradiol (E2) rapidly induces a rearrangement of the actin cytoskeleton in a nongenomic and time-dependent manner in human endothelial cells  as well as in ER+ T47-D breast cancer cells . In the present study, steroid- and serum-deprived T47-D breast cancer cells were treated with RAL (10 nM) for different time-points and the actin cytoskeleton was visualized with immunofluorescence after staining with phalloidin linked to Texas Red. Actin fibres in unstimulated controls were arranged longitudinally through the major axis of the cells, which displayed regular cell membrane borders (Fig. 1A). Treatment with RAL from 2 to 20 min. led to a rapid change of the organization of the actin fibres. Actin translocated from the cytoplasm toward the edge of the cell membrane, forming a typical cortical actin complex, in association with the formation of specialized membrane structures, such as pseudopodia (Fig. 1A). These morphological changes were time-dependent and transient, being maximal after 15–20 min. and then showing a progressive reversal to the basal arrangement between 30 and 60 min. (Fig. 1A). In addition, treatment with RAL at different concentrations (from 0.1 to 100 nM) resulted in a remodelling of actin fibres that was present at all concentrations tested (Fig. 1B). Consistently with our previous data , E2 (10 nM) also induced a rapid actin remodelling in T47-D cells (Fig. 1B).
RAL rapidly activates the actin-regulatory protein, moesin
Based on our previous finding that the actin-regulatory protein moesin mediates the cytoskeletal effects of oestrogen [9, 12], we studied whether RAL stimulated actin rearrangement through this protein. Indeed, the cellular level of Thr558-phosphorylated moesin, which is functionally activated , rapidly increased in T47-D cells treated with RAL (10 nM) from 2 to 15 min. and then declined to the basal level after 20–60 min. (Fig. 2A), time-consistently with the kinetics of RAL-induced actin rearrangement. Exposure of T47-D cells to different concentrations of RAL (from 0.1 to 100 nM) for 15 min. resulted in an activation of moesin that was similar for all the RAL doses, and comparable to that of E2 (10 nM) (Fig. 2B). The total cell content of moesin did not change during this period (Fig. 2A and B).
RAL effects on breast cancer cell migration
To relate the RAL-induced actin cytoskeleton remodelling and moesin activation with increased cell motility, we performed horizontal migration assays with T47-D cells. In order to distinguish cell migration from cell proliferation, Cytosine β-D-arabinofuranoside hydrochloride (10 μM), a selective inhibitor of DNA strand separation that does not block RNA synthesis, was used to arrest cell proliferation. Treatment with different concentrations of RAL (from 0.1 to 100 nM) significantly increased the number of cells that migrated through the starting line, as well as the mean length of migration compared to control (Fig. 3A). E2 as well resulted in a significant increase of horizontal migration of T47-D cells (Fig. 3A). Interestingly, E2 and RAL-treated T47-D cells displayed different patterns of migration. In the presence of E2, these migrating cells formed a ‘compact’ migration front, whereas in the presence of RAL they displayed a variable tendency to move. Some migrating cells moved forward more rapidly than others thus creating a discontinuous migration front. Overall, T47-D cells treated with RAL displayed longer median migration distances, but the number of migrating cells was numerically lower than that in the presence of E2 (Fig. 3B and C).
RAL counteracts the E2-induced remodelling of the actin cytoskeleton
As a SERM, RAL can compete with oestrogens by binding to ORs and differently alter oestrogen action depending on the tissue. T47-D cells exposed to E2 (10 nM) together with increasing concentrations of RAL revealed that the E2-induced actin rearrangement and formation of cell membrane structures are attenuated in the presence of RAL, in a dose-related fashion (Fig. 4A).
RAL interferes with E2-induced moesin phosphorylation and cell migration
The co-administration of RAL with E2 (10 nM) resulted in a concentration-related repression of the moesin activation induced by estradiol. The effect of E2 was completely blunted in the presence of an equimolar amount of RAL (Fig. 4B). The effect of the co-administration of E2 and RAL changes depending on the amount of E2. Indeed, in the presence of lower E2 concentration (1 nM or 0.1 nM, corresponding to follicular peak or to postmenopausal E2 levels, respectively) a potentiation of moesin phosphorylation can be seen with the addition of low RAL concentrations, whereas a progressive inhibition of moesin is obtained in the presence of higher RAL amounts (Fig. 4C and D).
As expected, E2 significantly increased the migration of T47-D breast cancer cells (Fig. 5A–C). When the two drugs were co-administrated, RAL blocked E2-induced cell migration in a concentration-related manner. RAL affected both the absolute number of migrating cells as well as the median migration distance (Fig. 5A–C).
RAL induces rearrangements of the actin cytoskeleton, moesin activation and cell migration through an ER and G protein-dependent signalling pathway
To determine the signalling cascades involved in the action of RAL, we examined the effect of different inhibitors of intracellular pathways that have been associated with rapid oestrogen signalling in human cells [12, 14]. The RAL-induced actin rearrangement as well as moesin activation were significantly reduced by the ER antagonist ICI 182,780, by the G protein inhibitor pertussis toxin (PTX) and by the inhibitor of the kinase that is responsible for moesin phosphorylation, RhoA-associated kinase (ROCK-2), Y-27632. On the opposite, PD98059, an inhibitor of the ERK 1/2 mitogen-activated protein kinase (MAPK) cascade did not influence the effect of RAL (Fig. 6A and B). These findings suggest that both ER and G proteins are required for moesin phosphorylation.
In agreement, the horizontal migration of T47-D breast cancer cells induced by RAL was abrogated by the addition of ICI 182,780, of PTX and Y-27632, indicating that an ER/G protein/ROCK cascade is implicated in the signalling of RAL to T47-D cell migration. No significant reduction of cell migration was found in the presence of PD98059 (Fig. 7A–C).
RAL effects on T47-D cell invasion
Finally, we investigated the effects of RAL and E2 on breast cancer cell invasion of a three-dimensional matrix and compared the actions of these compounds with those of tamoxifen (TAM), a SERM commonly used to counteract the progression of ER+ breast cancer in clinical practice. E2, RAL or TAM all promoted cell invasion when used alone (Fig. 8A). E2 was the most potent among these compounds. The effects of RAL on cell invasion were blocked by interfering with ERs, G proteins and ROCK. Both TAM as well as RAL reduced the E2-induced invasion of breast cancer cells, and the effects of RAL were concentration-related (Fig. 8A and B).
Oestrogen plays a major role in mammary gland development and function. On the other hand, endogenous and exogenous oestrogens influence the development or the progression of breast cancer, as suggested by epidemiological studies and large-scale HRT clinical trials. Hence, oestrogen has been recently indicated as a carcinogen in the breast , although this is controversial.
SERMs block the action of oestrogen in the breast and other tissues by occupying ERs and inducing conformational changes that prevent the interaction of the receptors with coactivators . The current understanding of the mechanism of action of these drugs is that given the differences in the expression of the co-activators or co-repressors of ERs in different cells and tissues, each SERM will turn into a specific spectrum of bodily effects. Tamoxifen was first introduced into the clinics in the 1970s and is used to treat women with hormone-sensitive breast cancer. In addition to its anti-oestrogenic function in the breast, TAM is characterized by an unwanted stimulatory action on the endometrium, which turns into a higher rate of endometrial hyperplasia and cancer .
RAL has been initially approved for the prevention and treatment of osteoporosis in postmenopausal women. However, it also prevents breast cancer in osteoporotic women or in women at high risk for breast cancer without increasing the incidence of endometrial cancer .
Local invasiveness and distant metastasis are the foremost causes of breast cancer-related morbidity and death. TAM has been previously found to alter breast cancer cell adhesion, migration and invasion . In this study, we find that RAL, like tamoxifen, slightly increases T47-D breast cancer cell migration and invasion when administered alone. However, in the presence of estradiol, which potently increases breast cancer cell movement, both SERMs act as anti-oestrogens, resulting into a marked blockade of both horizontal cell migration and invasion of three-dimensional matrices. These complex actions on breast cancer cell motility are plaid through the regulation of the actin cytoskeleton of breast cancer cells via a set of molecular steps that converge on the ERM protein, moesin.
The ERM actin-binding proteins, including moesin, function as linkers between the actin cytoskeleton and plasma membrane proteins. By interacting with actin, activated ERMs induce actin de-polymerization and re-assembly toward the cell membrane edge, supporting the formation of cortical actin complexes . These complexes help the formation of molecular bridges between the actin cytoskeleton, integrins and focal adhesion complexes within ruffles and pseudopodia and are critical for cell movement . These structures assist the attachment of cells to the extracellular matrix and generate the locomotive forces for cell movement.
Estradiol induces moesin phosphorylation (which corresponds to activation) in human endothelial  and breast cancer cells  through a set of rapid signalling actions. RAL acts in a similar fashion by rapidly activating moesin and thus promoting actin cytoskeletal remodelling and the formation of cortical actin complexes and specialized membrane protrusions. These actions explain the weak promotion of breast cancer cell movement induced by RAL.
However, when estradiol and RAL are provided simultaneously, RAL inhibits the estradiol-dependent recruitment of moesin in a concentration-related fashion. Notably, the higher RAL concentrations used in this study are in the range of the serum concentrations of this SERM found with the doses used in the clinical setting (60 mg/day) .
The existence of differential clinical effects of RAL related to the concentration are well-established. For instance, low concentrations of RAL inhibit the growth of leiomioma cells, whereas higher concentrations of RAL have promoting effects . In parallel, high RAL concentrations promote endometrial cancer cell growth, whereas lower concentrations are ineffective . To explain such variable responses it has been proposed that RAL may differently alter the expression of ERs, the ratio of ERα/ERβ as well as the expression of cofactors at different concentrations .
The blockade of the moesin-linked cytoskeletal remodelling induced by oestrogen may be relevant for the efficacy of RAL in the setting of adjuvant hormonal therapy of ER+ breast cancer. This is supported by the evidence that ERM proteins are relevant regulators of the process of tumour metastasis . For instance, high ezrin expression is associated with early development of metastasis of breast cancer, osteosarcoma tumours or adenocarcinomas [23, 24], and we have recently found overexpression of moesin in aggressive breast cancers .
Raloxifene is effective in blocking the promotion of breast cancer motility and invasion induced by E2 at concentrations expected to be achieved during standard therapies with this drug. This might imply that in addition to reducing the number of breast tumors diagnosed in postmenopausal women, RAL could also limit the tendency of the cancers to spread locally or at a distance, particularly in women with significant concentrations of oestrogens in the breast, however this has not been investigated by clinical trials up to now.
This report establishes that the actin-binding protein moesin is a target of SERMs, mediating their effects on breast cancer cell migration and invasion. In addition, this mechanism of action potentially offers insight into a number of other processes regulated by SERMs that involve cell movement, such as wound healing, that is reportedly accelerated by RAL .
In this report, the regulation of moesin by RAL or oestrogen occurs within minutes and is quickly reverted thereafter, suggesting that it is enacted through an extra-nuclear signalling cascade of ERs . Indeed, the activation of moesin by RAL appears to be mediated by G proteins and by the RhoA-associated kinase (ROCK-2). G proteins and ROCK-2 are established regulators of ERM proteins and of cytoskeletal architecture. In particular, ROCK-2 directly targets moesin and triggers its threonine phosphorylation . Inhibitors of G proteins and of ROCK-2 consistently prevent the actions of RAL on moesin activation, horizontal migration and invasion of three-dimensional matrices, suggesting that these mediators are necessary for the effects of RAL on breast cancer cell motility investigated in this study. In agreement, in response to different stimuli (including oestrogens) ROCK-2 is recruited by a cascade involving the G protein Gα13 and the small GTPase RhoA, which then turns into the activation of ERM proteins and into cytoskeletal remodelling .
Taken together, our data suggest that the SERM raloxifene is able to alter ER+ breast cancer cell motility and invasive behaviour by regulating the actin cytoskeleton through the ERM protein moesin. Through this mechanism of action RAL appears to function as an antagonist against breast cancer cell migration and invasion induced by physiological concentrations of estradiol. This study, as every in vitro study, does not allow to predict if the effects of raloxifene on the motility of breast cancer cells will also be found in vivo. However, it provides a lead to follow for future investigations, giving potential molecular intermediates to target to modulate breast cancer using this SERM. In conclusion, this paper highlights novel extra-nuclear mechanisms of action of the SERM raloxifene that might be important for the effects of this drug on the progression of ER+ breast cancers.
We are grateful to Dr. Sandra Silvestri and to Dr. Donato Agnusdei, from Eli Lilly and the Company, for kindly providing raloxifene hydrochloride. This work has been supported by the PRIN grant 2004057090_007 by the Italian University and Scientific Research Ministry (MIUR) to T.S.
- 19The population pharmacokinetics of raloxifene in post- menopausal women. Pharmaceutical Science. 1998; 1: 518., , , et al .