Comparative actions of progesterone, medroxyprogesterone acetate, drospirenone and nestorone on breast cancer cell migration and invasion

T47-D and MCF-7 breast cancer cells were incubated in DMEM (GIBCO) containing 10% fetal calf serum (FCS) and 0.2 UI/mL insulin, L-glutamine, penicillin and streptomycin under a 5% CO₂ atmosphere at 37°C. MDA-MB-468 breast cancer cells were incubated in L-15 medium (Leibovitz)(GIBCO) containing 10% fetal calf serum (FCS) and L-glutamine, penicillin and streptomycin. Before treatments, cells were kept 48 hours in DMEM containing steroid-deprived FBS. Before experiments investigating non-transcriptional effects, the cells were kept in DMEM containing no FBS for 8 hours. Whenever an inhibitor was used, the compound was added 30 minutes before starting the treatments. Progesterone, medroxyprogesterone acetate, 17β-estradiol, PTX, Y-27632, PD98059 and wortmannin were from Sigma-Aldrich (Saint-Louis, MO). Drospirenone was a kind gift of Dr. Heiner Fritzemeier (Bayer Schering Pharma, Berlin, Germany), Nestorone was provided by Dr. R. Sitruk-Ware and ORG 31710 was a kind gift of Dr. Lenus Kloosterboer, from Organon Akzo Nobel (Oss, The Netherlands).

T47-D breast cancer cells were grown on coverslips and exposed to treatments. The cells were fixed with 4% paraformaldehyde for 30 min and permeabilized with 0.1% Triton for 5 min. Blocking was performed with PBS containing 1% bovine serum albumin for 30 min. Then cells were incubated with Texas Red-phalloidin (Sigma) for 10 min. After washing the nuclei were counterstained with 4'-6-diamidino-2-phenylindole (DAPI) (Sigma) and mounted with Vectashield mounting medium (Vector Laboratories, Burlingame, CA). Immunofluorescence was visualized using an Olympus BX41 microscope and recorded with a high-resolution DP70 Olympus digital camera. Cell membrane thickness and the gray level of extracellular area, cell membrane as well as cytoplasm were quantitated using Leica QWin image analysis and image processing software (Leica Microsystems, Wetzlar, Germany).

Cells were harvested in lysis buffer including100 mM Tris-HCl (pH 6.8), 4% SDS, 20% glycerol, 1 mM Na₃VO₄, 1 mM NaF, and 1 mM PMSF. Cell lysates were separated by SDS-PAGE. The antibodies used were: moesin (clone 38, Transduction Laboratories, Lexington, KY), Thr⁵⁵⁸-P-moesin (sc-12895, Santa Cruz Biotechnology, Santa Cruz, CA). Primary and secondary Abs were incubated with the membranes with standard technique [21]. Immunodetection was accomplished using enhanced chemiluminescence.

Plasmids for CMV human progesterone receptor A (hPR-A, # 95) and B (hPR-A, # 90) were provided by Dean P. Edwards (Baylor college of medicine, USA). Both plasmids (15 μg) were transfected into MDA-MB-468 breast cancer cells using the Lipofectamine (Invitrogen) according to the manufacturer's instructions. Cells (60–70% confluent) were treated 48 h after transfection and prepared according to the experiments to be performed.

Validated antisense phosphorotioate oligonucleotides (S-modified) (PONs) complementary to the 1–15 position of the human moesin gene coding region were obtained from Dharmacon. The sequence was 5'-TACGGGTTTTGCTAG-3' for moesin antisense PON. The complementary sense PON was used as control (5'-CTAGCAAAACCCGTA-3'). Transfections were performed on subconfluent T47-D cells. PONs were resuspended in serum-free medium with Lipofectamine (Invitrogen) and added to the culture medium every 12 h at the final concentration of 4 μM. Every 24 h, cells were washed and fresh medium supplemented with 4 μM PONs was added. Moesin silencing was assessed through protein analysis up to 48 h after transfection.

Cell migration was assayed with razor scrape assays as previously described [20]. Briefly, a razor blade was pressed through the confluent T47-D breast cancer cell monolayer into the plastic plate to mark the starting line. T47-D cells were swept away on one side of that line. Cells were washed, and 2.0 mL of DMEM containing steroid-deprived FBS and gelatin (1 mg/mL) were added. Cytosine β-D-arabinofuranoside hydrochloride (Sigma) (10 μM), a selective inhibitor of DNA strand separation which does not inhibit RNA synthesis was used 1 h before the test substance was added. Migration was monitored for 48 hours. Every 12 h fresh medium and treatment were replaced. Cells were digitally imaged and migration distance was measured by using phase-contrast microscopy.

Cell invasion were assayed following the standard method by using the BD BioCoatTM Growth Factor Reduced (GFR) Matrigel™ Invasion Chamber (BD Bioscience, USA). In brief, after rehydrating the GFR Matrigel inserts, the test substance was added to the wells. An equal number of Control Inserts (no GFR Matrigel coating) were prepared as control. 0.5 mL of T47-D cell suspension (2.5 × 10⁴ cells/mL) was added to the inside of the inserts. The chambers were incubated for 24 h at 37°C, 5% CO₂ atmosphere. After incubation, the non-invading cells were removed from the upper surface of the membrane using cotton tipped swabs. Then the cells on the lower surface of the membrane were stained with Diff-Quick stain. The invading cells were observed and photographed under the microscope at 100× magnification. Cells were counted in the central field of triplicate membranes. The invasion index was calculated as the % invasion test cell/% invasion control cell.

All values are expressed as mean ± SD. Statistical differences between mean values were determined by ANOVA, followed by the Fisher's protected least significance difference (PLSD).

Our first objective was to identify the effects of P, MPA, DRSP and NES on the spatial organization of actin fibers. We thus exposed steroid- and serum-deprived T47-D (ER⁺/PR⁺) breast cancer cells to these compounds and stained the actin cytoskeleton with phalloidin linked to a fluorescent dye (Texas Red). Based on the evidence that P, MPA and DRSP have comparable binding affinities for progesterone receptor (PR), while NES is about a 100-fold more effective than P in binding to PR [11], we used a 100-fold lower concentration of NES than other three progestins. At baseline, actin fibers were arranged longitudinally through the major axis of T47-D cells, which displayed regular cell borders (Fig. 1). In the presence of any of the four progestins (P, MPA, DRSP all 100 nM; NES, 1 nM), the vast majority of breast cancer cells displayed visible changes of actin fibers organization (within 10 to 15 min), with a rapid actin concentration at the cell membrane (Fig. 1). This was associated with a significant increase of the thickness of the cell membrane and of its fluorescence intensity, quantified by analyzing the pixel intensity including the cell membrane as well as the adjacent intra- and extra-cellular space (Table 1). Within the same time frame, cell membrane structures involved in cell adhesion and movement, including ruffles, focal adhesion complexes and pseudopodia were formed (Fig. 1). This process was transient, with actin fibers going back to the basal arrangement within 30 minutes, and was blocked by the PR antagonist ORG 31710 (1 μM), indicating the involvement of PR (Fig. 1).

Bild vergrößern

E2 (10 nM) rapidly induced actin rearrangement in T47-D breast cancer cells, consistent with its action on endothelial cells [20]. Each progestin, when added to E2, did not significantly change the effect of E2 itself, although the cells often display a somewhat more evident rearrangement of actin fibers (as shown by the quantification of the mean thickness and intensity of cell membrane, Table 1) and cell membrane structures formation as compared to treatment with the progestin alone (Fig. 2). The effect of the addition of the progestins to E2 was reduced by the PR antagonist ORG 31710 (1 μM) but not by the ER antagonist ICI 182,780 (100 nM) (Fig. 2).

Bild vergrößern

Likewise, in ER/PR positive MCF-7 breast cancer cells, P and MPA (both 100 nM) also provoked the rapid actin reorganization in the absence or presence of E2 (10 nM), which were indicated by the translocation of actin towards cell membrane and by the obvious formation of lamellipodia (Fig. 3, Table 1). However, in MDA-MB-468 ER^-/PR^- breast cancer cells, the same compounds failed to induce actin cytoskeleton remodeling (Fig. 3, Table 1). When MDA-MB-468 cells were transfected with plasmids encoding PRA and PRB, they became able to respond to progestin exposure with actin rearrangement (Fig. 3, Table 1), confirming the central role of PR.

Bild vergrößern

Treatment of T47-D cells with P, MPA, DRSP (all 100 nM) or NES (1 nM) resulted in rapid increases of Thr⁵⁵⁸-phosphorylation of moesin (which corresponds to activation) [22], with a first visible effect from 2 minutes, a phosphorylation peak at 15 minutes, and a progressive decline to basal levels between 30 minutes and 2 hours, which is time-consistent with the kinetics of actin rearrangement (Fig. 4A–D). The amount of phosphorylated moesin was related to the concentration of the compounds (Fig. 4E–H). In parallel, the cell content of wild-type moesin did not change during this time frame (Fig. 4A–H).

Bild vergrößern

When bioequivalent amounts of P, MPA, DRSP (all 100 nM) or NES (1 nM) were administered to T47-D cells, visible differences in moesin Thr⁵⁵⁸-phosphorylation were seen, with MPA inducing the strongest activation and DRSP the weakest (Fig. 5A). In order to more precisely compare the potency of each progestin on moesin activation, we performed dose-response curves of moesin phosphorylation with the four compounds by quantitatively analysing Thr⁵⁵⁸-phosphorylated moesin bands intensities with a chemiluminescence digital acquisition system. As shown in Fig 5B, MPA was consistently more potent in inducing moesin activation as compared to P and DRSP. DRSP was significantly less effective than P or MPA (Fig. 5B). As expected, the dose-response curve of NES was shifted to the left by two orders of magnitude compared with P, MPA or DRSP (Fig 5B). However, if bio-equivalent concentrations are compared (1 nM NES vs. 100 nM P or 0.1 nM NES vs. 10 nM P), NES shows comparable effects on moesin as P, being less active than MPA (Fig 5B).

Bild vergrößern

We recently identified that natural P triggers moesin phosphorylation through a PRA/Gα₁₃/RhoA/Rho-associated kinase (ROCK) pathway, while MPA recruits a PR/Src/phosphatidylinositol-3-kinase (PI3K)/RhoA/ROCK cascade (X.D. Fu et al, submitted). Moesin phosphorylation induced by DRSP (100 nM) or NES (1 nM) was prevented by the PR antagonist ORG 31710 (1 μM), by the G protein inhibitor pertussis toxin (PTX – 100 ng/mL) and by Y-27632 (10 μM), a specific inhibitor of ROCK-2, which is a known activator of moesin (Fig. 5C–D). However, distinct from the signaling pathways exploited by P and MPA, the MAPK inhibitor PD98059 (5 μM) and the PI3K inhibitor wortmannin (30 nM) also interfered with moesin activation by DRSP (100 nM) and NES (1 nM) (Fig. 5C–D), suggesting that MAPK and PI3K play a role in moesin activation by these progestins, and supporting the concept that each progestin might recruit partly distinct PR-dependent signaling cascades.

Estrogen signals to moesin through rapid, extra-nuclear signaling [20]. Moesin phosphorylation was slightly increased by the addition of E2 (10 nM) to each progestin compared to the progestins alone, although this was not statistically significant (Fig. 6A–C). Interestingly, the PR antagonist ORG 31710 (1 μM) inhibited the combined effect of E2 associated with each progestin on moesin (Fig. 6A–C).

Bild vergrößern

Our recent observation indicates that P and MPA promote T47-D cell horizontal migration through a complex cascade requiring PR, G proteins, MAPK, PI3K and the Rho-associated kinase, ROCK-2 (X.D. Fu et al, submitted). DRSP (100 nM) and NES (1 nM) increased T47-D cell migration, alike (Fig. 7A–D). DRSP- or NES-promoted cell migration were reduced by ORG 31710 (1 μM), by PTX (100 ng/mL), by the MAPK inhibitor PD98059 (5 μM), by the PI3K inhibitor wortmannin (30 nM) and by the ROCK inhibitor Y-27632 (10 μM) (Fig. 7A–D).

Bild vergrößern

We next compared the potency of these progestins on breast cancer cell migration. Among the tested compounds, MPA (100 nM) was the most potent, increasing cell migration by 54% compared to control (Fig. 8A–E). P (100 nM), DRSP (100 nM) and NES (1 nM) increased T47-D horizontal migration vs. control by 37%, 32% and 26% (Fig. 8A–E), respectively, which agrees with the order they rank on moesin activation.

Bild vergrößern

In order to check for the requirement of moesin for progestins-promoted cell migration we silenced moesin with antisense oligonucleotides (PONs) (Fig 8A, 8B, 8E). Transfection with moesin antisense PONs greatly reduced the action of both progesterone and MPA on cell migration, while sense PONs had no impact (Fig. 8A, 8B, 8E).

E2 promoted T47-D cell migration as well. However, no significant additive effects were found during co-treatment with any of the progestins. Interestingly, the PR antagonist ORG31710 (1 μM) significantly reduced cell migration associated with the four tested progestins both in the absence and in the presence of E2 (Fig. 8A–E).

Finally, we investigated the actions of the four progestins on breast cancer cell invasion of a three-dimensional matrix (matrigel). P, MPA, DRSP (all 100 nM) and NES (1 nM) all enhanced cell invasive behavior (Fig. 9). Consistently with the previous findings, the invasion indexes indicate that MPA is more effective compared to the other compounds (invasion index 4.68). Treatment with P, NES and DRSP resulted in invasion indexes of 3.91, 3.73 and 3.49, respectively. E2 (10 nM) was more potent than the progestins in driving breast cancer cell invasion of the matrix. When T47-D cells were exposed to E2 plus P, DRSP or NES the invasion indexes turned out to be significantly reduced vs. treatment with E2 alone. The addition of MPA to E2 also resulted in a reduction of the invasion index below that of E2 alone, but this did not reach statistical significance (Fig. 9).

Bild vergrößern