Alendronate Inhibits Lysophosphatidic Acid-induced Migration of Human Ovarian Cancer Cells by Attenuating the Activation of Rho

Introduction

Ovarian cancer is a highly metastatic disease characterized by widespread peritoneal dissemination and ascites, and is the leading cause of death from gynecologic malignancy (1) . This poor outcome appears to be correlated with the peritoneal dissemination of cancer cells (1) . Accordingly, one new therapeutic strategy is to clarify the mechanism of metastasis of cancer cells and to identify agents that prevent cancer cells from invading or migrating into the peritoneum. Among many growth-promoting factors known to be present in ovarian cancer ascites, LPA 3 is found there at significant levels (2–80 μm) and may play an important role in the development or progression of ovarian cancer (2) . LPA has been reported to induce many cellular effects including mitogenesis, the secretion of proteolytic enzymes (3) , and migration activity, which is accompanied by stress fiber formation and focal adhesion assembly 4 in ovarian cancer cells.

Cell migration is regulated by a combination of different processes: the contraction of actomyosin, the formation of stress fibers, and the turnover of focal adhesions (4) . Contraction of the actomyosin system is important for cell migration, and LPA induces MLC phosphorylation through the activation of the small GTP-binding protein Rho, leading to the stimulation of cell contractility and motility (5) . Another fundamental component affecting cell motility is the focal adhesion: cell-ECM adhesions can alter the capacity of the cell to attach and migrate through surrounding tissues (4) . Changes of the expression and activities of the components of focal adhesions could make an important contribution to cancer invasion (6) . At their cytoplasmic face, focal adhesions provide attachment for actin stress fibers. More than just sites of structural linkage between the ECM on the outside and the cytoskeleton on the inside, focal adhesions are regions of signal transduction. Components involved in multiple signal transduction pathways have been identified in focal adhesions, with most attention being directed toward tyrosine phosphorylation at these sites (7) . LPA stimulates cell motility by driving the formation of focal adhesions and elevating tyrosine phosphorylation of focal adhesion proteins such as paxillin, FAK, and Src in fibroblasts (5) and cancer cells (8) .

Mevalonate pathway is the synthesis pathway not only of cholesterol but also of various isoprenoids, such as FPP and GGPP. Both FPP and GGPP are important lipid attachments for the post-translational modifications of a variety of small GTPase proteins, such as Rho GTPases. Thereby, the inhibitors of the mevalonate pathway might have the opportunity to inhibit the invasion of cancer cells, as reported in 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors (9) .

BPs are the most potent and effective inhibitors of bone resorption in clinical use, and used world-wide against osteoporosis, Paget’s disease, bone metastases, and other bone diseases. Although the mechanism of action of all BPs had been considered to be the induction of apoptosis, the effect of N-BPs, such as alendronate, risedronate, and ibandronate, remains controversial. Recent reports demonstrated that N-BPs inhibit bone resorption by osteoclasts and tumor cell invasion without affecting cell viability at relatively low concentrations (10 , 11) and imply that some effect of N-BPs other than the induction of apoptosis is important at clinically used concentrations. Specifically, alendronate was shown recently to inhibit the cholesterol biosynthesis pathway, as well as isoprenylation (farnesylation and geranylgeranylation) by inhibiting either isopentanyl diphosphate synthase or the downstream enzyme, FPP synthase, or both (12) . Protein targets of isoprenylation include small G proteins such as Rho, Ras, Rac, and Rab, which require the post-translational modification to undergo a series of changes that lead to their attachment to the plasma membranes and their full function. Accordingly, N-BPs have the potential to inactivate small G proteins, which regulate cell growth, motility, and invasion, the same as with 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors (9) .

Although BPs are highly hydrophilic compounds and had been thought to accumulate specifically in bone, recent studies using ¹⁴C-labeled BPs proved that BPs were accumulated in the aortas of even healthy rabbits and in the human artery at relatively high concentrations (13) . In addition, chronic i.v. N-BP therapy was reported to increase high-density lipoprotein cholesterol and decrease low-density lipoprotein cholesterol in postmenopausal women (14) . These facts suggest that BPs, especially N-BPs, may directly affect organs other than bone. Although there have been many previous reports about the effect of BPs on cancer cells, they were focused mainly on cell viability and proteolytic activity of matrix metalloproteinases ( 11 , 15 , and 16 ), and few data have been obtained as yet on the exact mechanisms of the effects of BPs on cancer cell migration and metastasis.

In this study, we analyzed the cellular effects of alendronate, and showed that alendronate markedly inhibited LPA-induced migration of human ovarian cancer cells by attenuating the activation of Rho, which results in changes of cell morphology, loss of stress fiber formation, and focal adhesion assembly, and the suppression of phosphorylation of proteins, which are essential processes for cell migration.

Materials and Methods

Materials.

We used alendronate (sodium hydrate) injection (TEIJIN LIMITED) that has already been marketed in Japan. Alendronate (monosodium 4-amino- 1-hydroxybutylidene-1, 1-diphosphonate trihydrate) was solubilized in citrate solution at a concentration of 10 mm and adjusted to neutral. Anti-Ser19-phosphorylated MLC monoclonal antibody (pLC1) was a generous gift from Dr. Minoru Seto (Asahi Chemical Industry Company, Shizuoka, Japan). Anti-FAK polyclonal antibody, anti-paxillin monoclonal antibody, glutathione S-transferase-RBD and anti-Rho (-A, -B, -C) polyclonal antibody were purchased from Upstate Biotechnology (Lake Placid, NY). Anti-phospho-FAK (Tyr 397) polyclonal antibody, anti-phospho-paxillin (Tyr 31) polyclonal antibody, and anti-paxillin polyclonal antibody were obtained from BioSource International (Camarillo, CA). Anti-RhoA polyclonal antibody, horseradish peroxidase-conjugated antimouse and antirabbit IgG were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Polyvinylidene difluoride membranes (Hybond-P) and enhanced chemiluminescence Western blotting detection reagents were obtained from Amersham (Arlington Heights, IL). The CellTiter 96AQ kit to monitor cell proliferation was purchased from Promega (Madison, WI). Rhodamine-labeled phalloidin, Alexa Fluor 488-labeled goat antimouse and antirabbit antibodies, and Hoechst 33342 were purchased from Molecular Probes (Eugene, OR). Y-27632, a specific inhibitor of Rho-associated kinase, was kindly provided by WelFide Corporation (Osaka, Japan). Protein concentrations were determined using a Bio-Rad Dc kit (Hercules, CA) with BSA as a standard. DMEM was purchased from Life Technologies, Inc. (Gaithersburg, MD). LPA (oleoyl-sn-glycero-3-phosphate), antivinculin monoclonal antibody, and all of the other chemicals and reagents were purchased from Sigma (St. Louis, MO).

Cell Culture and Treatment.

The human ovarian cancer cell line, Caov-3, was purchased from American Type Culture Collection (Rockville, MD), cultured in DMEM supplemented with 10% (v/v) fetal bovine serum, penicillin (10 units/ml)-streptomycin (10 μg/ml) in 95% air, and 5% CO2 at 37°C, and was used within 15 passages after the initiation of culture.

Evaluation of Apoptosis by Fluorescence Microscopy.

Cells were plated on eight-well chamber slides coated with type I collagen and allowed to attach for 6 h, and then cultured under serum-free conditions [DMEM containing 0.1% (w/v) BSA] with various concentrations of alendronate for 24 h. After incubation, cells were fixed with 3.7% paraformaldehyde in PBS for 30 min and permeabilized with 0.5% Triton X-100, and cell nuclei were stained with the specific chromatin dye Hoechst 33354 (Molecular Probes) at 37°C for 30 min. After washing, samples were observed using a Zeiss confocal photomicroscope LSM510 (Zeiss, Thornwood, NY).

Cell Viability Assessment.

Cell viability was assessed by the MTT assay using a CellTiter 96AQ kit (Promega) according to the manufacturer’s instructions. Briefly, the cells (3 × 10³/well) were plated in 96-well plates and allowed to attach for 6 h, and then cultured under serum-free conditions with various concentrations of alendronate for 48 h. The number of surviving cells was determined by measuring the absorbance at 590 nm (A_{590 nm}) of the dissolved formazan product after addition of 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium inner salt for 1 h as described by the manufacturer. All of the experiments were carried out in quadruplicate, and viability was expressed as the ratio of the number of viable cells with alendronate treatment to the number without treatment.

Analysis of Migration.

Chemotactic directional migration was evaluated using a modified Boyden chamber. Porous filters (8-μm pores) were coated on the underside by passive adsorption of type I collagen (Sigma). Cells (1.5 × 10⁵/well) were plated in the upper chamber in the presence of various agents (LPA, alendronate, Y-27632, and so forth) as indicated and allowed to migrate for 4.5 h. Nonmigrating cells were removed from the upper chamber with a cotton swab, and migrating cells adherent to the underside of the filter were fixed, stained with Mayer’s hematoxylin solution, and enumerated using an ocular micrometer, and at least 10 fields/filter were counted. All of the experiments were independently performed in triplicate.

Analysis of Adhesion.

Caov-3 cells were incubated with or without various agents for 24 h. The cells were detached from dishes and resuspended in DMEM containing 0.1% BSA with or without various agents. The cells (3 × 10⁵/well) were allowed to attach for 30 min at 37°C and then washed with PBS three times and fixed with 3.7% paraformaldehyde in PBS for 30 min. Attached cells were enumerated using an ocular micrometer and at least 10 fields/filter were counted. All of the experiments were independently performed in triplicate.

Rho Pull-Down Assay.

The Rho pull-down assay was performed as described previously (17) . Briefly, cells (3 × 10⁵/well) were plated, allowed to attach for 6 h, and then cultured under serum-free conditions with or without various agents for 24 h. After incubation, cells were stimulated with 25 μm LPA for 1 min, washed twice with PBS, and lysed in radioimmunoprecipitation assay buffer. Cell lysates were clarified by centrifugation, and equal volumes of lysates were incubated with Rhotekin RBD-agarose beads (30 μg) at 4°C for 45 min. The beads were washed three times with wash buffer. Bound Rho proteins were detected by Western blotting using a polyclonal antibody against Rho (-A, -B, -C). Western blotting of the total amount of Rho in cell lysates was performed for the comparison of Rho activity (level of GTP-bound Rho) in different samples.

Immunocytochemical Study.

Cells were plated on eight-well chamber slides coated with type I collagen, allowed to attach for 6 h, and then cultured under serum-free conditions with or without various agents for 24 h. After incubation, cells were stimulated with 25 μm LPA for 30 min, fixed with 3.7% paraformaldehyde in PBS for 30 min, permeabilized with 0.1% Triton X-100, and stained with anti-paxillin (1:500) at 4°C overnight. After washing, samples were incubated with Alexa Fluor 488-labeled goat antimouse IgG (1:1000). Specimens were double-stained with rhodamine-labeled phalloidin (1:200) for 30 min at room temperature. The images were recorded and analyzed using a Zeiss confocal photomicroscope LSM510. For RhoA staining, fixed and permeabilized cells were incubated with anti-RhoA (1:200) at 4°C overnight. After washing, samples were incubated with Alexa Fluor 488-labeled goat antirabbit IgG (1:1000) and examined under a fluorescence microscope.

Western Blot Analysis.

Cells (3 × 10⁵/well) were plated and allowed to attach for 6 h, and then cultured under serum-free conditions with or without various agents for 24 h. After incubation, cells were stimulated with 25 μm LPA for the times indicated in the text, washed twice with PBS, and lysed in ice-cold lysis buffer [20 mm Tris (pH 7.4), 150 mm NaCl, 1% Triton X-100, 1 mm EDTA, 1 mm EGTA, 2.5 mm sodium PP_I, 1 mm glycerolphosphate, 1 mm sodium orthovanadate, 1 μg/ml leupeptin, and 1 mm phenylmethylsulfonyl fluoride]. Equal amounts of samples (20 μg) were resolved by SDS-PAGE and transferred to Hybond-P. The transferred samples were incubated with the antibody indicated in the text and then incubated with the corresponding secondary horseradish peroxidase-conjugated IgG, and the immunoblotted proteins were visualized with enhanced chemiluminescence reagents.

Statistical Analysis.

All of the statistical analyses were carried out using Student’s t test.

Results

Effects of Alendronate on Cell Morphology and Viability.

We first studied the effect of alendronate on the morphology of human ovarian cancer cells. Typical shapes of Caov-3 cells are shown in Fig. 1A ⇓ , panel a. Untreated Caov-3 cells were flat and well spread, but treatment of the cells with 30 μm alendronate induced cell retraction from the substratum and loss of contacts between neighboring cells, resulting in spindle-shaped morphology (Fig. 1A ⇓ , panel c). These morphological changes were reversed 24 h after the removal of alendronate. The cells treated with alendronate in the presence of GGOH (30 μm) were flat, like the control cells (Fig. 3d) ⇓ , while the cells treated with alendronate in the presence of FOH (30 μm) remained spindle-shaped (Fig. 3e) ⇓ . The same morphological changes were observed when Y-27632, an inhibitor of the Rho/Rho-associated kinase pathway, was used (Fig. 3f) ⇓ . These results are consistent with the possibility that alendronate changes the well-spread morphology of Caov-3 cells by inhibiting geranylgeranylation of Rho. To investigate whether alendronate induces apoptosis in ovarian cancer cells or not, morphological evaluation of apoptosis was performed by nuclear staining of Caov-3 cells. Apoptotic nuclei, as indicated by chromatin aggregation, marginalization, or clumping, were hardly seen in alendronate-treated cells (Fig. 1A ⇓ , panel d), and the morphology and condensation of nuclei of aledronate-treated cells did not differ from those of untreated cells (Fig. 1A ⇓ , panel b). To confirm that these morphological data were in accord with the viability of Caov-3 cells, the MTT assay was performed. Fig. 1B ⇓ shows that up to 30 μm alendronate did not affect cell viability. These results suggest that in ovarian cancer cells the disruption of the actin cytoskeleton and the change of morphology induced by alendronate do not require the apoptosis of cells, and these findings are in accord with those of the previous studies in osteoclasts (10) and cancer cells (11 , 15 , 16) .

LPA-induced Migration Was Inhibited by Alendronate, and the Inhibition Was Prevented by the Addition of GGOH.

The effect of alendronate on the migration and adhesion capacity of ovarian cancer cells was assessed using in vitro migration (Fig. 1C) ⇓ and adhesion (Fig. 1D) ⇓ assays. We studied previously that LPA promoted the migration of Caov-3 cells in a dose-dependent manner at concentrations of up to 25 μm, and we therefore adopted 25 μm as the concentration of LPA in this study. Alendronate significantly suppressed the LPA-induced migration and adhesion in a dose-dependent manner. The role of geranylgeranylated or farnesylated proteins was studied by treating cells with GGOH or FOH. Whereas neither GGOH nor FOH modified cell migration or adhesion without LPA stimulation (data not shown), the inhibitory effect of alendronate on the LPA-induced migration and adhesion was prevented by the addition of GGOH but not by the addition of FOH (Fig. 1C) ⇓ . LPA-induced migration has been reported to be regulated by Rho-mediated activation of actomyosin contractility in fibroblasts (5) and in cancer cells (18) . Thus, these results suggest that alendronate suppresses LPA-induced migration and adhesion by suppressing the prenylation of small GTP-binding protein Rho in ovarian cancer cells.

Activation of Rho by LPA and Inhibition of This Activation by Alendronate.

To confirm that the inhibitory effect of alendronate on LPA-induced migration is because of the inactivation of Rho, we measured the intracellular levels of the GTP-bound, active form of Rho using the pull-down assay system. Because the activation of Rho by LPA was reported to reach a peak after 1 min in fibroblasts (17) , we compared the activation of Rho after a 1-min treatment with LPA or other agents. As shown in Fig. 2A ⇓ , the level of the active form of Rho was elevated after the addition of LPA, and alendronate inhibited the elevation induced by LPA. The addition of GGOH in the presence of alendronate restored the activation of Rho. These results suggest that Rho activation by LPA is attenuated by alendronate via the inhibition of geranylgeranylation. The addition of FOH in the presence of alendronate partially restored the level of activation of Rho, and this may be because GGPP is the downstream metabolite of FPP, and FOH could be metabolized to GGPP to some extent.

For the activation of small G proteins including Rho, these proteins must be targeted to the plasma membrane. Therefore, to examine RhoA localization in Caov-3 cells, cells were immunostained with a RhoA-specific antibody. In control cells, RhoA immunostaining was confined largely to the cytosol (Fig. 2B ⇓ , panel a). In contrast, the stimulation of Caov-3 cells with LPA resulted in the translocation of RhoA to the membrane at the edges of lamellae (Fig. 2B ⇓ , panel b). The translocation of RhoA from the cytosol to the membrane in response to LPA was markedly blocked by pretreatment with alendronate (Fig. 2B ⇓ , panel c). The inhibitory effect of alendronate on LPA-induced translocation of RhoA was prevented by the addition of GGOH (Fig. 2B ⇓ , panel d). These results suggest that alendronate inhibits the activation of Rho by blocking Rho translocation to plasma membrane via the inhibition of geranylgeranylation.

LPA-induced Formation of Stress Fibers and Focal Adhesions Was Inhibited by Alendronate, and This Inhibition Was Prevented by the Addition of GGOH.

Cell migration begins with an initial protrusion or extension of the plasma membrane at the leading edge of the cell. The protrusion is driven by the polymerization of a network of cytoskeletal actin filaments and is stabilized through the formation of adhesive complexes (4) . To investigate the mechanism of the inhibitory effect of alendronate on LPA-induced migration and adhesion, actin stress fibers and paxillin, one of the major focal adhesion proteins, were visualized (Fig. 3) ⇓ . LPA treatment caused a drastic increase of actin bundles and changed the localization of paxillin to the edge of the actin bundles (Fig. 3b) ⇓ . Treatment with alendronate significantly decreased the LPA-induced formation of stress fibers and focal adhesions (Fig. 3c) ⇓ , similar to the treatment with Y-27632 (Fig. 3f) ⇓ . The inhibitory effect of alendronate on the formation of stress fibers and focal adhesions was almost completely prevented by the addition of GGOH (Fig. 3d) ⇓ but not FOH (Fig. 3e) ⇓ . These results suggest that in ovarian cancer cells alendronate inhibits the formation of stress fibers and focal adhesions by suppressing the prenylation of Rho.

The Inhibitory Effect of Alendronate on LPA-stimulated Phosphorylation of MLC, and Tyrosine Phosphorylation of Paxillin and FAK.

Cell migration is regulated by a combination of different processes: the contraction of actomyosin, the formation of stress fibers, and the turnover of focal adhesions (4) . The phosphorylation of MLC and tyrosine phosphorylation of focal adhesion proteins such as paxillin or FAK are especially essential processes in LPA-induced cell migration (7) . Therefore, we analyzed the effect of alendronate on the phosphorylation of these proteins by Western blotting (Fig. 4) ⇓ . In preliminarily experiments, we assessed the time course of the phosphorylation of MLC and the tyrosine phosphorylation of focal adhesion proteins induced by LPA. Those experiments indicated that the phosphorylation of MLC reached a peak after 15 s of treatment with LPA, and that tyrosine phosphorylation of focal adhesion proteins reached a peak after 30 min of treatment (data not shown). Therefore, we compared the phosphorylation of these proteins after treatment with LPA for those times. Alendronate inhibited the phosphorylation of MLC (Fig. 4A) ⇓ , and the tyrosine phosphorylation of paxillin (Fig. 4B) ⇓ and FAK (Fig. 4C) ⇓ . These inhibitory effects were prevented by the addition of GGOH, although FOH prevented those to a markedly lesser extent. The results of Western blotting showed that alendronate inhibits the Rho-mediated activation of actomyosin contractility and tyrosine phosphorylation of focal adhesion proteins by suppressing the prenylation of Rho.

Discussion

Patients with ovarian cancer have the highest mortality rate among women with gynecologic cancers. The mortality of ovarian cancer has not been dramatically improved despite recent advances in chemotherapy. During the progression of ovarian carcinomas, cancer cells released from the surface of the tumor may then form malignant ascites from which they can adhere, and invade tissues and organs in the peritoneal cavity. In this way, the poor outcome is, at least in part, because of the peritoneal dissemination caused by the aggressive migration activity of ovarian cancer cells (1) . Presently used anticancer drugs, even if they are very effective at killing cancer cells, can be used only at limited concentrations because of their toxicity to normal cells. Accordingly, it seems worthwhile to look for drugs that inhibit the progression of ovarian cancer without affecting cell viability.

Our results highlight the importance of the products of the mevalonate pathway in cancer cell migration in response to LPA and suggest that alendronate inhibits LPA-induced cancer cell invasion by preventing geranylgeranylation of Rho. A significant inhibitory effect of alendronate on LPA-induced Caov-3 cell migration was observed at a concentration of 1 μm, and half-maximal inhibition was estimated to occur at ∼3 μm (Fig. 1C) ⇓ . Previous reports about the effects of N-BPs on cancer cells mainly focused on cell viability and proteolytic activity, and the effective concentrations of N-BPs were relatively high (11 , 15 , 16) . According to those reports, the concentrations of pamidronate, zeladronate, or ibandronate, which induced apoptosis of breast cancer and prostate cancer cells, were ∼100 μm, which agrees with our MTT assay results (Fig. 1B) ⇓ . Considering that the concentration of alendronate, which significantly inhibits bone resorption, is 0.1 μm, which is the clinically used concentration of alendronate (19) , effects of alendronate other than induction of apoptosis of cells seems likely to be clinically important. A recent report demonstrated that alendronate inhibits in vitro invasion of prostate cancer cells at low concentrations, but at such low concentrations of alendronate, the secretion of matrix metalloproteinases is not inhibited (20) . Considering these facts, our findings here that the inhibitory effect of alendronate on migration occurs without enhanced apoptosis of cells and is caused by attenuating the activation of Rho, which results in changes of cell morphology, loss of stress fiber formation, and focal adhesion assembly, and the suppression of phosphorylation of MLC and tyrosine phosphorylation of focal adhesion proteins, appear likely to account for at least some of the clinical effects of alendronate.

In our present study, the addition of GGOH restored the inhibitory effect of alendronate. On the other hand, the addition of FOH did not restore the inhibitory effect of alendronate-induced inhibition of focal adhesion assembly and cell migration despite the partial restoration in Rho activation and MLC phosphorylation. This inconsistency suggests two possibilities: (a) FOH might be partially metabolized to GGPP, which activates Rho and MLC phosphorylation to a lesser extent than GGOH, and lesser phophorylation of MLC might not be enough to activate focal adhesion assembly and cell motility; and (b) the addition of FOH might have the possibility to induce farnesylation of signaling molecules, which activate MLC phosphorylation but do not promote migration activity, although we could not identify the farnesylated molecule. Also, these results suggest that the formation of focal adhesion assembly, as well as the tyrosine phosphorylation of focal adhesion components, is essential to keep cytoskeletal organization and the resultant onset of migration activity.

For the formation of peritoneal dissemination, ovarian cancer cells need to detach from the primary tumor, attach the ECM of other tissues, and migrate into stromal lesions. Once cancer cells migrate into stromal lesions, angiogenesis occurs from pre-existing capillaries or venules. The fact that alendronate accumulates in vessels (13) strongly suggests that alendronate has the potential to prevent metastasis of ovarian cancer cells at concentrations that might be relatively nontoxic to normal cells in comparison to most presently used anticancer agents.

In conclusion, alendronate markedly inhibits LPA-induced migration of human ovarian cancer cells by attenuating the activation of Rho, which results in changes of cell morphology, loss of stress fiber formation, and focal adhesion assembly, and the suppression of phosphorylation of MLC and tyrosine phosphorylation of focal adhesion proteins. This study may provide the basis for a new therapy that controls the metastasis of ovarian cancer.