AmotP130 regulates Rho GTPase and decreases breast cancer cell mobility

Abstract Angiomotin (Amot) is a newly discovered, multifunctional protein that is involved in cell migration and angiogenesis. However, the role of its isoform, AmotP130, in the regulation of cytoskeleton and metastasis of breast cancer, is unclear. The aim of this study was to investigate the role of AmotP130 in the reorganization of the actin cytoskeleton and the changes of morphology in breast cancer cells through the Rho pathway that influences the invasion and migration of cells. The results suggested that AmotP130 suppressed the invasion ability through remodelling the cytoskeleton of breast cancer cells, including the actin fibre organization and focal adhesion protein turnover. Global transcriptome changes in breast cancer cells following knockdown of AmotP130 identified pathways related with the cytoskeleton and cell motility that involved the Rho GTPase family. From database analyses, changes in the Rho GTPase family of proteins were identified as possible prognostic factors in patients with breast cancer. We have been suggested that AmotP130 suppressed the invasion ability through remodelling of the cytoskeleton of breast cancer cells, involving regulation of the Rho pathway. The cytoskeleton‐related pathway components may provide novel, clinically therapeutic targets for breast cancer treatment.

related to the cell migration, tube formation and angiogenesis promotion. 8,9 Two isoforms of Amot, AmotP80 and AmoP130 have been identified. Compared with AmotP80, AmotP130 contains an extended N-terminal domain. 10 Notably, AmotP130 was reported to associate with F-actin through a conserved F-actin-binding domain bundle, whereas the AmotP80 was not. To some extent, the binding was related with actin bundling and other rearrangements of the cytoskeleton. However, the role of AmotP130 in the regulation of the cytoskeleton and metastasis of breast cancer is unclear. 11 In vitro studies suggested that cell directional metastasis is composed of a series of processes, including cell protrusion formation, 12 focal adhesion (FA) establishment, 13 cell contraction and FA turnover. 14,15 FAs are composed of different proteins including vinculin, a-actinin and Arp2/3. 16 The adhesive structures bridge between the actin cytoskeleton and extra cellular matrix (ECM). 17 The actin architectures are integrated and produce the force and movement. 18 Together, a coordinated set of procedures regulates cell migration.
The mechanism whereby AmotP130 affects cell morphology, adhesion and invasion ability through actin filaments organizations is unknown. In this study, we showed that AmotP130 inhibited cell adhesion and invasion ability. The processes could be controlled by the cytoskeleton reorganization and were related to the Rho pathway components.

| Drugs
Zoledronic acid (ZOL) was kindly provided by Novartis (Shanghai, China). The drug was solubilized and stored at À20°C. It was diluted in medium before use. The overexpression sequences were found in the GenBank database (NM_001113490.1). Lentivirus vectors were also constructed, generated and purified by GeneChem (Shanghai, China). The methods of lentivirus infection were as previously published. 19

| Western blotting
The lysis buffer and the procedure for membrane and cytosolic protein purification were as previously published. 20 The cell lysates were sonicated to disrupt cell membranes and release cytosolic proteins. Lysates were ultracentrifuged at 105 9 g for 1 hour at 4°C.
The supernatant collected is the cytosolic fraction, and the resulting precipitate is the membrane fraction. RIPA buffer (50 mmol/L Tris [pH 7.5], 1 mmol/L EDTA, 100 mmol/L NaCl, 2.5 mmol/L sodium orthovanadate, 10 lL/mL protease inhibitor cocktail, 1 mmol/ LPMSF, 0.5% NP40 and 0.5% Triton X-100) were used to obtain the total protein fraction. The BCA assay (Pierce, Rockford, IL, USA) was used to quantify protein concentrations. Equal amounts of the cell lysate protein (30 lg) were subjected to 10% SDS-PAGE, and the gels were then electroblotted using nitrocellulose membranes (Millipore, Boston, MA, USA). The membranes were then blocked with 5% non-fat dry milk in Tris-buffered saline with Tween-20 for 2 hours and incubated with the indicated primary antibody at 4°C overnight. The bound antibodies were detected with horseradish peroxidase HRP-conjugated secondary antibodies (1:5000; Cell Signal Technology). The reactive bands were visualized by chemiluminescence with the Luminol reagent (Millipore). GAPDH was used as a loading control.

| Immunofluorescent staining
The procedures used for immunofluorescent staining have been previously reported. 19  The excess medium was removed and serum-free medium (50 lL per well) was further incubation at 37°C, for 30 minutes. Cells were removed and resuspended at 1 9 10 5 cells/mL. For each cell line, 100 lL of the cell suspension was added into three wells and incubated with Matrigel â . Cells (three samples) without Matrigel â served as a control. The plates were incubated at 37°C for 1 hour, and then the excess media and non-adhered cells were removed. The remaining adhered cells were quantified using the 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma-Aldrich) assay. The absorbance was measured at 492 nm using a multifunction microplate reader (POLARstar â OPTIMA; BMG Labtech, Ortenberg, Germany). Three independent experiments were performed for each cell line.    For orthotopic xenografts, 2 9 10 6 cells were injected into the mammary fat pads of both sides. At the end of the sixth week, all mice were killed and the lung tissues of mice were dissected to obtain lung metastases.

| Statistical analysis
The SPSS statistical software (SPSS, version 18.0; Chicago, IL, USA) was used for statistical analysis. A value of P < .05 was considered as statistically significant. The statistical tests were two-sided.  To explore the cytoskeleton structure at regions of intercellular contacts, cytoskeleton organizations were analysed under low density or high density ( Figure 1E,F). At high density, the actin fibres in the MCF7-sh cells were assembled as contractile bundles in the cytoplasm and were gathered at regions with the cells contacting each other, whereas no aggregated distribution appeared in MCF7 cells.
The same phenotype was observed in the MM231 and MM231OE cells.

| AmotP130 affected cell adhesion and focal adhesion complexes
The results showed that F-actin assembled in bundles and gathered to the borders of cells after AmotP130 knockdown. The cytoskeletal rearrangement has been reported to be mediated by cell adhesion and focal adhesion kinase (FAK) activation. 22,23 Thus, it was conceivable that AmotP130 affected the cytoskeleton rearrangement through interfering with the cell adhesion ability and FAK activation.
We analysed the cell adhesion ability under various levels of AmotP130 expression (Figure 2A It has been reported that FAK is a core component of adhesion complexes 24 and plays an essential role in FA turnover. 25 Figure 4A). An additional file shows the related pathways and molecules in more detail (see Table S1). All the listed pathways were activated when  We next inhibited the Rho signalling pathway using ZOL, which was reported to inactivate Rho GTPases. 29 MCF7-sh and MM231 cells were treated with 70 lmol L À1 ZOL for 24 hours. Western blot analyses showed that ZOL decreased the protein levels of Rac1, Cdc42, P-Rac1/Cdc42 and ROCK1, but increased RhoA protein levels ( Figure 5B).
Previous studies showed that the translocation from cytoplasm to membrane is an early step in RhoA/ROCK signal transduction cascade and the membrane translocation could be an activation form of Rho A. 30,31 The pathway inhibitor ZOL inhibited membrane RhoA activation. 32,33 Therefore, the translocation of RhoA to the membrane might be more reliable than the total protein level to indirectly measure the activation of RhoA. We therefore further examined the RhoA protein levels in the cytosol and membrane fractions CHEN ET AL. The predicted pathway showed the experimental data on the signal path of the signal transmission. The signal transduction process of different molecules and the regulation of the differential genes were given. The activation predictions of other genes were predicted by molecular activation prediction. The intensity of the node colour indicated the degree of up-(red) or down-(green) regulation following AmotP130 knockdown in MCF7 cells. D, The network analysis from the STRING website. The network showed results obtained upon entering proteins including AmotP130, RhoA, Rac1, Cdc42, Rock1, FAK and Vinculin. This was the prediction of protein-protein association. *P < .05, **P < .01, ***P < .001 and MM231 cells ( Figure 5E, lower panel). After treatment with 70 lmol L À1 of ZOL for 24 hours, the stress fibre formation was less than that in untreated cells. This was clearly observed using confocal microscopy; F-actin appeared scattered in the cytosol as granules (the asterisk labels scattered actin; Figure 5E). At the cell border, less stress fibres formed into bundles and the cell cortex was smoother in the treated cells (arrow labelling the fibres and cell cortex; Figure 5E). The disorganization of the actin cytoskeleton after Rho inhibition was similar to that of the AmotP130 overexpressing To extrapolate this study to clinical practice, we searched a single integrated database that was publicly available with supporting follow-up data. 34,35 This integrative data analysis tool was used to analyse the prognostic power of the cytoskeletal pathway-related genes. Figure 6C  However, the results were negative ( Figure S2C).

| DISCUSSION
Amot is an angiostatin-binding protein that promotes endothelial cell migration. 7 Our team has published a review about Angiomotin family member previously. 36 We concluded that in epithelial and F I G U R E 6 Schematic of the research and the clinical analysis of the cytoskeleton-related biomarkers. A, Cell forward movement included several steps. First, the cell formatted protrusions from the non-adhesion state. Then, the cell adhered to the extra cellular matrix (ECM) and the focal adhesion (FA) established. The FAs grasped the ECM and the tension produced by stress fibres contraction made the FAs dissemble and turnover. Finally, the cell contracted and the cell body was squeezed forward. B, From current study, we can infer that AmotP130 regulated the RhoA translocation from membrane to cytoplasm, inhibiting the Rac1/Cdc42 expression as well as its phosphorylation. Also, from our observation, AmotP130 affected the FAK and the focal adhesion formation. As a result, the cytoskeleton rearrangement and actin polymerization were controlled by AmotP130. C, Comparison of overall survival (OS) and relapse-free survival (RFS) rates by different level of RhoA and Rac1 mRNA expression in patients with breast cancer from integrated database endothelial cells, all Amot isoforms were localized to cell tight junctions and played important roles in the apical-basal polarity and stability of the cytoskeleton. However, the function of Amot was reported differently by different researchers even in the same cell line. We thought that the two isoforms, cell types and experimental conditions, could affect the function of Amot. In tumour cells, the role of AmotP130 is complicated and it has been assigned both oncogenic and tumour suppressor in cancer. But precisely because of the inconsistent reports on AmotP130 and the lacking study in breast cancer, we designed the experiments to explore the function of AmotP130 in breast cancer. In our study, Amot played an important role in promoting breast cancer cell proliferation and invasion. 37 Recently, studies have indicated that AmotP130 and AmotP80 play different biological functions in vitro. AmotP80 enhanced cell migration and AmotP130 interacted with actin and affected cell shape. 10,38 In this study, we showed that AmotP130 reduced the adhesion and invasion ability of breast cancer cells. Regarding the mechanism, we found that AmotP130 expression was related with the arrangement and polymerization of F-actin filaments. Furthermore, microarray analyses showed that Rho GTPase family members were associated with AmotP130, which was reported to regulate the cytoskeleton and cell motility. 28,[39][40][41] Cell motility processes consist of several steps, including protrusion at the leading edge, adherence to the ECM, FA formation and turnover, cell contraction and pushing the cell body forward (Figure 6A). The process is dependent on the dynamic cytoskeleton. [12][13][14][15]42 As a result, the cells display an aggressive phenotype. In this study, AmotP130 expression was related with a less "aggressive" shape and reduced adhesion ability. This phenomenon was driven by the reorganization of the cytoskeleton as was controlled by AmotP130 expression. It is well known that the actin cytoskeleton is linked with the ECM by FAs 43 and that cell adhesion is an important component of cell migration. 44 FA formation and turnover have been reported to be related with the dynamics of the cytoskeleton and regulate cell movement. 45,46 Here, we identified the proteins FAK and vinculin, which were the core members of the adhesion complex. 24,47,48 Vinculin is mainly locates at the cell-cell junctions and at extracellular matrix focal spots, which participates in cellular chemical and chemical signal transduction through the interaction with adhesin proteins, cytoskeletal proteins and cytoskeletal F-actin. FAK is a non-receptor tyrosine kinase, which is presented in the cytoplasm.
It's activated and then transferred to FAs to catalyze the tyrosine phosphorylation of the target protein on the FAs. We focused on the colocalization of the two proteins to represent the cell FA formation. However, the different signal distribution of two proteins also existed due to the functional differences in the protein.
AmotP130 not only decreased the protein levels, but also changed their expression pattern within the cytoskeleton. Therefore, we suggest that AmotP130 is associated with lower invasive potential.
AmotP130 reorganized the cytoskeletal proteins, including actin fibres and FAs, thus interfering with cell protrusion, adhesion formation and turnover during the migration process ( Figure 6A).
The cell migration observed in cancer cells differs from normal cells and may have mechanistic differences. 49,50 We speculated in our study that AmotP130 changed the organization of cytoskeleton so that suppressed the cell migration and motility. In videos of living cells, the migration pattern appeared different between cells.
AmotP130 knockdown caused the cells to move quickly. The orientation depended on pseudopodium formation and angle change. The effective contraction of the pseudopodia controlled the cell "forward" movement. Therefore, it is reasonable to assume that AmotP130 reduces cell invasion and motility, especially pseudopodium-dependent movement.
To explore the mechanism behind these phenomenons, we analysed microarray data and identified several functional pathways affected by AmotP130. Notably, the Rho GTPase family members, which were reported to co-ordinate together and rearrange the actin cytoskeleton, 51 were also identified in our study. RhoA, Rac1 and Cdc42 52-54 are three prototypical members that co-ordinate together temporally and spatially. Previous studies suggested that Amot controlled RhoA activity in vitro, 55 and consequently regulated the cytoskeleton, cell junctions and migration. 56 We focused on the role of AmotP130 in this regulation. We verified the inhibition of the Rho pathway through Rac1, Cdc42 and P-Rac1/Cdc42 expression, and RhoA membrane protein localization 12,57 using ZOL, an inhibitor of the Rho pathway that inhibits the mevalonate pathway. 58,59 Pathway inhibition reduced the cell invasion ability, scattered actin fibres at the cytoplasm, impaired FAs, and this reversed the effects of AmotP130 down-regulation. These results implied that AmotP130 regulated the Rho pathway and changed FA and cytoskeleton organization, finally influencing cell invasion and migration. Taking the results of previous articles and our microarray assay together, we pointed out the connection between AmotP130 and Rho GTPase initially ( Figure 6B). Using the database, the Rho family members were suggested to have an influence on the prognosis of patients with cancer ( Figure 6C). But the analyses about Amot without separating the isoforms were negative ( Figure S2C). This information regarding the regulation of cytoskeletal pathways may suggest novel therapeutic targets for cancer treatment. And more clinical researches about the relationship between expression of AmotP130 and the survival of the patients are needed.

| CONCLUSION S
In conclusion, the experimental results of this study confirmed that AmotP130 was related with suppressed invasion ability through remodelling of the cytoskeleton of breast cancer cells, including actin fibre organization and FA turnover. Notably, the mechanism involved AmotP130 regulation of the Rho pathway. It is tempting to speculate that the cytoskeleton-related pathway proteins could provide new therapeutic targets for clinical breast cancer diagnosis or treatment. CHEN ET AL.