Resistin, a fat-derived secretory factor, promotes metastasis of MDA-MB-231 human breast cancer cells through ERM activation

Resistin, an adipocyte-secreted factor, is known to be elevated in breast cancer patients. However, the molecular mechanism by which resistin acts is not fully understood. The aim of this study was to investigate whether resistin could stimulate invasion and migration of breast cancer cells. Here, we report that resistin stimulated invasion and migration of breast cancer cells as well as phosphorylation of c-Src. Inhibition of c-Src blocked resistin-induced breast cancer cell invasion. Resistin increased intracellular calcium concentration, and chelation of intracellular calcium blocked resistin-mediated activation of Src. Resistin also induced phosphorylation of protein phosphatase 2A (PP2A). Inhibition of c-Src blocked resistin-mediated PP2A phosphorylation. In addition, resistin increased phosphorylation of PKCα. Inhibition of PP2A enhanced resistin-induced PKCα phosphorylation, demonstrating that PP2A activity is critical for PKCα phosphorylation. Resistin also increased phosphorylation of ezrin, radixin, and moesin (ERM). Additionally, ezrin interacted with PKCα, and resistin promoted co-localization of ezrin and PKCα. Either inhibition of c-Src and PKCα or knock-down of ezrin blocked resistin-induced breast cancer cells invasion. Moreover, resistin increased expression of vimentin, a key molecule for cancer cell invasion. Knock-down of ezrin abrogated resistin-induced vimentin expression. These results suggest that resistin play as a critical regulator of breast cancer metastasis.


Resistin increases invasion and migration of breast cancer cells.
To determine whether resistin affects the migration and invasion of MDA-MB-231 human breast cancer cells, wound-healing and invasion assays were performed. As demonstrated by the scratch wound assays, MDA-MB-231 cell migration was significantly increased following resistin treatment (Fig. 1A). In addition, resistin also increased MDA-MB-231cells invasion at a concentration of 10 ng/ml. The degree of invasion was further increased at higher concentrations Scientific RepoRts | 6:18923 | DOI: 10.1038/srep18923 of 25 ng/ml and 50 ng/ml (Fig. 1B). Resistin had no significant cytotoxic effect on cell viability up to 50 ng/ml in MTT assays (Data not shown). Tumor cells are known to have heterogeneity. To confirm the metastatic effect of resistin on the breast cancer cells, we tested another breast cancer MCF-7 cells (Supplementary Fig. 1 A,B,C and D). Resistin increased both migration and invasion of MCF-7 cells in a dose-dependent manner (Fig. 1C,D). These results demonstrated that resistin promotes invasion and migration of breast cancer cells.

Intracellular calcium is required for resistin-induced phosphorylation of c-Src.
Calcium is a crucial regulator of cell migration and invasion 37 . To investigate the effects of resistin on intracellular calcium, calcium concentrations were measured in the presence of resistin. Results indicated that intracellular calcium concentration was increased 30 seconds after addition of resistin ( Fig. 2A). c-Src is a critical link between multiple signaling pathways that regulate proliferation, invasion, survival, metastasis, and angiogenesis [38][39][40] . To evaluate the possible involvement of c-Src during resistin treatment, we measured c-Src phosphorylation. Phosphorylation of c-Src increased in a dose-and time-dependent manner in the presence of resistin (Fig. 2B,C). It has been reported Western blots were performed on cell lysates (30 μ g), and the membrane was probed for phospho-c-Src with c-Src as the control. (C) Cells were incubated with 25 ng/ml resistin for the indicated times. Western blots were conducted on cell lysates, and the membrane was probed for phospho-c-Src with c-Src as the control. (D) After being serum-starved overnight, cells were treated with the indicated concentrations of resistin for 30 min. Western blots were conducted on cell lysates, and the membrane was probed for phospho-c-Src with c-Src as the control. (E) Cells were incubated for 30 min with 25 ng/ml resistin in the presence of BAPTA-AM. Western blots were conducted on cell lysates, and the membrane was probed for phospho-c-Src with c-Src as the control (F) Each column represents the average of 4 replicates of the Matrigel invasion assays. The results are displayed as mean ± SD from 3 independent experiments. *P < 0.05 vs. untreated condition. that calcium regulates the activity of c-Src 41 . We therefore assessed the effect of BAPTA-AM, a calcium-selective chelator, on c-Src phosphorylation. We found that intracellular calcium chelation decreased phosphorylation of c-Src in a dose-dependent manner (Fig. 2D). We next investigated the role of calcium in resistin-mediated phosphorylation of c-Src. Chelating of intracellular calcium inhibited resistin-mediated phosphorylation of c-Src, suggesting that calcium is required for c-Src phosphorylation (Fig. 2E). To assess the effect of c-Src on resistin-induced invasion, we pre-treated breast cancer cells with the c-Src inhibitor PP2, and analyzed the invasion ability of MDA-MB-231 cells via a transwell assay. Pre-treatment with PP2 decreased resistin-induced MDA-MB-231 cells invasion (Fig. 2F), suggesting that this process is mediated by c-Src. Our data indicated that resistin stimulates invasion and migration of MDA-MB-231 cells through calcium-dependent c-Src pathway.

PP2A is involved in resistin-induced breast cancer cells invasion.
The catalytic subunit of PP2A is regulated by phosphorylation of Tyr 307, resulting in inactivation of the enzyme 42,43 . To assess the effect of resistin on PP2A activity, we measured phosphorylation of PP2A. We found that resistin increased PP2A phosphorylation in a dose-and time-dependent manner (Fig. 3A,B), suggesting that resistin reduces PP2A activity. It is known that c-Src directly phosphorylates PP2A 44 . Therefore, to elucidate the mechanism underlying resistin-mediated inhibition of PP2A activity, we next examined the interaction between c-Src and PP2A. We found that PP2A interacted Western blots were conducted on cell lysates (30 μ g), and the membrane was probed for phospho-PP2A with PP2A serving as the control. (C) MDA-MB-231 cells were immunoprecipitated with anti-PP2A antibody, followed by western blots against c-Src and PP2A. (D) Cells were treated with 25 ng/ml resistin for the indicated times, and cell lysates were immunoprecipitated with anti-PP2A antibody. Western blots were performed to detect protein expression of PP2A and Src. (E) Cells were treated with PP2 (0, 1, 3, 10, or 30 μ M). Cell lysates (30 μ g) were analyzed by western blots using antibodies against phospho-PP2A and phospho-c-Src, while PP2A and c-Src served as controls. (F) Cells were pre-treated with PP2 (10 μ M) and incubated with 25 ng/ml resistin for 60 min. Cell lysates (30 μ g) were analyzed by western blot using antibodies against phospho-PP2A and phospho-c-Src, while PP2A and c-Src served as controls. (G) For the Matrigel invasion assay, the number of cells invading through the Matrigel was represented by each column (4 replicates). The results are displayed as mean ± SD from 3 independent experiments. *P < 0.05 vs. untreated condition. **P < 0.05 vs. resistin-treated condition.
with c-Src in MDA-MD-231 cells (Fig. 3C). Resistin increased this interaction in a time-dependent manner, suggesting that resistin facilitates interactions between c-Src and PP2A (Fig. 3D). To determine whether c-Src kinase regulates phosphorylation of PP2A, we measured the level of PP2A phosphorylation following treatment with the c-Src inhibitor PP2. Inhibition of c-Src decreased PP2A phosphorylation in a dose-dependent manner (Fig. 3E). In addition, inhibition of c-Src decreased resistin-mediated phosphorylation of PP2A, suggesting that c-Src is involved in resistin-induced inhibition of PP2A (Fig. 3F). To better understand the role of PP2A, we measured MDA-MD-231 cells cell invasion using the transwell assay. Results indicated that MDA-MD-231 cells pre-treated with the PP2A inhibitor okadaic acid exhibited enhanced resistin-induced invasion (Fig. 3G). These results suggest that resistin stimulates invasion of MDA-MB-231 cells by reducing PP2A activity.
PKCα is involved in regulation of resistin-induced PP2A activity. PKCα has been implicated in tumor growth and progression 45 . To evaluate the involvement of PKCα in PP2A activity, we first investigated whether resistin increases phosphorylation of PKCα . Indeed, resistin increased PKCα phosphorylation in a doseand time-dependent manner (Fig. 4A,B). It was previously reported that PKCα is dephosphorylated by PP2A 46 . To investigate the relationship between PKCα and PP2A, we performed immunoprecipitation assays, and demonstrated that PP2A interacted with PKCα (Fig. 4C). Moreover, resistin was found to enhance this interaction in a time-dependent manner (Fig. 4D), suggesting that resistin regulates the interaction between PKCα and PP2A. To determine whether PP2A regulates phosphorylation of PKCα , we measured phosphorylation of PKCα following treatment with okadaic acid. Inhibition of PP2A activity increased the phosphorylation of PKCα , suggesting that PP2A enhances PKCα activity under resistin-treated conditions (Fig. 4E). We next assessed the effect of resistin on PKCα phosphorylation in the presence of okadaic acid. Inhibition of PP2A increased PKCα phosphorylation, suggesting that PP2A is involved in resistin-induced augmentation of PKCα activity (Fig. 4F). To investigate the role of PP2A on cancer metastasis, we measured invasion activity of MDA-MD-231 cells pretreated with okadaic acid, and found that inhibition of PP2A increased the resistin-induced breast cancer cells invasion (Fig. 4G). To confirm the effect of resistin on PKCα activation, we examined the subcellular localization of MDA-MD-231 cells expressing PKCα -DS-RED by immunocytochemistry. Treatment with resistin led to translocation of PKCα -DS-RED (arrow indicated parts) from the cytosol to the plasma membrane (Fig. 4H). These results suggest that resistin stimulates invasion of MDA-MB-231 cells by regulating the PP2A-mediated PKCα pathway.

Resistin stimulates invasion of breast cancer cells through ERM. ERM proteins provide a physical
link from F-actin to membrane-associated proteins on the surface of cells. This linker function makes ERM essential for many fundamental cellular processes such as determination of cell shape, polarity, and cell adhesion. We first tested the effect of resistin on ERM activation. Resistin induced ERM phosphorylation in a dose-dependent manner in MDA-MB-231 cells (Fig. 5A). Resistin also increased ERM phosphorylation in a time-dependent manner (Fig. 5B). It was previously reported that ezrin is a downstream effector of PKCα that controls cell motility 47 . To confirm this, we investigated the interaction between PKCα and ERM, Results showed that ERM was able to interact with PKCα (Fig. 5C), and that resistin increased this interaction in a time-dependent manner (Fig. 5D). To confirm the involvement of PKCα in ERM activation, we measured ERM phosphorylation following treatment with Gö6976, a PKCα inhibitor. Inhibition of PKCα activity reduced ERM phosphorylation in a dose-dependent manner, suggesting that PKCα is involved in resistin-induced ERM activation (Fig. 5E,F). Furthermore, ERM phosphorylation failed to occur in PKCα kinase-defective (KD) mutants (Fig. 5G). Compare with cells transfected with non-target siRNA, the cells treated with ezrin siRNA showed a wider wound area after wound generation by resistin indicating a defect in migration (Fig. 5H). To define the contribution of ERM to the resistin-induced metastatic effect, an invasion assay was performed following ezrin knockdown. Ezrin siRNA-transfected cells significantly attenuated resistin-induced cell invasion (Fig. 5I). To further confirm the effect of resistin on ezrin activation, we examined subcellular localization of GFP-tagged ezrin following resistin treatment. Treatment with resistin led to translocation ezrin (indicated by arrow) from the cytosol to the plasma membrane (Fig. 5J). As shown in Fig. 5K, co-localization of ezrin (green signal) and PKCα (red signal) to the plasma membrane (indicate by arrow) was observed following resistin treatment. These results suggest that resistin promotes invasion and migration of MDA-MB-231 cells by regulating the PKCα -ezrin pathway.

Ezrin is involved in resistin-induced vimentin expression.
Vimentin is a filamentous protein that controls cell shape changes during epithelial-mesenchymal transitions (EMT), and are strongly associated with cell invasion and poor tumor prognosis 48 . To gain insight into the roles of resistin on EMT, we evaluated the effect of resistin on vimentin expression. Administration of resistin induced an increase in vimentin expression (Fig. 6A). On the other hand, knockdown of ezrin by siRNA transfection significantly reduced resistin-induced expression of ERM proteins (Fig. 6B). Combined with all our data, we suggested the schematic diagram of the mechanism of resistin. Resistin triggers the activation of ERM pathway via calcium-mediated Src/PP2A/PKCα , which in turn induces the expression of vimentin (Fig. 6C). These results indicate that resistin triggers cell migration and invasion via activation of ERM proteins.

Discussion
The objective of our study was to determine whether resistin directly regulates invasion of tumor cells, and if so, what are the molecular signaling pathways are involved in the process. Our principal finding was that ERM proteins promote resistin-mediated invasion of MDA-MB-231 cells. Association of ERM with enhanced tumor cell invasion raises several questions regarding the mechanism by which resistin can promote tumor cell invasion. The results of our study suggest that PKCα -mediated ERM activation plays a critical role in the progression of tumor Scientific RepoRts | 6:18923 | DOI: 10.1038/srep18923 cell invasion. Combined with a recent report showing that ERM is critical for tumor cell invasion, these findings implicate a role for ERM in resistin-mediated tumor cell invasion.
PKC has been implicated in a various types of tumorigenesis including cancer cell proliferation, migration, and invasion 49 . Among PKC isozymes, PKCα has been suggested to be important in breast cancers 50,51 . For example, PKCα expression is elevated in breast cancer tissues 41 . PKCα overexpression also increases the invasiveness of breast cancer cells 52 . However, the molecular mechanism underlying PKCα -mediated tumorigenesis is not known due to a lack of isozyme-specific tools. Our results showed that ERM induction was dependent on PKCα activity, suggesting that PKCα -controlled pathways represent an alternative mechanism by which ERM can stimulate cancer cell invasion. ERM proteins provide a physical link from F-actin to membrane-associated proteins, which are important for many cellular processes including cell motility, cytokinesis, and migration. Phosphorylation of ERM is important for its activation. Several protein kinase C kinases have been implicated in ERM activation [53][54][55][56] . In the present study, resistin increased phosphorylation of ERM through PKCα . Moreover, interactions between these two proteins were observed under resistin stimulation. Nevertheless, this study does not address the molecular Some studies have proposed potential mechanisms connecting resistin with breast tumors. Accumulating evidence suggest that resistin exerts its neoplastic effect via two mechanisms. First, it acts directly on cancer cells by stimulating specific signaling pathways via an unknown receptor. Second, it may act indirectly on target cells by regulating inflammatory responses, influencing tumor angiogenesis, and modulating insulin sensitivity 57 . More mechanistic studies need to be conducted in order to understand this relationship at a molecular level. Adipocytes represent the majority of the breast tissue, and function as an active endocrine organ by secreting bioactive adipocytokines. In addition, adipocytes also regulates pathological processes such as carcinogenesis. In obesity-associated cancer, excess body weight has been proposed as a trigger for subclinical low-grade inflammatory state, where activation of pro-inflammatory adipocytokines may lead to carcinogenesis 58,59 . Importantly, resistin is found in inflammatory zones 59 , and elevated resistin levels are found in several malignancies such as breast and colorectal cancers. In vitro studies showed that resistin stimulates cancer cell proliferation through PI3K and MAPK signaling 60,61 . Another study demonstrated that resistin promotes the development of cancer through upregulation of proinflammatory cytokines 62 . Induction of matrix metalloproteinases (MMPs) may participate in tumor migration and invasion. In this study, we found that treatment of MDA-MB-231 cells with resistin led to a robust increase in the expression of vimentin, whereas induction of vimentin was not observed under ERM knockdown conditions. These results suggested that ERM may be involved in resistin-induced vimentin upregulation. Combined with the observation that vimentin is critical for tumor metastasis, the findings of our current study indicated that resistin induces tumor invasion via ERM-mediated vimentin induction. However, more studies are required to clarify the mechanistic role of resistin in the association between obesity and breast cancer. The concentration of resistin in humans ranges from 2 to 20 ng/ml 63 . In this experiment, we used 25 ng/ml of resistin, suggesting that our data may have physiological relevance.
Some studies have also investigated the metastatic effect of resistin in vivo. It was reported that resistin promotes chondrosarcoma metastasis and MMP2 expression through activation of AMPK/p38 signaling pathway and down-regulation of miR-519d expression in mouse models 64 . Mice that received anti-resistin antibodies showed decreased incidence of cancer development and metastasis 65 . It was proposed that resistin promotes cancer dissemination by maintenance and promotion of cell adhesion molecules 66 . Moreover, resistin influenced tumor progression via induction of pro-angiogenic proteins 67 . Together, these observations suggest that resistin is associated with cancer cell metastasis effect in vivo, but many questions remain to be answered.
Although we have provided clinical implications of resistin through the aforementioned experiments, there are still limitations in our study. Our experiments were all performed in an in vitro setting due to the mechanistic nature of the study. While our study results have provided molecular insights into breast cancer metastasis, our data are lacking in their physiological relevance. In addition, as we did not verify the disease phenotype in an ERM knock-down animal, we cannot conclude that metastasis of breast cancer is solely contributed by ERMs.

Methods
Reagents. Antibodies against protein phosphatase 2 A (PP2A) were purchased from Merck Millipore  complete the de-esterification process. The cells were treated with resistin, and the cultures plates were placed on a temperature-controlled confocal microscope (Zeiss LSM 700 Meta; Zeiss, Oberkochen, Germany) at 10× magnification. The excitation and emission wavelengths for signal detection were 488 nm and 515 nm, respectively. Western blot analysis. Following experimental manipulations, the medium was removed, and the cells were washed with ice-cold phosphate-buffered saline (PBS). They were then lysed with 100 μ l of lysis buffer [50 mM Tri-HCl (pH 7.4), 1% Triton X-100, 0.25% sodium deoxycholate, 150 mM EDTA, 1 mM sodium orthovanadate (Na 3 VO 4 ), 1 mM NaF, 1 mM phenylmethylsulfonyl fluoride (PMSF)]. Proteins were resolved on 10% SDS-PAGE gels and transferred to nitrocellulose membranes. Membranes were blocked in Tris-buffered saline with 0.1% Tween 20 (TBS-T buffer) and 5% dry milk (w/v) for 1 h, and washed 3 times in TBS-T. Membranes were then incubated overnight at 4 °C with primary antibodies, and probed with a secondary antibody conjugated to HRP (Amersham International PLC, Buckinghamshire, UK) for 1 h. The blots were visualized via chemiluminescence using the Amersham Biosciences ECL Detection System (Amersham International PLC). Wound-healing assay. Confluent cells were serum starved for 12 h, and then a standardized cell-free area was introduced by scraping the monolayer with a sterile tip. Cells were photographed using a phase-contrast microscope (4× , Nikon TS100). After intensive washing, fresh medium supplemented with 1% FBS was added into cells with different concentrations of resistin. After incubation for 48 h, cells were photographed again at 3 random areas. The migrated cells were quantified by manual counting, and the inhibition ratio was expressed as percentage of the control.
Invasion assay. The upper chamber of a transwell insert (8 μ m in pore size) was coated with 100 μ l of a 1:6 mixture of Matrigel (BD Biosciences, Bedford, MA) and PBS (invasion assay), and dried for 30 min at 37 °C. Cell suspension (100 μ l; 4× 10 5 cells/ml) in serum-free medium was placed in the upper compartment of the chamber. The bottom chambers were supplemented with 500 μ l complete medium (10% FBS) containing the indicated concentrations of resistin. After incubation for 24 h, the non-migrant cells from the upper face were scraped using a cotton swab. The invaded cells on the lower face were fixed with 4% paraformaldehyde (PFA), and stained with crystal violet (Sigma-Aldrich, St. Louis, MO, USA). Random fields were counted, and representative images were photographed using the AxioCam HRC CCD camera (Carl Zeiss).