Breast cancer is the most frequently diagnosed cancer in females worldwide and according to data from the World Health Organization comprises 16% of all female cancers. A major challenge currently facing the scientific community is that a large number of breast cancer patients present with metastatic cancer or relapse and metastasize after initial response to standard cancer therapy. The epithelial-mesenchymal transition (EMT) is an essential process in multiple biochemical changes, including embryonic development, tissue remodeling and wound healing. EMT also plays a crucial role in tumor invasion and metastasis (Wang and Zhou, 2001) and is a reversible phenotypic conversion that often occurs at the invasive front of many metastatic cancers (Christofori, 2006). The zinc-finger transcription factors Snail and Slug have been characterized as key EMT regulators (Nieto, 2002). Snail and Slug expression triggers EMT in breast cancer cells by repressing E-cadherin expression (Zhou et al., 2004); therefore, the loss of E-cadherin expression is considered a defining feature of EMT.

GSK3β inhibits Snail and Slug expression by inhibiting their transcription (Christofori, 2006) and regulating their degradation and nuclear translocation (Zhou et al., 2004). The active, un-phosphorylated form of GSK3β maintains both (i) epithelial phenotypes by inhibiting the expression and stabilization of Snail and Slug and (ii) high E-cadherin expression (Boble and Woodgett, 2007) in resting epithelial cells. However, Snail interacts physically and functionally with lysyl oxidase like-2 (LOXL-2) (Peinado et al., 2005) that is associated with aggressive tumors and participates in tumor progression (Barker et al., 2011; Moreno-Bueno et al., 2011; Peinado et al., 2008). The functional collaboration of LOXL-2 with Snail to repress E-cadherin expression is strictly dependent on the presence in the Snail protein of two lysine residues, K98 and K137 (Peinado et al., 2005), which are also involved in the interaction with GSK3β and ubiquitination (Zhou et al., 2004). Based on these findings, a hypothetical model is emerging in which LOXL-2 counteracts the effect of GSK3β on Snail.

Cysteine-rich angiogenic inducer 61 (Cyr61) was found to be upregulated in several cancers, including breast cancer cells, and possibly plays a critical role in the induction of EMT-related genes that promote invasion and metastasis in many cancers (Haque et al., 2011; Hou et al., 2014; Lin et al., 2005; Tan et al., 2009). Cyr61 mRNA expression is also positively correlated with more advanced features in breast cancer patients, such as tumor size and lymph node metastasis (Xie et al., 2001). Therefore, searching for phytochemicals that attenuate Cyr61 signaling is a promising approach for cancer treatment. Baicalein, a bioactive flavonoid extracted from the roots of Scutellaria baicalensis or Scutellaria radix, has been used in Chinese herbal medicine to treat respiratory tract infection, hepatitis, and cancer. Previous investigations have showed that baicalein causes cell cycle arrest and suppresses the proliferation of cancer cells, and induces apoptosis in a variety of human cancer cell lines (Chen et al., 2000; Fox et al., 2012; Kuo et al., 2009; Li-Weber, 2009; Po et al., 2002). Although baicalein has been reported to inhibit the migration and invasion of cancer cells (Wang et al., 2010; Wu et al., 2011), additional research into the mechanism of its antimetastatic activities is needed to facilitate development of anticancer therapies.

In the present study, we sought to identify the mechanism of antimetastatic activities of baicalein, which is related to Cyr61 and LOXL-2. We found that baicalein downregulates Cyr61 and subsequently enhances GSK3β activity, which then stimulates Snail and Slug ubiquitination. Moreover, baicalein induces a dramatic decrease in LOXL-2 levels and subsequently attenuates the interaction between LOXL-2 and Snail or Slug, which results in an enhanced interaction between GSK3β and Snail or Slug. These results suggest baicalein to be a promising candidate for novel anticancer drug with antimetastatic activity.

Cell culture and reagents

MDA-MB231 cells were cultured in DMEM medium and MCF-7 breast carcinoma cells (ATCC, USA) were cultured in a mixture of F-12K and DMEM media. Culture media were supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 1% antibiotic. All cultures were maintained at 37°C in a humidified incubator in a 5% CO₂ atmosphere. RPMI 1640 medium, bovine serum albumin, trypsin/ethylenediaminetetraacetic acid, FBS, and Antibiotic-Antimycotic 100× were purchased from Invitrogen (USA). Dimethyl sulfoxide (DMSO), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), baicalein and anti-β-actin antibodies were purchased from Sigma Aldrich (USA). Anti-Slug, anti-vimentin, anti-E-cadherin, anti-GSK3β, anti-pGSK3β, anti-Akt, anti-pAkt antibodies and 5127S-mouse anti-rabbit IgG (conformation specific) antibody—which does not recognize the IgG heavy (50 kDa) or light (25 kDa) chain—were purchased from Cell Signaling Technology (USA). Anti-Snail, anti-LOXL-2, anti-ubiquitin antibodies and MG132 were purchased from Santa Cruz Biotechnology (USA). Polyvinylidene fluoride (PVDF) membranes for Western blotting were purchased from Millipore (USA).

Cytotoxicity

Cells survival was measured by a standard MTT viability assay as previously described (Tran et al., 2016).

Wound healing assay

The cells were cultured in six-well plates and grown in medium containing 10% FBS to an almost confluent monolayer, then a linear wound was created in the monolayer by scratching using a plastic pipette tip. The monolayer was washed twice with PBS to remove debris or detached cells and the drugs were then added. The cultures were incubated at 37°C and observed using a microscope (Olympus, UK) after 24 h. Under the microscope, the cells that migrated into the cell-free zone defined by the linear “wound” line were evaluated.

Invasion assay

Invasion assays were performed in Boyden chambers. The polycarbonate filters (8-μm pore size, Corning) were precoated with Matrigel Matrix (BD Biosciences). Cells (1 × 10⁵ cells/ml) in 100 μl medium (containing 0.1% FBS) pretreated with or without 25 μM baicalein for 24 h were seeded in the upper chamber. Then, 600 μl medium with 10% FBS were added to the lower chamber and served as a chemotactic agent. After 24 h of incubation, the cells in the upper chamber were fixed in 4% paraformaldehyde for 20 min. Next, the Matrigel was mechanically removed from the filter using a cotton swab. The cells adhering to the filter underside were stained with crystal violet and counted under a microscope. Each invasion assay was performed in triplicate.

RT-PCR

Total RNA from cell lines was extracted with TRIzol reagent (Invitrogen). Reverse transcription was performed using the reverse transcription system (Promega, USA). PCR primers for amplification were as follows: Snail, forward 5′-GAGGACAGTG GGAAAGGCTC-3′, reverse 5′-TGGCTTCGGATGTGCATCTT-3′; Slug, forward 5′-GAACTCACACGGAGAAG-3′, reverse 5′-ACACAGCAGCCAGATTCCTC-3′; vimentin, forward 5′-AATG GCTCGTCACCTTCGTGAAT-3′, reverse 5′-CAGATTAGTTTC CCTCAGGTTCAG-3′; E-cadherin, forward 5′-GGAAGTCAGTT CAGACTCCAGCC-3′, reverse 5′-AGGCCTTTTGACTGTAAT CACACC-3′; GAPDH, forward 5′-GAGAAGGCTGGGGCTCAT TT-3′, reverse 5′-AGTGATGGCATGGACTGTGG-3′. PCR was performed using Taq polymerase (iNtRON Biotechnology Inc., Korea). PCR was initiated by incubating the samples at 95°C for 5 min, followed by 35 cycles of 40 s denaturation at 95°C, 40 s annealing at 55.4°C (for Snail), 57.5°C (for Slug and GAPDH), 58°C (for vimentin), 58.7°C (for E-cadherin) and 5 min elongation at 72°C. Samples were analyzed by electrophoresis on 1.2% agarose gels containing 0.002% nucleic acid staining solution (RedSafe™; Biotechnology Inc., Korea).

Western blotting

Treated cells were harvested and lysed in radioimmunoprecipitation assay (RIPA) buffer and kept on ice for 30 min. Bicinchoninic acid (BCA) assays were performed using equal amounts of protein sample per well. Lysate aliquots were separated on SDS-polyacrylamide gels and transferred to PVDF membranes using a glycine transfer buffer. Immonublotting was performed using primary and secondary antibodies. Protein bands were detected using the WEST-ZOL^® plus Western Blot Detection System (iNtRON, Korea).

Immunoprecipitation

Lysates were incubated with primary antibodies overnight with gentle rocking at 4°C. Protein A/G agarose beads (25 μl of 50% bead slurry) were then added and samples were incubated at 4°C with gentle rocking for 3 h. Samples were then centrifuged at 14,000 rpm for 1 min at 4°C and the pellets were resuspended in 40 μl 2× SDS sample buffer and vortexed. Next, the pellets were heated at 100°C for 10 min, separated on SDS-PAGE gels (10–12%), and transferred to PVDF membranes. Membranes were then probed using the protocol recommended for Western blotting by Cell Signaling Technology.

Transient transfection

Plasmids expressing Cyr61 and empty vector were prepared. Cells were transfected with Cyr61 and empty vector using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions and treated as indicated. At 24 h post-transfection, the cells were harvested and the lysates were subjected to immunoblotting analysis.

Statistical analysis

Group comparisons were performed using the SPSS version 12.0 software with one-way analysis of variance (ANOVA) and Student’s t-test. P < 0.05 and P < 0.001 were considered to indicate statistically significant differences. All experiments were performed in triplicate.

Baicalein suppresses the EMT of MDA-MB231cells

Cells were treated with baicalein at the concentrations that had no significant effect on cell survival (Supplementary Fig. S1) and tested for their metastatic potential using a unidirectional wound-healing assay and by a Transwell^® cell migration assay. Both cell migration (Fig. 1A) and invasion (Fig. 1B) were significantly decreased by exposure to baicalein at concentrations of (0–25 μM). Our results revealed that baicalein treatment upregulated E-cadherin at both the protein and mRNA levels in MDAMB-231 cells. Increased E-cadherin protein levels along with a gradual decrease in protein levels of the mesenchymal markers, Slug, Snail, and vimentin, were observed in baicalein-treated cells (Fig. 1C). However, MDAMB-231 cells treated with baicalein had no significant decrease in levels of the mRNA of transcriptional repressors, Snail and Slug (Fig. 1D), suggesting that baicalein-mediated downregulation of Snail and Slug proteins occurred partly at the post-transcriptional level.

Ubiquitination of Snail and Slug by baicalein is induced via the activation of GSK3β

Because the stability of Snail and Slug are regulated via ubiquitin-mediated proteasomal degradation (Zhang et a., 2003), we hypothesized that downregulation of Snail and Slug by baicalein may be mediated by ubiquitination. To test this hypothesis, MDA-MB231 cells were treated with baicalein alone or followed by the proteasomal inhibitor MG132. The results showed that MG132 blocked the Snail and Slug degradation induced by baicalein treatment (Fig. 2A). As shown in Fig. 2B, the ubiquitination of Snail and Slug was induced by baicalein treatment, suggesting that baicalein downregulates Snail and Slug via proteasomal degradation. Next, we examined whether the levels of Snail and Slug reduced by baicalein were mediated via the phosphorylation of GSK3β. As shown in Fig. 2C, baicalein markedly decreased the levels of phosphorylated GSK3β, suggesting that baicalein induces degradation of Snail and Slug via the GSK3β. To confirm that GSK3β is the main kinase that phosphorylates and then induces the degradation of Snail and Slug, we examined the role of GSK3β following treatment with LiCl, a potent GSK3β inhibitor (Lee et al., 2012). LiCl induced GSK3β phosphorylation and cotreatment with LiCl and baicalein rescued Snail and Slug, and subsequently caused the loss of E-cadherin expression, displaying EMT (Fig. 2C). Moreover, LiCl treatment blocked the baicalein-induced ubiquitination of Snail and Slug (Fig. 2D), suggesting that baicalein induces ubiquitination of Snail and Slug via GSK3β.

Baicalein reduced the expression of LOXL-2 that counteracts the interaction between GSK3β and EMT repressors Snail and Slug

LOXL-2 has been reported to be a potential partner of Snail that promotes its stability and collaborates in EMT (Peinado et al., 2005). Therefore, we assessed the ability of baicalein to counteract GSK3β activity via the interaction between LOXL-2 and Snail. Notably, treatment with baicalein inhibited LOXL-2 expression (Fig. 3A) and reduced the interaction of LOXL-2 with Snail and Slug (Fig. 3B), possibly due to downregulation of LOXL-2. As shown in Fig. 3C, baicalein induced the interaction of GSK3β with Snail and Slug, whereas it reduced the associations between LOXL-2 and Snail and Slug. As expected, the inhibition of GSK3β induced by LiCl treatment reduced the binding of Snail and Slug to GSK3β (Fig. 3C). Taken together, these findings suggest that baicalein enhances the interaction of GSK3β with Snail and Slug, possibly by suppressing the interaction of LOXL-2 with Snail and Slug.

Baicalein inhibits the EMT by downregulating the Cyr61-mediated Akt/GSK3β pathway

Because Cyr61 reportedly contributes to the activation of Akt that is upstream of GSK3β (Grezeszkiewiz et al., 2002), we examined the effect of baicalein on Cyr61 expression. There were increased elevated levels of Cyr61 and the mesenchymal markers, such as vimentin, Snail and Slug, but lower levels of the epithelial marker, E-cadherin, in MDA-MB231 cells compared with MCF-7 cells (Fig. 4A). As shown in Fig. 4B, baicalein markedly decreased the levels of Cyr61 and phosphorylated GSK3β, as well as the phosphorylation status of Akt in a dose-dependent manner. To confirm the mechanism underlying the effect of baicalein on Cyr61-induced Akt activation and inhibition of GSK3β activity, MCF-7 cells transiently transfected with Cyr61, referred to as MCF-7-Cyr61, were established (Supplementary Fig. S2). As shown in Figure 4C, the Cyr61 protein level was increased and p-Akt and p-GSK3β levels were higher in MCF-7-Cyr61 cells compared with MCF-7 cells (transfected with empty vector pcDNA3.0). The high levels of p-Akt and p-GSK3β in MCF-7-Cyr61 cells were reduced to the levels in MCF-7-pcDNA3.0 cells following baicalein treatment. Moreover, baicalein induced the reversal of EMT in MCF-7-Cyr61-transfected cells (Fig. 4C), indicating that baicalein inhibited the Cyr61-induced EMT via the Akt/GSK3β pathway. Collectively, our results suggested that baicalein targets Cyr61 to inhibit Akt activation, subsequently enhancing the activity of GSK3β, which stimulates the ubiquitination of Snail and Slug (Fig. 4D).

Cyr61, the initially discovered member of the CCN family, is highly expressed in various cancer cell lines and it has been reported to mediate various cellular functions (Babic et al., 1998; Chung et al., 2015). A previous study reported that Cyr61 is a novel and potent regulatory factor that can target the PI3K/Akt/mTOR signaling pathway (Grezeszkiewiz et al., 2002). Baicalein is a flavonoid derived from the root of Scutellaria baicalensis and has long been widely used in oriental medicine. Chung et al. reported that baicalin and baicalein inhibit TGF-β1-mediated EMT by reducing the expression level of the EMT-related transcription factor, Slug, via the NF-κB pathway (Chuang et al., 2012; Chung et al., 2015). Because Cyr61 can promote NF-κB activity (De Craene et al., 2005; Grezeszkiewiz et al., 2002), and also contributes to the activation of Akt that is upstream of GSK3β (Grezeszkiewiz et al., 2002), Cyr61 plays an important role during the EMT in breast cancer cells. However, studies regarding the Cyr61-mediated molecular mechanism underlying the antimetastatic effect of baicalein on breast cancer are limited. In the present study, we demonstrated that baicalein regulates EMT by targeting Cyr61.

Snail and Slug are highly unstable proteins with short half-lives (Dominguez et al., 2003). Our data indicated that baicalein induces Snail and Slug degradation via post-translational regulation processes (Figs. 2A and 2B). Recently, GSK3β was characterized as a kinase responsible for the phosphorylation of Snail and Slug (Jänicke, 2009), and for the protein stability of Snail (Grille et al., 2003). In the present study, we showed that baicalein not only downregulated GSK3β phosphorylation but also inhibited activation of Akt, which is a downstream regulator of PI3K (Fig. 4B). Akt is frequently activated in various cancers and plays a critical role in the induction of TGFβ and EGF-dependent EMT (Bakin et al., 2000; Julien et al., 2007). Akt can also phosphorylate IKKα to increase Snail expression and induce the EMT (Ma et al., 2016).

Moreover, we found that baicalein regulated LOXL-2 expression (Fig. 3A). LOXL-2 is one of the five members of the lysyl oxidase (LOX) family of extracellular matrix proteins and mediates the crosslinking of stromal collagens and elastin (Hayashi et al., 2004). LOXL-2 protein levels were higher in poorly differentiated breast carcinomas and the LOXL-2 mRNA levels were elevated in invasive and metastatic breast cancer cell lines (Kirschmann et al., 2002). Furthermore, LOXL-2 physically interacts with, and apparently attenuates the GSK3β-dependent Snail degradation (Peinado et al., 2005). Our results showed that baicalein inhibits the EMT by attenuating the interaction of LOXL-2 with Snail and Slug while stimulating the interaction of GSK3β with Snail and Slug (Fig. 3C), leading to their ubiquitination (Fig. 2D).

In the present study, we demonstrated for the first time that baicalein suppresses the EMT by modulating Cyr61 expression and its downstream signaling pathway, Akt/GSK3β, in human breast cancer MDA-MB231 cells (Fig. 4D). Baicalein induces ubiquitination of Snail and Slug via stimulating interaction between GSK3β and Snail or Slug. Moreover, baicalein downregulates LOXL-2, thus attenuating the interaction between LOXL-2 and Snail or Slug. Although the detailed mechanisms warrant further investigation to understand the role of LOXL-2 in breast cancer metastasis, these observations suggest that baicalein can be developed as a potential treatment for breast cancer metastasis.