Extracts of Perilla frutescens var. Acuta (Odash.) Kudo Leaves Have Antitumor Effects on Breast Cancer Cells by Suppressing YAP Activity

Yes-associated protein (YAP)/WW domain-containing transcription factor (TAZ) is critical for cell proliferation, survival, and self-renewal. It has been shown to play a crucial oncogenic role in many different types of tumors. In this study, we investigated the antitumor effect of the extracts of Perilla frutescens var. acuta (Odash.) Kudo leaves (PLE) on Hippo-YAP/TAZ signaling. PLE induced the phosphorylation of YAP/TAZ, thereby inhibiting their activity. In addition, the treatment suppresses YAP/TAZ transcriptional activity via the dissociation of the YAP/TAZ-TEAD complex. To elucidate the molecular mechanism of PLE in the regulation of YAP activity, we treated WT and cell lines with gene knockout (KO) for Hippo pathway components with PLE. The inhibitory effects of PLE on YAP-TEAD target genes were significantly attenuated in LATS1/2 KO cells. Moreover, we found the antitumor effect of PLE on MDA-MB-231 and BT549, both of which are triple-negative breast cancer (TNBC) cell lines. PLE reduced the viability of TNBC cells in a dose-dependent manner and induced cell apoptosis. Further, PLE inhibited the migration ability in MDA-MB-231 cells. This ability was weakened in YAP and TEAD-activated clones suggesting that the inhibition of migration by PLE is mainly achieved by regulating YAP activity. Taken together, the results of this study indicate that PLE suppressed cell growth and increased the apoptosis of breast cancer (BC) cells via inactivation of YAP activity in a LATS1/2-dependent manner.


Introduction
Perilla frutescens is a member of the Liliaceae/Labiatae (mint) family of flowering plants. Its leaves are consumed as a popular functional food in East Asian countries [1]. It has also been widely used as a traditional medicinal herb for expelling pathogens from the superficies and relieving exterior syndrome. e plant is used in traditional Chinese medicine (TCM) for pain due to blood stagnation caused by cold congealing. It has been used for various disorders such as depression-related diseases, asthma, anxiety, and tumors [1,2]. Studies on P. frutescens have demonstrated its antioxidant [3], anti-inflammatory [4,5], and anti-allergic properties [5,6]. Several studies have reported that P. frutescens inhibits tumor growth and invasion in metastatic cancer cell lines and animal models [7][8][9]. Luteolin and triterpene acids, constituents of P. frutescens, have anti-inflammatory and antitumor effects [10,11]. Although P. frutescens has been suggested to be a useful medicinal herb with antitumor activity, its exact molecular mechanisms are not yet fully understood.
e Hippo signaling pathway was first discovered in Drosophila, where it functions as a regulator of tissue and organ size [12]. Components of the Hippo pathway are highly conserved from Drosophila to humans, and a few mutations have been associated with human cancers [13]. e mammalian Hippo kinase cascade includes the mammalian STE20-like1/2 (MST1/2) and large tumor suppressor kinase 1/2 (LATS1/2) as well as the downstream effector Yes-associated protein (YAP) and WW domain-containing transcription factor (TAZ) [14]. YAP and TAZ are transcription co-activators without any DNA-binding domains; they enhance transcription by binding to specific transcription factors and in turn induce the expression of genes regulating cell proliferation, apoptosis, and differentiation [15]. e Hippo pathway inhibits both YAP and TAZ in a phosphorylation-dependent manner [16]. Abnormal YAP/TAZ overexpression and dysregulation of core Hippo kinases have been frequently observed in various human cancers [13,17].
Triple-negative breast cancers (TNBCs) are an aggressive type of breast cancer (BC) that do not express receptors against estrogen or progesterone and HER2 [18,19]. ese tumors have a poorer prognosis than other types of BC due to the resistance to hormonal therapies or HER2 targeted therapies such as trastuzumab. us, intense studies are being carried out seeking the underlying mechanisms and new treatment approaches for TNBCs [20]. Recent studies reported that the Hippo pathway is highly activated in TNBCs and could be a driving force for the progression of TNBCs [21][22][23][24].
Numerous proteins have been implicated to act upstream of and regulate the Hippo pathway [14,25]. However, the medicinal herbs that regulate the Hippo pathway are mostly unknown. erefore, the aim of this study was to investigate the antitumor effects of P. frutescens on TNBC cells and its role in the Hippo-YAP signaling pathway.

Retrovirus Infection and Stable Cell Lines.
MDA-MB-231 cells were infected with retrovirus with 5 µg/ mL polybrene (Millipore Co.). Retroviral infection was performed as previously described [27]. Infected cells were selected with 2 mg/mL G418 (Gold Biotechnology, Inc., St Louis, MI, USA). e analyses were performed using Agilent 1260 Infinity II with a ZORBAX Eclipse Plus C18 (4.6 × 250 mm, 5 µm, 95Å; Agilent Technologies, Inc., Santa Clara, CA, USA). e mobile phase contained 0.1% phosphoric acid (solvent A) and acetonitrile (solvent B). e flow rate was 0.8 mL/min, and the injection volume was fixed at 10 μl. e column temperature was maintained at 35°C during the separation, and the optical density of the effluent was monitored at 330 nm. Commercial β-carotene, caffeic acid, luteolin, quercetin, rosmarinic acid, and linolenic acid were purchased from Sigma-Aldrich and used as standards.

Western Blot and Immunoprecipitation Assay.
Western blotting was performed using a standard protocol [27]. For immunoprecipitation, cells were lysed in a mild lysis buffer (10 mM Tris at pH 7.5, 100 mM NaCl, 1 mM EDTA, 1% NP-40, 50 mM NaF, 1 mM Na 3 VO 4 , and protease inhibitor cocktail [Sigma-Aldrich]) and centrifuged at 4°C at 12,000 × g. e supernatants were immunoprecipitated with appropriate antibodies for 2 h at 4°C, followed by the addition of Protein A/G magnetic beads ( ermo Fisher Scientific, Inc.) for an additional 1 h at 4°C. e immunoprecipitates were washed three times with the mild lysis buffer, and then the samples were eluted with SDS-PAGE sample buffer.

Luciferase
Assay. HEK293A cells were seeded on 12-well plates, and after 12 h, they were transfected with 5×UAS-Luc reporter, pRL reporter, Gal4-TEAD4, and Flag-YAP2. On the next day, the cells were treated with PLE for 10 h and lysed. Luciferase activity was measured using the Dual-Glo ® luciferase reporter assay kit (Promega, Madison, WI, USA) according to the manufacturer's instructions. e relative firefly luciferase activity was normalized to that of Renilla luciferase.

RNA Isolation and Real-Time Quantitative Polymerase
Chain Reaction (RT-qPCR). RNA isolation was performed using an RNeasy mini kit (QIAGEN, Germantown, MD, USA). Equal amounts of cDNA were synthesized from 1 µg of total RNA using reverse transcriptase (Promega) with random hexamers (Takara Bio Inc., Shiga, Japan) and necessary reagents for synthesis. qPCR was performed using the SYBR Green PCR master mix (Sigma-Aldrich) with an appropriate primer pair, following the manufacturer's protocol. e normalization of several YAP-dependent genes was done using hypoxanthine-guanine phosphoribosyltransferase 1 gene (HPRT). e primers used were as follows: For the wound healing assay, the cells were plated in 6-well plates and allowed to reach approximately 80% confluence. Wounds were then generated by scratching the surface using a Scar ™ scratcher (SPL Life Sciences Co., Ltd., South Korea), followed by washing with phosphate-buffered saline (PBS) to remove the cell debris. e cell medium was replenished with or without PLE and allowed to migrate to the wound area for 36 h. e wound regions were photographed every 12 h, and the images were captured under a phase-contrast microscope (OLYMPUS 1X71 DP controller, Tokyo Japan). e average wound widths at 0 and 36 h time points were quantified using the ImageJ program (National Institutes of Health, Bethesda, MD, USA).
2.11. Cytotoxicity Assay. Cell viability was assessed using the EZ-Cytox cell viability assay kit (Daeil Lab Service, Seoul, South Korea). Briefly, HEK293A, MCF10A, MDA-MB-231, and BT549 cells were added to each well of a 96well plate. After incubation for 24 h, the cells were treated with PLE for 24 h. en, 10 μL EZ-Cytox solution was added to each well of the plate and the cells were incubated at 37°C for 2 h. Subsequently, the absorbance was measured at 450 nm using a microplate spectrophotometer (BioTek Instruments, Winooski, VT, USA). Cell viability (%) was calculated using the formula [A s /A c ] × 100, where A s is the absorbance of the well-containing cells, culture medium, EZ-Cytox solution, and stimulants and A c is the absorbance of the well-containing cells, culture medium, and EZ-Cytox solution.
2.12. Sulforhodamine B (SRB) Assay. HEK293A, MCF10A, MDA-MB-231, and BT549 cells were seeded into each well of a 96-well plate. After incubation for 24 h, the cells were treated with DMSO and 100 μg/mL of PLE for up to 5 days; the cell medium with serially diluted PLE was refreshed every 2 days. e plates were harvested every 2 days at the same time and fixed with 10% trichloroacetic acid (TCA) solution (Sigma-Aldrich) at 4°C overnight. e plates were washed with distilled water more than three times. After that, the plates were stained with 0.4% SRB (Sigma-Aldrich) Evidence-Based Complementary and Alternative Medicine 3 in 1% acetic acid solution for 30 min at 25°C. e cells were then washed with 0.1% acetic acid solution and dried at 25°C for 1 h. Finally, 10 mM Trizma base (Sigma-Aldrich) was added, and the plates were incubated on a rocker for 1 h to resolve the stained cells. e SRB levels were measured using a microplate spectrophotometer (BioTek Instruments) at an absorbance of 540 nm.

Clonogenic Growth
Assay. MDA-MB-231 cells and TEAD1ΔC-YAP (AD)-expressing cells were seeded in a 6well plate. After 24 h, the cells were treated with 10 and 25 μg/mL of PLE for 10 days. en, the cells were fixed with 4% formaldehyde for 10 min and stained with 0.25% crystal violet for 12 h. After washing three times with distilled water, the cells were destained with 95% ethanol for 1 h. e absorbance of the 100 μL destaining solution was measured at 595 nm using a microplate spectrophotometer (BioTek Instruments).

Statistical Analysis.
Values are expressed as the mean ± standard error of the mean (SEM) of three samples obtained from three independent experiments. Student's ttest (unpaired, one-tailed) was performed to calculate a p-value using Prism 5.0 software (GraphPad Software, La Jolla, CA, USA) and considered at * p < 0.05, * * p < 0.01, and * * * p < 0.001.

Analysis of PLE using HPLC.
In this study, we screened medicinal herb extracts to identify novel inhibitors of YAP/ TAZ. rough a review of the literature, 13 medicinal herbs were selected to screen for possible inhibitory activity against YAP/TAZ through the Hippo tumor suppressor pathway. To determine the effects of the medicinal herb on YAP activation, we evaluated the YAP phosphorylation status using mobility shift on a phos-tag gel. A pilot screening of the medicinal plant extracts identified that PLE potently inhibited YAP by promoting its phosphorylation ( Figure S1). Phytochemical constituents such as catechin, ferulic acid, apigenin, luteolin, rosmarinic acid, and caffeic acid have been identified from P. frutescens [28]. erefore, in this study, commercially available β-carotene, caffeic acid, luteolin, quercetin, rosmarinic acid, and linolenic acid were used as standard compounds for HPLC-UV analysis. e quantitative profiles of the PLE composition are shown in Figure 1(a). Rosmarinic acid was exclusively detected in the range of 38-40 min and at 4.6 mg per gram of PLE (Figure 1(b) and Table 1).

PLE Increases YAP Phosphorylation and Reduces YAP-TEAD-Mediated Transcriptional Activity.
To investigate the molecular mechanisms associated with the Hippo-YAP pathway, we treated HEK293A cells with the PLE. e IC50 value of HEK293A was 584.3 μg/mL (Figure 2(a)), and the percentages of cell proliferation were 18.6% lower when treated with 100 μg/mL of PLE, compared to those with control ( Figure 2(b)). Expression of pYAP, YAP, and TAZ in PLE-treated HEK293A cells was determined by immunoblotting. As shown in Figure 2(c), HEK293A cells showed increased phosphorylation of YAP at Ser127 and Ser397 following PLE treatment. Activated LATS1/2-driven phosphorylation on both Ser127 and Ser397 is an important step for nuclear/cytosol translocation and protein degradation [16,29]. PLE significantly reduced YAP and TAZ levels in HEK293A cells in a concentration-dependent manner.
ese results are consistent with those of previous studies showing that phosphorylated YAP and TAZ are sequestered and degraded in the cytoplasm, thus revealing that PLE could suppress the function of YAP by upregulating the level of phosphorylation [30].
Many transcription factors, including TEA domain family members (TEADs), RUNX2, TBX5, Oct4, and ErbB4, have been implicated as YAP partners [31,32]. In mammalian cells, the TEADs (TEAD1-4) are major partners mediating YAP/TAZ biological functions [33]. When Hippo signaling is not activated, YAP and TAZ are not phosphorylated. ese unphosphorylated YAP and TAZ enter the nucleus and bind to the transcription factor TEAD to induce gene transcription [34]. To determine the effects of PLE on YAP/TAZ and TEAD binding, we conducted a coimmunoprecipitation assay in HEK293A cells. Consistent with YAP phosphorylation, PLE blocked the interaction between YAP and TEAD (Figure 2(d)). To ascertain the effects of PLE on the YAP-TEAD transcriptional complex, we evaluated its transcriptional activity using the luciferase reporter system (Figure 2(e)). PLE also effectively blocked the ability of YAP to activate TEAD-mediated transcription but did not abolish its ability to activate RUNX2 ( Figure S2).
To further confirm the negative regulation of YAP by PLE, we measured YAP-TEAD-driven target gene expression. We found that PLE significantly diminished the expression of YAP-TEAD target genes, including CTGF, CYR61, INHBA, CXCL1, FSTL1, and PAPPA, in a dosedependent manner (Figure 2(f )). Collectively, our data suggest that PLE inhibited YAP activity via disruption of the YAP-TEAD transcriptional complex.

PLE Inhibits YAP/TEAD Transcriptional Activity via LATS1/2-Dependent Mechanisms but Not MST1/2-Dependent Mechanisms.
Among the Hippo signaling core kinases that regulate YAP/TAZ activity, LATS1/2 is one of the most significant kinases for YAP. Phosphorylation of Ser909 and r1079 is essential for LATS1 kinase activity [35]. Following PLE treatment, the increase in YAP phosphorylation could be due to LATS1 kinase activity. To test this hypothesis, we monitored Ser909 phosphorylation in endogenous LATS1 in response to PLE. Phosphorylation of Ser909 in LATS1 was increased by PLE treatment (Figure 3(a)). Next, we attempted to determine whether PLE downregulates the transcriptional activity of YAP/TAZ via LATS1/2 using the LATS1/2 −/− (LATS1/2 KO) HEK293A cells generated by the CRISPR-Cas9 system Figure 3(b). e transcriptional activation of YAP target genes such as CTGF and CYR61 was suppressed by PLE treatment. However, no noticeable changes in the expression levels of YAP target genes were observed between PLE-treated and untreated groups in LATS1/2 KO cells (Figure 3(c)). is suggests that PLE regulates target gene expression via LATS1/2-dependent mechanisms.
MST1/2 is autophosphorylated on multiple sites in response to various stress stimuli, which results in the activation of MST1/ 2 [36]. Among these multiple autophosphorylation sites, phosphorylation at r183 and r180 is essential for Mst1 and Mst2 activation, respectively [37]. Following PLE treatment, the increase in LATS1/2 phosphorylation could be due to MST1/2 activation. To test this hypothesis, we monitored r183 phosphorylation. e phosphorylation of r183 in MST1 was slightly increased by PLE treatment (Figure 3(d)). Next, to confirm the role of MST1/2 in PLE-induced YAP regulation, we compared YAP target gene expression between wild-type and MST1/2 −/− (MST1/2 KO) HEK293A cells (Figure 3(e)). As expected, PLE treatment decreased the expression of CTGF and CYR61 (Figure 3(f)). However, PLE still reduced the expression of CTGF and CYR61 in MST1/2 KO cells, suggesting that PLE reduces YAP target gene levels independent of MST1/2.
Although MST1/2 is well characterized as an upstream activating kinase of LATS1/2 [35], accumulating evidence indicates that MST1/2 and MAP4Ks directly phosphorylate and activate LATS1/2 in a partially redundant manner [26,38,39]. Next, to confirm the role of MAP4Ks in PLE-induced YAP regulation, we compared YAP target gene expression between wild-type and MAP4K4/6/7 −/− (MAP4K4/6/7 KO) HEK293A cells. PLE treatment decreased the expression levels of CTGF and CYR61, but MAP4K4/6/7 deletion did not alter the expression levels of these genes in response to PLE treatment, suggesting that PLE could reduce YAP target gene levels dependent on MAP4K4/6/7 ( Figure S3). In this study, we discovered a novel LATS1/2-dependent mechanism for the inhibitory effect of PLE on YAP/TAZ phosphorylation and inactivation, but MAP4Ks had a redundant function in activating LATS1/2 kinases.

Induction of Cytotoxicity and Apoptosis by PLE in TNBC Cells MDA-MB-231 and BT549.
e Hippo-YAP signaling pathway has attracted much attention in cancer research since the Hippo-YAP signaling pathway, a well-known regulator of organ size, has been known to play an essential   Evidence-Based Complementary and Alternative Medicine role in cancer development [40,41]. e regulation of Hippo-YAP represents a promising strategy for the treatment of various cancers [21,42]. erefore, we further explored the effects of PLE on Hippo-YAP signaling in TNBC cells. Western blotting was performed to determine the levels of YAP and TAZ protein in 10 BC cell lines. YAP and TAZ were overexpressed in the MDA-MB-231, BT549, and MDA-MB-468 cell lines, whereas their expression levels were relatively low in the MCF7 and MDA-MB-436 cell lines. MDA-MB-231 and BT549 have been reported to have high invasion potential; therefore, these two cell lines were selected for further study ( Figure S4).
To investigate whether PLE has a better cytotoxic effect on TNBC compared to normal human mammary epithelial cell line, the cytotoxic activity of PLE was examined using MTT assay in MDA-MB-231, BT549, and MCF10A cells with various concentrations of PLE for 24 h. e results showed that PLE significantly inhibited the growth of MDA-MB-231 and BT549 cells to a greater extent compared with that of MCF10A cells in a concentration-dependent manner, with IC50 values of 268.9 μg/ mL (Figure 4(a)), 307.1 μg/mL (Figure 4(d)), and 680.5 μg/mL ( Figure S5(a)), respectively.
e IC50 values showed that MCF10A cells were more resistant to the cytotoxic effect of PLE than both TNBC cells. ese results showed that PLE has a stronger cytotoxic effect on TNBC cell lines than on nonmalignant MCF10A cells. We then examined the effect of PLE on cell apoptosis. Because the cleavage of several key proteins, such as Caspase-3 and PARP, is considered to be a hallmark of apoptosis, we measured the levels of cleaved Caspase-3 and PARP in MDA-MB-231 and BT549 cells at PLE concentrations of 50 and 100 μg/mL. Western blot analysis revealed significant upregulation of cleaved Caspase-3 and PARP in both BC cell lines in a dose-dependent manner (Figures 4(b) and 4(e)).
Since PLE decreased TNBC cell viability and increased cell apoptosis effectively, we next investigated the suppressive effect of PLE on TNBC cell proliferation. MCF10A, MDA-MB-231, and BT549 cells were treated with 100 μg/mL PLE for 5 days, and their proliferation ability was assessed using the SRB assay. As shown in Figures 4(c) and 4(f ), PLE markedly inhibited the proliferation of MDA-MB-231 and BT549 cells to a greater extent compared with that of MCF10A cells (Figure S5(b)). Collectively, these results indicate that PLE significantly inhibited the growth of MDA-MB-231 and BT549 cells. (e, f ) Wild-type and MST1/2 KO HEK293A cells were treated with PLE for 12 h and mRNA levels of CTGF and CYR61 were measured using RT-qPCR (error bars represent ± SEM from n � 3 per group). Cell lines were immunoblotted to analyze indicated antibodies. * p < 0.05, * * p < 0.01, and * * * p < 0.001; Student's t-test (unpaired, one-tailed) was used for statistical analysis In Figure 3(a)), the numbers of size maker depart from indicated lines.

PLE Blocks Tumor Progression of Human BC Cells by
Repressing YAP Activity. We further explored the effects of PLE on Hippo-YAP signaling in TNBC cells. We examined the levels of phosphorylated YAP in treated BT549 and MDA-MB-231 cells. As shown in Figures 5(a) and 5(b), PLE induced phosphorylation and degradation of YAP as well as TAZ in MDA-MB-231 and BT549 cells in a dose-dependent manner. At least one TEAD gene is expressed in most adult tissues [43][44][45], and all four TEADs are abundantly expressed in TNBC cells as indicated by the expression profiles of the TEAD extracted from the Cancer Cell Line Encyclopedia (CCLE) ( Figure S6). To determine the effects of PLE on YAP/TAZ and TEAD binding, we conducted a co-immunoprecipitation assay in MDA-MB-231 cells. Consistent with YAP phosphorylation, PLE blocked the interaction between YAP and TEAD ( Figure 5(c)), strengthening the notion of the link between PLE and YAP inhibition. YAP enhances several processes responsible for tumor growth and metastasis, such as cellular proliferation, migration, and invasion [46,47]. erefore, we performed a wound-healing assay to determine whether PLE (c, f ) MDA-MB-231 and BT549 cells were treated with DMSO and 100 μg/mL PLE, and the medium with diluted PLE was refreshed every two days. After treatment for another 5 days, cell proliferation was determined by using SRB assay at an absorbance of 540 nm. e proliferation of MDA-MB-231 and BT549 cells with PLE was decreased by 59.3% and 53.88%, respectively, compared with control cells. Bars represent mean ± SEM (n � 3). * * * p < 0.001, * * p < 0.01; Student's t-test (unpaired, one-tailed) was used for statistical analysis. 8 Evidence-Based Complementary and Alternative Medicine BT549 WT and LATS1/2 KO cells generated by the CRISPR-Cas9 system (Figure 3(b)). e MTT assay showed that BT549 LATS1/2 KO cells were more resistant to the cytotoxicity of PLE than BT549 WT cells after treatment with PLE at increasing concentrations for 24 h. e IC50 value of BT549 WT was 307.1 μg/mL (Figure 4(d)), and the BT549 LATS1/2 KO showed an IC50 value of 713 μg/mL ( Figure S8(a)). Western blot results showed that PLE increased the levels of cleaved PARP1 proteins in BT549 cells to a greater extent compared with those in BT549 LATS1/2 KO cells ( Figure S8(b)). ese findings imply that PLE requires LATS1/2 to repress YAP/TAZ signaling.

Discussion
BC is the most frequently occurring cancer among women, accounting for 25% of all cancer cases worldwide. Despite the substantial efforts made in the past decades to improve survival and quality of life, BC remains a deadly threat for patients [48]. TNBCs, characterized by the absence of estrogen, progesterone, and ErbB2 receptors, account for approximately 10-15% of all BCs and are an aggressive type of BC. ey have a poorer prognosis than other types of BCs. TNBC has an inadequate response to tamoxifen or trastuzumab, which has been developed to treat estrogen receptor-positive (ER+) BC or HER2 positive BC [49,50]. In addition to the limitations of targeted In the clonogenic growth assay, DMSO-treated cells served as a control. Each value is expressed as mean ± SEM from n � 3 per group. * p < 0.05, * * p < 0.01; Student's t-test (unpaired, one-tailed) was used for statistical analysis. In Figure 5(b), the numbers of size maker depart from indicated lines.
Evidence-Based Complementary and Alternative Medicine therapies, the appearance of drug resistance to monotherapy is a major obstacle to the treatment of TNBC [18][19][20]. us, there is a continuing need to explore new therapeutic targets in BC.
In addition, the discovery of therapeutic agents for these targets is still essential in chemotherapy for BC. e role of YAP in organ size control or tumor development has been confirmed in mammals using transgenic mouse models [12,51]. YAP and TAZ are cotranscription factors between the cytoplasm and nucleus. ey are known to induce cell proliferation and antiapoptotic genes by binding to several transcription factors, particularly TEADs [41]. When Hippo, the core upstream kinase of YAP/TAZ, is activated, it phosphorylates LATS1/2, which in turn inactivates YAP/TAZ via phosphorylation. Conversely, when the Hippo kinase system is off, unphosphorylated YAP/TAZ enters the nucleus and expresses target genes associated with cell proliferation [34].
Activation of YAP/TAZ has been reported in various cancers [13,52]. YAP/TAZ plays a particular role in the transformation of healthy epithelial cells into metastasized cancer cells through the EMT process [53]. In particular, YAP/TAZ activity is proportionally correlated with the grade and stage of BC [54], and nuclear expression of TAZ is strongly associated with TNBC [55]. YAP enhances tumor progression and metastasis in BC cells [47]. us, YAP/TAZ has an oncogenic role in the migration and invasion of BC cells [56]. Strategies to inhibit YAP/TAZ activity or activate Hippo signaling can be beneficial in treating various cancers, including TNBC.
Natural medicinal products are derived from living organisms, such as plants, animals, or microorganisms. Recently, natural compounds have gained importance with an improved understanding of their pharmacological or biological effects and contributions for medicinal use [57]. Clinical studies have revealed that natural compounds have considerable promise against many diseases [58].
us, natural compounds with physiological functions have been identified as targeted interventions that can regulate the activity of key molecules of signaling pathways. Among the numerous natural plant products, P. frutescens has been widely used as a source of herbal materials in TCM. PLE is a functional and medicinal herb that has been widely consumed in Asian countries. Previous studies have reported that PLE has anti-oxidative, anti-inflammatory, and antitumor effects [2][3][4][5][6][7]. Several studies have revealed the presence of anthocyanins, flavonoids, and phenolic acids in P. frutescens [3,[8][9][10][11]. In addition, recent studies suggest that various compounds in PLE have anticancer activities. Luteolin, a flavonoid isolated from P. frutescens leaves, has been reported to inhibit tumor proliferation and metastasis and induce apoptosis via various cellular signaling pathways [59][60][61]. Rosmarinic acid was suggested as an active ingredient mediating anticarcinogenic effects in a mouse skin papilloma model and exerting anti-inflammatory effects through its antioxidant activity [62][63][64].
Our results showed that PLE exerts an anticancer effect on MDA-MB-231 and BT549 cells. PLE not only inhibited the viability of TNBC cells (Figures 4(a) and 4(d)) but also reduced their proliferation and migration (Figures 4(c), 4(f ), and 5(d)). Examining the effect of PLE on Hippo-YAP signaling revealed that PLE inhibited the activation of YAP/ TAZ via phosphorylation (Figures 2(c), 5(a), and 5(b)). PLE also inhibited the binding between YAP and TEAD in HEK293A and MDA-MB-231 cells (Figures 2(d) and 5(c)). PLE downregulated the transcriptional activity of YAP-TEAD (Figures 3(c) and 3(f )). In particular, the inhibitory effects of PLE on the expression of YAP-dependent genes were also abolished in LATS1/2 KO cells (Figure 3(c)). e cytotoxicity and apoptosis-inducing ability of PLE were also significantly attenuated in the LATS1/2 knockout clone ( Figures S8(a) and S8(b)). ese results suggest that PLE exhibits anticancer effects by inhibiting the activity of YAP through LATS-dependent YAP phosphorylation.
In this study, the anticancer efficacy of PLE through inhibition of YAP activity was clarified using an in vitro model. Hence, further studies are needed to verify the efficacy of PLE in animal models and to develop new therapeutic strategies using PLE to regulate the growth and proliferation of TNBCs. WW domain-containing transcription factor TNBC:

Abbreviations
Triple-negative breast cancer YAP: Yes-associated protein.
Data Availability e data used in the study are available upon request to the corresponding author.

Disclosure
Cho-Long Kim and Yu-Su Shin are the co-first authors.  Figure S1. Perilla frutescens var. acuta (Odash.) Kudo leaf extract (PLE) induces YAP phosphorylation. PLE increased YAP phosphorylation. HEK293A cells were treated with various natural compounds (50 μg/mL) for 6 h. Dimethyl sulfoxide (DMSO) was used as a control treatment. Phosphorylated proteins were detected using a phos-tag gel. Vinculin was used as the loading control. Figure S2. PLE has no effect on YAP mediated RUNX2-luferase reporter activity. HEK293A cells were co-transfected with a 6 × OSE2luciferase reporter for RUNX2, with or without HA-YAP2. HEK293A cells were treated with PLE in a concentrationdependent manner. Cells were harvested 10 h after treatment. Representative results of a single experiment with n � 3 biological replicates; three independent experiments were carried out. Luciferase activity was measured and normalized to that of co-transfected Renilla. Figure S3. PLE inhibited mRNA expression of YAP target genes in a MAP4K4/6/7-dependent manner. Wild-type and MAP4K4/ 6/7 KO HEK293A cells were treated with PLE for 12 h, and then mRNA levels of CTGF and CYR61 were measured using RT-qPCR (error bars represent ± SEM from n � 3 per group). * p < 0.05, * * p < 0.01, and * * * p < 0.001; Student's t-test (unpaired, one-tailed) was used for statistical analysis. Figure S4. Expression of YAP and TAZ on human BC and normal cell lines was analyzed by western blot. Cell lysates were subjected to immunoblotting with the indicated antibodies. Both YAP and TAZ were expressed predominantly in MDA-MB-231 and BT549 human BC cell lines. Figure S5. Effects of PLE on the viability of MCF10A cells. (a) MCF10A cells were treated using the indicated concentrations of PLE for 24 h. Cell viability was measured by using the EZ-Cytox cell viability assay kit. IC50 values were calculated from concentration response curves using Prism 5.0 software. e IC50 value of MCF10A cells was 680.5 μg/mL. Error bars represent ± SEM from n � 3 per group. (b) MCF10A cells were treated with DMSO and 100 μg/mL of PLE for 5 days. Cell proliferation was measured by SRB assay at an absorbance of 540 nm. MCF10A cells with PLE were decreased by 14% compared with that of the control cells. Bars represent mean ± SEM (n � 3). * * * p < 0.001, * * p < 0.01; Student's t-test (unpaired, one-tailed) was used for statistical analysis. Figure S6. PLE disrupts YAP-TEAD interaction in MDA-MB-231 cells. (a) Levels of TEAD expression among several BC cell lines. Gene expression analysis of the Broad-Novartis Cancer Cell Line Encyclopedia (CCLE) dataset showed that TEAD1-4 were highly expressed in various BC cell types. e x-axis indicates cancer cell types; y-axis indicates the target expression levels of TEAD1-4 in BC cell lines. (b) MDA-MB-231 cells were treated with PLE (50 μg/mL) for 6 h. Endogenous YAP/TAZ was immunoprecipitated, and the co-precipitated TEAD was detected by western blot. Figure S7. TEAD1ΔC-YAP (AD), a fusion of the TEAD DNA-binding domain with the YAP transactivation domain, was stably expressed in MDA-MB-231 cells. Expression of HA-pPGS (empty vector) and HA-TEAD1ΔC-YAP (AD) was confirmed using western blot, and the mRNA level of CYR61 was determined using RT-qPCR (error bars represent ± SEM from n � 3 per group). Figure S8. LATS1/2 KO BT549 cells increased resistance to PLE inhibitory effects to a greater extent compared with wild-type cells. (a) BT549 LATS1/2 KO cells were treated with various concentrations (0.001, 0.01, 0.1, 1, 10, 100, and 1000 μg/mL) of PLE for 24 h. Cell viability was measured by using the EZ-Cytox cell viability assay kit. IC50 values were determined from concentration response curves using Prism 5.0 Software. e IC50 value of BT549 LATS1/2 KO cells was 713 μg/mL. Error bars represent ± SEM from n � 3 per group. (b) To measure levels of apoptosis-related protein, western blot was performed on wild-type and LATS1/2 KO BT549 cells treated with PLE at 50 and 100 μg/mL for 12 h. e lysates were immunoblotted with the indicated antibodies. (Supplementary Materials)