Piperlongumine suppresses YAP-ET-1-CXCL2 signaling to modulates aggressiveness of triple-negative breast cancer cells.

Introduction
Breast cancer is the most common malignancy and ranks second in cancer mortality in women [1]. Based on the molecular marker expressions, breast cancer is stratified into different subtypes, including luminal, human epidermal receptor 2 (HER2), and triple-negative breast cancer (TNBC) [2]. Although TNBC merely accounts for 10 %–20 % of all breast cancer cases, it is the most aggressive and lethal breast cancer subtype that responds poorly to targeted or hormone therapy and is prone to metastasis and recurrence [3,4]. Due to its aggressive features and a lack of efficient therapeutic targets, patients with TNBC have poorer survival outcomes than non-TNBC patients [5]. Therefore, novel therapeutic strategies are urgently needed to improve the current state of TNBC treatments.
Natural products play essential roles in cancer drug development, especially as alternative treatments that preserve patients’ quality of life without compromising therapeutic efficacy. Piperlongumine (PL) is a natural alkaloid isolated from Piper longum L and has been extensively studied for treating various diseases, including cancers [[6], [7], [8]], inflammatory disorders [9,10], and oxidative stress-related conditions [11]. PL was shown to possess selective cytotoxicity towards cancerous cells in in vitro models of lymphomas, melanomas, glioblastomas [[12], [13], [14]], as well as oral, head and neck, lung, breast, and liver cancers [[15], [16], [17]]. Mechanistically, PL was found to inhibit cancer cell proliferation via induction of reactive oxygen species (ROS) and activation of the p38/c-Jun N-terminal kinase (JNK) and mitogen-activated protein kinase (MAPK) pathways [18,19]. PL was also observed to reduce aggressiveness in prostate cancer cells via downregulating nuclear factor (NF)-κB DNA-binding activity, along with decreasing nuclear translocation of p50 and p65 [20]. Additionally, PL was shown to induce caspase-dependent apoptosis through downregulating Bcl-2 and activating caspase 3 in multiple studies [[21], [22], [23]]. More importantly, in vivo studies demonstrated that PL exhibited selective cytotoxicity towards glioblastomas and thyroid tumors while being highly safe in animal models [24].
The Hippo pathway is a signaling pathway whose dysregulation is associated with a variety of diseases including cancers. Yes-associated protein (YAP)/transcriptional coactivator with a PDZ-binding domain (TAZ) are key downstream transcriptional components of the Hippo pathway that bind to TEA domain transcription factor 1–4 (TEAD1–4) transcriptional factors to regulate expressions of many genes that mediate cell proliferation, apoptosis, and stem cell self-renewal. TNBC cells were found to upregulate YAP/TAZ nuclear translocation to promote metastatic characteristics such as the epithelial-to-mesenchymal transition (EMT), cancer stemness, drug resistance, and metabolic reprogramming [[25], [26], [27]]. Therefore, targeting YAP/TAZ’s transcriptional activity may be a promising avenue for TNBC treatment.
To realize this, we examined the antitumor activity of PL and investigated its underlying inhibitory mechanism in TNBC cell lines. We analyzed differentially expressed genes (DEGs) upregulated in metastatic TNBC cells and subjected them to machine learning to identify distinct metastatic gene signatures associated with worse prognoses in breast cancer patients. Among these, we discovered the overexpression of endothelin (ET)-1 was a downstream gene of YAP, that was specifically upregulated in metastatic breast tumor cells and conferred worse prognoses especially in TNBC patients. Moreover, PL treatment induced tumor apoptosis and suppressed the migration and invasion of TNBC cells. A molecular study revealed that PL suppressed ET-1 transcription through attenuating YAP activation, suggesting that PL could be a promising therapeutic strategy for TNBC patients.

Materials and methods

Cell culture and reagents
The human BT549 (RRID: CVCL_1092) and MDA-MB-231 (RRID: CVCL_0062) TNBC cell lines were purchased from American Type Culture Collection. Cells were maintained in Dulbecco’s modified Eagle medium (DMEM), supplemented with 1 % penicillin-streptomycin, 1 % L-glutamine, and 10 % fetal bovine serum (FBS) at 37°C in a humidified atmosphere of a 5 % CO2 incubator. PL was purchased from MedChemExpress. Antibodies for Bcl-2 (GTX100064), BAX (GTX109683), ET-1 (GTX22786), Snail (GTX125918), vimentin (GTX100619), caspase-3 (GTX129040), caspase-9 (GTX112888), and α-tubulin (GTX112141) were obtained from Gentex. Antibodies for YAP (#12395), phosphorylated (p)-YAP (S127) (#13008), p-YAP (S61) (#75784), p-AMPK (AMPK; #130429), AMPK (#103487), p-LATS1 (S909) (#9157), and LATS1 (#3477) were purchased from Cell Signaling Technology. The antibody for Twist1 (A3237) was purchased from Abclonal. Piperlongumine (PL) was purchased from MedChemExpress (MCE) (catalog no. HY-N2329, ≥98 % purity). PL was dissolved in DMSO to prepare a 10 mM stock solution and stored at –20°C. Working concentrations were freshly prepared by diluting the stock solution in culture medium immediately before use.

Cell-viability assay and colony-formation assay
In total, 2 × 104 cells were seeded in 24-well plates and treated with indicated concentrations of PL for 24, 48, and 72 h. After treatment, culture medium was added with cell counting kit 8 (CCK8) reagent (Donjido) for another 3-h incubation, cell viability was measured at an absorbance of 540 nm on a spectrophotometer, and results are presented as percentages relative to untreated cells. BT549 and MDA-MB-231 cells were seeded at a density of 5 × 103 cells in six-well plates. Cells were refreshed with complete DMEM and supplemented with PL at the indicated concentrations for another 10 days of incubation, after which cell colonies were fixed with methanol, stained with 0.5 % crystal violet, and photographed.

Flow cytometry for cell apoptosis
2.5 × 104 cells were seeded in 24-well plates and treated with indicated concentrations (0, 5, and 10 μM) of PL for 24 h. After 24 h, cells were harvested, washed in ice-cold phosphate-buffered saline (PBS), and resuspended in binding buffer. Then, cell pellets were mixed with an annexin V/7-ADD kit (Donjido) for 15 min at 4°C in the dark. Cell apoptosis was evaluated with a BD FACSVia flow cytometer, and the rate of cell death was analyzed using a BD FACSVia flow cytometric internal software system (BD Biosciences).

Cell-proliferation assay
4 × 104 cells were seeded into 24-well plates and treated with indicated concentrations (0, 5, and 10 μM) of PL for 24 h. After treatment, cells were washed with PBS and incubated with 10 μM EdU reagent (Donjido) for another 6 h. Cells were then fixed with a 4 % paraformaldehyde (PFA) solution for 15 min, followed by permeabilization with 0.5 % Triton X-100 in PBS for 20 min. An appropriate amount of reaction solution and 4′,6-diamidino-2-phenylindole (DAPI) were added. The fluorescent signal was visualized on a fluorescence microscope (Zeiss).

Transwell migration and invasion assays
Cells were suspended in serum-free medium supplemented with PL and seeded at a density of 5 × 104 cells/insert into 8-μm-pore transwell inserts (Corning, New York, NY, USA) coated with Matrigel Matrix (Corning) or without. The lower chamber was supplemented with complete DMEM with 10 % FBS as a chemoattractant. After a 24-h incubation, tumor cells in the upper chamber were removed with a cotton swab, and cells that had invaded or migrated to the lower chamber were fixed with methanol at 4°C for 15 min, followed by crystal violet staining and microscopic observations.

CAF coculture assay
2 × 104 MDA-MB-231 cells were seeded in 24-well plates and treated with PL for 24 h. The next day, 2 × 104 human lung fibroblast WI38 cells were seeded into the 8-μm-pore trasnwell insert and cocultured with MDA-MB-231 cells for another 24 h. The migration of WI38 cells were stained with crystal violet and observed in an inverted microscope.

Western blotting
Total cellular proteins were extracted with iced radioimmunoprecipitation assay (RIPA) buffer containing a protease and phosphatase inhibitor cocktail (Roche) and quantified using the Bradford reagent. Equal amounts of protein were separated by denaturing sodium dodecylsulfate (SDS)-polyacrylamide gel electrophoresis (PAGE), followed by transfer to polyvinylidene difluoride (PVDF) membranes. After blocking with 1 % bovine serum albumin (BSA)/TBST blocking buffer at room temperature for 30 min, the PVDF membranes were incubated with specific primary antibodies at 4°C overnight. Membranes were washed three times with TBST buffer, followed by incubation with a horseradish peroxidase-conjugated secondary antibody at room temperature for 60 min. Protein bands was visualized on an enhanced chemiluminescence (ECL) system (Millipore, Bedford, MA, USA). Protein bands were quantified by ImageJ software.

Real-time quantitative polymerase chain reaction (RT-qPCR)
Total RNA was isolated with a GENzolTM TriRNA Pure kit (Geneaid), and 1 μg of an RNA sample was reverse-transcribed with M-MLV reverse transcriptase (Promega) and amplified with GoTaq qPCR Master Mix (Promega) using an Eco Real-Time PCR system (Illumina) with specific primers (Origene). All qPCRs were carried out at least three times, and results were calculated using the ∆∆CT equation and are expressed as multiples of change relative to the control sample. Primers are listed as following: YAP1 (forward), TAC TGA TGC AGG TAC TGC GG YAP1; YAP1 (reverse), TCA GGG ATC TCA AAG GAG GAC; ET-1 (forward), AAG GCA ACA GAC CGT GAA AAT; ET-1 (reverse), CGA CCT GGT TTG TCT TAG GTG; CXCL2 (forward), CCA ACC ACC AGG CTA CAG G; CXCL2 (reverse), GCG TCA CAC TCA AGC TCT G 18sRNA (forward), GCA ATT ATT CCC CAT GAA CG; 18sRNA (reverse), GGG ACT TAA TCA ACG CAA GC.

Plasmids and transfection
MDA-MB-231 and BT549 cells were seeded into six-well plates and transfected with small hairpin (sh)RNA against human YAP shRNA#1 (TRCN0000107265) and shRNA#2 (TRCN0000107266) in pLKO.1 vectors (National RNAi Core Facility; Academia Sinica) using the PolyJet transfection reagent (SignaGen Laboratories, Ijamsville, MD, USA). Target sequences of shRNA#1 and shRNA#2 were CCCAGTTAAATGTTCACCAAT and GCCACCAAGCTAGATAAAGAA, respectively. shRNA targeting LacZ was used as a negative control shRNA.

Immunofluorescence (IF) staining
MDA-MB-231 and BT549 cells were seeded onto eight-well Millicell EZ slides (Millipore) coated with poly-L lysine and allowed to attach overnight. Breast cancer cells treated with indicated concentration of PL (0 and 5 μM) for 24 h were fixed with 4 % PFA at room temperature for 15 min, followed by permeabilization, blocking, and incubating with an EDN1 antibody (GTX22786, Genetex, Irvine, CA, USA) at 4°C overnight. The next day, cells were washed with iced PBS three times, followed by incubation with a Daylight anti-mouse 568 secondary antibody at room temperature for 30 min. Nuclei were counterstained using UltraCruz mounting medium (Santa Cruz Biotechnology).

Animal studies
All animal studies were conducted in accordance with the guidelines and approval of the Animal Care and Use Committee of Taipei Medical University (LAC-2024-0259). Mice were maintained in a specific pathogen-free facility under a 12:12-h light/dark cycle, with ad libitum access to autoclaved chow and water. To evaluate the inhibitory effect of PL on tumorigenesis and metastasis, a syngeneic mouse model was established. Six-week-old female BALB/c mice were obtained from the National Laboratory Animal Center (Taipei, Taiwan). Briefly, 4T1 cells (1 × 106) were resuspended in 100 μL of PBS mixed with Matrigel and subcutaneously injected into the dorsal flank of BALB/c mice. After 10 days, when tumors reached approximately 100 mm3, the mice were randomized into two groups (n = 6 per group) and administered PL (10 mg/kg) every two days. Mice were sacrificed on day 26. Primary tumors, lungs, and livers were collected for further analysis. During the experiment, body weight and tumor growth were monitored every two days. Tumor volume was calculated using the formula: length × width2 × 0.52. Tissue samples from tumors, lungs, and livers were fixed, sectioned, and stained with hematoxylin and eosin (H&E), followed by light microscopy to evaluate the presence and distribution of metastatic nodules in secondary organs.

Immunohistochemical (IHC) staining
Breast cancer tissue microarrays (TMAs; BR10010-L87) were purchased from US Biomax. The TMA was deparaffinized, rehydrated, and heated in an antigen-unmasking solution (Vector Laboratories) for 1 h. Slides were blocked with 3 % bovine serum albumin (BSA) and incubated overnight with anti-YAP anti-ET-1 primary antibodies at 4°C. Tissues were incubated with SignalStain Boost IHC detection reagent (Cell Signaling Technology) for 1 h at room temperature the next day. Slides were washed with PBS three times and stained with 3,3′-diaminobenzidine (DAB) peroxidase substrate (Vector Laboratories).

Differentially expressed gene (DEG) analysis
Primary tumor and its lung metastasis were isolated from a previously estbalished orthotopic mouse model and sent for RNA sequencing (RNA-Seq). RNA expressions in fragments per kilobase million (FPKM) were used in a fold change analysis, and genes with an adjusted p value of <0.05 and log2 (fold change) of >1 or <0.5 were identified as being differentially expressed. In the total of 269 differentially expressed genes, 142 were upregulated and used in later screening for poor prognosis driver genes.

Survival analysis
Recurrence-free survival (RFS) prognoses of the ET-1 gene in luminal, HER2, and basal-like TNBC breast cancer patients were obtained from the KM plotter website (https://kmplot.com). High- and low-expression groups of patients were stratified according to optimal gene expression cutoff values.

Gene set enrichment analysis (GSEA)
Analyses of associations of EDN1 with metastasis-related biological functions were carried out using a GSEA algorithm with Gene Ontology Biological Process gene sets (https://www.gsea-msigdb.org). A TNBC cohort from GSE76275 was downloaded from a Gene Expression Omnibus (GEO) database. Patients were categorized into low- and high-EDN1 groups based on the median expression level.

Statistical analyses
Values are expressed as the mean ± standard deviation (SD) of three independent experiments. Statistical significance was determined by an unpaired, two-tailed Student’s t-test unless otherwise specified. A correlation coefficient was analyzed by Pearson's test. * p<0.05, ** p<0.01, and *** p<0.001 were regarded as significant. All statistical analyses were carried out using GraphPad Prism 6.0 software (La Jolla, CA, USA).

Results

Identification of ET-1 elevation in metastatic breast cancer cells
To investigate a distinct metastatic signature for predicting survival prognoses, we conducted an RNA-Seq analysis from primary (PT) and lung metastatic (LM) tumor cells, which were isolated from spontaneous lung metastatic tumors in an orthotopic mouse 4T1 mammary tumor model [28]. RNA-Seq data revealed 142 differential expression genes (DEGs) with a fold change > 2 that were significantly upregulated in 4T1-LM cells (Fig. 1A). To further validate clinical relevance of these metastasis-associated genes, we analyzed patient survival outcomes from the Metabric database with these 142 DEGs by using Lasso-Cox regression analysis to compute risk score for each patient. Then, 25 genes were narrowed down that correlated with the survival probability (Fig. 1B). Patients were then divided into low- and high-risk groups based on their median risk score (Fig. 1C). Based on risk group gene expression distributions, we depicted 6 genes (ET-1, ANGPTL4, MTHFD2, GAREM1, PHACTR2, and LIN28) that are highly expressed in high-risk groups (Fig. 1D), which were validated to be upregulated in 4T1-LM cells comparing to PT cells.
We previously identified that elevation of YAP signaling plays a crucial role in tumor metastasis, cancer stemness, and drug resistance of breast, lung, and colon cancers [[17], [18], [19],21,22], we therefore examined the correlation between YAP and the selected 6 genes. Results showed that 4 genes, including ET-1, ANGPTL4, PHACTR2, and GAREM1, were positively associated with YAP. Among these 4 genes, ET-1, which is known as endothelin-1, showed significant correlation with YAP (Pearson’s r=0.453, p<0.01) (Fig. 1E). To further validate the clinical significance of the YAP-ET-1 axis in breast cancer metastasis, mouse 4T1 tumor tissues were stained with YAP and ET-1 antibodies. Results of the IHC assay showed that YAP and ET-1 expressions were higher in lung metastatic tumor tissues, compared to primary tumors (Fig. 1F). Consistently, clinical examination of a breast tumor tissue microarray (TMA) showed that YAP and ET-1 expressions were concomitantly upregulated in patients with metastatic tumors (Fig. 1F). Furthermore, the KM plotter database revealed that ET-1 overexpression conferred a poor survival outcome to breast cancer patients, particularly those with the triple-negative breast cancer (TNBC) subtype (p=0.01), but not in luminal (p=0.60) or HER2 (p=0.53) patients (Fig. 1G).

PL reduces the growth and invasiveness of TNBC cells
Next, we sought to investigate potential chemopreventive agents capable of suppressing breast tumor metastasis. Specifically, we treated breast tumor cells with piperlongumine (PL), a bioactive compound derived from long pepper that has been reported to exert broad-spectrum antitumor activities; however, its impact on the aggressiveness of TNBC remains unclear. The human TNBC cell lines BT549 and MDA-MB-231 were treated with increasing concentrations of PL (0–25 μM) for 24–72 h. Results showed that PL reduced cell viability in a dose- and time-dependent manner, as indicated by progressively lower IC50 values with longer treatment durations (Fig. 2A-B). Additionally, the EdU staining assay for DNA replication indicated that treatment with PL at 5 μM significantly inhibited cell proliferation, and 10 μM of PL drastically reduced cell numbers (Fig. 2C). We thus examined cell apoptosis using annexin V/7AAD staining, and results indicated that treatment with PL at 5 μM for 24 h increased tumor apoptosis by 10 %–20 %, and PL at 10 μM drastically induced tumor apoptosis (Fig. 2D). Western blot analyses further validated the induction of apoptotic markers including upregulation of the BAX-to-BCL2 ratio with an increase of cleaved-caspase 3 and cleaved-caspase 9 expressions in the presence of PL (Fig. 2E).
We next examined the effect of PL on tumor aggressiveness, TNBC cell lines including MDA-MB-231 and BT549, and mouse TNBC 4T1 cells were pretreated with PL (0, 1, and 5 μM) for 24 h, and cell mobility was assessed by a transwell assay. Results showed that 1 μM PL significantly impeded cell migratory and invasive capabilities (Fig. 3A). Consistently, results of the wound-healing assay depicted that 1 μM PL substantially reduced cell migration to the wound region (Fig. 3B). Because cytoskeletal remodeling and cell polarization are essential for cancer cell migration and invasion, which are critical processes in metastasis, we performed immunofluorescence staining. Treatment with 1 μM PL induced actin filament depolarization in both MDA-MB-231 and BT549 cells (Fig. 3C). Therefore, we thus examined the effect of PL on the epithelial-mesenchymal transition (EMT) markers. The results showed that PL dose-dependently downregulated the mesenchymal markers including Snail, Twist, Vimentin, and N-cadherin, while it modestly upregulated epithelial marker E-cadherin protein level (Fig. 3D). These findings demonstrated the distinct antitumor properties of PL against the growth and invasiveness of TNBC cells.

PL inhibits YAP through both hippo-dependent and -independent mechanisms
The Hippo-YAP signaling axis plays a critical role in tumor progression. We previously demonstrated that YAP promotes tumor aggressiveness and stemness in colon and lung cancers [26,29], and drives metabolic reprogramming and therapy resistance in TNBC cells [25,30]. We therefore examined whether PL exerts an inhibitory effect on Hippo-YAP signaling. BT549 and MDA-MB-231 cells were treated with PL, and Western blot analysis revealed a marked reduction in total YAP protein levels following PL treatment (Fig. 4A). Time-course analyses further showed that PL induced an increase in YAP phosphorylation at Ser61 and Ser127 during the early phase of treatment (1-4 h), accompanied by a modest reduction in total YAP protein levels (Fig. 4B). Notably, prolonged PL exposure resulted in a progressive and more pronounced decrease in total YAP protein abundance, accompanied by reduced p-YAP levels at later time points (Fig. 4B). Given that Ser127 phosphorylation of YAP is mediated by the canonical Hippo kinase LATS1, we next examined upstream Hippo signaling and found that PL treatment time-dependently increased LATS1 phosphorylation (Fig. 4C). In addition, previous studies have reported that AMPK can facilitate YAP Ser61 phosphorylation [31], and consistent with this notion, PL treatment markedly elevated AMPK phosphorylation (Fig. 4C). These results suggest that PL modulates YAP signaling through both Hippo-dependent and Hippo-independent pathways in a time-dependent manner. Consistent with the biochemical findings, immunofluorescence staining demonstrated that PL treatment significantly reduced nuclear YAP localization in both TNBC cell lines (Fig. 4D), further confirming the suppression of YAP signaling activity by PL.

ET-1 is a downstream of YAP and is suppressed by PL
Since we observed a positive association between YAP and ET-1 expression in metastatic breast cancer (Fig. 1), we next examined whether YAP regulates ET-1 expression in TNBC cells. YAP was silenced in BT549 and MDA-MB-231 cells using two independent shRNAs, and qPCR analysis demonstrated that YAP knockdown significantly reduced ET-1 mRNA expression in both cell lines (Fig. 5A). To validate these findings at the protein level, Western blot analyses were performed under the same experimental conditions. Efficient depletion of YAP protein was confirmed in both TNBC cell lines, accompanied by reduced phosphorylated YAP levels and a marked decrease in ET-1 protein expression (Fig. 5A). These results indicate that YAP positively regulates ET-1 expression at both the transcriptional and protein levels. We next investigated whether pharmacological inhibition of YAP signaling by PL recapitulates the effects of genetic YAP depletion. PL treatment significantly suppressed ET-1 mRNA expression in both BT549 and MDA-MB-231 cells (Fig. 5B). Consistently, Western blot analyses revealed that PL treatment reduced total YAP protein levels in a dose-dependent manner, accompanied by alterations in YAP phosphorylation status and a concomitant decrease in ET-1 protein levels (Fig. 5C). In line with these biochemical results, immunofluorescence staining further demonstrated that PL treatment markedly reduced ET-1 protein expression in TNBC cells (Fig. 5D). Collectively, these findings establish ET-1 as a downstream target of YAP and demonstrate that PL suppresses ET-1 expression, at least in part, through attenuation of YAP signaling in TNBC cells.

ET-1 correlates with CXCL2 expression and CAF-associated features
ET-1 has been implicated in tumor progression and stromal remodeling through its ability to modulate inflammatory and chemokine signaling. To further investigate the potential role of ET-1 in the tumor microenvironment, we analyzed its association with immune- and stroma-related signatures in breast cancer datasets. Correlation analyses revealed that ET-1 expression was positively associated with multiple stromal cell populations, including CAFs, as well as with established CAF-related markers such as FAP, MFAP5, TNC, and PDGFA (Fig. 6A). In addition, ET-1 expression showed positive correlations with several IFNγ-inducible immune-related genes, among which CXCL2 exhibited one of the strongest associations (Fig. 6B). Given the reported role of CXCL2 in shaping inflammatory and protumorigenic stromal environments, we next examined whether modulation of YAP signaling affects CXCL2 expression. Genetic depletion of YAP significantly reduced CXCL2 mRNA levels in MDA-MB-231 cells, and pharmacological inhibition with PL produced a similar suppressive effect (Fig. 6C). To assess the functional consequences of these alterations, we performed Transwell migration assays using WI-38 fibroblasts cultured with conditioned media derived from control, YAP-depleted, or PL-treated MDA-MB-231 cells. Conditioned media from YAP-depleted or PL-treated tumor cells significantly impaired fibroblast migration compared with controls (Fig. 6D). Consistently, both YAP knockdown and PL treatment led to reduced α-SMA expression in WI-38 fibroblasts, indicating attenuation of fibroblast activation-associated features (Fig. 6E). Together, these data suggest that YAP-mediated ET-1 expression promotes CXCL2 expression induction and CAF activation, which can be attenuated by PL treatment.

PL suppresses breast tumor growth and modulates CAF-associated features in the tumor microenvironment
Finally, to evaluate the in vivo effects of PL, we established an orthotopic breast cancer model in Balb/c mice by injecting murine TNBC 4T1 cells, followed by treatment with PL (10 mg/kg, every two days). Throughout the treatment period, the body weights of PL-treated mice remained comparable to those of PBS-treated controls, indicating good tolerability (Fig. 7B). Notably, PL treatment significantly suppressed primary breast tumor growth from day 18 onward compared with PBS-treated mice (Fig. 7A). Histological examination of major organs revealed no overt pathological alterations in the liver or kidney following PL administration, suggesting minimal systemic toxicity (Fig. 7C). Lung tissue sections were examined histologically, and H&E staining showed that PL treatment markedly reduced lung metastasis compared with PBS-treated controls (Fig. 7C). Immunohistochemical analyses of primary tumors demonstrated that PL treatment markedly reduced the expression of YAP, Ki67, and ET-1, while increasing cleaved caspase-3 levels, indicating reduced tumor cell proliferation and enhanced apoptosis (Fig. 7D). Additionally, PL treatment significantly decreased the expression of the cancer-associated fibroblast markers α-SMA and FAP, suggesting attenuation of CAF-associated features within the tumor microenvironment (Fig. 7D). Collectively, these findings demonstrate that PL exerts potent anti-tumor and anti-metastatic effects in vivo, highlighting its therapeutic potential for TNBC.

Discussion
Over the past few decades, PL was reported to possess a variety of biological activities including anti-diabetes, anti-inflammation, neuroprotection, and especially antitumor properties [32,33]. The anti-neoplastic activities of PL were demonstrated in several cancer models. PL triggers tumor cell death via inducing apoptosis, pyroptosis, and ferroptosis [[34], [35], [36]]. In addition, autophagic cell death induction by PL was also identified [[37], [38], [39]]. In breast cancer, PL inhibits antioxidant enzymes and promotes DNA damage [40], and PL induces apoptosis and synergizes with doxorubicin by inhibiting the Janus kinase 2 (JAK2)-signal transduction and activator of transcription 3 (STAT3) pathway in TNBC [41]. However, the effect of PL on TNBC aggressiveness and its underlying inhibitory mechanism remain unexplored. In the present study, we demonstrated that the therapeutic window of PL against TNBC cells was a dose of approximately 1–10 μM. Our data showed that PL at 10 μM drastically suppressed cell growth and induced caspase-dependent apoptosis, while PL at 1 μM greatly suppressed migration and invasion. Previous studies showed that PL exhibited specific cytotoxicity against tumor cells, while showing limited cytotoxicity against healthy cells. Those in vitro and in vivo studies confirmed that PL does not affect normal cells in concentrations which are highly toxic toward cancer cells and exhibits no obvious side effects in mouse models [36,42]. Thus, PL might be promising for further development as an anticancer agent for treating TNBC.
Mechanistically, we identified that PL induced YAP phosphorylation and inhibited YAP nuclear translocation via activating upstream mediators, including LATS1 and AMPK. LATS1, a Hippo pathway kinase, phosphorylates YAP at serine127, which promotes the association of YAP with 14-3-3 and triggers proteosomal degradation. Accordingly, phosphorylation of YAP at Ser61 is attributed to AMPK which is governed by intracellular energy levels [31,43]. Notably, YAP also plays a crucial role in promoting tumor glycolysis, which underpins cell growth, and cancer stemness and metastasis [44,45]. In line with our findings, recent studies showed that PL impaired glycolytic metabolism in TNBC and lung cancer cells [17,46], and PL reversed cisplatin resistance via attenuating Hippo-YAP in oral squamous cell carcinoma [47]. Hippo-YAP signaling controls multiple metabolic pathways which coordinate the availability of energy and metabolites and are associated with tumor progression. We previously identified YAP upregulation in metastatic colon and lung cancers [26,27,29], and YAP plays a critical role in modulating intracellular redox homeostasis, which further supports metabolic adaptation and consequently promotes breast cancer metastasis [30]. Thus, these data underscore the antitumor potential of PL being involved in modulating tumor metabolism and could be a novel therapeutic agent.
ET-1 and its receptors (ET-1Rs), also known as ETA receptor (ETAR) and ETB receptor (ETBR), belong to the G protein-coupled receptor (GPCR) superfamily and play physiological roles as potent vasoconstrictors. They were proven to promote tumor progression in a variety of cancer types [48]. Elevation of ET-1 and ET-1R orchestrates tumor metastasis via their roles in regulating the EMT, cell invasion, and angiogenesis [49]. It was demonstrated that ET-1R overexpression induces dephosphorylation and nuclear accumulation of YAP [50], and activation of the ET-1 axis fosters YAP-induced chemotherapy escape in ovarian cancer [51]. Moreover, ET-1 mediated the activation of YAP/ZEB1 circuit to further drive cellular plasticity, invasion and EMT in ovarian cancer [52]. Herein, we identified that ET-1 was significantly upregulated in metastatic tumor tissues of mouse mammary tumors and human breast cancer patients, and ET-1 overexpression conferred a risk for predicting poor survival probabilities for breast cancer patients. Moreover, ET-1 was identified as downstream of YAP, while suppression of YAP diminished ET-1 transcription. ET-1 expression was associated with oncogenic pathways involved in tumor mobility and metastasis. Importantly, our data showed for the first time that treatment with PL effectively inhibited ET-1 expression, suggesting that downregulation of YAP-ET1 expression by PL contributed to its inhibitory activity against tumor aggressiveness. Moreover, PL attenuated actin polarization and suppressed EMT markers expression and tumor invasion, suggesting its inhibitory action on YAP-ET1. Nevertheless, the in vivo antitumor metastatic capability by PL as a preclinical evaluation and the combined effect of PL with conventional therapies for treating TNBC are worth investigating in the future.
CAFs represent a major stromal component of the tumor microenvironment and play critical roles in supporting tumor growth, immune modulation, and metastatic progression. Emerging evidence has highlighted the importance of chemokine-mediated crosstalk between tumor cells and CAFs in shaping a protumorigenic niche [53]. Among these factors, CXCL2 has been implicated in the recruitment of myeloid cells, induction of inflammatory signaling, and establishment of a permissive stromal environment [54,55]. Importantly, our in vivo analyses further demonstrated reduced expression of CAF-associated markers, including α-SMA and FAP, in primary tumors following PL treatment. These observations support the involvement of the YAP-ET-1-CXCL2 axis in modulating CAF-associated features within the tumor microenvironment and are consistent with our in vitro functional findings. Though the present study primarily focuses on histological and functional readouts, the precise molecular mechanisms by which ET-1-CXCL2 signaling influences CAF activation states, recruitment dynamics, and stromal remodeling remain to be fully elucidated and warrant further investigation. Addressing these questions will benefit from future studies incorporating comprehensive in vivo functional assays, expanded CAF marker panels, and advanced approaches such as single-cell transcriptomic profiling to resolve CAF heterogeneity. Collectively, our findings highlight a tumor-intrinsic signaling axis linking YAP activity to chemokine-mediated stromal modulation and provide a conceptual framework for future investigations aimed at targeting tumor-stroma crosstalk in TNBC.

Funding
This study was supported by grants from the Ministry of Science and Technology, Taiwan (NSTC112-2320-B-038-046-MY3, MOST111-2320-B-038-019, and MOST111-2314-B-038-081 to Lin C.W.), Chi-Mei Medical Center (114TMU-CM-04 to Yang C.C.), and Far-Eastern Memorial Hospital (FEMH-2024-C-029, FEMH-2025-C-032 to Sheung P.W.).

Ethics approval and consent to participate
Not applicable.

Patient consent for publication
Not applicable.

Data availability statements
The data generated in the present study may be requested from the corresponding author.

CRediT authorship contribution statement
Ming-Yi Hsieh: Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Ching-Chieh Yang: Resources, Methodology, Investigation, Funding acquisition. Wen-Jing Hsu: Resources, Formal analysis, Data curation. Zei-Wei Liu: Resources, Formal analysis. Hsin-Ying Lu: Methodology. Ming-Chen Chiang: Resources. Cheng-Jui Huang: Resources, Formal analysis. Ching-Yun Liao: Formal analysis. Cheng-Wei Lin: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Funding acquisition. Pei-Wei Shueng: Resources, Funding acquisition.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.