Mechanism by which Icariin suppresses pulmonary metastasis in triple-negative breast cancer through downregulation of the SPDL1/JAK2/STAT3 signaling pathway

Abstract

Triple-negative breast cancer (TNBC), a distinct subtype of breast cancer, is characterized by the concurrent absence of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) in tumor tissues. This molecular phenotype renders current endocrine and targeted therapies ineffective, underscoring the urgent need to explore novel therapeutic targets. This study focused on icariin (ICA), a bioactive flavonoid derived from the traditional Chinese herb Epimedium, and systematically elucidated its mechanism in suppressing TNBC progression on the basis of its unique antitumor pharmacological properties. In vitro cell models and an in vivo pulmonary metastasis mouse model revealed that ICA significantly inhibited the malignant biological behaviors of TNBC cells (proliferation, invasion, and metastasis). RNA sequencing revealed significant alterations in the expression of key molecules of the JAK-STAT signaling pathway. Mechanistic investigations demonstrated that ICA suppressed the phosphorylation and activation of the JAK2/STAT3 signaling axis by downregulating spindle apparatus coiled-coil protein 1 (SPDL1). These experimental data confirmed that modulation of the SPDL1/JAK2/STAT3 signaling pathway constitutes the molecular basis of the anti-TNBC effects of ICA. This study elucidated the anti-TNBC mechanism of ICA at the signal transduction level, providing an innovative theoretical foundation for the development of natural product-based targeted therapies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12935-026-04216-3.

Introduction

Breast cancer, a predominant gynecological malignancy, has drawn significant attention because of the distinct clinicopathological features of TNBC and its molecular subtypes. Epidemiological evidence indicates that TNBC patients exhibit a younger median age of onset (45–55 years on average), high tumor histological grade, aggressive metastatic potential, and elevated postoperative recurrence rates, collectively contributing to a markedly lower 5-year survival rate than other breast cancer subtypes [1]. Current clinical management relies primarily on radical surgery combined with neoadjuvant chemotherapy supplemented with adjuvant radiotherapy, immune checkpoint inhibitors (anti-PD-1/PD-L1 monoclonal antibodies), and PARP inhibitors [2, 3]. Additionally, novel potential drugs, such as the MNK1/2 degrader VNLG-152R and its deuterated analogs, have emerged as promising therapeutic options in the fight against TNBC [4]. Notably, approximately 30% of early-stage TNBC patients develop distant metastases posttreatment, with metastatic lesions frequently acquiring resistance to conventional chemotherapeutic agents, representing the leading cause of therapeutic failure [5]. At the molecular level, epithelial‒mesenchymal transition (EMT), a critical biological process enabling tumor cell migration and invasion, orchestrates TNBC metastatic cascades. Fundamental studies have revealed that the upregulation of EMT-associated transcription factors (vimentin and N-cadherin) facilitates tumor cell detachment from primary sites by remodeling cytoskeletal structures and degrading extracellular matrices, underscoring the importance of identifying agents that target TNBC invasion and metastasis suppression [6].

Icariin (ICA), the principal bioactive constituent of the medicinal herb Epimedium, has demonstrated antitumor activity across multiple cancer types. For example, ICA induces apoptosis in hepatocellular carcinoma cells via the PPARG/FABP4 signaling pathway and suppresses proliferation in colorectal cancer through epigenetic silencing of the circβ-catenin axis [7, 8]. Additionally, ICA reduces prostate cancer cell viability by inhibiting androgen receptor (AR) activity [9]. However, the therapeutic role and molecular mechanisms of ICA in TNBC remain to be fully elucidated.

The JAK2/STAT3 signaling axis is constitutively activated in various human malignancies. For example, interleukin-6 (IL-6) promotes hepatocellular carcinoma cell proliferation by upregulating the IL-6 receptor (IL-6R) through JAK2/STAT3 pathway activation [10]. Notably, KIAA0040 overexpression is positively correlated with nervous system gliomas via the JAK2/STAT3 signaling axis, establishing it as a prevalent therapeutic target [11]. JAK2 and STAT3 are indispensable for cellular proliferation and survival under pathophysiological conditions, demonstrating functional interdependence. Phosphorylation serves as a critical activation mechanism, with STAT3 phosphorylation-mediated activation driving protumorigenic effects [12]. Small-molecule inhibitors targeting this pathway have been developed.

SPDL1, also designated CCDC99, is a cell cycle-regulating gene that encodes the spindly protein, a critical mediator of spindle checkpoint silencing during mitotic progression [13]. The overexpression of SPDL1 is correlated with poor clinical outcomes in TNBC patients and facilitates epithelial‒mesenchymal transition (EMT) in breast cancer cells [14]. Although SPDL1 has positive regulatory effects on TNBC proliferation and EMT, its functional involvement in ICA-mediated tumor suppression remains undefined. In this study, we demonstrated that SPDL1 overexpression activated the JAK2/STAT3 signaling axis and amplified EMT phenotypes, thereby increasing the invasive and metastatic capacities of breast cancer cells. Moreover, ICA treatment attenuated these effects, consisitent with suppression of SPDL1-driven pathway activation. Further, SPDL1 silencing augmented the ICA-mediated inhibition of this signaling axis, supporting their mechanistic interplay.

Materials and methods

Cell lines and culture

The MDA-MB-231 and MDA-MB-468 human triple-negative breast cancer cell lines were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). The cells were cultured in DMEM (Gibco, USA) supplemented with 10% fetal bovine serum (FBS; BI, USA), 1% penicillin‒streptomycin solution (100 × ; Beyotime, China), and 1% nonessential amino acids, and the cells were maintained at 37 °C in a humidified 5% CO₂ incubator. Icariin (ICA, purity > 99%) was obtained from Selleck and dissolved in 100% dimethyl sulfoxide (DMSO) to prepare a stock solution, which was stored at -20 °C until use.

Cell viability assay

MDA-MB-231 and MDA-MB-468 cells in logarithmic growth phase were digested to prepare single-cell suspensions. The cells were seeded uniformly into 96-well plates at a density of 1 × 10⁵ cells/well (100 μL/well) and incubated for 24 h to ensure adhesion. The cells were subsequently treated with graded concentrations of ICA (0, 15, 30, 60, or 120 μmol/L) and incubated for 24, 48, or 72 h under standard culture conditions (37 °C, 5% CO₂). Following treatment, 10 μL of CCK-8 solution (Beyotime, China) was added to each well, followed by additional incubation at 37 °C for 2 h. The absorbance was measured at 450 nm using a microplate reader.

Colony formation assay

MDA-MB-231 and MDA-MB-468 cells were seeded into 6-well plates (1,000 cells/well) and allowed to adhere for 24 h. The cells were then treated with ICA for 48 h, followed by continuous culture for an additional 14 days. After incubation, the colonies were fixed with 4% paraformaldehyde for 15 min and stained with 0.1% crystal violet for 15 min. Images were acquired via a digital camera, and the number of colonies was quantified with ImageJ software. The experiments were independently performed in triplicate.

Cell apoptosis assay

Cell apoptosis was assessed using an Annexin V-FITC Apoptosis Detection Kit (Beyotime, China). Following ICA treatment, the cells were stained with Annexin V-FITC and propidium iodide (PI) and analyzed by flow cytometry. The experiments were independently repeated three times.

Wound-healing assay

The cells from culture flasks were digested and seeded into 6-well plates at a density of 1 × 10⁶ cells/well, followed by incubation at 37 °C until they reached 90% confluence. Under sterile conditions in a laminar flow hood, a straight wound was created in each well using a 200 μl pipette tip along the central axis to ensure uniformity. The wells were gently washed twice with phosphate-buffered saline (PBS) to remove detached cells. The cells were then cultured in ICA-containing medium for 48 h in a humidified incubator at 37 °C and 5% CO₂. Wound closure was monitored at 0, 24, and 48 h using a phase-contrast microscope (Ti-E; Nikon, Japan), and migration distances were quantified with ImageJ software. Three independent experiments were performed with triplicate technical replicates.

Cell migration and invasion assays

To evaluate the inhibitory effects of ICA on TNBC cell migration and invasion, Transwell chambers (Corning, NY, USA) were placed in 24-well plates to establish upper/lower compartment systems. For the migration assays, MDA-MB-231 and MDA-MB-468 cells pretreated with ICA for 24 h were digested, centrifuged, and resuspended in PBS to minimize residual medium interference. The lower chambers were filled with 700 μL of medium containing 20% FBS (BI, USA), while 300 μL of cell suspension (2 × 10⁵ cells/chamber) was added to the upper chambers. After 48 h of incubation (37 °C, 5% CO₂), the chambers were gently washed twice with PBS, fixed with 4% paraformaldehyde, and stained with 0.1% crystal violet. For the invasion assays, cells were seeded in Matrigel-coated upper chambers, and the same procedures used for the migration assays were then performed. The migrated/invaded cells were imaged under an inverted microscope (DMi8; Leica, Germany) and quantified using ImageJ software. Three independent experiments with triplicate technical replicates were performed.

Quantitative RT‒PCR

MDA-MB-231 and MDA-MB-468 cells from the experimental groups were harvested. Total RNA was extracted using the RNAiso Plus reagent, and the RNAconcentration/purity was determined spectrophotometrically. Reverse transcriptionwas performed with the PrimeScript™ RT Reagent Kit (Takara) to synthesize cDNA. Quantitative real-time PCR (qRT‒PCR) was conducted using TB Green Premix Ex Taq II (Takara) on a QuantStudio 6 Flex system. The following primer sequences were used: r18S-F, 5′-GTAACCCGTTGAACCCCATT-3′; r18S-R,5′-CCATCCAATCGGTAGTAGCG-3′;h-SPDL-1-F, 5′-CGCCTCAGAAAGAGGAGACAC-3′; and h-SPDL1-R, 5′-GCTTTGATTCAGCAGCTAACC-3′.

Western blot (WB) analysis

After the cells were treated with the indicated concentrations of ICA, the cellular proteins were extracted with a protein extraction kit (Beyotime, China), and the protein concentrations were determined with a concentration determination kit (Beyotime, China), followed by denaturation in a metal bath. A total of 30–40 mg of mouse lung tissue was then immersed in RIPA buffer kit and minced with a tissue homogenizer, and the protein concentration was determined. After loading, the protein samples were separated by electrophoresis with a 10% SDS‒PAGE gel and transferred to PVDF membranes, which were then blocked with 5% BSA. Various primary antibodies were then added to the membranes and incubated overnight at 4 °C. The next day, the corresponding secondary antibodies were added and incubated for 1 h at room temperature. After washing, the membranes were developed in an apparatus (Bio-Rad, USA) using highly sensitive enhanced chemiluminescence reagents (Beyotime, China). The antibodies against Bcl2 (3498), caspase-3 (9662), cleaved caspase-3 (9664), JAK2 (3230), STAT3 (9139), phospho-STAT3 (9145), and GAPDH (2118), as well as the horseradish peroxidase (HRP)-labeled secondary antibody (7076), were purchased from Cell Signaling Technology (Massachusetts, USA). The antibodies against N-cadherin (ab18203) and SPDL1 (ab99344) were acquired from Abcam (Cambridge, USA). The antibodies against Bax (WL01637), vimentin (WL01960) and MMP9 (WL03096) were acquired from WanLeibio (China).

RNA sequencing

Total RNA was extracted from tissue using TRIzol, and RNA quantity and integrity were confirmed before library construction. mRNA was enriched by poly(A) selection, fragmented, converted to double-stranded cDNA, end-repaired, adapter-ligated, size-selected, PCR-amplified, and the libraries were sequenced on a NovaSeq X Plus platform to generate paired-end reads. Raw reads were trimmed and filtered with fastp, then high-quality reads were aligned to the reference genome using HISAT2 and assembled with StringTie. Gene expression levels were quantified by RSEM, differentially expressed genes were identified with DESeq2 or DEGseq, and their biological functions and pathways were explored by GO and KEGG enrichment analyses using Goatools and Python scipy.

Cell infection assay

After trypsinization, resuspension, and cell counting, MDA-MB-231 and MDA-MB-468 cells were seeded in 24-well plates at a density of 3 × 10⁴ cells per well and incubated in a constant-temperature chamber to allow natural adherence. Viral transfection was subsequently performed using the LV-OE-SPDL1 and LV-sh-SPDL1 lentiviral vectors (Genechem, China) according to the manufacturer’s protocol. The HitransG viral infection enhancer (Genechem, China) was added to optimize transduction efficiency.

Immunofluorescence assay

The cells were cultured on glass coverslips, and the ICA-treated group was exposed to ICA for 48 h. Following three washes with PBS, the cells were fixed with 4% paraformaldehyde and blocked with 10% goat serum for 30 min. Subsequent incubations with species-specific primary and secondary antibodies were performed, followed by nuclear counterstaining with DAPI. Fluorescence images were captured via a microscope equipped with appropriate optical configurations.

Molecular docking

To evaluate the binding energy and molecular interactions between ICA and SPDL1, structure-based molecular docking was performed using AutoDock Vina 1.5.6 according to previously reported docking protocols [15]. The 3D structure of ICA (PubChem CID: 5,318,997) was obtained from the PubChem Compound Database, and the crystal structure of human SPDL1 (PDB ID: 8ARF) was downloaded from the Protein Data Bank. Before docking, all water molecules and non-essential heteroatoms were removed from the receptor, and polar hydrogen atoms were added. A cubic docking grid was defined to cover the putative ligand-binding pocket of SPDL1 and its surrounding residues. The receptor was kept rigid, whereas all rotatable bonds in ICA were set as flexible. Docking was carried out with an exhaustiveness of 8, and other parameters were kept at their default values in AutoDock Vina. The top-ranked binding poses, based on the predicted Gibbs free energy of binding (ΔG, kcal/mol), were analyzed and visualized using PyMOL 3.1 and Discovery Studio 2019 (DS2019).

In vivo lung metastasis model

All animal procedures were conducted in accordance with the ‘Guidelines for Animal Welfare in China’ and approved by the Institutional Animal Care and Use Committee of Shanghai University of Traditional Chinese Medicine, Putuo Hospital (DWEC-A-2024–18-1–88). Female BALB/c nude mice (4 weeks old) were intravenously injected via the tail vein with 200 μL of MDA-MB-231 or MDA-MB-231/SPDL1 cell suspensions (1 × 10⁷ cells/mL) to establish a breast cancer pulmonary metastasis model. The animals were randomized into the following three groups: CTRL, OE-SPDL1, and OE-SPDL1 + ICA (oral dose: 200 mg/kg/day). Drug administration via oral gavage commenced one week postinoculation, with weekly body weight monitoring. All mice were humanely sacrificed after two weeks of treatment. Lungs were harvested for gross assessment of pulmonary weight and metastatic nodule enumeration, followed by fixation in 10% neutral-buffered formalin for subsequent histopathological analyses.

Hematoxylin and eosin staining

Cardiac, hepatic, and splenic tissues were fixed in 4% paraformaldehyde for ≥ 24 h and processed through dehydration, paraffin embedding, and sectioning to generate 5-μm-thick slices. Subsequent procedures included dewaxing, hematoxylin staining, gradient alcohol dehydration, and xylene-based clearing. The sections were mounted with neutral balsam and microscopically examined for histoarchitectural analysis.

Immunohistochemistry

Immunohistochemical (IHC) analysis was performed through the following standardized procedures. First, lung tissue sections were deparaffinized with xylene and rehydrated through a graded ethanol series. Antigen retrieval was achieved by heating the sections in boiling sodium citrate buffer (10 mM, pH 6.0) for 15 min. Subsequently, endogenous peroxidase activity was blocked by incubation with 3% hydrogen peroxide solution for 10 min at room temperature. After three 5-min washes with phosphate-buffered saline (PBS), the sections were incubated with 5% normal goat serum (Vector Laboratories) for 30 min to block nonspecific binding. The sections were then incubated with primary antibodies diluted in antibody dilution buffer (1:200) overnight at 4 °C in a humidified chamber. Following three additional PBS washes, the sections were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (1:500 dilution) for 30 min at room temperature. After thorough washing, immunoreactivity was visualized using 3,3’-diaminobenzidine (DAB) substrate (Dako) using a microscope. Finally, the nuclei were counterstained with Mayer’s hematoxylin for 45 s, followed by dehydration through a graded alcohol series and xylene clearing. Digital images were acquired using a Leica DM4000B microscope equipped with a CCD camera under consistent illumination conditions.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 8.0. All the quantitative data are expressed as the means ± standard deviations (SDs) from at least three independent biological replicates. Intergroup comparisons were conducted with unpaired Student’s t tests for two-group analyses, while one-way analysis of variance (ANOVA) with Tukey’s post hoc test was applied for multigroup comparisons. A threshold of P < 0.05 was predefined to establish statistical significance across all experiments.

Results

ICA suppresses TNBC cell proliferation and induces apoptosis

To investigate the cytotoxic and inhibitory effects of icariin (ICA) on triple-negative breast cancer (TNBC) cells, MDA-MB-231 and MDA-MB-468 cells were exposed to increasing concentrations of ICA (0–120 μmol/L) for 24–72 h. The CCK-8 assay results demonstrated dose- and time-dependent inhibition of TNBC cell proliferation by ICA, with IC50 values of 24.38 μM for MDA-MB-231 cells and 24.58 μM for MDA-MB-468 cells after 48 h of treatment (Fig. 1A-B). ICA at a concentration of 15 μM was selected for subsequent experiments on the basis of preliminary optimization. Colony formation assays revealed significant suppression of TNBC cell proliferation in the ICA-treated groups compared with the vehicle control groups (Fig. 1C-D). Flow cytometric analysis using Annexin V-FITC/PI staining revealed that 15 μM ICA treatment for 48 h significantly increased the percentage of apoptotic MDA-MB-231 and MDA-MB-468 cells (Fig. 1E-F). Compared with no treatment, ICA treatment upregulated the expression of proapoptotic proteins (Bax and cleaved caspase-3) but downregulated the expression of antiapoptotic proteins (Bcl-2 and caspase-3) (Fig. 1G-H). These results indicated that ICA effectively inhibits TNBC cell proliferation and induces apoptosis through the modulation of Bcl-2 family proteins and caspase activation.

Fig. 1.

ICA suppresses TNBC cell proliferation and induces apoptosis. (A-B) CCK-8 assay demonstrating the dose-dependent inhibition of TNBC cell proliferation by ICA (15–120 μmol/L) over 48 hours. (C-D) Colony formation assays showing the reduced clonogenic capacity of TNBC cells after treatment with 15 μM ICA. (E-F) Flow cytometric quantification of apoptotic cells using Annexin V/PI double-staining following treatment with 15 μM ICA. The data represent three independent experiments. (G-H) Western blot analysis of the levels of apoptosis-related proteins (Bax, Bcl-2, caspase-3, and cleaved caspase-3) in TNBC cells treated with 15 μM ICA. Statistical significance versus the control group: *P < 0.05, **P < 0.01, and ***P < 0.001

ICA inhibits TNBC cell invasion and metastasis

To investigate the regulatory effects of ICA on triple-negative breast cancer (TNBC), RNA sequencing was performed on MDA-MB-231 cells treated with 15 μM ICA for 48 h. The results from the GO enrichment analysis suggested that ICA may influence the regulation of TNBC cell migration (Fig. 2A). Compared with no treatment, the wound-healing assays revealed that ICA treatment (15 μM) significantly suppressed the migratory capacity of TNBC cells at both 24 and 48 h (Fig. 2B). To further investigate invasive behavior, Transwell migration and Matrigel-based invasion assays were conducted, which revealed marked reductions in both the migratory and invasive activities of TNBC cells following 24 and 48 h of ICA treatment (Fig. 2C-D). Immunofluorescence staining confirmed decreased vimentin expression in ICA-treated cells (Fig. 2E). Subsequently, Western blot analysis of epithelial–mesenchymal transition (EMT)-related markers revealed that ICA treatment significantly downregulated N-cadherin, MMP9, and vimentin protein expression compared with that in the control groups (Fig. 2F). These findings indicated that ICA effectively inhibits TNBC cell invasion and metastasis by modulating key molecular mediators of EMT.

Fig. 2.

ICA inhibits TNBC cell invasion and metastasis. (A)GO enrichment analysis (B) Wound-healing assay showing the migratory capacity of TNBC cells treated with 15 μM ICA at 24 and 48 hours. (C-D) Transwell migration and Matrigel-based invasion assays demonstrating the effect of ICA treatment on the migratory and invasive activities of TNBC cells after 24 and 48 hours. (E) Immunofluorescence staining and quantitative analysis of vimentin expression in TNBC cells after ICA treatment. (F) Western blot analysis of EMT-related proteins (N-cadherin, vimentin, and MMP9) in TNBC cells exposed to 15 μM ICA . The data represent three independent experiments. Statistical significance versus the control group: *P < 0.05, **P < 0.01, ***P < 0.001

ICA downregulates SPDL1 expression

To investigate the molecular mechanism underlying the ICA-mediated suppression of triple-negative breast cancer (TNBC), RNA sequencing was performed. Comparative analysis revealed significant downregulation of SPDL1 (spindle apparatus coiled-coil protein 1), a mitotic regulator and novel prognostic marker in TNBC, in ICA-treated cells compared with control cells (Fig. 3A-B). Validation experiments further confirmed consistent downregulation of SPDL1 at both the transcriptional and protein levels (Fig. 3C). To further elucidate the interaction between ICA and SPDL1 at the structural level, we performed molecular docking using AutoDock Vina [16]. Inspection of the top-ranked docking poses showed that ICA consistently occupied a cavity on SPDL1 in which the ligand established specific interactions with several residues in the binding pocket. In the lowest-energy pose, ICA formed two hydrogen bonds with ASN35 and GLN29, indicating that these residues may contribute to anchoring ICA within the pocket. The calculated binding energy for this pose was − 6.6 kcal/mol, which is compatible with a moderate but specific protein–ligand interaction in the context of virtual screening. In addition to hydrogen bonding, the three-dimensional docking model indicated that ICA engaged in π-related hydrophobic contacts with LEU30 and VAL31, and was further stabilized by surrounding van der Waals contacts from neighboring residues within the pocket. Examination of several of the next best-scoring poses revealed similar orientations of ICA and recurrent interactions with ASN35 and GLN29, supporting the robustness of this binding mode rather than a single isolated solution. Taken together, these docking results suggest that ICA can form a reasonably stable complex with SPDL1 via a combination of hydrogen bonding and hydrophobic interactions (Fig. 3D-F). Furthermore, Western blot analysis demonstrated a significant reduction in SPDL1 protein expression in ICA-treated cells compared to untreated control cells (Fig. 3G-H). These results suggest that the downregulation of SPDL1 contributes to the antitumor effects of ICA in triple-negative breast cancer.

Fig. 3.

ICA suppresses SPDL1 expression in TNBC cells. (A-B) RNA sequencing (RNA-seq) analysis showing differential gene expression profiles (heatmap) and significantly downregulated SPDL1 (volcano plot) in TNBC cells treated with 15 μM ICA for 48 hours. (C) RNA-seq quantification confirming reduced SPDL1 mRNA levels in ICA-treated cells compared with those in control cells. (D) Chemical structure of ICA. (E) Molecular docked pose of ICA in the ligand binding domain of SPDL1 showing the interacting amino acid residues. Gibb’s free energy of binding (ΔG°) = −6.6 kcal/mol. (F) Protein surface and ligand-binding groove of the SPDL1 bound with ICA. (G-H) Western blot analysis with quantitative validation demonstrating decreased SPDL1 protein expression following ICA treatment (15 μM, 48 hours). The data represent three independent experiments. Statistical significance versus the control: **P < 0.01 and ***P < 0.001

ICA suppresses TNBC cell proliferation and metastasis through SPDL1 downregulation

To investigate the functional role of SPDL1 in ICA-mediated antitumor effects, stable overexpressing-SPDL1 (OE-SPDL1) and SPDL1 knockdown (shSPDL1) MDA-MB-231 and MDA-MB-468 cell lines were established via lentiviral transduction followed by puromycin selection. Successful overexpression and knockdown of SPDL1 was confirmed by qRT‒PCR (Fig. 4A-B) and Western blot (Fig. 4C-D) analyses. Functional analyses were performed among the CTRL, OE-SPDL1, sh-SPDL1, OE-SPDL1 + ICA, and sh-SPDL1 + ICA groups. Colony formation assays revealed increased proliferative capacity in OE-SPDL1 cells, which was significantly attenuated by ICA cotreatment. In contrast, sh-SPDL1 cells showed a marked reduction in proliferative ability compared to the control, and this effect was further diminished upon ICA treatment (Fig. 4E-F). Transwell migration and Matrigel-based invasion assays demonstrated that SPDL1 overexpression increased metastatic potential, whereas ICA administration effectively attenuated this prometastatic phenotype. In contrast, sh-SPDL1 cells exhibited significantly reduced migratory and invasive capabilities compared to control cells, and ICA treatment did further alter these properties (Fig. 4G-H). Mechanistically, Western blot analysis revealed that OE-SPDL1 upregulated epithelial‒mesenchymal transition (EMT)-related proteins (N-cadherin, vimentin, and MMP9), whereas ICA treatment downregulated their expression to baseline levels. In sh-SPDL1 cells, the expression of these EMT markers was already significantly reduced compared to control cells, and ICA treatment did further affect their expression (Fig. 4I-J). Immunofluorescence staining further confirmed that ICA suppressed SPDL1-induced vimentin expression (Fig. 4K-L). These results suggest that SPDL1 may contribute to TNBC progression through EMT activation, and that ICA potentially exerts its therapeutic effects by modulating SPDL1-mediated oncogenic signaling.

Fig. 4.

ICA suppresses TNBC cell proliferation and metastasis via SPDL1 downregulation. (A-B) Validation of SPDL1 overexpression efficiency by qRT‒PCR (C-D) Western blot analyses in TNBC cells transduced with overexpressing-SPDL1 (OE-SPDL1) and knockdown-SPDL1 lentivirus (sh-SPDL1). (E-F) Colony formation assay comparing the proliferative capacity among the various experimental groups. (G-H) Transwell migration assay (upper) and Matrigel-based invasion assay (lower) evaluating metastatic potential across the various experimental groups. (I-J) Western blot analysis of EMT-related proteins (N-cadherin, vimentin, and MMP9) in the various experimental groups. (K-L) Immunofluorescence staining and quantitative analysis of vimentin expression in the various experimental groups. The data represent three independent experiments. Statistical significance: *P < 0.05, **P < 0.01, ***P < 0.001

ICA inhibits the JAK2/STAT3 signaling pathway through SPDL1 downregulation

Transcriptomic profiling via RNA sequencing (RNA-seq) revealed some enrichment of the JAK-STAT signaling pathway in ICA-treated TNBC cells (Fig. 5A). Given the known involvement of JAK-STAT signaling in tumor proliferation and metastasis, key components of this pathway were further assessed by Western blot analysis [17]. ICA treatment reduced JAK2 and phosphorylated STAT3 (p-STAT3) protein expression, while total STAT3 protein levels remained unchanged (Fig. 5B). To explore the potential role of SPDL1 in regulating this pathway, OE-SPDL1 cells were analyzed, which revealed upregulated JAK2 and p-STAT3 protein expression compared with that in control cells, without affecting total STAT3 expression (Fig. 5C). Since IL-6 is known to activate JAK2/STAT3 signaling [18], we also investigated the effects of IL-6 stimulation combined with SPDL1 knockdown (shSPDL1) and ICA treatment. The experimental groups included control, IL-6 (20 ng/mL, 24 h), IL-6 + shSPDL1, and IL-6 + shSPDL1 + ICA (15 μM) treatments. Western blot analysis showed that IL-6 stimulation increased the levels of JAK2 and p-STAT3, but this effect was attenuated by SPDL1 knockdown. Furthermore, ICA treatment appeared to further reduce the expression of SPDL1, JAK2, and p-STAT3 proteins (Fig. 5D). These findings suggest that SPDL1 may play a role in enhancing JAK2/STAT3 activation in TNBC, and that ICA disrupt this pathway, potentially through SPDL1 downregulation.

Fig. 5.

ICA suppresses the JAK2/STAT3 signaling pathway via SPDL1 downregulation. (A) KEGG pathway enrichment analysis of differentially expressed genes in ICA-treated TNBC cells. (B) Western blot analysis of JAK2, STAT3, and phosphorylated STAT3 (p-STAT3) protein levels in TNBC cells treated with 15 μM ICA. (C) Western blot validation of SPDL1-mediated JAK2/STAT3 pathway regulation in the various experimental groups. (D) Western blot analysis of SPDL1, JAK2, STAT3, and p-STAT3 protein expression in TNBC cells under the following four experimental conditions: CTRL, IL-6 stimulation (20 ng/mL, 24 hours), IL-6 + shSPDL1, and IL-6 + shSPDL1 + ICA (15 μM). The data represent three independent experiments. Statistical significance: *P < 0.05, **P < 0.01, ***P < 0.001

ICA exerts antitumor effects in vivo

To evaluate the therapeutic consistency of ICA between in vitro and in vivo models, a pulmonary metastasis model was established by tail vein injection of MDA-MB-231 cells and SPDL1-overexpressing MDA-MB-231 cells into BALB/c nude mice. The animals were divided into the following three groups: CTRL, OE-SPDL1, and OE-SPDL1 + ICA (200 mg/kg, daily oral gavage for 14 days). Gross examination revealed an increased pulmonary metastatic burden in the SPDL1-OE group compared with the control group, which was significantly attenuated by ICA cotreatment (Fig. 6A). Compared with those of control mice, quantitative analyses revealed increased lung weights and metastatic tumor counts in SPDL1-OE mice. ICA intervention reduced these parameters compared with those of the SPDL1-OE group (Fig. 6B-C). No significant differences in body weight or splenic weight were observed among the groups (Fig. 6D-E). Histopathological assessment via H&E staining confirmed the absence of ICA-induced cardiotoxicity, hepatotoxicity, splenic toxicity or nephrotoxicity (Fig. 6F). These findings demonstrated that ICA effectively inhibits TNBC metastasis in vivo through SPDL1 downregulation while maintaining a favorable safety profile.

Fig. 6.

ICA inhibits TNBC progression in vivo with favorable safety. (A) Lung Metastasis Assessment in Treated Animals: Results from the CTRL, OE-SPDL1, and OE-SPDL1+ICA (200 mg/kg) treatment groups. (B-D) Quantitative analysis of the number of metastatic tumors, lung weight and spleen weight across the experimental groups. (E) Body weight changes in tumor-bearing mice during the 14 day treatment period. (F) Histopathological evaluation by H&E staining showing no toxicity in heart, spleen, liver and Kidney tissues from ICA-treated mice. The data represent three independent experiments. Statistical significance versus the control: **P < 0.01, ***P < 0.001

ICA suppresses SPDL1-mediated JAK2/STAT3 signaling and metastasis in vivo

Protein lysates from the lung tissues of the three experimental groups (CTRL, OE-SPDL1, and OE-SPDL1 + ICA) were evaluated by Western blot analyses. Compared with controls, SPDL1-overexpressing lung metastases presented increased expression of SPDL1, JAK2, phosphorylated STAT3 (p-STAT3), and metastasis-associated proteins (N-cadherin, vimentin, and MMP9). ICA cotreatment attenuated these SPDL1-associated changes without altering total STAT3 levels (Fig. 7A–7D). Immunohistochemical (IHC) analysis further confirmed elevated SPDL1, JAK2, and p-STAT3 protein expression in OE-SPDL1 metastases, which was attenuated by ICA intervention (Fig. 7E-7F). These results demonstrated that ICA inhibits TNBC metastasis in vivo by targeting the SPDL1/JAK2/STAT3 axis and suppressing EMT-related protein expression, suggesting that its antitumor activity is mechanistically linked to JAK2/STAT3 pathway inhibition.

Fig. 7.

ICA suppresses SPDL1-mediated metastasis and JAK2/STAT3 signaling in vivo. (A-B) Western blot analysis with quantitative validation of SPDL1, JAK2, STAT3, and phosphorylated STAT3 (p-STAT3) protein expression in lung metastases from the CTRL, OE-SPDL1, and OE-SPDL1+ICA groups. (C-D) Western blot quantification of metastasis-associated proteins (N-cadherin, MMP9, and vimentin) in lung tissues. (E-F) Immunohistochemical (IHC) analysis of SPDL1, JAK2, and p-STAT3 expression in pulmonary metastatic nodules (200× magnification; brown cytoplasmic staining indicates positive expression). The data represent three independent experiments. Statistical significance versus the control: *P < 0.05, **P < 0.01, and ***P < 0.001

Discussion

Triple-negative breast cancer (TNBC), which is characterized by the absence of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) expression, represents a clinically aggressive breast cancer subtype with limited therapeutic options [19]. The lack of actionable molecular targets precludes TNBC patients from benefiting from endocrine therapies or HER2-targeted treatments, resulting in poor clinical outcomes [20]. Epidemiological studies have indicated that approximately 15% of breast cancer patients present with regional lymph node or distant organ metastasis at initial diagnosis, with pulmonary metastasis occurring in up to 30% of advanced cases [21, 22]. TNBC exhibits a particularly high metastatic propensity, attributed to its enrichment of cancer stem cells and activation of prometastatic signaling pathways, such as the Wnt/β-catenin, NF-κB or AMPK/mTOR pathways [23–25]. The epithelial‒mesenchymal transition (EMT), a critical biological process conferring migratory and invasive capabilities to cancer cells, is characterized by downregulation of E-cadherin and concomitant upregulation of mesenchymal markers (N-cadherin and vimentin) and matrix-degrading enzymes (MMP2/MMP9). Emerging evidence suggests that EMT modulation is a promising therapeutic strategy to suppress TNBC metastasis and recurrence [26].

ICA, a bioactive flavonoid compound isolated from the traditional Chinese herb Epimedium brevicornum, has demonstrated broad-spectrum antitumor activity against various solid malignancies, including colorectal and hepatocellular carcinomas [27–29]. Building upon these findings, the present study systematically investigated the therapeutic potential of ICA in triple-negative breast cancer (TNBC) via the use of MDA-MB-231 and MDA-MB-468 cell models. In vitro experiments revealed that ICA treatment significantly suppressed TNBC cell proliferation and induced apoptosis. Furthermore, ICA administration attenuated metastatic potential, as evidenced by reduced migratory/invasive capacities and concomitant downregulation of EMT-associated proteins (N-cadherin, MMP9, and vimentin). These findings suggested that ICA may exert its anti-TNBC effects through targeted modulation of EMT progression, potentially via unknown molecular mediators.

SPDL1 (spindle apparatus coiled-coil protein 1), a critical regulator of mitotic progression, is closely associated with genomic instability and malignant tumor phenotypes. Previous studies have revealed significant overexpression of SPDL1 at both the mRNA and protein levels in TNBC tissues, with elevated SPDL1 expression correlating with reduced overall survival in TNBC patients, suggesting its potential as an independent prognostic biomarker [14]. Despite these findings, therapeutic strategies targeting SPDL1 remain largely unexplored. Motivated by this therapeutic gap, we hypothesized that ICA might inhibit TNBC progression through SPDL1 downregulation. To test this hypothesis, RNA sequencing and molecular docking analyses were performed in MDA-MB-231 cells with or without ICA treatment. RNA-seq revealed significant downregulation of SPDL1 in ICA-treated cells, while docking simulations demonstrated a moderate binding affinity between ICA and SPDL1. Functional validation using overexpressing-SPDL1 (OE-SPDL1) and SPDL1-knockdown (sh-SPDL1) TNBC cell models demonstrated that SPDL1 overexpression was associated with an increased metastatic phenotype, while SPDL1 knockdown reduced proliferative and metastatic capacities. ICA cotreatment attenuated the prometastatic effects associated with SPDL1 overexpression. notably, the inhibitory phenotype observed in sh-SPDL1 cells was further enhanced following ICA treatment. Mechanistically, SPDL1 overexpression appeared to activate prometastatic signaling cascades, which were mitigated by ICA treatment. Taken together, these results indicate that SPDL1 contributes to TNBC aggressiveness and that ICA may exert antitumor activity at least in part via modulation of SPDL1-mediated signaling.

To delineate the molecular basis of SPDL1-mediated TNBC metastasis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis was performed, which revealed significant enrichment of JAK-STAT signaling in ICA-treated MDA-MB-231 cells. This finding aligned with established evidence demonstrating constitutive JAK2/STAT3 pathway activation in TNBC, where its pharmacological inhibition suppresses cancer cell proliferation and migration [30]. Notably, traditional herbal agents, such as Magnolol, have been shown to induce TNBC apoptosis through blockade of JAK2/STAT3 phosphorylation [31]. Western blot validation confirmed that SPDL1 overexpression increased JAK2 and STAT3 phosphorylation (p-STAT3) levels, whereas ICA treatment effectively suppressed this increase. Rescue experiments employing IL-6 stimulation (20 ng/mL, 24 h), a canonical JAK2/STAT3 pathway agonist, demonstrated that SPDL1 knockdown (shSPDL1) attenuated IL-6-induced pathway activation, with ICA cotreatment producing synergistic inhibition. These findings suggest a potential regulatory mechanism whereby ICA inhibit TNBC metastatic behavior, possibly mediated through SPDL1-dependent modulation of JAK2/STAT3 signaling.

To substantiate the mechanism of action of ICA and its antitumor efficacy in vivo, pulmonary metastasis models were established via tail vein injection of MDA-MB-231 cells (CTRL) and overexpressing-SPDL1 (OE-SPDL1) MDA-MB-231 cells in immunocompromised mice. Systemic administration of ICA (200 mg/kg, daily oral gavage for 14 days) significantly attenuated SPDL1-driven metastatic progression. Histopathological assessment via H&E staining confirmed the absence of ICA-induced cardiotoxicity, hepatotoxicity, splenic toxicity or nephrotoxicity across all treatment groups, underscoring its favorable safety profile. These findings demonstrated that ICA exerts potent anti-TNBC effects through dual mechanisms. The ability of the compound to simultaneously disrupt oncogenic signaling while maintaining systemic tolerability positions ICA as a promising phytotherapeutic candidate for clinical translation in TNBC management.

While the present study sheds light on the anti-TNBC activity of ICA via SPDL1-JAK2/STAT3 axis modulation, several limitations should be noted. First, the use of only MDA-MB-231 and MDA-MB-468 cell lines limits the generalizability of the findings across different TNBC molecular subtypes. Additionally, the use of a single ICA concentration (15 μM) in vitro restricts the analysis of dose‒response relationships and temporal dynamics. Although the regulation of the JAK2/STAT3 pathway by SPDL1 was identified, potential interactions with epigenetic mechanisms (e.g., histone acetylation or DNA methylation) or other oncogenic pathways, remain unexplored [24, 32–35]. Moreover, while molecular docking studies were conducted to explore the binding interactions of ICA with target proteins, the results were not followed by biochemical validation, limiting their interpretability. Furthermore, recent evidence highlighting the role of tumor microenvironment (TME) components—especially immune cells like M2 macrophages and regulatory T cells—in TNBC metastasis indicates a gap in this study, which did not evaluate the potential immunomodulatory effects of ICA, as seen with traditional agents like Ruyiping, which target macrophage polarization [36–38]. Additionally, the pharmacokinetics and pharmacodynamics of ICA were not investigated, meaning its safety and efficacy as a therapeutic agent are yet to be fully assessed. Future research will address these gaps by integrating multi-omics approaches, including DNA methylation arrays and chromatin immunoprecipitation (ChIP), to explore the epigenetic regulatory roles of SPDL1, alongside phosphoproteomic profiling to map ICA-dependent signaling networks. Moreover, immune-competent humanized mouse models will be used to assess the dual impact of ICA on immune remodeling (e.g., macrophage polarization and checkpoint molecule expression) and SPDL1-JAK2/STAT3-EMT axis coordination. In addition, the biochemical validation of molecular docking results will be incorporated, using techniques such as surface plasmon resonance and molecular dynamics simulations, to confirm the binding interactions and the stability of ICA-target protein complexes. By incorporating a broader representation of TNBC subtypes through patient-derived organoids and accounting for microenvironmental complexity, these efforts aim to enhance the translational potential of ICA as a multitarget phytotherapeutic against the heterogeneous landscape of TNBC.

Author contributions

Wang and Li conceived and designed the experiments; Xiao, Yang, Li performed the experiments and drafted the manuscript, and revised it critically for important intellectual content. Chen, Li, He performed part of the experiments and analyzed the data. All authors read and approved the final manuscript.

Funding

This work was supported by National Natural Science Foundation of China(81973625), Key Medical Discipline Project of Shanghai Municipal Health Bureau(2024ZDXK0046), “Xinglin Scholar” Discipline Talent Program of Chengdu University of Traditional Chinese Medicine(YYZX2022164) and the Shen Hongquan’s Academic Experience Research Studio of Shanghai Famous Veteran TCM Expert(SHGZS-202224).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

All animal experiments were conducted in accordance with guidelines and protocol approved by the institutional animal care and use committee of Putuo Hospital, Shanghai University of Traditional Chinese Medicine, China.(DWEC-A-2024–18-1–88).

Competing interests

The authors declare no competing interests.