Ferulic acid (C₁₀H₁₀O₄; MW: 194.18; shown in Fig. 1A) was purchased from Sigma Chemical Co. (Sigma, St. Louis, MO, USA). Human breast cancer cell line MDA-MB-231 was obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA), and cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. All the cells were incubated at 37°C in a humidified atmosphere containing 5% CO₂.

Cells seeded onto a 96-well plate at a density of 4×10³ cells/well were treated with 100 µl medium plus dimethyl sulfoxide (DMSO; vehicle control, final concentration <0.1%) or ferulic acid (final concentration: 3, 10, 30 and 100 µM), and incubated for 24, 48 and 72 h, respectively. At the end of the drug exposure duration, cell viability was assessed using the Cell Counting Kit-8 (CCK-8) assay according to the manufacturer's protocols (Dojindo, Gaithersburg, MD, USA). Six parallel replicates were measured for each sample. Each plate included a control well that contained medium only, which was used to obtain a background spectrometric absorbance value that was subtracted from the test sample readings. The data are expressed as the mean ratio of live cells (treated vs. control) ± SD for 3 replicates.

For apoptosis analysis, human breast cancer cell line MDA-MB-231 was treated with ferulic acid (final concentration: 10, 30 and 100 µM) and DMSO (vehicle control, final concentration <0.1%) for 48 h, incubated with fluorescein isothiocyanate-Annexin V and propidium iodide (BD Biosciences) for 15 min, and then analyzed by flow cytometry.

Caspase-3 activity was measured using colorimetric assay kits (Beyotime Institute of Biotechnology, Jiangsu, China) according to the manufacturer's instructions. Briefly, after isolation by two-step collagenase digestion, the breast cancer cells were mixed with 50 µl cold lysis buffer. The supernatant (10 µl) was then mixed with 10 µl caspase substrate and 80 µl reaction buffer and incubated at 37°C for 2 h in the dark. The fluorescence intensity of the caspase substrate at 405 nm was measured using a microplate reader. The caspase activity was calculated as the OD405/100 µg protein.

For the wound healing assay, we drew horizontal lines across the back of the wells of 6-well plates with a marker pen. The cells (5×10⁵ cells/well) were plated into the 6-well plates. On the following day, the confluent cell monolayers were carefully wounded (perpendicular to the horizontal lines) with sterile pipette tips and washed with phosphate-buffered saline (PBS) twice to remove cellular debris. Serum-free medium was added into the wells. Then, the wounded cell monolayers were cultured for another 48 h treat with or without different concentrations of ferulic acid. Images were captured under a microscope to observe the distribution of the cells at the scratch zone at the last time point. The degree of wound closure was then quantitatively evaluated using Image-Pro Plus software. Four fields from each well were documented, and each experiment was performed in triplicate.

A cell migration assay was performed using Transwell chambers with a pore size of 0.8 µm. Breast cancer cells were trypsinized, washed and suspended in medium without FBS. To the lower wells of the chambers, migration-inducing medium (with 10% FBS) was added. A total of 2×10⁴ cells were seeded in serum-free medium in the upper chamber, containing various concentrations of ferulic acid (10, 30 and 100 µM) or vehicle solvent (DMSO). After incubating for 24 h at 37°C, the cells in the upper chamber were carefully removed with a cotton swab, and the cells that had migrated to the reverse face of the membrane were fixed in methanol, stained with Giemsa solution and counted.

For the immunofluorescence experiments, cells were prepared and analyzed under a fluorescence microscope following the procedures previously described (25). Briefly, cells treated with or without ferulic acid were incubated with primary antibody against vimentin (1:100; 5741) or E-cadherin (1:200; 3195) (both from Cell Signaling Technology), and then incubated with DyLight 594 or 488 (CW Biotech, Beijing, China) secondary antibody against rabbit IgG. Cells were then counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (Roche, Basel, Switzerland) and imaged with the fluorescence microscope.

The cell lysate was prepared using a protein extraction kit (Active Motif Company, Carlsbad, CA, USA) according to the manufacturer's instructions. Protein concentrations were determined by BCA protein assay kit (Thermo Fisher Scientific, Inc., Waltham, MA, USA), protein was loaded onto 10–12% Tris-acrylamide gels, electrophoresed and then transferred to nitrocellulose membranes. After blocking non-specific binding sites, the membranes were probed with primary antibodies, including anti-E-cadherin (1:1,000; 3195), anti-claudin-1 (1:1,000; 13255), anti-N-cadherin (1:1,000; 4061), anti-vimentin (1:1,000; 5741), anti-Snail (1:1,000; 3897) (all from Cell Signaling Technology), anti-Slug (1:1,000; ab27568; Abcam) and anti-β-actin (1:1,000; sc47778; Santa Cruz Biotechnology). Labeled proteins were visualized by an ECL chemiluminescence kit (Millipore, Billerica, MA, USA).

MDA-MB-231 cells were treated with various concentrations of ferulic acid (3, 10, 30 and 100 µM). At 48 h after treatment, the total RNA was extracted using TRIzol reagent (Invitrogen), respectively. Reverse transcription reactions were performed using 2 µg of total RNA and were subsequently processed using a reverse transcription kit (TransGen Biotech, Beijing, China). Specific products were amplified by qPCR using the following primers: E-cadherin, 5′-CGAGAGCTACACGTTCACGG-3′ (forward) and 5′-GGGTGTCGAGGGAAAAATAGG-3′ (reverse); claudin-1, 5′-CCTCCTGGGAGTGATAGCAAT-3′ (forward) and 5′-GGCAACTAAAATAGCCAGACCT-3′ (reverse); vimentin, 5′-GACGCCATCAACACCGAGTT-3′ (forward) and 5′-CTTTGTCGTTGGTTAGCTGGT-3′ (reverse); N-cadherin, 5′-AGCCAACCTTAACTGAGGAGT-3′ (forward) and 5′-GGCAAGTTGATTGGAGGGATG-3′ (reverse); Snail, 5′-TCGGAAGCCTAACTACAGCGA-3′ (forward) and 5′-AGATGAGCATTGGCAGCGAG-3′ (reverse); Slug, 5′-CGAACTGGACACACATACAGTG-3′ (forward) and 5′-CTGAGGATCTCTGGTTGTGGT-3′ (reverse). Verification of the expression levels of the genes was performed by qPCR using TransStart Green qPCR SuperMix kit (TransGen Biotech).

Mouse tumors, lungs and liver were fixed with paraformaldehyde, embedded in paraffin, and then sectioned. Some sections were incubated with hematoxylin and eosin (H&E), dehydrated and mounted. For immunohistochemical analysis, sections were boiled in retrieval solutions to expose the antigens, and incubated at 4°C overnight with anti-Ki67 (1:100; ab16667; Abcam) and anti-cleaved caspase-3 (1:300; 9661; Cell Signaling Technology). The section-affixed slides were counterstained with hematoxylin, dehydrated and mounted. The immunostaining results were independently evaluated by two pathologists.

Five-week-old female BALB/c nude mice were purchased from the Institute of Laboratory Animal Science (Chinese Academy of Medical Science, Beijing, China). All procedures for the animal experiments were conducted according to the Animal Ethics Committee of Chongqing Medical University. All the animals were randomly divided into two groups of 9 mice each and housed according to the national and institutional guidelines for humane animal care. At 6 weeks of age, the mice were perorally (p.o.) gavaged with either 100 µl ferulic acid (100 mg/kg weight) or normal saline (control) and the animals were gavaged daily during the experiment. At 7 weeks of age, the MDA-MB-231 cells were inoculated subcutaneously into the right flanks of the mice (1.5×10⁶ cells/mouse). Body weights were monitored weekly as an indicator of overall health. Tumor diameter was measured every week, and tumor volumes were calculated with the formula: tumor volume (mm³) = 0.5 × length (mm) × width² (mm²). The mice were sacrificed via CO₂ asphyxiation 6 weeks after the transplant. Tumors were then removed, weighed and sent for immunohistochemistry (IHC) analysis.

For the tumor metastatic assay, six-week-old female BALB/c nude mice were p.o. gavaged with 100 µl ferulic acid (100 mg/kg weight) or normal saline (control) for one week. Then, they were administered MDA-MB-231 cells (2.5×10⁶ cells) via the tail vein (6 mice in each group), and the treatment continued until the end of the experiment. At the end of the experiment, on day 28, the mice were sacrificed via CO₂ asphyxiation. The lungs and livers were removed, and sent for H&E staining.

Statistical analyses were performed using SPSS 13.0 software. Data from all the experiments are presented as means ± SD and represent 3 independent experiments. One-way analysis of variance (ANOVA) or Student's t-test was used to compare means between treatment groups. A P-value of <0.05 was considered to indicate a statistically significant result.

Breast cancer cell line MDA-MB-231 was treated with various concentrations of ferulic acid (3, 10, 30 and 100 µM). As shown in Fig. 1B, we observed that ferulic acid inhibited the cell proliferation of breast cancer cell line MDA-MB-231 in a dose-dependent manner. To further explore the effects of ferulic acid on cell apoptosis, MDA-MB-231 cells were used for the analysis of apoptosis by FACS. Following treatment with ferulic acid (10, 300 and 100 µM) for 48 h, significantly increased apoptosis was observed when breast cancer cells were treated with ferulic acid, which may partly contribute to the decreased cell viability (Fig. 2A and B). Moreover, the effect of ferulic acid on apoptosis was further confirmed by analyzing the activation of caspase-3 in the apoptosis pathway. In addition, the results showed that ferulic acid considerably enhanced the activity of caspase-3 in breast cancer cells (Fig. 2C). Briefly, these data suggest that ferulic acid suppresses breast cancer cell proliferation and induces apoptosis.

To investigate the role of ferulic acid in breast cancer metastasis, we investigated the migratory capacity of high metastatic cell line MDA-MB-231. We firstly carried out wound healing assays to evaluate the effects of ferulic acid on breast cancer cell migration at the concentrations of 10, 30 and 100 µM. As shown in Fig. 3A and B, the migration across the wound edges was markedly slower after treatment with different concentrations of ferulic acid. From the Transwell assay, we also found that ferulic acid dose-dependently inhibited breast cancer cell migration (Fig. 3C and D). These results indicate that ferulic acid inhibits breast cancer migration even at a low concentration (10 µM) and that ferulic acid indeed possesses the ability to inhibit tumor metastasis in vitro.

EMT is thought to be a driver of invasion and metastasis in different types of epithelial cancers (26). The most comprehensive theory describing how initially quiescent tumor cells acquire metastatic capability is EMT (27). A group of pleiotropic transcription factors (TFs) have been found capable of orchestrating EMT programs. Most EMT-TFs are transcriptional repressors and many, such as Snail, Slug, ZEB1 and Twist, directly repress mediators of epithelial adhesion, the most important of which is E-cadherin, an integral component of adherens junctions and claudins, which are necessary for the assembly of tight junctions between adjacent cells (28–33). To evaluate the effect of ferulic acid on EMT, we treated mesenchymal-like breast cancer cell line MDA-MB-231 with ferulic acid at different concentrations (10, 30 and 100 µM). We first detected the expression of E-cadherin and vimentin, a mesenchymal intermediary filament by immunofluorescence. As shown in Fig. 4A, the expression of vimentin was markedly decreased after treatment with ferulic acid. However, we observed a significant increase of E-cadherin. Moreover, we investigated the effects of ferulic acid on the expression of these EMT markers at protein and mRNA levels (Fig. 4B and C). Consistent with immunofluorescence, increased expression of epithelial markers and decreased mesenchymal markers were demonstrated in ferulic acid-treated cancer cells. Collectively, these observations support that ferulic acid induces an effective switch from a mesenchymal to an epithelial phenotype in breast cancer cells.

Based on the in vitro findings described above, MDA-MB-231 xenograft model was employed to evaluate the antitumor potential of ferulic acid in vivo. Ferulic acid was orally initiated one week prior to tumor cell injection and then continued until the end of the experiment (Fig. 5A). Oral administration of ferulic acid had no detectable toxicity, as there were no differences in body weight between the control and treatment groups, and no signs of adverse health reactions, pain or distress were observed. As shown in Fig. 5B and C, the tumors observed from the xenografts treated with ferulic acid exhibited smaller tumor volumes and lower tumor weights as compared with the control mice. Consistent with the in vitro experiments, we detected a significant decrease in proliferation (Ki67 staining) and increase in apoptosis (active caspase-3 staining) in tumors from the ferulic acid-treated mice (Fig. 5D). These in vivo data were consistent with the in vitro results and confirmed that ferulic acid exerts a marked antitumor activity against breast cancer.

To investigate whether ferulic acid can affect breast cancer metastasis in vivo, we established a metastatic model by injecting human breast cancer MDA-MB-231 cells into the tail vein of BALB/c nude mice. The metastasis of MDA-MB-231 cells into the lungs and liver was measured by H&E staining 4 weeks after injection. As shown in Fig. 6A and B, the tumor nodules on the surface of the lungs and liver were significantly decreased in the ferulic acid-treated mice (100 mg/kg, daily). In summary, our data demonstrated that ferulic acid significantly inhibits breast cancer metastasis in vivo.