MCF-7 (ATCC: HTB-22) were purchased from American Type Culture Collection (ATCC, Manassas, VA, United States) and routinely maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 0.37% Na₂CO₃, 50 units/mL penicillin, and 50 μg/mL streptomycin. The cells were incubated at 37°C in a humidified atmosphere with 5% CO₂.

Complete MammoCult medium was freshly prepared with 45 mL MammoCult basal medium (Stemcell Technologies, Canada), 5 mL growth supplement (Stemcell Technologies, Canada), 200 μL heparin adjusted to a final concentration of 4 μg/mL (Stemcell Technologies, Canada), 250 μL hydrocortisone adjusted to a final concentration of 0.48 μg/mL (Sigma, St. Louis, MO, United States), and 500 μL 100× penicillin and streptomycin (Invitrogen, Carlsbad, CA, United States). Complete MammoCult medium was filter-sterilized with a 0.22-μm syringe filter.

MCF-7 cells were collected and seeded in ultra-low attachment six-well plates (BD Biosciences, San Jose, CA, United States) with complete MammoCult medium at a density of 20,000 cells/well. Bakuchiol {4-[(1E,3S)-3-ethenyl-3,7-dimethylocta-1,6-dienyl] phenol} purity 98% by HPLC was obtained from Enzo (Farmingdale, NY, United States). Cells were exposed to ethanol (vehicle control, 0.1% v/v, same amount in all the treatment conditions) or treated with 4 or 7 μg/mL bakuchiol, then allowed to grow for a week to form mammospheres (p1). Mammospheres from each group were collected and trypsinized separately. Cells from different groups were seeded and treated with ethanol or bakuchiol again for a week to form mammospheres (p2). Cell images were taken under an inverted microscope (model DMI3000 B, Leica Microsystems, Germany).

Cells were collected and washed with staining buffer (PBS containing 1% BSA). Then, 1 × 10⁶ cells were resuspended in 100 μL staining medium with 20 μL fluorochrome-conjugated antibodies for 30 min on ice. Cells were washed with staining buffer and filtered using a strainer cap tube (BD Bioscience, San Jose, CA, United States) to avoid cell clumps.

Cell phenotypes were determined by flow cytometry using a FACS Aria III cell-sorting flow cytometer (BD Biosciences, San Jose, CA, United States) with FACS Diva 6.1 software (BD Biosciences, San Jose, CA, United States). Several dot-plots of different parameters were created on a global worksheet. Then, gates were drawn on corresponding populations of breast cancer cells to identify the subpopulation of interest, including SSC-A versus FSC-A for live cells, SSC-A versus PE-A for CD24⁺ cells, SSC-A versus FITC-A for CD44⁺ cells, and PE-A versus FITC-A for CD44⁺/CD24^-/low cells. The CD44⁺/CD24^-/low cells were BCSCs.

A total of 1 × 10⁶ CD44⁺/CD24^-/low cells were seeded in a low-attachment, 60 mm plate in complete MammoCult medium with ethanol (vehicle control) or different bakuchiol concentrations. Cells were allowed to grow for 7 days to form mammospheres. Mammospheres were imaged at 8–10 random fields on days 4 and 7 with a phase contrast microscope.

CD44⁺/CD24^-/low cells in low-attachment plates were exposed to ethanol (vehicle control) or different bakuchiol concentrations for 4 days. Aldehyde dehydrogenase (ALDH) activity was determined by using ALDEFLUOR Kit (Stemcell Technologies, Canada) according to manufacturer’s instructions. Briefly speaking, cells were incubated in ALDEFLUOR assay buffer containing ALDH substrate. In each treatment condition a sample of cells was stained under identical conditions with diethylaminobenzaldehyde (DEAB), a specific ALDH inhibitor, as negative control. Following incubation, the cells were collected, resuspended in ALDEFLUOR assay buffer, and analyzed with flow cytometry (Becton Dickinson, Mountain View, CA, United States).

CD44⁺/CD24^-/low cells in low-attachment plates were exposed to ethanol (vehicle control) or different bakuchiol concentrations for 4 days. Annexin-V/propidium iodide (PI) (Promega, Fitchburg, WI, United States) staining was conducted as described by the manufacturer.

CD44⁺/CD24^-/low cells in low-attachment plates were exposed to ethanol (vehicle control) or different bakuchiol concentrations for 4 days. JC-1 staining (Invitrogen, Carlsbad, CA, United States) was conducted in accordance with the manufacturer’s instructions.

The production of reactive oxygen species (ROS) was determined by using a ROS/superoxide detection kit (Enzo, Farmingdale, NY, United States). CD44⁺/CD24^-/low cells were exposed to ethanol (vehicle control) or different bakuchiol concentrations for 4 days. In accordance with the manufacturer’s instructions, staining buffer with oxidative stress detection reagent (green fluorescence) and superoxide detection reagent (orange fluorescence) were added to washed, trypsinized, and resuspended cells. The stained cells were analyzed by flow cytometry (Becton Dickinson, Mountain View, CA, United States). The ROS inhibitor (N-acetyl-L-cysteine) was used as negative control.

Intracellular lipid content was quantified using AdipoRed Assay Reagent (Lonza, Walkersville, MD, United States). AdipoRed is a solution of the hydrophilic stain Nile Red, which can stain intracellular lipid droplets. CD44⁺/CD24^-/low cells were treated with ethanol (vehicle control) or different bakuchiol concentrations for 4 days. Cells were collected, washed, and stained with AdipoRed in accordance with the manufacturer’s recommendations. After 10 min of staining, fluorescence was measured using a SpectraMax Microplate Reader (Molecular Devices, Hong Kong) under excitation at 485 nm and emission at 572 nm.

After CD44⁺/CD24^-/low cells were exposed to ethanol (vehicle control) or different bakuchiol concentrations for 4 days, RNA extraction, genomic DNA decontamination, and semi-quantitative real-time PCR were performed as described in our previous study (Li et al., 2016). Primers are listed in Table 1.

Wild-type AB zebrafish embryos at 48 hpf were dechorionated prior to cell injection. Approximately 500 CM-DiI (Invitrogen, Carlsbad, CA, United States)-labeled cells were injected into embryonic yolk sacs. Then, the embryos were exposed to ethanol (solvent control) or different bakuchiol concentrations. The exposed embryos were maintained at 28°C for 1 h and then placed in an incubator at 34°C for 7 days. The incubation medium was replaced every 2 days. All procedures with fish were conducted following the License to Conduct Experiments approved by the Government of the Hong Kong SAR Department of Health [Refs. (15–10) in DH/HA&P/8/2/5 Pt.3].

To determine the percentage of metastasis (number of embryos with tail or head metastasis/number of total embryos), xenotransplanted embryos were examined using a lightsheet microscope (Lightsheet Z.1, ZEISS, Germany) equipped with 5× water lens. Images were taken using Z-stack functions. The results were projected with the maximum projection by using lightsheet software (Zen, ZEISS, Germany).

Statistical analyses were conducted using SigmaPlot 10.0 and SPSS 13.0. Experiments were conducted in triplicate and data were presented as mean values ± standard deviation. Data were analyzed with Student’s t-test and one-way ANOVA. A p-value < 0.05 was considered statistically significant.

MCF-7 cells were seeded in ultra-low attachment plates and treated with ethanol (vehicle control) or with 4 or 7 μg/mL bakuchiol for a week to form mammospheres. Mammosphere formation occurred in the control and 4 μg/mL bakuchiol treatment groups (Figure 1A) but not in the 7 μg/mL bakuchiol treatment group (data not shown). Mammospheres in control and treatment groups exhibited significant morphological differences: mammospheres in the treatment groups were smaller, looser, and less spherical than those in the control group (Figure 1A). The percentages of mammosphere formation (number of mammospheres/number of seeded cells) were downregulated by bakuchiol in both p1 and p2 passages (Figure 1B).

The subpopulation of CD44⁺/CD24^-/low cells accounted for less than 1% of the total cells (Figure 2A). We used this subpopulation of cells for the following mammosphere formation assay. Cells were imaged on days 1, 4, and 7 to monitor mammosphere growth after CD44⁺/CD24^-/low cells were treated with different bakuchiol concentrations. The majority of cells were separated on day 1 (Figure 2B). On days 4 and 7, control group mammospheres were compact and spherical, and treatment group mammospheres became less compact, fewer in number, and smaller as bakuchiol concentration increased (Figure 2B). When cells were exposed to 7 μg/mL bakuchiol, mammospheres appeared only on day 4 but not on day 7 (Figure 2B).

The background ALDH activity varied in different treatments. Thus, we investigated the background ALDH activity of each treatment condition by using DEAB as manufacturer suggested. After 4 days of treatment, bakuchiol significantly reduced the ALDH activity in a dose-dependent manner (Figure 3).

We exposed CD44⁺/CD24^-/low cells to ethanol (vehicle control) or to 4 or 7 μg/mL bakuchiol for 4 days. We then analyzed cell apoptosis using Annexin V/PI staining and mitochondrial membrane potential using JC-1 staining. As bakuchiol concentration increased, the percentage of early apoptotic cells increased (Figure 4A) and mitochondrial membrane potential decreased (Figure 4B) in CD44⁺/CD24^-/low cells. The mRNA levels of pro-apoptotic genes, BNIP3 and DAPK2 were significantly upregulated in cells after bakuchiol treatment (Figure 4C).

After 4 days of bakuchiol treatment, staining with ROS detection dye revealed that bakuchiol induced total ROS production (Figure 5A) and superoxide production (Figure 5B). Although total ROS production and superoxide production were upregulated in both treatment groups, total ROS production and superoxide production in the 7 μg/mL treatment group were lower than that in 4 μg/mL group (Figures 5A,B).

We analyzed lipid content and FASN mRNA expression levels in CD44⁺/CD24^-/low cells that were exposed to different bakuchiol concentrations for 4 days. AdipoRed staining revealed that lipid content increased in cells that were exposed to 4 μg/mL bakuchiol but decreased in cells that were exposed to 7 μg/mL bakuchiol (Figure 6A). Lipid contents in the treatment and control groups were not significantly different, whereas those in the two treatment groups were significantly different (Figure 6A). FASN mRNA level was elevated in the 4 μg/mL bakuchiol treatment group and downregulated in the 7 μg/mL bakuchiol treatment group (Figure 6B). The trend of upregulation at 4 μg/mL and downregulation at 7 μg/mL is consistent with that for lipid content.

Cell metastasis was observed on day 7 after cell injection in 48-hpf zebrafish embryos that were exposed to 0, 0.5, and 1 μg/mL bakuchiol. The representative images of embryos with and without cell metastasis are shown in Figure 7A. The percentages of embryos without metastasis increased in a dose-dependent manner, and the numbers of fluorescent particles decreased in a dose-dependent manner (Figure 7B). Mortality in the control and treatment groups was not significantly different, as is consistent with our previous result (Li et al., 2016).

Considering that bakuchiol inhibits cell metastasis in zebrafish embryos, we investigated how bakuchiol affected the expression levels of metastasis-related genes in CD44⁺/CD24^-/low cells. Treatment with different concentrations of bakuchiol upregulated CK18 and downregulated Notch3 in a dose-dependent manner (Figure 8A). Genes in the transforming growth factor-β (TGF-β) signaling pathway, such as TGFBR1 and ACVR1B, decreased upon exposure to bakuchiol. When cells were co-treated with the TGF-β signaling pathway inhibitor SB-431542, the expression levels of TGFBR1 and ACVR1B further decreased (Figure 8B). FASN expression exhibited a similar trend (Figure 8B).