The MDA-MB-231 human breast cancer cell line was obtained from American Type Culture Collection (ATCC; Manassas, VA, USA) and maintained according to ATCC's recommendations. The AS-B244 human breast cancer cell line was established from the tissue of a female patient (Academia Sinica, Taipei, Taiwan) with triple negative breast cancer and maintained as previously described (19). Hinokitiol was purchased from Sigma-Aldrich (St. Louis, MO, USA), dissolved in ethanol (EtOH; Avantor Performance Materials, Center Valley, PA, USA) as a 100 mM stock solution and stored at −20°C. MG132 was purchased from Tocris Bioscience (Bristol, United Kingdom), dissolved in dimethyl sulfoxide (Sigma-Aldrich) as a 100 mM stock solution and stored at −20°C.

Cells were resuspended in Dulbecco's Modified Eagle's Medium: Nutrient Mixture F-12 (Thermo Fisher Scientific Inc., Waltham, MA, USA) containing 1% methyl cellulose (Sigma-Aldrich) to avoid cell aggregation, fibroblast growth factor-basic (20 ng/ml; PeproTech, Inc., Rocky Hill, NJ, USA), human epidermal growth factor (20 ng/ml; PeproTech, Inc.,), insulin (5 µg/ml; Sigma-Aldrich) and B27 supplement (dilution, 1:50; Gibco®; Thermo Fisher Scientific, Inc.). Cells were seeded at 10⁴ cells per dish into ultralow-attachment 10 cm dishes (Corning Life Sciences BV, Amsterdam, Netherlands). Following 7 days of incubation, the mammospheres were collected using a 100 mm cell strainer (BD Biosciences, San Jose, CA, USA) and dissociated by HyQTase (Hyclone; GE Healthcare, Logan, UT, USA) to achieve a single cell suspension for additional experiments.

For the cytotoxicity assay, 1×10⁴ mammosphere cells/well were seeded in 96-well plates (Corning Life Sciences BV) and cultured for 48 h with or without hinokitiol or 0.1% EtOH. Cell viability was determined by WST-1 reagent (BioVision, Inc., Milpitas, CA, USA) using a microplate reader (SpectraMax Plus 384; Molecular Devices, LLC, Sunnyvale, CA, USA), according to the manufacturer's protocol. The IC₅₀ value was calculated by GraFit version 7 software (Erithacus Software Ltd., East Grinstead, UK).

Wells of µ-microslide (ibidi GmbH, Martinsried, Germany) were coated with 10 µl of Matrigel (BD Biosciences) at a concentration of 8 mg/ml, and incubated at 37°C overnight. Mammosphere cells were suspended at 2×10⁴ cells/50 µl in M200 medium (Gibco; Thermo Fisher Scientific, Inc.) containing 1X low serum growth supplement (Gibco; Thermo Fisher Scientific, Inc.) and loaded into one Matrigel-coated well. The slide was then incubated at 37°C in a 5% CO₂ incubator and the VM structures were recorded by inverted microscopy (AE30; Motic Electric Group Co., Ltd., Xiamen, China) at 6 h post-seeding. Images of the wells were analyzed with the Tubeness function on Image J version 1.50e software (National Institutes of Health, Bethesda, MA, USA), and VM scores were calculated according to a formula described by Aranda et al (20) as follows: VM = [(no.spouting cells × 1) + (no. connected cells × 2) + (no. polygons × 3)] / total no. cells.

Cells were harvested using trypsin (Gibco; Thermo Fisher Scientific, Inc.), lysed in mammalian protein extraction reagent (Pierce; Thermo Fisher Scientific Inc.) and the protein concentration was determined by bicinchoninic acid reagent (Pierce; Thermo Fisher Scientific Inc.). In total, 25 µg of extracted protein was separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane (Immobilon-P; Merck Millipore, Darmstadt, Germany). The membranes were then blocked with 5% skimmed milk (Sigma-Aldrich), dissolved in Tris-buffered saline with 0.05% Tween-20 (TBST; Sigma-Aldrich) at room temperature for 1 h, followed by incubation with primary antibodies [mouse anti-human EGFR monoclonal antibody was purchased from Santa Cruz Biotechnology Inc. (Dallas, TX, USA; catalog no. sc-377229); mouse anti-human tubulin monoclonal antibody was purchased from Novus Biologicals, LLC (Littleton, CO, USA; catalog no. NB100-690)] at 4°C overnight. Hydrogen peroxidase-conjugated anti-rabbit or anti-mouse immunoglobulin G polyclonal antibody (catalog no 7076; Cell Signaling Technology, Inc., Danvers, MA, USA) were used as secondary antibodies. The membranes were washed 3 times for 5 min each time with TBST following blocking, primary antibody incubation and secondary antibody incubation. Developed chemiluminescence signals from catalyzed ECL substrate (PerkinElmer Inc., Waltham, MA, USA) were detected by a Luminescence-Image Analyzer (ImageQuant LAS 4000 mini; GE Healthcare Bio-sciences, Pittsburgh, PA, USA).

Total RNA was extracted from AS-B244 and MDA-MB-231 cells and purified using an RNA extraction kit (Quick-RNA™ MiniPrep kit; Zymo Research Corporation, Irvine, CA, USA) and complementary DNA (cDNA) was synthesized with a first strand cDNA synthesis kit (RevertAid First Strand cDNA Synthesis kit; Fermentas; Thermo Fisher Scientific, Inc.). The expression of genes was detected with specific primers and the KAPA SYBR™ fast qPCR kit (Kapa Biosystems, Inc., Wilmington, MA, USA) with the ABI StepOne™ Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.). The primer sequences used in the present study were synthesized by Integrated DNA Technologies Pte., Ltd. (Singapore, China) and were as follows: EGFR, forward 5′-CAGCGCTACCTTGTCATTCA-3′ and reverse 5′-TGCACTCAGAGAGCTCAGGA-3′; mitochondrial ribosomal protein L19, forward 5′-GGGATTTGCATTCAGAGATCAG-3′ and reverse 5′-GGAAGGGCATCTCGTAAG-3′. The cycling conditions were as follows: 50°C for 2 min, 95°C for 10 min, followed by 40 cycles of 95°C for 10 sec and 60°C for 1 min. The end-point used in quantification was calculated by the StepOne™ software (v2.2.2; Applied Biosystems; Thermo Fisher Scientific, Inc.), and the quantification cycle number (Cq value) for each analyzed sample was calculated. EGFR expression was normalized to MRPL19, which has been reported as one of the most stable internal control genes (21–23), to derive the change in Cq value (∆Cq). The primer sequence for MRPL19 was as follows: Forward, 5′-GGGATTTGCATTCAGAGATCAG-3′; and reverse, 5′-GGAAGGGCATCTCGTAAG-3′. The relative gene expression differences between groups was calculated using 2-^∆∆Cq (24). The PCR experiments were repeated three times.

Cells were seeded into a 12-well plate (Corning Life Sciences BV) at a density of 1×10⁵ cells/well and incubated with 50 µg/ml cycloheximide (Sigma-Aldrich) for 1, 3, 6, 9 or 12 h with 0.1% EtOH or 10 µM hinokitiol. Cells were harvested with trypsin at 1, 3, 6, 9 or 12 h post-seeding and the protein expression was detected by western blot analysis.

Quantitative data are presented as the mean ± standard deviation. Comparisons between two groups were analyzed with the Student's t-test. Comparisons among multiple groups (>3) were analyzed with one-way analysis of variance, and post-hoc tests were performed with Tukey Multiple Comparison analysis. P<0.05 was considered to indicate a statistically significant difference.

Hinokitiol, alternatively known as β-thujaplicin, is a tropolone-associated natural compound with antimicrobial, anti-inflammatory and antitumor activity. The present study initially examined the cytotoxic effect of hinokitiol on mammosphere cells derived from AS-B244 or MDA-MB-231 human breast cancer cells. The results revealed that the IC₅₀ values of hinokitiol for AS-B244 or MDA-MB-231 cells were 33.6±8.8 and 46.6±7.5 µM, respectively (Fig. 1A and B). Furthermore, the present study determined whether hinokitiol exhibited an inhibitory effect on the VM activity of BCSCs at concentrations below the IC₅₀. The mammosphere cells derived from AS-B244 or MDA-MB-231 cells were treated with 1 or 10 µM hinokitiol for 24 h and seeded into Matrigel-coated microwells for the analysis of in vitro VM activity in the presence of hinokitiol. Treatment with hinokitiol dose-dependently inhibited the VM activity of AS-B244 (Fig. 1C; 1 µM, P=0.041; 10 µM, P=1.18×10⁻⁹) or MDA-MB-231 (Fig. 1D; 1 µM, P=0.0102; 10 µM, P=1.30×10⁻⁸) mammosphere cells. These results indicate that hinokitiol may inhibit VM activity of BCSCs.

A previous study demonstrated that the VM activity of BCSCs is mediated by the activation of EGF/EGFR (8). The present study examined the effect of hinokitiol on EGFR expression. In the western blot analysis, the protein level of EGFR in AS-B244 or MDA-MB-231 mammosphere cells was downregulated in a dose-dependent manner by hinokitiol (Fig. 2A). The present study examined whether hinokitiol affected the mRNA expression of EGFR. In the RT-qPCR analysis, the mRNA level of EGFR in AS-B244 or MDA-MB-231 mammosphere cells did not change with hinokitiol treatment (Fig. 2B; AS-B244, P=0.1088; MDA-MB-231, P=0.0502). Previous studies have reported that surface EGFR expression may be regulated by the proteasomal protein degradation pathway (25,26). The present study examined the protein stability of EGFR following hinokitiol treatment, and the results indicated that hinokitiol decreased EGFR protein stability in AS-B244 mammosphere cells (Fig. 3).

The present study determined whether the downregulation of EGFR protein expression by hinokitiol in BCSCs was mediated by the proteasomal degradation pathway. Following co-treatment in combination with MG132, a proteasome inhibitor, hinokitiol lost the ability to inhibit EGFR protein expression in AS-B244 or MDA-MB-231 mammosphere cells (Fig. 4A). The VM inhibition activity of hinokitiol in AS-B244 mammosphere cells was additionally reversed by MG132 treatment (Fig. 4B; P=2.64×10⁻⁴). These results suggest that the VM inhibition activity of hinokitiol in breast cancer stem/progenitor cells may be mediated by proteasomal degradation of EGFR.