The oral cancer cell lines SAS and CGHNC8 (30) were used. SAS cells were kindly provided by Professor S.C. Lin, Yang Ming University (Taipei, Taiwan) and CGHNC8 cells were derived from oral cancer squamous cell carcinomas of patients from Chang Gung Memorial Hospital (Taoyuan, Taiwan). These cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS; Biological Industries, Cromwell, CT, USA) and antibiotics (100 U/ml penicillin, 100 U/ml streptomycin and 0.25 g/ml amphotericin B) (Invitrogen; Thermo Fisher Scientific, Inc.) at 37°C in 5% CO₂.

PL (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) was dissolved in 100% dimethyl sulfoxide stock solution and subsequently diluted using the culture medium. The following primary antibodies were used in the present study: Anti-E-cadherin (cat. no. 24E10, dilution, 1:1,000), anti-N-cadherin (cat. no. D4R1HN, dilution, 1:1,000), anti-vimentin (cat. no. D21H3, dilution, 1:1,000; Cell Signaling Technology, Inc., Danvers, MA, USA), anti-Snail (cat. no. SC-28199, dilution, 1:1,000), anti-Slug (cat. no. SC-10436, dilution, 1:1,000), anti-Oct-4 (cat. no. SC-9081, dilution, 1:1,000), and anti-CK18 (cat. no. SC-6259, dilution, 1:1,000; Santa Cruz Biotechnology, Inc., Dallas, TX, USA), anti-NANOG (cat. no. ab109250, dilution, 1:1,000; Abcam, Cambridge, UK), anti-SOX2 (cat. no. AB5603, dilution, 1:1,000; EMD Millipore, Billerica, MA, USA), and anti-GAPDH (cat. no. GTX100118, dilution, 1:8,000; GeneTex, Inc., Irvine, CA, USA).

For tumor sphere formation, cells were cultured in DMEM/nutrient mixture F-12 (F12) serum-free medium (Gibco; Thermo Fisher Scientific, Inc.) containing 10 ng/ml epidermal growth factor, 10 ng/ml human recombinant basic fibroblast growth factor, and B27 supplement (Invitrogen; Thermo Fisher Scientific, Inc.). A diluted cell suspension (3,000 cells/ml) was seeded in ultralow attachment plates (Corning Incorporated, Corning, NY, USA) then incubated at 37°C for 2 weeks and fresh medium was added to each well every 4 days. After 2 weeks, the tumor spheres were visualized and enumerated through a Nikon ECLIPSE Ti inverted light microscope.

Cell migration ability was evaluated using an in vitro wound healing assay. Cells were seeded in ibidi^® culture inserts (Applied BioPhysics, Inc., Troy, NY, USA) on 6-well plates. Following incubation at 37°C for 8 h, the culture insert was detached, forming a 500 µm cell-free gap in the cell monolayer. Subsequently, the medium was replaced with new medium containing 1% FBS and 2.5 and 5 µM PL, and cell migration towards the gap area was photographed every 5 h by Nikon ECLIPSE Ti inverted light microscope. The cell invasion assay was performed using BD BioCoat Matrigel invasion chambers (BD Biosciences, Franklin Lakes, NJ, USA) and Millicell invasion chambers (EMD Millipore). The membrane of the Millicell upper chamber insert (pore size, 8 mm) was placed in a 24-well plate and coated with Matrigel. Cells in DMEM with 1% FBS were seeded in the upper chambers. The lower chambers contained DMEM with 10% FBS in order to trap the invading cells. Cells (1×10⁵) were seeded in the upper chambers, with 2.5 and 5 µM concentrations of PL for 24 h and incubated at 37°C. Cell invasion ability was determined by staining the cells with crystal violet for 30 min at room temperature that successfully passed through the Matrigel-coated membrane to the opposite chamber and invaded cells was photographed by Nikon ECLIPSE Ti light microscope.

To examine whether PL serves as a radiosensitizer, a clonogenic assay was conducted. A total of 10³ cells were seeded in 6-well dishes and incubated for 16 h. Following treatment with 0.625 µM PL for 24 h at 37°C, the cells were exposed to 2 Gy of radiation. Subsequently, the medium was replaced with DMEM with 10% FBS. The cells were incubated at 37°C for 5–7 days and stained using crystal violet for 30 min at room temperature. The results of the radiosensitivity assay were similar to those of the clonogenic survival assay. The number of surviving colonies (defined as a colony with ≥50 cells) were counted. The survival percentage was calculated using the following equation: Survival percentage = (number of colonies of treated cells)/(number of colonies of control cells) ×100. The colonies were counted using ImageJ software (version 2.0; National Institutes of Health, Bethesda, MD, USA). To determine chemosensitivity, a total of 3,000 cells/well were seeded in 96-well plates. After 16 h, the medium was removed, and 0.03 µM 5-fluorouracil (5-FU) (Sigma-Aldrich; Merck KGaA) was added alone or in combination with 2.5 µM PL in the DMEM medium. After 48 h incubated at 37°C, cell viability was determined using the MTS assay (Promega Corporation, Madison, WI, USA) according to the manufacturer's protocol and measured the absorbance at 490 nm. The relative cell viability was calculated using the following equation: Cell viability = (absorbance of treated cells)/(absorbance of control cells) ×100.

Cells were seeded at a density of 5×10⁴ in 6-well plates for 16 h and subsequently treated with 5 µM PL for 24 h and 48 h at 37°C. The extent of cell growth was determined daily using a hemocytometer. For colony formation analysis, 10³ cells were seeded in 6-well plates for 16 h, treated with 0.625 and 2.5 µM PL, and allowed to grow uninterrupted for 7 days at 37°C in DMEM containing 20% FBS. Cell colonies were counted following staining with 5% crystal violet for 15 min at room temperature.

Total RNA was extracted using the Quick-RNA MiniPrep kit (Zymo Research Corp., Irvine, CA, USA) according to the manufacturer's protocol. The concentration, purity, and amount of total RNA were quantified using the Nano-Drop ND-1000 ultraviolet spectrophotometer (NanoDrop Technologies; Thermo Fisher Scientific, Inc.). Reverse transcription was conducted as previously described (31). qPCR was performed using the Fast SYBR^®-Green Master Mix (Invitrogen; Thermo Fisher Scientific, Inc.) on the ABI Step One Real-Time PCR system (ABI Corporation, Lee's Summit, MO, USA), according to the manufacturer's protocol. The thermal cycling conditions were an initial denaturation step at 95°C for 10 min, 40 cycles at 95°C for 15 sec and 60°C for 1 min. A comparison RT-PCR method was used to assay the relative expression of the genes in the control and PL treated groups, the expression level of each gene was determined using the ΔΔCq method as described previously (32). The following oligonucleotide sequences were used: SOX2 forward, 5′-GCCGAGTGGAAACTTTTGTCG-3′ and reverse, 5′-GGCAGCGTGTACTTATCCTTCT-3′; NANOG forward, 5′-TTTGTGGGCCTGAAGAAAACT-3′ and reverse, 5′-AGGGCTGTCCTGAATAAGCAG-3′; Oct-4 forward, 5′-CTTGAATCCCGAATGGAAAGGG-3′ and reverse, 5′-GTGTATATCCCAGGGTGATCCTC-3′; and GAPDH forward, 5′-CGGAGTCAACGGATTTGGTCGTATG-3′ and reverse, 5′-AGCCTTCTCCATGGTGGTGAAGA-3′. GAPDH expression was used as a reference.

Protein extraction and western blotting were performed as previously described (33). Cells were collected and lysed in CHAPS lysis buffer (10 mM Tris, pH 7.4, 1 mM MgCl₂, 1 mM EGTA, 150 mM NaCl, 0.5% CHAPS and 10% glycerol; Sigma-Aldrich; Merck KGaA) containing a protease inhibitor. For protein separation, 30 µg of protein was electrophoresed on 10% sodium dodecyl sulfate-polyacrylamide gel and transferred onto a polyvinylidene difluoride membrane (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The membranes were blocked and subsequently hybridized with specific primary antibodies, followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG secondary antibody (cat. no. 21234; dilution, 1:5,000) or goat anti-mouse IgG secondary antibody (cat. no. 31430; dilution, 1:5,000) (Invitrogen; Thermo Fisher Scientific, Inc.).

Proteins were visualized using Western Blotting Plus Chemiluminescence reagents (Thermo Fisher Scientific, Inc.) and autoradiography. The density of each protein band was determined following normalization with the GAPDH control band by using a gel image system and ImageJ software (version 2.0; National Institutes of Health).

Animal procedures were ethically approved by the Institutional Animal Care and Use Committee of Chang Gung University (Taoyuan, Taiwan) and conformed to the guidelines of the research council for the care and use of laboratory animals at our institution. All animals were allowed free access to chlorinated water and irradiated food. The mice were housed in individually ventilated cages with a ventilation rate of 65 exchanges of fresh air per hour, a controlled temperature of 22°C and a 12/12-h light-dark cycle. The human end points for tumor model was approved by IACUC in this study included: Tumor volume reaching 2,500 mm³ and more than 10% body weight loss. A total of 5 five-week-old male BALB/cAnN.Cg-Foxn1^nu/CrlNarl mice (18–22 g; National Laboratory Animal Center, Taipei, Taiwan) were used, and 10⁶ SAS cells were subcutaneously injected into the upper portion of the hind limb. After 7 days, PL (2.4 mg/kg) was intraperitoneally administered daily for 13 consecutive days. The tumor size was monitored, and the volume was calculated using the following equation: Tumor volume: (Short length of the tumor² × long length of the tumor)/2. Mice were sacrificed 6 weeks after the first PL administration, and the xenograft tumors were removed and weighed. Xenografted tumors were fixed in 3.7% formaldehyde solution for 24 h at room temperature and embedded in paraffin. Each sample was counterstained with a hematoxylin and eosin stain (H&E) according to the manufacturer's protocol (Zymed; Thermo Fisher Scientific, Inc.).

Student's unpaired t-test was used to compare the difference between two groups. Statistical comparisons between multiple groups were performed using a one-way analysis of variance and Tukey's post hoc test. The results were presented as the mean ± the standard deviation (SD). All of the figures were plotted using GraphPad Prism (version 6.0; GraphPad Software Inc., La Jolla, CA, USA). P<0.05 was considered to indicate a statistically significant difference.

In the present study, a tumorsphere formation assay was used to investigate the function of PL in regulating CSC properties following treatment of SAS and CGHNC8 cells with PL (0.0, 2.5 and 5.0 µM). PL inhibited the number and size of tumor spheres formed by the tested cell lines in a concentration-dependent manner. Following treatment with 2.5 and 5.0 µM PL, the number of tumorspheres was reduced to ~67.28 and 91.58%, respectively, compared with control cells, and similar effects were observed in the CGHNC8 cells (Fig. 1A). The average diameters of the tumorspheres were 418.33 and 190.88 µm in the control SAS and CGHNC8 cells, respectively. By contrast, treatment with 5 µM PL induced the formation of small spheres with diameters of 48.27 and 95.62 µm in the SAS and CGHNC8 cells, respectively (Fig. 1B). Furthermore, the present study examined the alteration of the expression of CSC-associated genes and proteins following PL treatment. Transcription factors including Oct-4, NANOG and SOX2 regulate the pluripotency of embryonic stem cells, and CK18 participates in cell differentiation (34,35). Following PL treatment, the mRNA levels of SOX2, NANOG, and Oct-4 decreased to 42, 60 and 36% respectively in the SAS cells and to 86, 96 and 87% respectively in the CGHNC8 cells, compared with control cells (Fig. 2A). Furthermore, following PL treatment, the CK18 expression levels increased 1.6 and 1.8 times in the SAS and CGHNC8 cells, respectively. Similarly, PL reduced Oct-4, NANOG, and SOX2 protein levels and increased CK18 protein levels (Fig. 2B). These results suggest that PL inhibits tumor sphere-forming ability via the suppression of Oct-4, NANOG, and SOX2 expression and promoting cell differentiation.

As CSCs are invasive, the present study examined the potential effect of PL on cell invasion. Cell migration and invasion were determined using wound healing and Matrigel Transwell assays, respectively. Although different cell lines migrated at different rates, the control cells covered the majority of the wound area; in comparison, fewer PL-treated cells migrated to the wound area (8% in the SAS cells and 61% in the CGHNC8 cells) after 10 h (Fig. 3A). The Matrigel invasion assay revealed that after 18 h of PL treatment, the number of invasive cells for the SAS and CGHNC8 cell lines decreased compared with that observed for the control cells (Fig. 3B). PL reduced the cell migration and invasion ability of the two cell lines in a concentration-dependent manner. EMT is another biological process that enables the metastasis of tumor cells through the bloodstream. Similar to the cancer stemness pathway, this process mechanistically initiates cancer metastasis and confers treatment resistance (36). Therefore, the function that PL serves in EMT was investigated. Compared with the control cells, the PL-treated cells exhibited an increased expression level of E-cadherin but a decreased expression level of N-cadherin and vimentin. PL reduced the levels of transcription factors, including Snail and Slug that participate in EMT (Fig. 3C). These results suggest that PL suppressed cell migration and invasion ability, and so regulated EMT in oral cancer cells.

Surgery, radiotherapy, chemotherapy or combinations of these methods are the main therapeutic methods for oral cancer. The main reason for therapeutic failure is acquired resistance, which is attributed to CSCs. Therefore, whether PL serves as a sensitizer was investigated in order to improve therapeutic efficacy. The clonogenic survival assay was used to examine the radiosensitization effect of PL following the treatment of cells with PL alone, radiation alone or a combination of PL and radiation. The radiation sensitivity levels of SAS and CGHNC8 cells treated with the combination of PL and radiation increased by 47.5 and 25.63%, respectively (Fig. 4A). 5-FU, which is commonly used for treating oral cavity cancers, was selected to test chemosensitivity. As presented in Fig. 4B, ~60 and 70% of SAS and CGHNC8 cells, respectively, were killed following treatment with 5-FU for 48 h. The same effect was observed in PL-treated cells. The SAS and CGHNC8 cells that were treated with 5-FU and PL exhibited enhanced drug sensitivity (Fig. 4B). Thus, PL enhanced the radio- and chemosensitivity levels of oral cancer cells.

PL induces apoptosis in various types of cancer (19,20). Therefore, whether PL regulates the tumor growth ability of oral cancer cells was investigated. PL treatment (5 µM) for 48 h reduced the growth of the SAS and CGHNC8 cells to 11.2 and 21.9%, respectively, compared with the control cells (Fig. 5A). The colony formation assay was used to examine the long-term effect of PL by treating cells with PL (0.625 and 2.500 µM) for 7 days. Consistently, PL significantly suppressed colony formation, and no or few colonies were formed following PL treatment (2.5 µM; Fig. 5B). Furthermore, PL reduced the growth of oral cancer cells and their colony-forming ability. In order to investigate the effects of PL on tumor growth in vivo, xenograft SAS tumors were established in BALB/c nude mice. Following injection with SAS tumor cells, the mice (5 per treatment group) were intraperitoneally administered with PL or control saline and monitored for 6 consecutive weeks. Tumor growth was significantly slower in the PL-treated groups than in the control groups (Fig. 5C). On day 42, tumor growth decreased by 63% in the PL-treated group (P<0.05). Once the mice were sacrificed, the tumor weight was revealed to be decreased by 66% in the PL-treated group (Fig. 5D). Tumor tissues were confirmed using H&E staining (Fig. 5E). The results indicate that PL inhibited tumor growth in vitro and in vivo.