Background
Oral squamous cell carcinoma (OSCC) is the most common type of head and neck cancer, which is estimated over 200,000 new cases and 120,000 deaths worldwide [
1]. In Taiwan, OSCC has emerged as one of the major malignancies with high increasing rate of both incidence and mortality in the past decade [
2]. First-line combination chemotherapy with docetaxel, cisplatin and 5-flurouracil (TPF) nowadays has been the most commonly used induction regimen for the treatment of advanced diseases (stages III and IV), but the side effects are severer than single-drug chemotherapy [
3,
4]. Despite the improvements of surgical and radiation techniques, the 5-year survival rate of oral cancer has remained unchanged at about 50 % over the past 30 years [
5]. Local recurrence and distant metastases are two critical influencing factors on survival of OSCC. Therefore, it is urgent to develop more effective agents for the improvement of clinical outcome.
According to the model of cancer stem cells (CSCs), increasing evidence suggests that tumor recurrence and metastases are caused exclusively by a rare subpopulation of tumor-initiating cells with stem cell properties [
6‐
9]. CSCs exhibit capacities of self-renewal, tumorigenicity and differentiating into non-stem cancer cells that constitute the bulk of tumors [
10,
11]. Thus, targeting the CSCs population has become a novel strategy to prevent tumor recurrence or metastasis. How to eradicate the existing CSCs to improve the survival of patients with OSCC after surgery and radio- or chemo-therapy becomes a challenging issue.
Isolation of CSCs from solid tumors has been successfully achieved through several methods based on the properties of CSCs [
7,
12]. One common method is the side population (SP) technique based on the ability of these cells to efflux a fluorescent DNA-binding dye Hoechst 33342, as first described by Goodell [
13]. The SP cells are a subset of cells harboring stem cell-like properties that show a distinct low Hoechst 33342 dye staining pattern [
14]. Some studies demonstrated that SP cells isolated from various cancer cell lines showed high expression of stemness markers and the ability to initiate tumor formation as well as resistance to chemotherapy [
14,
15]. Thus, it is postulated that SP cells are enriched of CSCs and represent an important potential target for novel anticancer drug development. Several reports had shown that SP cells possessing properties of CSCs could be isolated from OSCC cell lines [
16‐
18], however, little is known about the eradication of these CSCs. Based on our previous studies, natural products and phytochemicals are the potential source of CSC targeting agents [
19‐
22].
Honokiol is a bioactive compound purified from the bark of traditional Chinese herbal medicine
Magnolia species. Evidences from in vitro and animal models had demonstrated that honokiol possessed a variety of pharmacological effects, such as anti-inflammation, anti-angiogenesis, anti-arrhythmic and antioxidant activity [
23,
24]. It had also been shown to exert various protecting effects against hepatotoxicity, neurotoxicity, thrombosis and angiopathy [
23]. The anticancer activity of honokiol had been demonstrated in a variety of cancer cell lines, including breast, lung, ovary, prostate, gastrointestinal and oral cancer cells as well as in xenograft animal models [
24‐
26]. Our previous work and the study by Ponnurangam et al. had demonstrated the eliminating effect of honokiol on the CSCs-like population in OSCC and colon cancer cells through inhibition of Wnt/β-catenin [
20] and Notch [
27] pathway, respectively. In addition to the above stemness-associated pathways, several well-known survival/proliferation pathways such as JAK/STAT [
28], PI3K/Akt [
29,
30] and MEK/Erk [
30,
31] had been shown to govern the maintenance and survival of CSCs. However, the effects of honokiol on these pathways of CSC are remained to be elucidated. Hence, it is interesting and worth to investigate honokiol-mediated elimination of CSCs in association with inhibition of these pathways.
In this study, we investigated honokiol-mediated suppression on these survival/proliferation signaling pathways in CSCs-enriched SP from OSCC cells and examined the in vivo effectiveness by xenograft mouse model and immunohistochemical tissue staining. As expected, our results showed that honokiol inhibited these pathways in SP spheres from SAS oral cancer cells and reduced the growth and immunohistochemical staining of xenograft tumor.
Methods
Cell lines and sphere culture
Eight human oral squamous cell carcinoma (OSCC) cell lines (FaDu, KB, OE, OECM-1, SAS, SCC4, SCC25 and YD10B) were maintained in RPMI 1640 with 10 % FBS and 1 % penicillin/streptomycin at 370C, 5 % CO2, in a humidified chamber. After sorting, the side population cells were seeded at a density of 500 cells/well in 6-well ultra-low attachment plates (Corning Life Science, Corning, NY, USA) with HEscGro medium (Millipore, Billerica, MA, USA) containing epidermal growth factor (EGF, 10 ng/ml) plus basic fibroblast growth factor (bFGF, 8 ng/ml) but without any serum. The spheres were harvested after 14 days of culture for subsequent assays. The non-SP cells were incubated with serum-containing RPMI medium.
Chemicals and reagents
Honokiol (purity >98 %) was kindly provided by Dr. Jack L. Arbiser, Emory University, USA. It was dissolved in dimethyl sulfoxide (DMSO) and further diluted in sterile culture medium for in vitro experiments. The final concentrations of DMSO in cell cultures were all less than 0.05 %. The antibodies against Bax (B-9, mouse monoclonal antibody, sc-7480), Bcl-2 (100, mouse monoclonal antibody, sc-509), Erk (K-23, rabbit polyclonal antibody, sc-94), phospho-Erk (E-4, mouse monoclonal antibody, sc-7383) and STAT3 (F-2, mouse monoclonal antibody, sc-8019) were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA). The antibodies against caspase 3 (5A1E, rabbit monoclonal antibody, #9664), Akt (5G3, mouse monoclonal antibody, #2966), phospho-Akt (587 F-11, mouse monoclonal antibody, #4051), JAK2 (D2E12, rabbit monoclonal antibody, #3230), phospho-JAK2 (D4A8, rabbit monoclonal antibody, #8082) and phospho-STAT3 (D3A7, rabbit monoclonal antibody, #9145) were obtained from Cell Signaling Technology (Beverly, MA, USA).
Identification and purification of side population
The side population (SP) cells were analyzed and sorted by Hoechst 33342 (Sigma) staining and FACSAria™ III sorter (BD Biosciences, San Jose, CA, USA). Cells were detached from dishes with Trypsin-EDTA (Invitrogen, Grand Island, NY, USA) and suspended at 1 × 106 cells/mL in Hanks balanced salt solution (HBSS) supplemented with 3 % fetal calf serum and 10 mM HEPES. These cells were then incubated at 37 °C for 90 min with 2.5 μg/mL Hoechst 33342, either alone or in the presence of 50 μM reserpine (Sigma), a nonspecific inhibitor of drug-resistance ATP-binding cassette (ABC) pumps. The diminishment of SP cells in the presence of reserpine was used to define the flow cytometry gate for sorting SP cells. After 90-minute incubation, the cells were centrifuged for 5 min at 300 x g, 4 °C and resuspended in ice-cold HBSS. The cells were kept on the ice to inhibit efflux of Hoechst dye and 1 μg/mL propidium iodide (BD) was then added to discriminate dead cells. Finally, these cells were filtered through a 40 μm cells trainer (BD) to obtain single suspension cells for the analysis and sorting on FACSAria III flow cytometer.
In vivo tumorigenicity assay
Dispersed cells were re-suspended in PBS. A 100 μL suspension containing various numbers of SP or non-SP cells were injected subcutaneously into the right flanks of 4- to 5-week-old male NOD/SCID mice, obtained from Taiwan University Animal Center (Taipei, Taiwan). The animal study protocols were approved by the institutional animal care and use committee of National Heath Research Institutes, Taiwan. Tumor volume was measured on a weekly basis by a digital caliper and calculated using the following formula: 0.52 × L × W2 (L, longest diameter; W, shortest diameter). The experiment was terminated 10 weeks after tumor cells inoculation and mice were euthanized. The tumor’s wet weight was then measured.
The spheres were collected by gentle centrifugation, dissociated with trypsin-EDTA and then mechanically pipetted. The resulting single cells were re-centrifuged to remove trypsin-EDTA and re-suspended in SP medium to allow spheres re-formation. The spheres were passaged every 5–7 days before they reached a diameter of 100 μm. For the sphere formation assay, the SP and non-SP cells were seeded at a low density of 20 cells/μL in the SP medium as described above. Ten days after plating, the number of spheres (>50 μm) formed was counted under a microscope.
Cells were plated at a density of 500 cells/well on 6-well plates and cultured in serum-containing RPMI media at 37 °C in 5 % CO2 for 2 weeks. The number of colonies was counted after crystal violet staining (Sigma).
Reverse transcription polymerase chain reaction (RT-PCR)
Trizol reagent was used to extract the mRNAs from the SAS SP and parental cells according to the manufacturer’s recommended protocol. Two μg RNA was added to RT-PCR reactions containing primers at a concentration of 0.5 μM. After a 42 °C/60-min reverse transcription step, 25–36 cycles of PCR amplification were performed at 94 °C for 30 s, 55 °C for 50 s, and 72 °C for 50 s. PCR products were run on 1.5 % agarose gels for identification. Primers used were, for ABCG2, forward: 5′-CATCAACTTTCCGGGGGTGA-3′ and reverse: 5′-TGTGAGATTGACCAACAGACCA-3′; for EpCAM, forward: 5′-CTGCCAAATGTTTGGTGATG -3′ and reverse: 5′-ACGCGTTGTGATCTCCTTCT-3′; for Oct-4, forward: 5′-GGAGAGCAACTCCGATGG-3′ and reverse: 5′-TTGATGTCCTGGGACTCCTC-3′; for Nestin, forward: 5′-CTCTGACCTGTCAGAAGAAT-3′ and reverse: 5′-GACGCTGACACTTACAGAAT-3′; for GAPDH, forward: 5′-ACCACAGTCCATGCCATCAC-3′ and reverse: 5′-TCCACCACCCTGTTGCTGTA-3′.
Apoptosis analysis by Annexin V and Propidium iodide (PI) double staining
The Annexin V-FITC Apoptosis Detection Kit (BD Biosciences, San Jose, CA, USA) was used. In brief, the harvested cells were re-suspended in 1x binding buffer at a density of 1 × 106 cells/mL and cells of each 100 μl aliquot were stained with Annexin V-PI labeling solution (containing 5 μl Annexin V-FITC and 5 μl propidium iodide) at room temperature in the dark for 15 min. Finally, binding buffer (400 μl) was added and the cells were analyzed by flow cytometer.
Western blot analysis
The SP-derived spheres were collected and lysed in RIPA buffer containing protease inhibitors. Protein concentrations were measured by using the BCA protein assay kit (Thermo Scientific Biosciences, Rockford, IL, USA). Quantified protein lysates were separated by SDS-PAGE, transferred onto PVDF membrane (Millipore, Billerica, MA, USA) and immunoblotted with the primary antibodies. After incubation with HRP-conjugated secondary antibody, immunoreactive bands were visualized by enhanced chemiluminescence detection system (Millipore, Billerica, MA USA). The protein bands were quantified by AlphaEaseFC™ software.
Knockdown of STAT3
STAT3 siRNA was purchased from Cell Signaling (SignalSilence® Stat3 siRNA II #6582). The mismatch siRNA oligonucleotide 5′-UCGGCUCUUACGCAUUCAA-3′ was used as a siRNA control. Cells were transfected with siRNA oligonucleotide using Oligofectamine reagent according to the manufacturer’s instructions (Invitrogen, Grand Island, NY, USA) and analyzed 72 h post-transfection.
Wound healing assay
SAS cells were seeded into a 6-well plate. After growing to confluence, straight scratches were made across the monolayer by using a white tip along plate cover. Then, IL-6 (50 ng/ml) or honokiol (5 μM) was added into wells as indicated and recorded by photography 24 h later.
Xenograft assay
NOD/SCID mice were inoculated subcutaneously with 5 × 10
3 SAS SP cells into the flank and allowed to grow. Mice were randomly divided into four groups (
n = 5): vehicle control (1 % carboxymethyl cellulose, CMC, Sigma) and honokiol-treated groups at different dose (20, 40, 80 mg/kg). Three weeks after inoculation, honokiol (diluted in 1 % CMC immediately prior to administration) was given intraperitoneally to mice thrice a week until week 10. At the end, mice were sacrificed and the tumors were paraffin embedded for the immunohistochemical staining of PCNA (PC10, mouse monoclonal antibody, #2586, Cell Signaling Technology, Beverly, USA) and CD31 (JC/70A, mouse monoclonal antibody, ab9498, Abcam, Cambridge, UK). The PCNA labeling index was calculated as the percentage of positively stained nuclei in a total of 600 cells in 3 different areas. The vascular density was determined by counting the number of CD31-positive microvessels per high-power field (x200) [
32].
Statistical analysis
Quantitative data were shown as mean ± SD. Differences between control and honokiol-treated groups were analyzed by Student’s t-test. A p-value of <0.05 was considered statistically significant in each experiment.
Discussion
The resistance of OSCC to conventional chemotherapy or radiation therapy might be due to existence of CSCs [
37]. Consequently, agents capable of eliminating this CSC population are desirable for improving the clinical outcomes of OSCC treatments. Many preclinical studies had shown the anticancer activities of honokiol [
24]. Recently, our group and Ponnurangam et al., had reported the elimination of CSC-like population by honokiol in OSCC and colon cancer cells through Wnt/β-catenin [
20] and notch pathway inhibition [
27], respectively. This study now further demonstrated its inhibitory effects on the survival/proliferation signaling such as JAK2/STAT3, AKT, and ERK in the CSC-like SAS sphere cells and confirmed the in vivo effectiveness in xenograft animal model.
Generally, SP has been proposed as a practical method to enrich and isolate CSCs from many tumor tissues and cell lines [
14]. Several studies had demonstrated that SP isolated from OSCC cell lines indeed possesses the properties of CSCs and higher tumorigenicity [
16‐
18]. However, Broadley et al. had shown controversial results that the SP isolated from glioblastoma multiforme cells did not have enhanced stem-like property and tumor initiating activity over the non-SP cells, suggesting that the CSCs enriched by SP technique should be further confirmed by animal experiment [
38]. In our results, the SP percentage in OECM-1 (20.5 %) is much higher than that in SAS (2.9 %) cells. This phenomenon is in accordance with the report by Chiou et al. that OECM-1 expressed higher ABCG2 compared to SAS cells [
33]. However, the SAS cells are much more tumorigenic and metastatic than the OECM-1 cells [
34]. Considering this controversy, we performed an animal experiment to confirm that the SAS SP did have much higher tumorigenicity (approximately ten thousand times higher) than the non-SP. Therefore, we used SAS SP xenograft as a model to evaluate the effectiveness of honokiol.
The effects of honokiol on the increase of Bax to Bcl-2 ratio and subsequent apoptosis induction had been reported in various types of cancer cells [
39]. The significance of Bax to Bcl-2 ratio on the progression of several diseases or malignant tumors had been investigated by several studies [
40]. This ratio may serve as a predictive marker to evaluate prognosis in patients with rectal carcinomas who have undergone elective colectomy and received post-surgery adjuvant treatment [
41]. Our results further demonstrated the increased Bax to Bcl-2 ratio in the CSC-like SAS sphere cells after treatment with honokiol, indicating the potential of honokiol to improve OSCC therapy via apoptosis induction of CSCs. Compared to OECM-1 spheres, the honokiol-induced late apoptosis was more dominant in SAS sphere cells, suggesting the application of honokiol in the high-grade and aggressive OSCC might be more useful. Further clinical investigation is warranted.
Honokiol had been shown to induce apoptosis in various types of cancer cells through inhibition of several well-known survival/proliferation signaling pathways such as JAK/STAT, PI3K/Akt and MEK/Erk [
42‐
45]. As these pathways also govern the CSC maintenance and survival [
28‐
31], the honokiol-mediated inhibition of these pathways and apoptosis induction in CSC-like sphere cells would provide further mechanisms underlying its CSCs elimination potential.
The STAT3 signaling also mediates IL-6 induced EMT (epithelial-mesenchymal transition) to promote the metastasis of head and neck tumor cells [
46]. The inhibitory effect of honokiol on EMT by targeting STAT3 signaling was recently reported [
47]. In line with this, we also observed inhibitory effects of honokiol on the migration of SAS cells induced by IL-6 and on the STAT3 activity in SAS sphere cells. Furthermore, the contribution of STAT3-mediated EMT on CSC-like phenotype had also been noted [
48,
49]. It is possibile that honokiol also suppressed the STAT3-EMT-promoted CSC-like traits in the microenvironment within the xenograft tumor. Further investigation is needed.
Constitutive activation of the STAT3 is associated with not only cell proliferation and metastasis but also angiogenesis [
50,
51]. It is known that anti-angiogenesis via STAT3 inactivation also plays an important role in the honokiol-mediated anticancer activities [
52]. In agreement with this, our immunohistochemical results show that not only PCNA but also CD31 (endothelial marker) were markedly suppressed in honokiol-treated xenograft tumor tissues, indicating that honokiol may be regarded as a useful antiangiogenic agent for the treatment of OSCC.
Acknowledgements
This work was supported by National Health Research Institutes (Grant CA-101-PP-37 and CA-102-PP-37), Wan Fang Hospital, Taipei Medical University (Grant 103-wf-eva-08), and Health and Welfare Surcharge of tobacco products, Taiwan (MOHW104-TDU-B-212-124-001). We thank Dr. Jack L. Arbiser, Emory University, USA for providing bulk honokiol compound. We also thank the staffs at the Laboratory Animal Center of the National Health Research Institutes (NHRI, Taiwan) for technical support and Dr. Ying-Ying Shen at the Pathology Core Laboratory of NHRI for pathology consultation.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
JSH, CJY, SEC and GML conceived this study and wrote the manuscript. CTY participated in the design of the study and worked with JSH and WJC to carry out the experiments and analyze the data. LML, RMC and JWP provided important suggestions for data processing and manuscript editing. JSH and CJY contributed equally to this paper. All authors read and approved the final manuscript.