Lupeol Suppresses Cisplatin-Induced Nuclear Factor-κB Activation in Head and Neck Squamous Cell Carcinoma and Inhibits Local Invasion and Nodal Metastasis in an Orthotopic Nude Mouse Model

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common human neoplasm, with an estimated annual worldwide incidence of 500,000 new cases (1, 2). Despite advancements in surgery, chemotherapy, radiotherapy, and combinations of treatment modalities, the long-term survival of patients with HNSCC has remained 50% for the past 30 years (3). Therefore, investigations of potential alternative treatments for HNSCC with greater efficacy and fewer associated toxicities are urgently needed.

Human papilloma virus infection and alcohol and tobacco use are each known to play a role in the pathogenesis of HNSCC (4). Further, the combined effects of alcohol and tobacco use have a multiplicative risk in the development of HNSCC (5, 6). Cigarette smoke and alcohol consumption can cause significant oxidative stress (7, 8), which increases DNA damage (9) and, consequently, leads to the malignant transformation of normal cells (10). Epidemiologic studies have attributed the lower incidence of HNSCC in the United States to the high intake of fruits and vegetables, suggesting that some dietary substances have a protective effect against HNSCC (11, 12). In recent years, substances obtained from fruits and vegetables have gained considerable attention for the prevention and/or treatment of certain cancers (13–15). Lup-20(29)-en-3β-ol (lupeol), a triterpene found in fruits (e.g., olive, mango, strawberry, grapes, and figs), vegetables, and medicinal plants, recently showed antitumor activity in a two-stage model of mouse skin carcinogenesis and apoptotic effects in androgen-dependent prostate carcinomas (16, 17). Lupeol possesses strong antioxidant, anti-inflammatory, antiarthritic, antimutagenic, and antimalarial activities in in vitro and in vivo systems by inhibiting Ras and Fas signaling pathways (18–21). It has also been shown that lupeol induces differentiation and inhibits growth of mouse melanoma and human leukemia cells (22, 23). In view of the wide range of pharmacologic activities of lupeol, we exploited the therapeutic efficacy of lupeol in the treatment of HNSCC.

Cisplatin-based chemotherapy concurrent with radiation therapy was commonly used in the treatment of locally advanced HNSCC (24). However, the result is unsatisfactory due to the acquisition of chemoresistance by tumor cells. Multiple mechanisms for chemoresistance have been described in various cancer cells (25–27). Chemotherapeutic compounds that induce apoptosis are also known to activate nuclear factor-κB (NF-κB) in the protection of genotoxic stress. However, the role of NF-κB–related chemoresistance remains open and controversial (28). Therefore, we set out first to understand the molecular mechanism governing the chemoresistance of HNSCC and then to examine the potential chemosensitization effect of lupeol in combination with cisplatin treatment with the aim of developing an effective treatment for this deadly disease.

In the present study, lupeol selectively induced substantial HNSCC cell death and exhibited minimal effects on a normal tongue epithelial cell line in vitro. In vitro, down-regulation of NF-κB is the major mechanism of the anticancer properties of lupeol against HNSCC, as indicated by chemosensitization of the tongue carcinoma cell line CAL27. We observed that lupeol dramatically decreased tumor volume and suppressed local metastasis in our animal model. Cisplatin alone did not satisfactorily shrink the tumor and it resulted in toxicity by causing a decrease in body weight. Combination therapy consisting of a low dose of cisplatin and lupeol showed a synergistic effect on the suppression of tumor growth and metastasis, accompanied by inhibition of NF-κB without toxicity. Our results provide solid evidence that lupeol may be a novel therapeutic agent that is clinically applicable for the treatment of cancers in which NF-κB plays a significant role.

Cell lines and cell culture. The human oral squamous cell carcinoma cell lines TU159 and MDA1986 were obtained from the laboratory of Dr. Gary L. Clayman (The University of Texas M. D. Anderson Cancer Center, Houston, TX; ref. 29). The tongue fibroblast cell line Hs 677.Tg (CRL-7408) and the human tongue carcinoma cell line CAL27 were obtained from American Type Culture Collection (Manassas, VA). They were maintained in DMEM with high glucose (Life Technologies, Inc.) supplemented with 10% heat-inactivated fetal bovine serum (Life Technologies), 100 mg/mL penicillin G, and 50 μg/mL streptomycin (Life Technologies) at 37°C in a humidified atmosphere containing 5% CO₂.

Plasmids. IKKβWT was provided by Dr. D.Y. Jin (Department of Biochemistry, University of Hong Kong, Hong Kong, China; ref. 30).

Antibodies and reagents. Antibodies against β-actin, NF-κB/p50, NF-κB/p65, histone H, Sm-actin, and vimentin were purchased from Santa Cruz Biotechnology. Antibodies to E-cadherin and α-cadherin were purchased from Zymed Laboratories. Antibodies against caspase-3 and cleaved caspase-3 were purchased from Cell Signaling Technologies. Lupeol was purchased from Sigma. Cisplatin was purchased from David Bull Laboratories.

Treatment of cells. A stock solution of lupeol (30 mmol/L; MW, 426.72) was prepared by resuspension in warm alcohol and dilution in DMSO at a 1:1 ratio. For dose-dependent studies, the cells (50% confluent) were treated with lupeol (1–30 μmol/L) for 48 h in complete cell medium. The final concentrations of DMSO and alcohol were 0.25% and 0.075%, respectively, in all treatment protocols. After 48 h of treatment with lupeol, the cells were harvested and cell lysates were prepared and stored at −80°C for later use.

Cell transfection. For transient transfection, CAL27 cells were transfected with 2 μg of plasmid IKKβWT or empty vector as a control using FuGENE 6 according to the manufacturer's protocol (Roche). For generation of CAL27 cells harboring luciferase, a lentiviral vector harboring the luciferase gene was constructed and transfected using the Lentiviral RNAi Expression System (Invitrogen). Stable transfectants were generated from a pool of >20 positive clones, which were selected with blasticidin at a concentration of 2 μg/mL.

Immunofluorescence. Cells cultured on chamber slides were permeabilized with 0.1% Triton X-100 and fixed with 4% paraformaldehyde in PBS. The cells were incubated with monoclonal antibody against E-cadherin. The secondary antibody was TRITC-conjugated goat anti-mouse immunoglobulin G (Molecular Probes), and the cells were counterstained with 4′,6-diamidino-2-phenylindole. For phalloidin staining, the same fixation procedure was used as described above except that cells were incubated with fluorescein phalloidin (Molecular Probes) in 1% bovine serum albumin (dilution factor of 1:50) at 37°C for 1 h. All images were visualized by confocal microscopy and photographs were taken at ×600 magnification.

Cell cycle analysis. After lupeol treatment, the DNA content and cell cycle distribution of CAL27 cells grown in six-well plates were determined by flow cytometry. The cells were plated at a low density (5 × 10⁴ per well) and were harvested at 0, 6, 12, 24, and 48 h. Another set of controls that lacked lupeol treatment was used. Cells were trypsinized and washed once in PBS. They were then fixed in cold 70% ethanol and stored at 4°C. Before testing, ethanol was removed and the cells were resuspended in PBS. The fixed cells were then washed with PBS, treated with RNase (1 μg/mL), and stained with propidium iodide (50 μg/mL) for 30 min at 37°C. Cell cycle analysis was done in an EPICS profile analyzer using ModFit LT2.0 software (Coulter Electronics).

Immunostaining. Immunohistochemistry was done as previously described (31). To detect NF-κB/p65 and E-cadherin, the antibodies described above were used.

Luciferase promoter assay. Cells were plated onto 24-well culture plates and allowed to grow for 24 h; the NF-κB-luciferase–based construct and pRL-CMV-Luc were cotransfected into CAL27 cells using FuGENE 6 reagent (Roche Diagnostics). Cells were lysed 48 h after transfection and were assayed for luciferase activity using the Dual-Luciferase Reporter Assay System (Promega). Firefly luciferase activity was measured at 48 h after transfection and the reading was normalized to Renilla luciferase activity, which served as an internal control for transfection efficiency. Each experiment was done at least thrice in duplicate wells and each data point represented the mean and SD. The mean percentage increase (or decrease) in luciferase activity was presented as the final result and the SD of the means was used as the error bar.

Western blot. The Western blots were developed using the enhanced chemiluminescence kit (Amersham Biosciences) and was done with the antibodies described previously (32).

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay. CAL27, TU159, MDA1986, and CRL-7408 were seeded onto 96-well plates and appropriate concentrations of lupeol, ranging from 5 to 125 μmol/L, or cisplatin, ranging from 0.2 to 5 μg/mL, were then added. After 4 to 24 h, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) dye, at a concentration of 5 mg/mL (Sigma-Aldrich), was added and the plates were incubated for 12 h in a moist chamber at 37°C. Absorbance was determined by eluting the dye with DMSO (Sigma-Aldrich) and the absorbance was measured at 570 nm. At least three independent experiments were done.

Quantification of apoptosis. The terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) technique was done to detect apoptotic cells using the in situ cell death detection kit (Roche Diagnostics). Briefly, paraffin-embedded tissues were fixed with 15 μg/mL proteinase K in 10 mmol/L Tris-HCl (pH 7.4). The slides were then incubated with the TUNEL reaction mixture for 1 h at 37°C. After washing, the slides were incubated with horseradish peroxidase–conjugated anti-fluorescein antibody for 30 min at 37°C. For cytofluorometric apoptosis analysis, CAL27 cells (5 × 10⁵) were inoculated into each well of a six-well plate and treated with 1% DMSO and different doses (1, 5, and 10 μmol/L) of lupeol in 10% fetal bovine serum-DMEM for 48 h. The cells were then labeled with Annexin V-FITC (BD Biosciences Pharmingen) and analyzed with FACSCalibur (Becton Dickinson Immunocytometry Systems). Unstained cells were used as a negative control.

Wound healing assay and invasion assay. Cell migration was assessed by measuring the movement of cells into a scraped acellular area created by a 200-μL pipette tip (time 0) and the speed of wound closure was monitored after 24 h. Invasion assays were done with 24-well BioCoat Matrigel Invasion Chambers (Becton Dickinson) using 5 × 10⁴ cells in serum-free DMEM that were plated onto either control or Matrigel-coated filters. Conditioned medium from CAL27 cells was placed in the lower chambers as a chemoattractant. After 22 h in culture, cells were removed from the upper surface of the filter by scraping with a cotton swab. Cells that had invaded through the Matrigel and were adherent to the bottom of the membrane were stained with crystal violet solution. The cell-associated dye was eluted with 10% acetic acid and the resultant absorbance at 595 nm was determined. Each experiment was done in triplicate and the mean values (± SE) are presented.

Animal studies. A total of 5 × 10⁵ CAL27 cells were harvested from subconfluent cultures and injected directly into anterior tongue using a 1-mL tuberculin syringe (Hamilton Co.). A week later, the nude mice were randomized into four groups and each group consisted of 10 nude mice. They were treated with i.p. injection of lupeol (2 mg/animal) in 0.2 mL of corn oil, 5 mg/kg cisplatin, lupeol (2 mg/animal) plus 1 mg/kg cisplatin, or corn oil alone as the control group, twice per week for 30 days. The administration protocol of lupeol was followed according to Saleem et al. (16, 17).

Animal charge-coupled device experiments. A total of 5 × 10⁵ CAL27-luciferase–expressing cells in 30-μL HBSS were injected directly into the anterior tongue using a 1-mL tuberculin syringe (Hamilton Co.) with a 30-gauge hypodermic needle. The mice were imaged on days 0, 7, and 37 after tumor inoculation. Mice were anesthetized with a ketamine/xylazine mix (4:1). Imaging was done using a Xenogen IVIS 100 cooled charge-coupled device (CCD) camera (Xenogen). The mice were injected i.p. with 200 μL of 15 mg/mL D-luciferin for 15 min before imaging, after which they were placed in a light-tight chamber. The acquisition time ranged from 3 s to 1 min. The images shown are pseudoimages of the emitted light in photons/s/cm²/Sr, superimposed over the gray-scale photographs of the animal.

Histologic analysis. Sections of different tissues (4 μm) were cut and stained with H&E as previously described (32).

Statistical analysis. Continuous data were expressed as median and range and compared between groups using the Mann-Whitney U test. Categorical variables were compared using the χ² test (or Fisher's exact test where appropriate). A Pearson test was used for bivariate correlation comparison. All statistical analyses were done using statistical software (SPSS 9.0 for Windows, SPSS, Inc.). P < 0.05 was considered statistically significant.

Growth inhibition of HNSCC cell lines with lupeol. CAL27 is a tongue squamous cell carcinoma cell line, whereas TU159 and MDA1986 are primary and metastatic oral squamous cell carcinoma cell lines, respectively. A human tongue fibroblast cell line, CRL-7408, served as the control. The MTT assay was done to determine its IC₁₀ and IC₅₀ doses. The MTT assay showed that lupeol treatment induced dramatic cell death in HNSCC cell lines in a dose-dependent manner (Fig. 1A). There was no significant difference in drug sensitivity (IC₁₀ range, 13.7–22.2 μmol/L) among the three HNSCC cell lines (CAL27, MDA1986, and TU159). To examine its toxicity, the effect of lupeol on CRL-7408 cells was also investigated. The results showed that CRL-7408 is less sensitive to lupeol, with an IC₁₀ up to 41.2 μmol/L. The effect of lupeol on cell cycle distribution was evaluated by flow cytometry. When lupeol was administered at the IC₁₀ dose, CAL27 cells exhibited G₁ arrest (from 53.2% to 69.9%) in a time-dependent manner (Fig. 1B). The effect of lupeol on cell apoptosis was evaluated by cytofluorometric apoptosis analysis. Lupeol induced CAL27 cell apoptosis at doses that ranged from 15 to 30 μmol/L in CAL27 cells (Fig. 1C). The effect of lupeol on apoptosis and cell cycle distribution in another HNSCC cell line, MDA1986, was found to be similar (data not shown).

Reduced NF-kβ expression in CAL27 cells on lupeol treatment. To determine whether the growth-suppressive effect of lupeol is mediated through inhibition of NF-κB, we first examined the NF-κB promoter activity in CRL-7408 and three HNSCC cell lines (TU159, MDA1986, and CAL27). By promoter assay, CAL27 showed the highest NF-κB activity whereas CRL-7408 showed the lowest (Fig. 2A). NF-κB promoter activity was evaluated in CAL27 and MDA1986 cells with the highest NF-κB activity treated with lupeol at the dose of IC₁₀ for different time periods (from 6 to 48 h; Fig. 2B). There was a dramatic decrease in NF-κB after 6 h, which was correlated with increased growth inhibition shown in Fig. 1B. Consistently, decreased nuclear p65 and p50 protein levels accompanied by inhibition of IκBα phosphorylation were found by Western blot (Fig. 2C). To further confirm the central role of NF-κB in lupeol-induced growth suppression, we transiently transfected CAL27 and MDA1986 cells with either empty vector or IKKβWT and examined the difference in the effects of lupeol on growth suppression. As expected, NF-κB promoter activity was increased ∼3.5- and 3.1-fold, respectively (Fig. 2D). By MTT assay, IC₁₀ and IC₅₀ were increased in IKKβWT when compared with the empty vector control, suggesting that IκBα phosphorylation plays a central role in growth suppression of CAL27 and MDA1986 cells (Fig. 2D).

Lupeol and cisplatin synergistically inhibited cell growth through inhibition of NF-κB activity. Given the potential role of NF-κB in chemoresistance of some tumor types (28), we examined whether lupeol sensitized the effect of cisplatin. First, we determined the IC₁₀ and IC₃₀ of CAL27 and MDA1986 cells in response to cisplatin by the MTT assay. Cisplatin was found to inhibit CAL27 and MDA1986 cell growth in a dose-dependent manner (Fig. 3A). The respective treatments of CAL27 and MDA1986 cells that have high levels of NF-κB activity were either cisplatin (0.15 and 0.57 μg/μL; 0.21 and 0.68 μg/μL) or lupeol (13.7 and 32.3 μmol/L; 15.1 and 36.4 μmol/L). These treatments resulted in 10% to 30% inhibition of cell growth. When lupeol and cisplatin were combined at their IC₃₀ doses, growth was inhibited by 88% and 80% in CAL27 and MDA1986 cells, respectively. This effect was similar to that exerted by an 8.4-fold and 6.2-fold higher concentration of cisplatin (4.8 and 4.2 μg/mL, respectively; Fig. 3B). A weaker effect was seen in TU159 cells that showed low NF-κB activity (data not shown). The CAL27 cell line was chosen to study whether lupeol and cisplatin synergistically inhibited cell growth through inhibition of NF-κB activity. In Fig. 3C, it is shown that cisplatin induced NF-κB activity of CAL27 cells in a time-dependent manner. Cisplatin increased NF-κB promoter activity from 6 to 48 h at the IC₁₀ dose, accompanied by an increase in nuclear p65 and p50 protein levels. At the IC₁₀ dose, cleaved caspase-3 was barely detectable. When lupeol and cisplatin were combined, NF-κB promoter activity decreased from 6 to 48 h. Consistently, nuclear p65 and p50 protein levels decreased in a time-dependent manner. Apoptosis was evidenced by an increased cleaved caspase-3 protein level at 24 h after combined treatment with lupeol and cisplatin.

Lupeol suppressed cell motility and invasiveness by reversal of epithelial-to-mesenchymal transition. Increased invasion and cell motility is a major factor contributing to a poor prognosis of HNSCC (2). To examine the antimetastatic effect of lupeol on CAL27 and MDA1986 cells, we carried out assays for wound healing and invasion. As shown in Fig. 4A, similarly sized wounds were introduced into monolayers of CAL27 and MDA1986 cells at 0 h. In the control cells, the gap of the wound was filled gradually by migrating cells, and at 24 h after wound induction, the gap was almost closed (solid arrow). In contrast, after exposure to lupeol at doses ranging from 5 to 15 μmol/L, the speed of wound closure was much slower and the wound was still widely open at 24 h after exposure. As the speed of wound closure reflects the migration ability of cancer cells, these results indicate that lupeol treatment inhibits cell migration. Lupeol also exhibited a dramatic effect on the suppression of cell invasion. As indicated by an invasion assay, lupeol suppressed cell invasion by 1.7- to 6.6-fold and by 1.69- to 5.05-fold at doses ranging 5 to 15 μmol/L in CAL27 and MDA1986 cells, respectively (Fig. 4B). Cell motility is dependent on rearrangement of the actin cytoskeleton in invasive cells (33). In the untreated CAL27 cells, bundles of actin filaments were well organized as evidenced by the strong staining of stress fiber formation, intact network, and protruding morphology (Fig. 4C). However, after exposure of CAL27 cells to lupeol, a reduction and disruption of the actin stress fiber network and a lack of filopodia were observed (Fig. 4C). EMT, transdifferentiation from epithelial type to mesenchymal phenotype, is one of the major events during acquisition of the invasive phenotype in tumors of epithelial origin (34, 35). This process is often accompanied by expression of mesenchymal markers and loss of epithelial markers, especially E-cadherin (36–38). We therefore studied whether the inhibitory effect of lupeol on cell invasion of CAL27 and MDA1986 cells was mediated through reversal of EMT. We found that the morphology of lupeol-treated cells changed from a more elongated fibroblast-like morphology to a round and packed appearance of epithelial cells (Fig. 5A). In addition, expression of epithelial markers such as E-cadherin and γ-cadherin was increased by lupeol at IC₁₀ concentrations in a time-dependent manner (Fig. 5B). In contrast, expression of mesenchymal markers such as α-smooth muscle actin and vimentin was reduced (Fig. 5B). Most importantly, we found intense membrane staining of E-cadherin in lupeol-treated HNSCC cells compared with the controls, indicating activation of this protein, whereas weakly positive E-cadherin staining was mainly observed in the cytoplasm of untreated controls (Fig. 5C).

IKKβWT abolishes lupeol-induced chemosensitization and EMT reversal. To examine whether down-regulation of NF-κB is the major mechanism for lupeol-induced chemosensitization and reversal of EMT, we also transfected CAL27 cells with an NF-κB activator, IKKβWT, to examine whether the effect of lupeol can be reversed. Introduction of IKKβWT stimulated NF-κB activity by 3.5-fold as described above. The expression of IKKβWT abolished the chemosensitization effect of lupeol. Cleaved caspase-3 was observed at 48 h after the addition of both cisplatin and lupeol at IC₁₀ doses to CAL27 control cells, but it was absent in IKKβWT transfectants (Supplementary Fig. S1). To examine whether reversal of EMT by lupeol is NF-κB dependent, we examined alternations of both epithelial and mesenchymal markers in either pcDNA-CAL27 or IKKβWT-CAL27 transfectants after 48 h of lupeol administration at the IC₁₀ dose. By Western blot, there were no apparent changes in epithelial or mesenchymal markers in IKKβWT-CAL27 transfectants, suggesting that lupeol-induced reversal of EMT is NF-κB dependent (Supplementary Fig. S1).

Combined lupeol and cisplatin treatment significantly suppressed tumor growth and metastasis in an orthotopic metastatic nude mouse model of HNSCC. The in vivo therapeutic effect of lupeol was examined using the metastatic orthotopic tongue carcinoma nude mouse model. Based on the significant correlation between CCD camera signal and tumor size in vivo (39), the CCD camera provided a novel and noninvasive tool for evaluation of tumor size in vivo. CAL27 cells were first stably transduced with the luciferase gene by lentiviral infection and 5 × 10⁵ cells were injected directly into the anterior tongue of nude mice, as shown in Fig. 6A. On day 37 after tumor inoculation, local invasion of the lower jaw and regional lymph node metastasis developed and were detected by CCD camera (Fig. 6A). To confirm local invasion and regional metastasis, the animals were sacrificed and dissected (Fig. 6B). The tissues of the lower jaw and lymph node were further examined by H&E staining, which showed massive tumor cell infiltration into these tissues (Fig. 6B). Using this animal model, we examined the effect of lupeol alone and in combination with cisplatin using continuous lupeol administration at a dose of 2 mg/animal (group B), cisplatin alone (5 mg/kg; group C), lupeol (2 mg/animal) + cisplatin (1 mg/kg; group D), or corn oil (group A) as a control. The treatment began 7 days after the orthotopic implantation of CAL27 cells (Fig. 6C). During the experiment, there was a significant decrease in body weight in the control group (15 ± 2.4 g) and cisplatin alone group (15.2 ± 3.2 g) when compared with the lupeol (23.9 ± 2.5 g) and lupeol + cisplatin (22.9 ± 3.2 g) groups. This result indicated that both lupeol-treated and lupeol combined with low-dose cisplatin groups showed no signs of toxicity (infection, diarrhea, or loss of body weight). Histologic sections of normal organs like tongue, heart, liver, spleen, lung, and kidney showed no substantial cell death after H&E staining (data not shown). The tumor volumes in these four groups of animals were documented with a CCD camera. Figure 6C shows the optical CCD signals from representative groups on day 30 after treatment and Fig. 6D shows graphs of the results from cohorts using the same photonic scale. Lupeol was found to decrease the tumor volume significantly in a manner that is much more effective than cisplatin treatment (P < 0.05). Lupeol exerted a synergistic effect with low-dose cisplatin resulting in dramatic shrinkage of the tumor. By TUNEL assay, there was a remarkable difference in the number of apoptotic nuclei and patchy necrosis in tumor tissues treated with lupeol (Supplementary Fig. S2B) when compared with the untreated samples (Supplementary Fig. S2A). Combined lupeol and cisplatin treatment significantly induced tumor cell apoptosis and necrosis when compared with cisplatin alone (Supplementary Fig. S2C and D). Lupeol not only shrunk the tumor volume but also suppressed local invasion of jaw and nodal metastasis. Five of five nude mice exhibited local invasion of the jaw and/or nodal metastasis in the control group. However, an absence of jaw invasion and nodal metastasis was found in both lupeol and lupeol + cisplatin groups. Although cisplatin can exert a limited apoptotic effect on tumor cells, two of five nude mice exhibited local metastasis. From our in vitro study, we found that lupeol suppressed NF-κB–dependent EMT. By immunostaining, NF-κB expression was found to be abundant in both the nucleus (active form) and cytoplasm in the control group (Supplementary Fig. S3). After lupeol treatment, less nuclear and cytoplasmic staining of NF-κB protein was observed in the CAL27 treatment group (Supplementary Fig. S3). Accompanied with decreased NF-κB protein in the treatment group, increased membranous staining of E-cadherin was also observed, which indicates a functionally active protein (Supplementary Fig. S3).

Despite multiple modalities of treatment, such as surgery, radiation, and chemotherapy, head and neck cancers continue to have one of the lowest 5-year survival rates (3). The recent preclinical success of green tea in growth suppression of HNSCC cell lines suggested the potential use of dietary substances in the treatment of HNSCC (40). The major advantage of dietary substances over the conventional chemotherapy is its minimal toxicity to the body. In addition, natural agents that induce apoptosis may provide an opportunity for minimal acquired drug resistance and decreased mutagenesis (41). In the present study, we found that lupeol effectively and selectively inhibited cellular proliferation of head and neck cancer in vitro and in vivo. Used alone or in combination with cisplatin, lupeol induced tumor cell apoptosis and suppressed metastasis through modulation of NF-κB activity and was well tolerated in mice.

A striking observation from Fig. 1A suggests a selective response of HNSCC cell lines to lupeol compared with CRL-7408 cells. Our data are significant because, in recent years, emphasis has been placed on natural diet-based agents that selectively or preferentially eliminate cancer cells by inhibiting cell cycle progression and/or by causing apoptosis. Studies have shown that inhibition of NF-κB could result in suppression of tumor growth (28, 42). One major mechanism of NF-κB activation is through inhibition of IκB phosphorylation (43). Lupeol prevented this phosphorylation and thus resulted in reduced NF-κB activation. Our data support the idea that lupeol exerts its effects in HNSCC through inhibition of the NF-κB pathway. This finding explains the selective elimination of HNSCC cells, which showed elevated NF-κB activity when compared with normal tongue fibroblast cells.

Recent data suggested that chemotherapy including cisplatin activates the NF-κB activity, which results in chemoresistance (28). To test this hypothesis experimentally in HNSCC, we examined NF-κB activity in CAL27 cells on cisplatin treatment. Interestingly, cisplatin was able to increase NF-κB promoter activity in a time-dependent manner, accompanied with an increase in nuclear p65 and p50 protein levels. Activation of NF-κB by cisplatin could contribute to the acquired chemoresistance of HNSCC to cisplatin in a clinical situation (44). The role of NF-κB in chemoresistance of HNSCC to cisplatin was further confirmed by transfection of IKKβDN into CAL27 cells, resulting in chemosensitization (data not shown). In Fig. 3B, lupeol synergistically augmented the growth inhibitory effect of cisplatin. The underlying mechanism most probably involves down-regulation of NF-κB activity. The role of NF-κB in lupeol-induced chemosensitization was further confirmed by transfection of IKKβWT into CAL27 cells (Supplementary Fig. S1). This finding provided solid confirmation that lupeol can potentially be used for combination therapy with chemotherapeutic agents and radiation that activate the NF-κB activity.

Recently, NF-κB was shown to be involved in tumor progression via EMT, a central process governing both morphogenesis (45) and carcinoma progression in multicellular organisms (31). Due to the inhibitory effect of lupeol on NF-κB activation, we examined whether lupeol inhibited CAL27 cell motility and invasion by suppression of NF-κB–dependent EMT. In Fig. 4A and B, lupeol inhibited cell motility and invasion in a dose-dependent manner, which is supported by disruption of F-actin structure. Accompanied with decreased invasiveness, lupeol induced morphologic changes from fibroblastic to epithelial appearance, which was accompanied with a gain of the epithelial marker vimentin and loss of mesenchymal markers. In addition, these changes were accompanied with increased translocation of E-cadherin from the membrane to the cytoplasm, indicating functional activation. To further examine whether reversal of EMT is NF-κB dependent, we transfected IKKβWT into CAL27 cells. Activation of NF-κB by IKKβWT offset the effect of lupeol on reversal of EMT. Our work is the first demonstration that lupeol suppresses tumor cell invasiveness by reversal of EMT via the down-regulation of NF-κB activity.

To gain further support for our hypothesis, the in vivo effects of lupeol were examined in an orthotopic metastatic nude mouse model of oral tongue squamous cell carcinoma. In this animal model, tongue carcinoma cells invade the lower jaw and metastasize to the neck lymph node after 37 days of tumor inoculation. On day 37, the body weight of nude mice is decreased due to a feeding problem. Lupeol was i.p. administered on day 7 after tumor inoculation rather on the day 0 time point. Lupeol shrunk the tumor volume by induction of apoptosis and necrosis as shown in Supplementary Fig. S2. Lupeol showed no sign of toxicity. Interestingly, no local metastasis was detected in lupeol-treated nude mice, likely due to reversal of NF-κB–dependent EMT. Cisplatin alone (5 mg/kg) was able to partly suppress tumor growth; however, only three of five nude mice showed no jaw invasion or nodal metastasis. In addition, cisplatin has severe side effects in terms of decreasing body weight. Our data showed that lupeol is surprisingly more potent than cisplatin by ∼3-fold in terms of decrease in tumor volume. Lupeol combined with low-dose cisplatin (5-fold less than cisplatin alone) can effectively suppress tumor growth and metastasis. Most importantly, combination therapy with a low dose of cisplatin is ∼40-fold more potent than cisplatin alone and has no side effects in this animal model.

Taken together, our present study showed, both in vitro and in vivo, the anticancer and antimetastatic efficacy of lupeol, which acts by down-regulating NF-κB activity, against HNSCC with high NF-κB expression without any toxicity in nonneoplastic tongue fibroblast cells. Our experiments using our animal model showed that lupeol is much more potent than cisplatin in the treatment of HNSCC in terms of efficacy, specificity, and toxicity. Lupeol exerted a significant synergistic cytotoxic effect when combined with low-dose cisplatin without side effects. Our tongue model system provides a strong basis for the development of lupeol as a single agent or for use in combination for the treatment of different types of cancers in which NF-κB plays a significant role.

Grant support: Betty and Kadoorie Cancer Research Fund.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Prof. Hasan Mukhtar for critical reading of the manuscript.