The in vitro and in vivo studies were divided into two parts. The first part was to investigate whether curcumin has a radiosensitizing effect and the experiments were subdivided into four groups: the control group (control or vehicle), curcumin-treated group (curcumin alone), radiation-treated group (radiation alone) and combined treatment group (combination). The other part was to verify the effect of the IκBαM vector, which represses constitutive or radiation-induced NF-κB activity, to radiosensitize the OSCC cell line. This experiment was also subdivided into four groups: empty vector-transfected group (empty vector), p-IκBαM-transfected group (IκBαM vector), radiation-treated group (radiation alone) and p-IκBαM-transfection plus radiation group (combination). The Experimental Animal Care and Use Committee of the National Yang-Ming University Office of Research and Development approved the experiment.

Stable clones of SAS/luc cells were established as described in our previous study (18). In brief, SAS cells were transfected with the pC1-luc vector using jetPEI™ (Polyplus Transfection, Sélestat, France). Cells (2×10⁵) were seeded in a 100-mm diameter dish containing 3 ml Dulbecco’s modified essential medium (DMEM; Gibco-Life Technologies, Grand Island, NY, USA) 24 h before transfection. pC1-luc (5 μl) and 10 μl of jetPEI^TM solution were diluted in 500 μl of 150 mM NaCl, respectively, then mixed together and incubated for 30 min at 37°C. The jetPEI/plasmid vector mixture was added to SAS cells and incubated at 37°C for another 24 h. Cells were then trypsinized and cultured in 100-mm dishes with DMEM containing 1 mg/ml G418 (Merck, Darmstadt, Germany) supplemented with 10% fetal bovine serum (FBS; Biological Industries, Beit Haemek, Israel) for 2 weeks. The surviving clones were isolated and transferred to 96-well plates for growth. The Luc protein expression in each clone was assayed using BLI. The recombinant bioluminescent cell clone was named SAS/luc.

The IκBα mutant vector (p-IκBαM; Clontech, Mountain View, CA, USA) was used to transfect SAS cells using the same protocol as in our previous study (19). The inhibitory effect of p-IκBαM super repressor on NF-κB activation was confirmed with electrophoretic mobility shift assay (EMSA). p-IκBαM-transfected SAS cells were used as a negative control of NF-κB activation.

SAS/luc cells were cultured in DMEM supplemented with 10% FBS and 100 U/ml penicillin plus 100 μg/ml streptomycin (Gibco-Life Technologies) at 37°C in a humidified atmosphere of 5% CO₂. G418 (500 μg/ml) was added to the medium to maintain the stable expression of the luc gene.

SAS/luc cells were seeded in 96-well plates (3,000 cells/well), incubated for 24 h, and then treated with different concentrations (0–50 μM) of curcumin (Sigma-Aldrich, St. Louis, MO, USA) for another 24 or 48 h. Cells were then washed with fresh medium. Subsequently, 100 μl of 5 mg/ml 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) solution was added into each well and incubated for 4 h at 37°C. After removing the MTT solution, cells were exposed to 100 μl of dimethyl sulfoxide for 5 min, and plates were scanned with an enzyme-linked immunosorbent assay (ELISA) reader (PowerWave X340; BioTek, Winooski, VT, USA) using a test wavelength of 570 nm and a reference wavelength of 630 nm.

SAS/luc cells (1×10⁶) were seeded in 100-mm dishes (2 dishes per group), incubated for 24 h, and then treated with or without 30 μM curcumin (i.e. half maximal inhibitory concentration, IC₅₀). Cells were irradiated immediately with various doses (2–12 Gy) using an X-ray irradiator (RS 2000; Rad Source Technologies, Suwanee, GA, USA) at a dose rate of 1.03 Gy/min, 80 cm source-to-skin distance (SSD) and field size of 30×30 cm². After irradiation, cells were incubated for 24 h. After that, cells were trypsinized, washed and resuspended in fresh medium, and then seeded in 100-mm diameter dishes. Twelve days later, the cell colonies were fixed with methanol:acetic acid (3:1) solution and stained with 2% crystal violet. The cell colony was deemed surviving when there were >50 cells. The surviving fraction (SF) was defined as the ratio of surviving colonies to the number of plated cells with correction for plating efficiency. The radiosensitization effect of curcumin on SAS/luc cells was evaluated according to the formulas (as listed below) suggested by Valeriote and Lin (20) and Carpentier et al (21).

where SF_R is the surviving fraction of SAS/luc cells treated by radiation alone; SF_C is the surviving fraction of SAS/luc cells treated with IC₅₀ curcumin alone (SF_C = 0.5); SF_R+C is the surviving fraction of SAS/luc cells treated by radiation plus IC₅₀ curcumin.

The surviving fraction of SAS/luc cells transfected with the empty or p-IκBαM vector was obtained using the same procedure as mentioned above.

After treatment with 6 Gy irradiation, 30 μM curcumin for 24 h, or a combination of both, a nuclear extraction kit (Chemicon, EMD Millipore, Billerica, MA, USA) was used to isolate the nuclear fractions of SAS/luc cells. The protocol for NF-κB/DNA binding activity evaluation followed the instructions provided with the kit (LightShift Chemiluminescent EMSA kit; Thermo Scientific Pierce Protein Biology Products, Rockford, IL, USA). The following DNA sequences were synthesized for EMSA analysis: sense, AGTTGAGGGGACTTTCCCAGGC and antisense, GCCTGGGAAAGTCCCCTCAACT. Nuclear extracts were incubated with the biotin-labeled DNA probe for 20 min at room temperature. The DNA-protein complex was separated from the free oligonucleotides on a 5% polyacrylamide gel, transferred to a nylon membrane and cross-linked by UV light. The membrane was incubated with streptavidin-horseradish peroxidase and detected by enhanced chemiluminescence (ECL; Thermo Scientific Pierce Protein Biology Products).

SAS/luc cells (1×10⁶) were grown in 6-well plates and treated with 6 Gy irradiation, 30 μM curcumin for 24 h or a combination of both. After treatment, a genomic DNA purification kit (Axygen, Tewksbury, MA, USA) was used to extract fragmented DNA following the protocol provided by the manufacturer. DNA laddering was analyzed with 1.5% agarose gel electrophoresis.

SAS/luc cells (2×10⁶) were seeded into 100-mm diameter dishes and treated with 30 μM curcumin, 6 Gy irradiation or a combination of both. The cells were then lysed with 100 μl lysis buffer (50 mM Tris-HCl, pH 8.0, 120 mM NaCl, 0.5% NP-40, 1 mM phenylmethanesulfonyl fluoride). Total proteins (40 μg) were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene fluoride membrane (EMD Millipore). The membranes were blocked with 5% non-fat milk for 1 h, and probed with the appropriate primary antibody (VEGF, MMP-9, BCL-2, XIAP, C-FLIP, cleaved-caspase-3, caspase-8, cytochrome c and β-actin; EMD Millipore) overnight at 4°C. The membranes were then washed in Tris-Tween buffer saline and incubated with horseradish peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) at room temperature for 1 h. Final quantification was achieved by using an ECL detection system (EMD Millipore). A cytosol/nuclear extraction kit (Chemicon International, Temecula, CA, USA) was used to isolate cytosolic cytochrome c following the instructions provided with the kit (22).

Six-week-old male NOD/SCID mice were purchased from the National Laboratory Animal Center, Taipei, Taiwan. SAS/luc cells (1×10⁶) in 0.2 ml serum-free medium were inoculated subcutaneously into the right legs of NOD/SCID mice. When the tumor volume reached 70 mm³, mice were randomized into four groups (n=5 for each group) according to the treatment modality: i) control group (phosphate-buffered solution with 1% dimethyl sulfoxide by gavage daily); ii) curcumin alone group (70 mg/kg/day by gavage from day 1 for 3 weeks); iii) radiation alone group (single dose 6 Gy on day 1); and iv) combination group (70 mg/kg/day by gavage from day 1 for 3 weeks plus 6 Gy irradiation on day 1). The treatment protocol for mice is depicted in Fig. 4. Tumor volume was measured by caliper weekly and calculated by the following formula (23): Volume = 1/2 × (4/3)π × length/2 × width/2 × thickness = 0.523 × (length × width × thickness).

The body weight of mice was monitored after treatment together with tumor volume tracking. Mice were sacrificed on day 21 for pathologic examination of the liver. All animal experiments complied with institutional animal care guidelines.

SAS/luc tumor-bearing mice received 150 mg/kg D-luciferin injection intraperitoneally. Before imaging, mice were anesthetized using 1–3% isoflurane for 15 min and continuously exposed to 1–2% isoflurane to maintain a sedative state during imaging. The photons emitted from the tumors were detected in vivo by IVIS 50 Imaging system (Xenogen Corp., Alameda, CA, USA). Image acquisition time was 10 sec. Regions of interest (ROIs) were drawn around the tumor and quantified as photons/second (p/sec) using Living Image software (Xenogen). In vivo BLI was performed once a week for three weeks, and the serial bioluminescent signals were quantified and displayed over time.

Mice were sacrificed on day 21 of the treatment period, and tumors were removed for nuclear protein extraction by using a nuclear extraction kit. The NF-κB/DNA binding activity was evaluated using LightShift Chemiluminescent EMSA kit.

Mice were sacrificed on day 21 of the treatment period and tumors were removed for protein extraction using the T-PER kit (Thermo Scientific Pierce Protein Biology Products). Equal amounts of protein (40 μg) were loaded on SDS-PAGE for electrophoresis and then transferred to nitrocellulose membranes. The membranes were incubated with primary antibodies specific for VEGF, MMP-9, BCL-2, XIAP, C-FLIP, cleaved caspase-3, caspase-8, cytochrome c and β-actin, followed by incubation with horseradish peroxidase-conjugated secondary antibodies. Protein expression was determined by ECL. All reagents and antibodies were purchased from EMD Millipore, except XIAP, which was purchased from Abcam (Cambridge, UK).

Data are presented as means ± standard deviation (SD). Student’s t-test was used for analysis, and P<0.05 was considered to indicate a statistically significant result.

Fig. 1A shows that curcumin killed more SAS/luc cells, as depicted by decreased cell viability, with increasing concentrations (10–50 μM) and longer treatment time (48 h). The IC₅₀ for curcumin was about 30 μM, which was selected for subsequent experiments. DNA laddering, which represents cell apoptosis, became more obvious in the SAS/luc cells when the cells were treated with curcumin for 48 h as compared to that at 24 h (Fig. 1B). Curcumin suppressed NF-κB/DNA binding activity also in a time-dependent manner (Fig. 1C). Expression of apoptotic proteins (i.e. cleaved caspase-3, cleaved caspase-8 and cytochrome c) increased while NF-κB-regulated effector proteins (i.e. angiogenesis protein VEGF, invasion protein MMP-9 and anti-apoptosis proteins BCL-2, XIAP and C-FLIP) decreased with curcumin treatment (Fig. 1D), both in a time-dependent manner.

SAS cells were transfected with p-IκBαM, a super-repressor of NF-κB, as a negative control of NF-κB activation. The radiation-induced NF-κB/DNA binding activity was inhibited by p-IκBαM (Fig. 2A). Transfection of p-IκBαM significantly increased radiation cytotoxicity at all doses (2–12 Gy) as compared with the transfection of the empty vector (Fig. 2B). Radiation-induced apoptosis, represented by DNA fragmentation (Fig. 2C) or increased apoptosis-related protein levels (Fig. 2D), was also enhanced following the transfection of p-IκBαM. Finally, the expression of radiation-induced NF-κB downstream effector proteins (i.e. VEGF, MMP-9, BCL-2, XIAP and C-FLIP) was inhibited by the transfection of p-IκBαM (Fig. 2D).

Curcumin inhibited radiation-induced NF-κB/DNA binding activity (Fig. 3A). The combination of curcumin with radiation significantly decreased the survival fraction at all doses (2–12 Gy) when compared with the radiation alone group (Fig. 3B). Synergistic effects were noticed with radiation doses >2 Gy (Table I). In addition, curcumin increased radiation-induced apoptosis as shown by DNA fragmentation (Fig. 3C) and increased apoptotic proteins levels (i.e. cleaved caspase-3, cleaved caspase-8 and cytochrome c in Fig. 3D). Fig. 3D indicates that curcumin suppressed radiation-induced NF-κB downstream effector protein expression (i.e. VEGF, MMP-9, BCL-2, XIAP and C-FLIP). These findings were identical to those observed in the aforementioned p-IκBαM-transfected group.

Fig. 4 depicts a flow chart of the in vivo experiment. SAS/luc tumor-bearing mice were divided into four groups: the vehicle-treated control (vehicle) group, the curcumin-treated (curcumin alone) group, the radiation-treated (radiation alone) group, and the combined treatment (combination) group. Mice were treated according to the grouping when the tumor volume reached 70 mm³. Fig. 5A depicts the tumor growth curve measured by caliper during the experiment. All the treatment groups showed significantly enhanced inhibition of tumor growth when compared with the control group. The combined treatment group (combination) exhibited optimal tumor control and a significantly smaller tumor volume on day 21 in comparison with the curcumin-treated (curcumin alone) or radiation-treated (radiation alone) group. BLI was also used to evaluate the therapeutic efficacy. On day 21 as shown in Fig. 5B and C, photons emitted from the tumors of the treatment groups were significantly lower than photons from the tumors of the control group. Photon flux of the combination group was significantly lower when compared with the photon flux in the curcumin or radiation alone group. As predicted in the in vitro data, ex vivo EMSA and western blotting showed that curcumin inhibited radiation-induced NF-κB/DNA binding activity, increased radiation-induced apoptotic protein levels, and suppressed radiation-induced NF-κB downstream effector protein expression (Fig. 5D and E). We traced body weight changes of the mice to evaluate the potential toxicity caused by the treatments. There were no significant changes (i.e. within 20% of the original weight) observed in all four groups (Fig. 6A). Pathological examination showed no difference in the mouse liver morphology among all the groups (Fig. 6B).