The human head and neck carcinoma cell line HSC3 (tongue carcinoma) was obtained from the Health Science Research Resources Bank, Osaka, Japan. HSC3 cells were cultured in Eagle's minimal essential medium (EMEM). Medium contained 10% heat-inactivated fetal bovine calf serum and 1% penicillin-streptomycin. The human fibroblast cell line TIG-1 was kindly supplied by the Laboratory of Physiological Chemistry, Faculty of Pharmaceutical Sciences at Kagawa, Tokushima Bunri University, Sanuki, Japan. TIG-1 cells were cultured in DMEM with 10% FBS. Cells were incubated in a humidified 5% CO₂ atmosphere at 37°C.

D-allose was supplied by the Department of Biochemistry and Food Science, Faculty of Agriculture, Kagawa University, Kagawa, Japan. Docetaxel was obtained from Sanofi S.A. (Paris, France) and stored in frozen aliquots. Before use, it was thawed and diluted to the desired concentrations in the cell culture medium or normal saline. The growth inhibitory effect of D-allose was compared with that of D-glucose or medium only. TIG-1 cells were seeded in 96-well plates at 1.0×10³ cells/100 µl and cultured for 24 h. The medium was then removed, and fresh medium containing D-allose or D-glucose was added. The cells that were seeded in 5 separate wells in each group were incubated for an additional 24–72 h. To investigate the effects of radiation, cells were treated with 25 mM D-allose or D-glucose 6 h before irradiation with X-rays (0, 4, 8 Gy), and then incubated for a further 72 h. The cells were irradiated with a dose of 0.59 Gy/min using an X-ray irradiator (HITEX type HW 260, 200 kV, mA, Osaka, Japan). The net number of viable cells was then determined using a Cell Counting Kit-8 (CCK-8; Dojindo Laboratories, Kumamoto, Japan) according to the manufacturer's instructions. The absorbance was measured by a microplate reader at 450 nm after 2 h incubation. Values are the mean of 3 independent experiments.

A terminal deoxynucleotidyl transferase d-UTP nick-end labelling (TUNEL) assay was performed using the Apoptosis Detection System Fluorescein kit (Promega Corporation, Madison, WI, USA). Briefly, treated TIG-1 cells were spread on a poly-L-lysine slide (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany), fixed with 4% paraformaldehyde, and permeabilized with 0.2% Triton X-100. Cells were incubated in 50 µl of TdT incubation buffer (nucleotide mix [fluorescein-12-dUTP] and TdT enzyme prepared according to the manufacturer's protocol) for 60 min at 37°C in a humidified chamber. The reaction was terminated by washing the cells in 2X saline sodium citrate buffer followed by 2 washes in PBS. Cells were counterstained with 1 µg/ml propidium iodide and then washed in distilled water. Staining was observed under a fluorescence microscope. Green fluorescence indicated DNA fragmentation due to fluorescein-12-dUTP labeling.

To investigate the effects of D-allose on the expression of TXNIP and TRX, cells were cultured in 6 cm dishes with 25 mM D-allose for 24 h. To investigate the effects of radiation on the expression of TXNIP and TRX, D-allose treated or untreated cells were incubated at 37°C for 6 h, and then exposed to a single 8 Gy X-ray dose. The cells were then incubated for a further 24 h. Quantitative polymerase chain reaction (qPCR) was carried out using TaqMan gene expression assay primers and the ABI7700 Real-Time PCR system. Each reaction was performed in duplicate. The GAPDH gene was used to normalize expression across assays and runs, and a quantification value (Cq) for each sample was used to determine the expression level of the gene.

HSC3 cells were used in a xenograft model with female athymic nude mice (BALB/c nu/nu, 5–6 weeks old). A suspension of 1×10⁶ cells in 0.1 ml EMEM was injected subcutaneously into both sides of the posterior flank using a 1-cc syringe with a 27G needle. Tumors were grown for 10 days until attaining an average size of 100 mm³ (day 0). A total of 42 nude mice were assigned to 7 treatment groups (including the control group), each consisting of 6 mice. The control group mice were injected with 0.1 ml normal saline at the same time points (group 1). For the two different D-allose treatment groups, 0.1 ml of 25 mM D-allose was injected into the tumor region once a week (group 2) or 5 times a week (group 3). For the low-dose weekly docetaxel and radiation treatment group which is established as clinical model, 3 mg/kg docetaxel (20% of the maximum tolerable dose) was injected intraperitoneally, and the mice were also irradiated on days 1 and 4 (Group 4). For the combined D-allose and radiation treatment, D-allose (with the same dosing regimen as Group 3) plus radiation with a 4 Gy dose on days 1 and 4 (Group 5). The docetaxel, radiation, and D-allose group was treated with the same regimen (group 5) and 0.1 ml of 25 mM D-allose was injected into tumor tissue on day 1 (group 6) or 5 times a week (group 7). These treatments were repeated for 3 weeks. This study was approved by the Animal Care and Use Committees of Kagawa University.

For the histological studies, one mouse in each treatment group was euthanized 3 weeks after the initiation of treatment. The posterior flank skin specimens were fixed in phosphate-buffered paraformaldehyde (4%), embedded in paraffin, and cut into 4 µm thick sections. The immunohistochemistry was performed using the Vectastain ABC rabbit IgG kit (Vector Laboratories, Inc., Burlingame, CA, USA) following the manufacturer's instructions. The following primary antibodies were used: anti-tumor necrosis factor (TNF)-α (NBP1-19532) polyclonal (Novus Biologicals, LLC, Littleton, CO, USA), anti-TXNIP rabbit polyclonal (Sigma-Aldrich; Merck KGaA), and anti-TRX (C63C6) rabbit monoclonal (Cell Signaling Technology, Inc., Danvers, MA, USA).

Intensity of staining were divided into four groups-no staining, weak staining, moderate staining and strong staining. One pathologist evaluated all pathological sections without the information of experimental design.

Protein was extracted from untreated normal skin, normal skin treated with 25 mM D-allose for 2 weeks, untreated tumor tissue, and tumor tissue treated with 25 mM D-allose for 2 or 3 weeks. For the Western blot analyses, proteins were separated on 10% SDS-PAGE gels, transferred to nitrocellulose membranes, blocked with 5% (w/v) non-fat dried milk in PBS, and incubated with anti-TXNIP (MBL, Nagoya, Japan), anti-TRX (MBL), and anti-GAPDH (14c10) antibodies (Cell Signaling Technology, Inc., Tokyo, Japan). Membranes were probed with a horseradish peroxidase-conjugated anti-mouse IgG (Amersham, Tokyo, Japan), and signals were detected using an enhanced chemiluminescence system (Amersham).

Comparisons between groups for cell growth assay and mRNA expressions were compared using the Kruskal-Wallis test. Post-hoc test was carried out using the Tukey's test. Pretreatment mRNA expressions between TIG cell and HSC3 cell were compared using the Student's t-test. Significant difference between in vivo experimental groups was estimated using the Kruskal-Wallis test. Post-hoc test was carried out using the Mann-Whitney U test with Bonferrioni's correction. P<0.05 was considered to indicate a statistically significant difference.

Compared to untreated cells, the growth of TIG-1 cells exposed to 25 mM D-allose or D-glucose increased significantly, by 116.8% (P<0.001) and 112.1% (P<0.01), respectively. The growth promoting effect of D-allose was dose dependent (Fig. 1A and B). No significant reduction in cell number was observed following irradiation with 4 Gy (94.6%, P=0.2) or 8 Gy (93.9%, P=0.2) in the control cells. D-glucose treated cells were also unaffected by 4 Gy irradiation (101%, P=0.84), although their growth was marginally suppressed after an 8 Gy irradiation (93%, P=0.051). D-allose treated cells were unaffected by either of the radiation doses (4 Gy: 102%, P=0.62; 8 Gy: 101%, P=0.784) (Fig. 1C).

The TUNEL assay was carried out on the TIG-1 cell line exposed to each sugar at 25 mM for 48 h, but no apoptotic changes were observed (Fig. 2).

To assess the effect of radiation on normal cells, 8 Gy was selected as the irradiation dose in this study. The mRNA expression of TRX and TXNIP is summarized in Table I. In untreated TIG-1 cells, the ratio of TRX and TXNIP (TRX/TXNIP) was 6.4. The mRNA expression of TXNIP after D-allose treatment increased approximately 2-fold, and as a result, TRX/TXNIP decreased to 2.2. No apparent changes were observed in either TXNIP or TRX mRNA expression after 8 Gy irradiation (ratio to control: 0.97 and 0.92, respectively), and the TRX/TXNIP ratio was only slightly lower (6.1). The effect of D-allose plus radiation was the same as that of D-allose treatment alone. Compared with TIG-1 cells, the mRNA expression level of TXNIP in HSC3 cells was relatively low (50.4 vs. 1.5) and TRX/TXNIP was very high (61.7). The mRNA expression of TXNIP after D-allose treatment had increased about 74-fold and TRX/TXNIP dramatically reduced to 1.4. The change of TXNIP mRNA expression after radiation treatment was no greater than 2.6-fold and TRX/TXNIP was 25.4. Combined D-allose and radiation treatment enhanced TXNIP mRNA expression (ratio to control: 135.6), and TRX/TXNIP was reduced to 1.1.

In order to determine the appropriate dose for tumor treatment, several different doses were tested, and we found that 25 mM D-allose had the same antitumor effect as 50 mM or even higher D-allose concentrations (data not shown).

We then examined the growth inhibitory effect of D-allose with or without radiation or docetaxel in this model. The treatment schedules are shown in Fig. 3A. Although docetaxel plus radiation treated mice on average suffered a ~5% decrease in body weight compared with normal saline treated mice, the difference in weight was not statistically significant (Fig. 3B). Overall, drug treatment was well tolerated, with no apparent toxicity, and organ macroscopic examinations were normal at sacrifice. The tumor growth curves are shown in Fig. 3C. The mean tumor volumes in all of the treated groups were significantly lower than that of the control group at day 49 (P<0.0005). The greatest tumor inhibition was observed in group 7 and then group 5, with weaker inhibition in groups 3, 4 and 6. A moderate inhibition was achieved in group 2. The mean tumor volume in the group treated with multiple-doses of D-allose, weekly-docetaxel, and radiation (group 7) was significantly lower than in mice treated with weekly-docetaxel plus radiation (group 4) (P<0.001). The changes in tumor volume ratios are shown in Fig. 3D. The tumor volume had increased 18-fold in the saline treated group (group 1) at day 49, but only 10.6- and 6-fold in the mice treated with D-allose once or 5 times a week (group 2 and 3, respectively). The treatment effect in group 3 was the same as that achieved with docetaxel plus radiation (group 4, 5.8-fold). Treatment with D-allose 5 times a week and radiation (group 5) reduced the tumor volume significantly (3.1-fold). Compared with group 4, additional D-allose treatment once a week did not enhance the tumor inhibitory effect at day 49 (group 6: 5.8-fold). However, the growth inhibitory effect in group 6 persisted 11 weeks after the initiation of treatment, while the tumors in the group 4 and group 5 mice had re-grown. Half of the tumors disappeared in the group treated with multiple-doses of D-allose, weekly low-dose docetaxel, and radiation (group 7) (Fig. 4).

Histopathological findings of normal skin are shown in Fig. 5Aa. Weak to moderate increased TNF-α expression was observed in untreated epithelium. Radiation exposure resulted in an increase in epidermal thickening and hyperkeratosis (Fig. 5Ab). Strong increased TNF-α expression was also observed in the irradiated epithelium. Weak increased TNF-α expression was found in the D-allose treated epithelium (Fig. 5Ac). D-allose treatment suppressed TNF-α expression and epidermal thickening, whilst, hyperkeratosis followed the combined use of D-allose and radiation treatment (Fig. 5Ad).

Western blot analyses revealed that no apparent change was observed about the expression of TXNIP in normal skin by D-allose treatment. The expression of TXNIP in transplanted tumor tissue after 3 weeks of D-allose treatment was markedly increased in comparison to tumors treated with D-allose for only 2 weeks (Fig. 5B). TRX expression increased slightly after 3 weeks of D-allose treatment.