For immunoblot analysis of NF-κB, MAb NF-κB p65/RelA (p65) antibody (Santa Cruz Biotechnology, CA, USA) was used as a primary antibody. Mouse anti-human NCAM monoclonal antibody (MAb NCAM antibody; CD56) was also purchased from Santa Cruz Biotechnology. MAb ICAM-1 was purchased from Medical & Biological Laboratories (Nagoya, Japan). Cimetidine was purchased from Sigma Chemical Co (MI, USA). H2R blockers were dissolved in PBS. Recombinant human TNF-α (R & D Systems, Inc., Minneapolis, MN) was used for the stimulation of HSG cells. MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide] were obtained from Sigma.

The HSG cell line, derived from a human submandibular salivary gland, was established by Shirasuna et al. 27 This cell line was maintained in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum (FBS), 100 IU/mL penicillin and 100 mg/mL streptomycin and grown to confluency in 25 cm² culture flasks at 37°C in a humidified 5% CO₂ incubator until required. Normal human neural progenitor (NHNP) cells and neural progenitor cell maintenance medium (NPMM) were purchased from Sanko Junyaku Co., Ltd. (Tokyo, Japan). NHNP cells were maintained with supplemented NPMM on 10 cm² polyethyleneimine (PEI) coated glass plates until they differentiated to neural cells.

HSG cells (1 × 10⁵ cells/ml) were added onto a semiconfluent monolayer culture of neural cells, incubated for 20 min at 37°C with rotation at 120 rpm, and washed extensively to exclude nonspecific cell attachment. The number of attached cells was counted directly under a microscope as reported previously 16. The cells were then co-cultured for 24 h, after which they were further co-incubated with various concentrations of cimetidine (10^-8 to 10^-4 M) for 24 h. The co-cultured cells were washed extensively to exclude nonspecific cell attachment. The numbers of attached cells, and the morphological changes of each cell type, were observed directly under a confocal laser microscope. For antibody-mediated blocking of cell adhesion, the neural cells were incubated with antibody to either NCAM or ICAM-1 (final dilution of 1:200 for both antibodies) for 3 h at 37°C in humidified CO₂ incubator, and then HSG cells were added.

Total RNA was extracted from monolayer HSG cells (1 × 10⁶ cells/ml) treated with TNF-α (10 ng/ml) or various concentrations of cimetidine (10^-8 to 10^-4 M) by the acid-guanidinium-phenol-chloroform (AGPC) method reported previously 17. To confirm the expression patterns of up-regulated or down-regulated NCAM and NF-κB genes after TNF-α or cimetidine treatment, real-time quantitative RT-PCR analyses were performed using a Bio-Rad iCycler system (Bio-Rad, Tokyo, Japan) and an iScript One-Step RT-PCR kit with SYBR Green I (Bio-Rad) according to the manufacturer's instructions. Briefly, the mRNAs were reverse-transcribed into cDNAs at 50°C for 10 min and reverse transcriptase was inactivated at 95°C for 5 min. PCR cycling and detection were followed by running for 45 cycles at 95°C for 10 s and 56°C for 30 s. PCR primers were designed and synthesized by Sigma-Aldrich, Inc. (Ishikari, Japan) with consideration of the special design criteria for real-time PCR primers. The following primer sequences were used in the PCR reactions: NCAM forward, GAA TGC CAC CGC CAA CCT C; NCAM reverse, GTC TTC CTC TTG CTC TAT CTG TTC C; NF-κB forward, AGG CGA GAG GAG CAC AGA TAC; NF-κB reverse, CGG CAG TCC TTT CCT ACA AGC; GAPDH forward, CAG CCT CAA GAT CAT CAG CA; GAPDH reverse, ACA GTC TTC TGG GTG GCA GT. Each sample was amplified in triplicate and the corresponding no-RT mRNA sample was included as a negative control. The relative mRNA level of each sample for each gene was normalized to the mRNA level of GAPDH, a housekeeping gene. The results were analyzed with the Bio-Rad iCycler Software 3.0 and Microsoft Excel 97 and presented as fold induction compared with the quantity of GAPDH mRNA (set at 1). The specificities of PCR products were analyzed by melting curve data and agarose gel electrophoresis to determine product size and to confirm that no by-products were formed.

To examine RelA translocation to the nucleus, we used a subcellular proteome extraction kit (S-PEK; Calbiochem, Darmstadt, Germany) according to the manufacturer's instructions to extract cytoplasm, cell membrane and nucleus fractions of HSG cells. Cells were treated with 10 ng/ml of TNF-α and/or 10^-4 M cimetidine, pelleted (5 × 10⁶ cells), washed twice, resuspended in 1 ml of ice-cold Extraction I containing 5 μl of protease inhibitor mixture, and then incubated for 10 min at 4°C with gentle agitation. The suspension was centrifuged at 1000 × g at 4°C for 10 min. The supernatant was used as the cytoplasm fraction; the pellet was resuspended in 1 ml of ice-cold Extraction II containing 5 μl of protease inhibitor mixture and incubated for 30 min at 4°C. It was then centrifuged at 6000 × g at 4°C for 10 min, and the supernatant was used as the cell membrane fraction; the pellet was resuspended in 500 μl of ice-cold Extraction III containing 5 μl of protease inhibitor mixture and 1.5 μl of Benzonase^® and incubated for 10 min at 4°C with gentle agitation. It was then centrifuged at 7000 × g at 4°C for 10 min, and the supernatant was used as the nucleus fraction. Each sample was subjected to immunoblot analysis.

For the detection of RelA protein by gel electrophoresis, 10 μg protein samples were mixed with an equal volume of SDS-PAGE sample buffer and boiled for 5 min. Each sample was loaded into a single lane and separated on a polyacrylamide gel of appropriate percentage. The proteins were then electroblotted onto nitrocellulose membranes. Subsequent immunoblot analysis was carried out according to the method reported previously 17. Filters were scanned and computer-generated images were analyzed with the National Institutes of Health IMAGE program to obtain densitometric values. For each series of samples (cytoplasm, cell membrane and nucleus), the relative density of each image was calculated and expressed as a percentage of the value (arbitrarily set at 100) indicated by an asterisk.

HSG cells (1 × 10⁵ cells/ml) were cultured for 12 h in 24-well culture plates containing RPMI1640 supplemented with 10% FBS. pTKκB2luc, a thymidine kinase (TK) luciferase construct containing five copies of the κB motif from the CXCL10/IP-10 gene, was kindly provided by Professor Y. Ohmori 28. Cells were transiently transfected with pTKκB2luc and pRL-TK reference Renilla luciferase plasmids (Promega, Madison, WI, USA) using FuGene transfection reagents (Roche, Nutley, NJ, USA), according to the manufacturer's instructions. At 24 h after transfection, the cells were treated with TNF-α for various periods (0, 1, 4, 8 and 24 h). Firefly and Renilla luciferase activities were assayed using reagents provided by Promega, according to the manufacturer's instructions. For standardization of transfection efficiencies, the luciferase activity from pTKκB2luc was normalized to the Renilla luciferase activity. The pGL3 control luciferase plasmid was purchased from Promega.

Small interfering RNAs (siRNAs) specific for human NCAM, NF-κB and scrambled (control) were synthesized by Sigma-Aldrich, Inc. (Ishikari, Japan). The sense and antisense strand sequences of the oligonucleotides were as follows: NCAM siRNA sense, GCA AUA UCA AGA UCU ACA ATT; antisense, UUG UAG AUC UUG AUA UUG CTT; control siRNA sense, AAU CAC AAU UGC GCA AUA ATT; antisense, UUA UUG CGC AAU UGU GAU UTT; NF-κB siRNA sense, GGG UAU AGC UUC CCA CAC UTT; antisense, AGU GUG GGA AGC UAU ACC CTT; control siRNA sense, AGU ACU CUC GUA CGC UCG ATT; antisense, UCA UGA GAG CAU GCG AGC UTT. Before transfection, FuGene 6 transfection reagent was mixed with 100 nM NCAM, 100 nM NF-κB or 100 nM control siRNA (3:3.4 μl) in serum-free medium, to a total volume of 500 μl and incubated for 30 min at room temperature. For NF-κB knockdown, HSG cells (1 × 10⁵ cells/ml) were rinsed with serum-free medium and transfected in 24-well plates with either a NF-κB siRNA duplex or a control siRNA, using FuGene 6 transfection reagents for 48 h at 37°C. Cells were treated with TNF-α for various periods (0–24 h) and subjected to a Real-Time Quantitative RT-PCR. For NCAM knockdown, HSG cells (1 × 10⁵ cells/ml) were rinsed with serum-free medium and transfected in 24-well plates with an NF-κB-dependent luciferase reporter plasmid and either a NCAM siRNA duplex or a control siRNA for 48 h at 37°C. Cells were treated with TNF-α for 4 h and subjected to a luciferase reporter assay and immunoblot analysis.

All of the experimental procedures were performed with the approval of the Animal Experimentation Committee of the Meikai University School of Dentistry. Specific pathogen-free athymic BALB/c female mice, 3–4 weeks of age, were kept under sterile conditions in a laminar flow room in cages with filter bonnets and fed a sterilized mouse diet and water. The mice were anesthetized by inhalation of diethyl ether. HSG cells (1 × 10⁶ cells) in 100 μl of PBS were injected subcutaneously into the back of each mouse with a 27-gauge needle. Tumor size was measured daily with calipers, and the HSG tumor masses in all mice, movable and elastic-hard, grew to approximately 10 mm in diameter by three weeks after HSG cell inoculation. To examine the effects of H2R antagonists, the mice were treated with 10^-4 or 10^-2 M/day cimetidine or saline (control) by intratumoral injection for seven consecutive days after the HSG tumor mass had reached approximately 10 mm in diameter. Subsequently, cimetidine or saline (control) was administered every other day for an additional 5 weeks (a total of 9 weeks after HSG cell implantation).

At nine weeks after HSG cell injection, the animals were sacrificed and the status of the tumor mass was evaluated quantitatively (Table 1). The tumor volume (V, mm³) was calculated as 0.5 × L × W², where L = length (mm) and W = width (mm), respectively. The percentage of tumor growth inhibition was expressed as the mean (± SD) tumor volumes (calculated in each group of ten mice) relative to the volume of tumors injected with control saline. The back skin was excised and cut into 2–3-mm thick slices. Then, a formalin-fixed, paraffin-embedded specimen was obtained for subsequent H&E staining and apoptosis analysis. Terminal deoxynucleotidyl transferase (TDT) dUTP nick end labelling (TUNEL) method was performed for evaluation of apoptotic cells using an in situ apoptosis detection kit (Takara, Shiga, Japan). Briefly, the deparaffinized sections were treated with 20 μg/ml proteinase K for 15 min and immersed in absolute methanol containing 0.3% H₂O₂ for 10 min at room temperature to block endogenous peroxidase activity. After washing in PBS, sections were incubated with TDT enzyme in a humidified chamber at 37°C for 1 h. After washing with PBS, the slides were incubated with diluted streptavidin-peroxidase (1:1000) for 30 min at room temperature. Positive cells were visualized using a diaminobenzidine substrate and counterstained with hematoxylin. TUNEL positive cells showed dark gray staining of the nucleus suggestive of internucleosomal DNA cleavage.

The correlations between the results of experimental treatments were evaluated by the two-tailed Student's t test. Differences in probability values (p-values) of less than 0.05 were considered statistically significant in all analyses. All analyses were performed with StatView statistical software (version 5.0; SAS Institute Inc., Cary, NC).

The study was approved by the Research Ethics Committee of the Meikai University School of Dentistry, Saitama, Japan (reference number: A0801).

To examine the effects of cimetidine on HSG cell adhesion to neural cells, a monolayer cell adhesion assay was carried out. As shown in Fig. 1A, HSG cells adhered strongly to neural cells. Using this model, we investigated the effects of various H2R antagonists at a range of non-cytotoxic concentrations (from 10^-8 to 10^-4 M 89; also confirmed by MTT assay in a previous study 12). Cultured neural cells were pretreated with cimetidine or other H2R antagonists for 24 h before the addition of HSG cells. After HSG cells were added, the cell adhesion assay was performed. As shown in Fig. 1A, adhesion of HSG cells to neural cells was inhibited by cimetidine in a dose-dependent manner. However, the other H2R antagonists, famotidine and ranitidine, had no inhibitory effect (data not shown). Aithough HSG cells and neural cells were further co-incubated with various concentrations of cimetidine (10^-8 to 10^-4 M) for 24 h, morphological change to apoptotic cells was not observed in neural cells. However, HSG cells obviously underwent apoptosis. This finding has shown in a previous study 12. To confirm that HSG adhesion to neural cells was a result of a cognate interaction between NCAM (on HSG) and NCAM (on neural cells), specific antibodies were added to block these molecules before the monolayer cell adhesion assay. Preincubation of HSG cells with an MAb NCAM abolished their ability to adhere to neural cells (Fig. 1B). Similarly, when neural cells were preincubated with the MAb NCAM, HSG adhesion to neural cells was also blocked (Fig. 1B). In contrast, pre-incubation of HSG cells with anti-ICAM-1 antibody did not significantly reduce cell adhesion (Fig. 1B, photo and graph). These findings indicate that homophilic NCAM binding is the primary mediator of cell adhesion between HSG cells and neural cells in this assay system.

To investigate the effect of cimetidine on NCAM and NF-κB mRNA levels, real-time quantitative RT-PCR analysis was carried out. Total RNA purified from HSG cultures treated with various concentrations of cimetidine at non-cytotoxic concentrations (from 10^-8 to 10^-4 M) was quantitated by real-time RT-PCR using specifically designed primer pairs. Surprisingly, the relative quantities of both NCAM and NF-κB mRNA rapidly decreased after treatment with 10^-8 M cimetidine for 24 h, and then continued to decrease further at higher doses of cimetidine (Fig. 2A). As a second step, we performed a time-course analysis on NCAM and NF-κB mRNA expression in HSG cells treated with 10^-4 M cimetidine (Fig. 2B). The relative quantity of NCAM mRNA rapidly decreased in the first 15 min, and then gradually decreased in a time-dependent manner. However, the relative quantity of NF-κB mRNA showed almost no variation up to 8 h, but had decreased after 24 h. Thus, while the rate of induction of NCAM mRNA declined rapidly, induction of NF-κB mRNA only declined slowly. These findings indicate that cimetidine time- and dose-dependently down-regulates the expression of NCAM mRNA and also reduces NF-κB mRNA expression in HSG cells in a dose-dependent fashion. As a control experiment, we performed a time-course analysis of NCAM and NF-κB mRNA expression in HSG cells treated with 10 ng/ml of TNF-α (Fig. 2C). At first, the relative quantity of NF-κB mRNA was constitutively induced in HSG cells and was gradually increased upon stimulation with TNF-α, an effect that reached a peak at 60 min and then time-dependently decreased up to 24 h. With an increasing quantity of NF-κB mRNA, the relative quantity of NCAM mRNA was gradually increased upon stimulation with TNF-α up to 24 h in a time-dependent manner. These results strongly suggest that NCAM mRNA is induced by the activation of NF-κB and may play a critical role in proliferative activity of HSG cells. To confirm whether NF-κB activation indeed increased NCAM mRNA in HSG cells, we performed a siRNA approach to reduce the expression of NF-κB and determined the effects on the basal and TNF-α-induced activity of NF-κB. As shown in Fig 2D, NF-κB knockdown by NF-κB siRNA greatly reduced NCAM mRNA expression in HSG cells, in comparison with transfection using NF-κB scrambled siRNA. These data indicate that NCAM mRNA is indeed induced by the activation of NF-κB in HSG cells.

To examine how RelA expression is regulated in HSG cells upon stimulation with TNF-α, immunoblot analysis was carried out followed by densitometric analysis. RelA protein was localized in the cytoplasm of HSG cells and was transported to the nucleus upon stimulation with TNF-α, an effect that reached a peak at 30 min and then decreased at 60 min (Fig. 3A). When the time course of RelA protein expression was analyzed up to 24 h, nuclear translocation of RelA was further decreased (data not shown). Regulation of RelA expression in HSG cells after stimulation with cimetidine was also analyzed. RelA was primarily localized in the cytoplasm with small quantities also detected in the cell membrane. The intensity of RelA expression in HSG cell membranes was decreased in a time-dependent manner by treatment with 10^-4 M cimetidine up to 60 min, whereas the intensity of RelA expression in cytoplasm of HSG cells gradually increased up to the same period (Fig. 3B). Interestingly, although the quantity of NCAM mRNA was time-dependently down-regulated in HSG cells treated with 10^-4 M cimetidine, real-time quantitative RT-PCR showed that the relative amount of NF-κB mRNA was not markedly down-regulated early after treatment (Fig. 2B). Taken together, these results show that cimetidine treatment of HSG cells resulted in complete inhibition of nuclear translocation of NF-κB, even though the relative quantity of NF-κB mRNA is sufficiently observed. Thus down-regulation of NCAM is induced via suppression of NF-κB activity by cimetidine. This effect may play a critical role in cimetidine-induced apoptosis of HSG cells.

We also examined the effect of cimetidine on TNF-α-induced nuclear translocation of RelA. RelA transport to the nucleus was detectable in HSG cells stimulated with 10 ng/ml of TNF-α and 10^-4 M cimetidine for 30 min (Fig. 3C). However, transport was markedly inhibited when compared with treatment with TNF-α alone. In contrast, the other H2R antagonists failed to block the nuclear translocation of RelA induced by TNF-α (data not shown). To investigate the effect of TNF-α on NF-κB-dependent transcriptional activity in HSG cells, a luciferase reporter assay was performed. TNF-α caused strong induction of luciferase activity (Fig. 3D, left). Maximum NF-κB-dependent transcription was observed at 4 h, with a 5-fold increase in luciferase activity compared with cells not exposed to TNF-α, and no further increase with time was observed. However, constitutive NF-κB activity was not observed in HSG cells. The increase in luciferase activity was completely dependent on the presence of κB sites, since the control plasmid lacking the κB elements did not respond to TNF-α (Fig. 3D, right). Furthermore, luciferase activities in HSG cells were time-dependently decreased in response to 10^-4 M cimetidine from 4 h after cimetidine stimulation (Fig. 3D, left). Half-maximal inhibition of luciferase activity was detected at 24 h after stimulation with 10^-4 M cimetidine.

To test whether cimetidine can suppress induction of luciferase activity after TNF-α stimulation, a luciferase reporter assay was also carried out. Induction of luciferase activity by TNF-α was inhibited by cimetidine in a time-dependent fashion (Fig. 3D, left). The decrease in luciferase activity was completely dependent on the presence of κB sites, since the control plasmid lacking the κB elements did not respond to cimetidine (Fig. 3D, right). Together, these data indicate that cimetidine has a specific and potent inhibitory effect on NF-κB-dependent transactivation.

To determine whether endogenous NCAM indeed functions as a regulator for NF-κB activation by TNF-α in HSG cells, we used a siRNA approach to reduce the expression of NCAM and determined the effects on the basal and TNF-α-induced activity of NF-κB. As expected, the NCAM protein was markedly reduced by NCAM siRNA (Fig. 4A). We then assessed the effect of NCAM siRNA on NF-κB-dependent transcriptional activity by TNF-α for 4 h. As shown in Fig 4B, NCAM knockdown by NCAM siRNA greatly reduced NF-κB activation by TNF-α in HSG cells, in comparison with transfection using NCAM scrambled siRNA. However, NCAM siRNA had no effect on constitutive NF-κB activity in HSG cells. These data indicate that NCAM functions as a regulator for NF-κB activation induced by TNF-α in HSG cells.

We then examined the effects of cimetidine on tumor development in vivo in a nude mouse model. Mice were inoculated with HSG (1 × 10⁶ cells) subcutaneously (Fig. 5A), and the effects of cimetidine on the extent of the HSG tumor mass were evaluated. Nine weeks after the injection of HSG cells, mice were sacrificed and the status of the tumor mass was examined histopathologically. As demonstrated in Table 1, cimetidine not only prevented the growth of the HSG tumor mass, but also induced a dose-dependent reduction in tumor volume. One mouse in the control group (1/10) died unexpectedly. As shown in Fig. 5B, histopathological findings revealed that the tumor mass was composed of several nodular tumors. The small nodes were composed of large round tumor cell cords clearly demarcated from the stroma. Inflammatory small round cell infiltration was found in the stroma. At 10^-4 M cimetidine, HSG tumor mass in 6 of 10 mice (60%) was resolved. Although remaining tumor cells were identified in 4 of 10 mice (40%) in which an HSG tumor mass had formed, there was clear regression of HSG tumor mass (Table 1) with evidence of extensive apoptotic cell death as observed by TUNEL in HSG tumors treated with 10^-4 M cimetidine (Fig. 5C). The significant morphologic changes, including shrinkage of cytoplasm, fragmented nuclei and release of apoptotic vesicles in HSG tumor mass, were also observed by the microscopic analysis (Fig. 5C, right). At the highest dose of cimetidine (daily doses of 10^-2 M), the HSG tumor mass was completely resolved when 1 × 10⁶ HSG cells were injected subcutaneously (none of the ten mice had detectable tumor mass; Table 1 and Fig. 5D). Famotidine or ranitidine at doses equivalent to that of cimetidine had no inhibitory effect on the growth of the HSG tumor mass (data not shown). In a repeat experiment using a greater number of HSG cells (1 × 10⁷) for inoculation, cimetidine demonstrated similar effects on the prevention of tumor growth (data not shown). Finally, immunohistochemical examination revealed that the expression of NCAM in HSG tumor mass was decreased by cimetidine in a dose-dependent fashion (Fig. 5E). These findings indicate that cimetidine effectively reduces the expression of NCAM, and ultimately induces apoptosis to HSG tumor masses in nude mice.