WK2 was established from a lower gingival carcinoma from a 58-year-old male patient who had not undergone preoperative therapy (5). WK3 was established from secondary cervical lymph node metastases of the lower gingival carcinoma from the same patient 7 months later. However, WK3 did not have stable growth. In this study, WK3F was established from WK3 by repeatedly selecting colonies that grew stably. This study design was approved by the appropriate ethics review boards of Kyushu University. This patient provided written, informed consent.

WK2 and WK3F, which was established from WK3, were used for various experiments. In addition, we used the human tongue oral squamous cell carcinoma cell lines HSC3 and SAS and human lung fibroblast cell line VA13 which WI38 was transformed by SV40 virus.

These cell lines were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS; Filton Pty, Brooklyn, Australia), 2 mM l-glutamine (Wako Pure Chemical Industries, Osaka, Japan), penicillin G (Meiji Seika Pharma, Tokyo, Japan), and streptomycin (Meiji Seika Pharma) in a 5% CO₂ incubator at 37°C. The cell lines were passaged using 0.05% trypsin/EDTA (Gibco, Grand Island, NY, USA) and phosphate-buffered saline (PBS).

WK2, HSC3, SAS, VA13 cells were transfected with the pAcGFP-N1 vector (Clontech, Palo Alto, CA, USA), while WK3F cells were transfected with the pDsRed-Express2-N1 vector (Clontech) using Lipofectamine LTX (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Colonies that exhibited resistance to geneticin (G418; Sigma-Aldrich) were pooled from the individual transfection experiments. In order to obtain cells that homogeneously expressed green fluorescent protein (GFP) and red fluorescent protein (RFP), a pool of geneticin resistant colonies was selected using green and red fluorescence. All selected cells were maintained in DMEM containing 10% fetal bovine serum and 1 μg/ml geneticin.

Each cell line was seeded onto a 96-well plate at a density of 5×10³ cells per well and incubated in DMEM containing 10% FBS for 8 h. After incubation, each cell line was analyzed by CellTiter 96 Aqueous Non-radioactive Cell Proliferation Assay G4000 (Promega, Madison, WI, USA) according to the manufacturer’s instructions. The absorbance of samples at 590 nm was measured with a microplate reader (model 680; Bio-Rad Laboratories, Hercules, CA, USA) with the results from the first day used as the baseline (absorbance = 1). Analysis of cell proliferation was performed on each day for 14 days. All experiments were carried out in triplicate and repeated 3 times.

Cell doubling time was calculated by utilizing the known mathematical formula for tumor doubling (6).

where T_i is the interval time, D_i is the absorbance from the second day, and D_o is the absorbance from the third day.

Each cell line (3×10⁵) was seeded on a 6-well plate, and 24 h later ‘wounds’ were scratched with a 200-μl pipette tip, washed with medium, and observed under a fluorescence microscope (BZ-8000; Keyence, Osaka, Japan). The wound regions were photographed again after 3, 6, 9, 12, 15, 18, 21 and 24 h, in order to measure the wound area. Wound area was standardized by using the following formula: Wound area (% of control) = Wound area after the indicated period ×100/initial wound area. All experiments were carried out in triplicate and repeated 3 times.

Evaluation of tumor dissemination from the primary cancer nest was modified from the procedure developed by Abe et al (7). In brief, 1×10⁶ cells from each fluorescently labeled cell line were mixed in 10 μl collagen type I gel to form a solid cell cluster. The collagen-embedded cell lines were allowed to solidify for 30 min at 37°C in a 100-μl microcentrifuge tube. The collagen-embedded cell lines were embedded in collagen type I gel (1×10⁵ cells/ml) containing fibroblasts and solidified. The fibroblast was derived from the WK2 and WK3F patient. Growth medium was placed over the collagen gels and cultured. Tumor dissemination was observed under a fluorescence microscope (BZ-8000; Keyence). The grade of tumor dissemination from the tumor cell pellet was evaluated by measuring the moving distance of farthest cells from the edge of the nest in each of the regions. The divided region is 12 regions by 5 degrees from the center of the aggregate. The evaluation was done after 7 days. All experiments were carried out in triplicate and repeated 3 times.

Evaluation of self-replication was performed as previously described (8). WK2 and WK3F were seeded at a density of 5×10⁴ cells/ml in 60-mm noncoated dishes with serum-free DMEM containing 40 ng/ml basic fibroblast growth factor (bFGF; Invitrogen) and 20 ng/ml epidermal growth factor (EGF; Invitrogen) for floating cultures. The cells were cultured in a humidified incubator in an atmosphere of 5% CO₂ at 37°C, in the presence of bFGF and EGF, which were added to the medium every other day. After 10 days, the diameters of the cell clusters that had developed were measured, and cell clusters with a diameter >100 μm were counted as spheres. For passaging, primary spheres (day 10) were treated with 0.05% trypsin/EDTA and dissociated into single cells, after which the cells were added to 24-well culture plates at a density of 1×10⁴ cells/ml in serum-free medium. The cells were cultured for a further 10 days in serum-free medium to obtain secondary spheres. In the same way, the diameters of the developed cell clusters were measured, and cell clusters with a diameter >100 μm were counted as spheres.

All animal experimental protocols used in this study were approved by the Animal Care and Use Committee of Kyushu University. Experiments were conducted using 8-week-old female athymic nude mice (BALB/cA Jcl-nu/nu) purchased from Kyudo (Fukuoka, Japan). The mice were housed in laminar flow cabinets under specific pathogen-free conditions in facilities approved by Kyushu University. For the experimental metastasis studies, 1×10⁶ cells in 40 μl PBS were injected into the tongue using a syringe with a 27-gauge disposable needle (TOP Plastic Syringe, Tokyo, Japan) under inhalation anesthesia using sevoflurane. The primary tumor volumes were measured weekly, and mice were sacrificed when the primary tumor volume reached 100 mm³ or when 60 days had passed after the injection. After sacrifice, tumors of the tongue and metastases from tongue tumors in the cervical lymph nodes were visualized macroscopically under light excitation.

Total RNA was isolated from WK2 and WK3F using TRIzol Reagent (Invitrogen) and was purified using an SV Total RNA Isolation System (Promega) according to the manufacturer’s instructions. RNA samples were quantified with an ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA), and the quality was confirmed with an Experion System (Bio-Rad Laboratories).

The cRNA was amplified, labeled, and hybridized to a 60K Agilent 60-mer oligomicro-array (SurePrint G3 Human Gene Expression Microarray 8×60 K, v2; Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s instructions. All hybridized microarray slides were scanned by an Agilent scanner. Relative hybridization intensities and background hybridization values were calculated using the Agilent Feature Extraction software (9.5.1.1).

Raw signal intensities and flags for each probe were calculated from hybridization intensities (gProcessedSignal), and spot information (e.g., gIsSaturated), according to the procedures recommended by Agilent. Flag criteria for the GeneSpring Software: Absent (A): ‘Feature is not positive and significant’ and ‘Feature is not above background’. Marginal (M): ‘Feature is not uniform’, ‘Feature is saturated’, and ‘Feature is a population outlier’. Present (P): others. The raw signal intensities of two samples were log₂-transformed and normalized by quantile algorithm with the ‘preprocessCore’ library package (9) by using the Bioconductor software (10).

We selected probes that called the ‘P’ flag in both samples. To identify up- and downregulated genes, we calculated Z-scores (11) and ratios (non-log scaled fold-change) from the normalized signal intensities of each probe for comparison between control and experiment sample. Then, we established criteria for regulated genes: upregulated genes, Z-score ≥2.0 and ratio ≥5-fold, downregulated genes, Z-score ≤−2.0 and ratio ≤0.2. The extracted data were analyzed using DAVID (12) that is a database provided by National Institute of Allergy and Infectious Diseases (NIAID). Moreover, microarray date analysis was supported by Cell Innovator.

All data are shown as mean ± SD, analyzed via analysis of variance and Student’s t-test, and processed by the statistical software SPSS 13.0. Statistical significance was assumed at P<0.05.

WK2, which was derived from a lower gingival carcinoma and WK3, which was established from a secondary cervical lymph node metastases from the same patient, were established in our laboratory. In this study, WK3F was established from WK3 by repeatedly selecting colonies that grew stably because WK3 was not able to grow stably.

In the images obtained using phase-contrast microscopy, WK2 and WK3F cells were found to be arranged in a cobblestone pattern (Fig. 1A and B). In addition, a few of the WK3F cells had an elongated fibroblast-like structure (Fig. 1B).

Histopathological findings from our examination of the primary tumor (Fig. 1C) indicated that the border of the tumor was unclear and that the large and small cancer nests were scattered. Moreover, the primary tumor had the shape of a well-differentiated carcinoma with a cancer pearl. On the other hand, images of the cervical lymph node metastases indicated that it had undergone funicular shaped proliferation (Fig. 1D).

MTT assays were then used to evaluated the proliferation ability of each cell line. The proliferation ability of WK3F cells was found to be about 1.2 times higher than that of WK2 cells at 7 days, at which point the WK3F cells had reached confluence (Fig. 2). We also calculated the doubling time during the logarithmic growth phase. The doubling time for WK2 was 32 h, whereas the doubling time for WK3F was 24 h. WK3F doubled as compared to WK2 in two-thirds the time (Table I).

Migratory ability was examined by using a wound healing assay (Fig. 3). The wound area in WK2 cell cultures was ∼50% after 24 h as compared to 0 h. On the other hand, the wound area of the WK3F cell cultures was ∼40% after 24 h (Fig. 3K). The migratory ability of WK3F was almost the same as that of the SAS and HSC3. These results demonstrate that WK3F had increased migratory ability when compared to WK2 (P<0.05).

As mentioned in Materials and methods, the evaluation of cancer cell invasion ability of the primary tumor was examined by a procedure that was developed by Abe et al (7). In this assay, invasion of such cell line is visualized as small green and red fluorescent spots escaping from the cell pellet, which mimics the primary cancer nest. Each cancer cell invasion was evaluated by using the distance of the most invasive cells from the artificial primary cancer nest in each region shown in Fig. 4F.

As shown in Fig. 4A, invasive cells were hardly observed in WK2-GFP. In contrast, WK3F-RFP cells demonstrated aggressive cell invasion into artificial stromal tissue (Fig. 4B). Fig. 4G shows a comparison of the invasiveness of each cell line. The distance of the invasive cells was higher by about two-fold in WK3F-RFP as compared to WK2-GFP (P<0.001).

WK2 and WK3F were evaluated for their ability to self-replicate, a cancer stem cell characteristic, by using the sphere-forming assay. WK2 cells were found to have weak sphere-forming capacity, with respect to both diameter and number (Fig. 5A and B), whereas WK3F cells showed significant sphere-forming capacity (Fig. 5C and D). The sphere diameter of WK3F was approximately twice that of WK2 in the primary and secondary spheres (Fig. 5E). Furthermore, the number of spheres was more different in the secondary spheres than in the primary spheres. The number of spheres formed by WK3F was ∼3 times higher than that of WK2 (Fig. 5F). These data suggest that WK3F contains many cells with self-replication ability.

WK2-GFP and WK3F-RFP tumori-genesis and metastasis in vivo were examined. Fig. 6 shows a tumor that formed in the tongue (A and E), its GFP and RFP excitation (B and F) and the submandibular lymph node metastasis (C and G) and its GFP and RFP excitation (B and F). Tumorigenesis (30%) and metastasis (33.3%) were observed in WK2-GFP. In contrast, tumorigenesis (50%) and metastasis (60%) were observed in WK3F-RFP (Table II). WK3F-RFP had higher tumorigenicity and metastatic ability than WK2-GFP.

Changes in WK2 and WK3F gene expression were comprehensively analyzed with the microarray method, using 50,599 gene probes to search for metastasis-related genes. A scatter plot of the signal values of WK2 and WK3F gene expression shows data with little dispersion and high reliability (Fig. 7). Genetic differences were recognized when increased expression with a Z-score of ≥2.0 and a ratio ≥5.0 or reduced expression with a Z-score of ≤−2.0 and a ratio of ≤0.2 were noted. Of these, we selected 15 upregulated and 15 downregulated candidate genes that are related to cancer or metastasis and whose functions are known (Fig. 8, Tables III and IV). Among these, the metastasis-related candidate genes with higher expression in WK3F than in WK2 were MAGEC1 (88.0-fold), which is a cancer antigen for malignant melanoma; matrix metallopeptidase 7 (MMP7) (18.6-fold), which is a matrix metalloproteinase; SNAI1 (6.6-fold), which is involved in the repression of E-cadherin transcription; and MACC1 (6.2-fold), which is related to the metastasis of colon cancer. High expression levels were observed in genes unrelated to metastasis, including EEF1A2 (60.1-fold) and BIRC3 (6.4-fold), which are related to anti-apoptosis; MAGEA11 (52.3-fold), which is a malignant melanoma antigen; and CLDN3 (9.3-fold) and PMP22 (6.3-fold), which are involved in cell adhesion.

Metastasis-related candidate genes with lower expression in WK3F than in WK2 were HTRA1 (0.012-fold), which regulates cell growth, and THBS2 (0.075-fold), which inhibits angiogenesis. Low expression levels were observed in genes unrelated to metastasis, such as CLCA2 (0.023-fold), which is related to cell adhesion.