A total of 16 advanced stage treatment naive oral squamous cell carcinoma (OSCC) samples were collected from patients at the department of Head and Neck Oncology, Tata Memorial Centre (Mumbai, India; TMH and ACTREC biorepositories) between May and October 2016. The ages of the patients ranged between 33 and 70 years. Of the 16 patients, 15 were males and 1 was female. Patients with human immunodeficiency virus or hepatitis B virus infections were excluded from the study. Fresh tumor samples were removed and collected by surgical resection from advanced-stage treatment-naive oral cancer and were stored in sterile vials. These samples were transported on ice and were treated with a 10% povidone-iodine solution (Wokadine™) for disinfection, followed by washing with sterile PBS to remove any traces of the disinfectant.

The samples were minced into fine pieces (2–3 mm) by using surgical blades. These pieces were then placed on 35-mm (diameter) tissue culture plates (Thermo Fisher Scientific, Inc., Waltham, MA, USA) containing medium for explant culture. The medium contained minimum essential medium (MEM; Thermo Fisher Scientific, Inc.) supplemented with 10% horse serum (Thermo Fisher Scientific, Inc.) and 10% fetal bovine serum (HiMedia Laboratories, Mumbai, India). The explants were maintained at 37°C with 5% CO₂ atmosphere. The explant cultures were regularly monitored for cell growth. Excessive fibroblast growth was avoided by passaging with differential trypsinization, and keratinocytes were passaged when they reached 75–80% confluency.

The cell lines were trypsinized using 0.25% trypsin (EDTA) with glucose. The cell monolayers were washed twice with PBS and trypsinized for 2–3 min. Trypsin was neutralized using complete medium. The cells were frozen and stored in liquid nitrogen in medium containing 10% dimethyl sulfoxide (DMSO). The vials containing the cell suspension and DMSO were cooled to −80°C at a rate of 1°C/min prior to being transferred to liquid nitrogen.

Cell lines were inoculated in 60-mm wide tissue culture plates (cat. no. 353002; Corning Incorporated, Corning, NY, USA). The cells were incubated at 37°C at 5% CO₂ atmosphere for up to 96 h. The cells were counted at 24-h intervals. Doubling time was calculated using the formula: T.ln2/ln(Xe/Xb) where T=time of incubation, Xe=number of cells at the end of incubation time and Xb=number of cells at the beginning of the incubation time.

Immunocytochemistry was performed on all three cell lines using primary antibodies against keratin 8 (1:100; cat. no. C5301; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) and keratin 14 (1:250; cat. no. MCA890; Serotech; Bio-Rad Laboratories, Inc., Hercules, CA, USA). Briefly, the cells were fixed at −20°C for 20 min using absolute methanol and were permeabilized using a 0.3% solution of Triton in methanol. This was followed by blocking with 5% fetal bovine serum for 1 h at room temperature. Furthermore, the cells were incubated with the primary antibodies overnight at 4°C, followed by incubation with a fluorescein isothiocyanate-conjugated (1:400; cat. no. 115095003) and cyanine-3-conjugated (1:400; cat. no. 715165151; both Jackson ImmunoReseach Laboratories, Inc., West Grove, PA, USA) secondary antibody for 1 h at room temperature. Immunocytochemistry for CD44 was conducted using a mouse monoclonal antibody (1:600; cat. no. 156-3C11; Cell Signaling Technology, Inc., Danvers, MA, USA). The cells were fixed at room temperature with 4% formaldehyde for 20 min, permeabilized via 0.3% Triton X-100 in PBS, and blocked using 5% normal goat serum (cat. no. 005000121; Jackson ImmunoReseach Laboratories, Inc.) for 1 h at room temperature. Following blocking, the cells were incubated with the primary antibody overnight at 4°C. The cells were then treated with the cyanine-3-conjugated secondary antibody for 1 h at room temperature. The images were recorded using a confocal microscope at ×400 magnification.

Cells from all lines were harvested following trypsinization and resuspended in PBS containing RNase (cat. no. R5000; Sigma-Aldrich; Merck KGaA) (0.2 mg/ml) and propidium iodide (0.08 mg/ml) and incubated at 37°C for 30 min for DNA staining, and the DNA content was compared with that of human mononuclear cells from peripheral blood, (cat. no. 690PB-100A; Sigma-Aldrich; Merck KGaA), which served as a control for diploid human genomic DNA content. The cells were analyzed using a BD FACSCalibur system to determine their fluorescence. The data was analyzed by using ModFit 2.0 software (Verity Software House, Inc., Topsham, ME, USA). The mean channel of cells that were in the G₀ phase was divided by that of the lymphocytes to determine DNA index of each cell line. The DNA index was then used to predict ploidy number of the cells.

DNA was extracted from each cell line using a lysis buffer containing 0.4 M Tris (pH 8.0), 0.1 M NaCl, 0.005 M EDTA, 0.2% SDS and proteinase K. The isolated DNA was screened for HPV infection by performing polymerase chain reaction (PCR), using a KAPA Taq PCR kit (cat. no. KK1015; Sigma-Aldrich; Merck KGaA) with MY09 (5′-CGTCCMARRGGAWACTGATC-3′) and MY11 (5′-GCMCAGGGWCATAAYAATGG-3′) primers (27). The following thermocycling conditions were used: Initial denaturation at 95°C for 5 min was followed by 40 cycles of denaturation at 95°C for 1 min, annealing at 55°C for 1 min and elongation at 72°C for 1 min. At the end of 40 cycles, the final elongation was carried out at 72°C for 10 min. The MY09 and MY11 primers were used to amplify a 450-bp region in the HPV genome. DNA isolated from the HeLa cell line was obtained from Dr Manoj Mahimkar (ACTREC, Tata Memorial Center, Mumbai, India) to be used as the positive control for the HPV detection PCR. DNA isolated from MCF7 cell line (obtained from Dr Santosh Kumar, Bhaba Atomic Reseach Centre, Mumbai, India) was used as the negative control for the assay.

Chromosome analysis was performed by using a modification of the standard procedures used for karyotyping from fibroblast cultures (28). The metaphases tended to overspread. Therefore, the trypsinization time was reduced for G-banding. For cytogenetic analysis, 50 metaphase spreads were screened for ploidy in all three cell lines.

To evaluate the capability of the three oral cancer cell lines to produce spheroids in vitro, each cell line was subjected to the spheroid formation assay. In total, 5,000 cells were grown in complete MammoCult medium (MammoCult™ kit; basal human medium; cat. no. 05621; Stemcell Technologies, Inc., Vancouver, BC, Canada) in 6-well flat bottom ultra-low attachment plates (cat. no. 3471; Costar; Corning Incorporated) and were incubated at 37°C in a CO₂ incubator. Primary orospheres were obtained following 3–4 days of culture; these primary orospheres were further trypsinized and subjected to secondary spheroid culture to assess the ability of cells to form secondary orospheres.

To assess the tumorigenic potential of the three OSCC cell lines, 2 million cells of each cell line were resuspended in 200 µl culture medium supplemented with 25% Matrigel (cat. no. 356230; Corning Incorporated) and were xenografted subcutaneously into the NOD/SCID mice. Animal protocols were approved by the Institutional Animal Ethics Committee Advanced Centre for Treatment, Education and Research in Cancer (ACTREC), Tata Memorial Centre (Navi Mumbai, India). NOD/SCID mice for tumorigenesis assay were procured from the Institutional Animal Facility at ACTREC. For the tumorigenesis assay, 15 female NOD/SCID mice with 5 replicates (n=5) per group for ACOSC3, ACOSC4 and ACOSC16 were used. The mice were aged 6–8 weeks and weighed 20–22 g. The measurements of tumor diameters and volumes were recorded once per week using a Vernier caliper. Mice were sacrificed when the tumors volume reached a maximum diameter of ~20 mm. The dynamics of subcutaneous tumor growth were evaluated in ACOSC4 (n=5) and ACOSC16 (n=3) cell lines. The histograms of tumor volume were plotted by using GraphPad Prism version 5 software (GraphPad Software, Inc., La Jolla, CA, USA). Mice were maintained at a temperature of 22±2°C, humidity of 50±5%, and under positive air pressure with 12–15 air changes/h. A light/dark cycle of 12-h was maintained. Mice were provided food and water ad libitum.

STR profiles of the cell lines were determined using 16 STR markers. The cells of the three OSCC cell lines were centrifuged at 162 × g for 5 min, and their DNA was extracted using GenElute™ Mammalian Genomic DNA Miniprep Kit (cat. no. G1N350; Sigma-Aldrich; Merck KGaA). To protect the identity of the patients, only eight of these markers have been provided in the data. Data analysis was performed using GeneMapper^® ID-X software version 1.4 (Thermo Fisher Scientific, Inc.). The STR profiling was performed using PowerPlex^® 16 HS System (Promega Corporation, Madison, WI, USA). The profiles were compared to the DSMZ STR profiles database (29).

Tissue samples from advanced-stage treatment-naive patients with oral cancer were used for explant culture and establishing the cell lines (Table I). The processed tissue samples were inoculated in complete MEM, which was changed every 2 days. Over 2–3 weeks, the explants exhibited the growth of keratinocytes with a compact, clustered cell morphology and fibroblasts with an elongated morphology surrounding the explant tissues. The selective removal of the fibroblasts was performed by differential trypsinization to enrich the keratinocytes. During the initial passages, the growth of the cells was observed in discrete patches. Following several passages, the pure keratinocyte cultures were obtained from 3 different tissue samples, which were sub-cultured for >40 passages. Importantly, all three cell lines exhibited and maintained the polygonal cell shape, which is typical of epithelial cell morphology (Fig. 1). Furthermore, the rate of cell division of all three cell lines was measured by calculating their doubling time. The data indicated that the doubling times of the cell lines ACOSC3, ACOSC4, and ACOSC16 were 45.42, 33.78, and 34.00 h respectively. These results suggest that the established OSCC cell lines had long-term expansion potential and stable cell morphology.

Epithelial cells derived from the different epithelial tissues displayed the distinct expression pattern of the cytokeratin isoforms. In addition, cancer cells originating from epithelial tissues exhibited abnormal expression of different cytokeratins. Immunocytochemistry was performed on the OSCC cell lines using the keratins 14 and 8, which are expressed in the basal cells of the epithelium and transformed keratinocytes, respectively (30,31). The data indicated that the ACOSC4 cell line exhibited uniform and bright staining for keratin 14, whereas the cell lines ACOSC3 and ACOSC16 exhibited heterogeneity in keratin 14 expression; certain cells exhibited bright staining for keratin 14, whereas other cells were completely negative for keratin 14 (Fig. 2A). In addition, keratin 8 expression analysis indicated that the three OSCC cell lines expressed different levels of keratin 8. ACOSC3 cells had the brightest staining for keratin 8, and ACOSC16 cells exhibited only slightly brighter staining for keratin 8 than ACOSC4 cells (Fig. 2B). IFA was performed using CD44 (CSC marker), which was expressed in all three cell lines (Fig. 2C). Collectively, these data suggest that all three OSCC cell lines are derived from basal keratinocytes (epithelial origin), have abnormal expression of the tumor-associated protein keratin 8, and contain a CSC subpopulation.

Abnormal DNA content in tumor cells due to defective mitosis or amplification of different DNA fragments is a leading hallmark of cancer. DNA content in all three OSCC cell lines was evaluated by performing ploidy analysis. Normal human lymphocytes were used as a control for diploid DNA content. The DNA index number was calculated by dividing mean channel for the cells in the G₀ phase by that of the diploid lymphocytes. The DNA indices for the ACOSC3, ACOSC4, and ACOSC16 cell lines were calculated to be 1.360, 2.577 and 2.377, respectively (Fig. 3A-D). ACOSC3 cells had slightly higher than diploid DNA content, whereas both ACOSC4 and ACOSC16 cell lines exhibited slightly higher than tetraploid DNA content. These results indicated that all three cell lines have abnormal DNA content; which may be further investigated to identify the potential role of abnormal DNA content in the transformation of the cells.

Infection by HPV 16 and 18 has raised concern as a risk factor for oral cancers. Approximately 25% of oropharyngeal cancers are attributed to HPV 16 infection, whereas 1–3% of all oropharyngeal cancers are attributed to HPV 18 infection (32). HPV-positive cancers are associated with a more favorable prognosis as compared with HPV-negative cancers (33). To examine the HPV status in all three oral cancer cell lines, PCR was performed. DNA from the HeLa cell line was used as the positive control for HPV detection, whereas DNA isolated from the MCF7 cell line was used as the negative control. All three OSCC cell lines were found to be HPV negative by PCR using the MY09/MY11 primer sets (Fig. 3E).

To precisely determine ploidy and gain insight into the chromosomal abnormalities in the cell lines, karyotyping of all three cell lines was performed. None of the cell lines exhibited metaphases with a normal karyotype (Fig. 4).

The ACOSC3 cell line was aneuploid, with hyperdiploidy (64–80 chromosomes) and near-diploidy (40–41 chromosomes) in 70 and 30% of the metaphases, respectively. The loss of the Y chromosome and two or three distinctive clonal markers was observed. One of these was a large submetacentric marker with sharp, equally spaced bands, which suggested amplification. Gains of chromosomes 1, 2, 3, 12, 19 and X were observed, whereas consistent losses of chromosomes 4, 7, 8, 13, 14, 21, 22 and Y were observed. A derivative chromosome 16 with very lightly stained short arms, suggesting t(16;19), was also observed. Seven metaphases were karyotyped.

The ACOSC4 cell line was aneuploid, with hyperdiploidy (63–83 chromosomes) in 85% of the metaphases and near-diploidy or pseudodiploidy in ~15% of the metaphases. Clonal structural anomalies, including a derivative chromosome 1 with the loss of the ‘p’ arm and a probable unbalanced t(1;12)(p11;p11), a suspicion of isochromosome 9q, deletion 6q and a Robertsonian t(15;21) were detected. Five or six copies of chromosomes 9 and 14 were observed in some metaphases. Frequent losses of chromosomes 18 and 19 was also observed. Six metaphases were karyotyped.

ACOSC16 was aneuploid, with hyperdiploidy (55–71 chromosomes) and loss of the Y chromosome in all metaphases. Clonal structural anomalies included additional material on 1p, deletion 3p, isochromosome 9q, additional material on 14q32 (immunoglobulin heavy chain locus rearrangement) and 11q23 (MLL rearrangement) together with medium-sized markers. Loss of chromosomes 4, 5, 7, 8, 13, 17 and 21 was frequently observed. Five metaphases were karyotyped.

Cancer cells exhibit anchorage-independent cell growth. To evaluate the ability of these cells to form spheres in vitro, equal numbers (10,000) of cells were cultured on a low-adherence plate for 5 days. The data suggested that the ACOSC3, ACOSC4, and ACOSC16 cell lines formed compactly rounded spheres, with high sphere-formation efficiency (Fig. 5A). Furthermore, these primary spheres were analyzed for their ability to give rise to secondary spheres. The primary spheres were trypsinized to prepare a single-cell suspension and were cultured in the low-adherence plate. Notably, when grown for 3–4 days, all three cell lines produced secondary orospheres. Collectively, these data suggest that the cells of all three OSCC cell lines are anchorage-independent, which is enhanced in stem cell populations.

Furthermore, to investigate the in vivo tumor formation efficiency of these cells, the NOD/SCID mice were subcutaneously injected with 2×10⁶ cells of all three cell lines (ACOSC3, ACOSC4, and ACOSC16) with Matrigel. The results indicated that all three cell lines exhibited capability to give rise to tumors in immunocompromised mice. This experiment was performed in 5 replicates for each cell line, and all the mice injected with these cells exhibited tumor formation within 10–15 days. Each injection of the cells resulted in the formation of single, irregularly shaped tumors (Fig. 5B). Specifically, the ACOSC4 cell line was more aggressive as compared with ACOSC3 and ACOSC16 cell lines as it produced tumors in a relatively decreased duration of time (Fig. 6). These data confirm the tumorigenic potential of all three OSCC cell lines.

STR profiling was performed on all three OSCC cell lines to confirm that their novelty and genetic distinctness from previously established cell lines (Table II). The STR profiles of the three cell lines were compared with the DSMZ database. The profiles of the three cell lines did not exhibit any significant match with those of any previously established cell lines. The STR profiles of the ACOSC3, ACOSC4 and ACOSC16 cell lines were distinct, thus confirming their uniqueness and the absence of cross-contamination. Collectively, these data confirm the novelty of the cell lines and that they are derived from the previously unreported tissue samples. It was confirmed that the cell lines were derived from the tumors donated by patients by determining the STR profiles of the tumors and matching them with the profiles of their respective cell lines.