The PTC cell line, BCPAP (harboring the BRAF mutation), was purchased from Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures GmbH (Braunschweig, Germany). In addition, the PTC cell line, K1 (harboring the BRAF mutation), was purchased from the Health Protection Agency Culture Collections (Salisbury, UK). BCPAP and K1 cells were cultured in RPMI 1640 medium. The PTC cell line, TPC-1 (harboring the RET/PTC mutation), was acquired from Dr Bryan R. Haugen of the Division of Endocrinology, Diabetes and Metabolism, University of Colorado Denver (Aurora, CO, USA) and cultured in high-glucose Dulbecco's modified Eagle's medium. All culture media were supplemented with 10% fetal bovine serum and antibiotics (100 U/ml penicillin and 100 µg/ml streptomycin) and cells were cultured in a humidified atmosphere containing 5% CO₂ at 37°C. All culture reagents were purchased from Life Technologies (Grand Island, NY, USA).

Cells (2×10⁵/ml) were seeded in a 12-well plate at 80% cell confluence, and stimulated with 20 ng/ml TNF-α (Invitrogen Life Technologies, Grand Island, NY, USA) and 50 U/ml IFN-γ (Roche Applied Sciences, New York, NY, USA) for 12 h, and then the culture medium was replaced with fresh medium. Cells treated only with medium were regarded as control groups. After 24 h, a scratch wound in the monolayer was created using a sterile 10 µl pipette tip. Phase contrast images were captured between 0 and 24 h using a DMi1 inverted microscope (Leica Microsystems, Wetzlar, Germany). Data are presented as the percentages of the remaining gap distance relative to the initial gap distance, and are expressed as the mean ± standard deviation (SD) measurements from three independent experiments.

Costar Transwell® chambers (pore size, 8 μm; Corning, Inc., Corning, NY, USA) were coated with 200 µl Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) at a 1:7 dilution and incubated overnight. The cells were co-cultured with 20 mg/ml TNF-α or 50 U/ml IFN-γ for 12 h, followed by incubation for 24 h in fresh culture medium. Next, the cells were seeded in the top chamber and medium containing 10% FBS was added to the lower chamber as a chemoattractant. After 24 h, the cells were fixed in 4% formaldehyde and stained with hematoxylin and eosin (Beyotime Institute of Biology, Suzhou, China). Cells that invaded through the pores to the lower surface of the filter were counted under a microscope (DMi1; Leica Microsystems). Data are expressed as the mean ± SD of triplicate measurements from three independent experiments.

Total RNA extraction, cDNA synthesis, and qPCR were performed as previously described (27). Briefly, total RNA was extracted from the cells using TRIzol reagent (Invitrogen Life Technologies) according to the manufacturer's instructions. RNA integrity was verified by 1.5% agarose gel electrophoresis, followed by staining with ethidium bromide (Sigma-Aldrich, St. Louis, MO, USA). The OD₂₆₀/OD₂₈₀ absorbance ratio (where OD is the optical density at 260 and 280 nm, respectively) was between 1.9 and 2.0 in each RNA sample. Next, 1 mg total RNA was used to prepare cDNA. A reverse transcriptase kit (PrimeScript™RT Reagent kit; Takara Biotechnology Co., Ltd., Dalian, China) was used for complementary DNA (cDNA) synthesis, at 37°C for 15 min, followed by 85°C for 5 sec, on an ABI 9700 GeneAmp® PCR system (Applied Biosystems Life Technologies, Grand Island, NY, USA). The transcripts were quantified using a Rotor-Gene 3000 Real-Time PCR system (Qiagen, Manchester, UK) and SYBR® Premix Ex Taq™ II (Takara Biotechnology Co., Ltd.), according to the manufacturer's instructions. Reactions began with a 10-sec hot-start activation of the Taq polymerase at 95°C, followed by 40–45 cycles of amplification in three steps (denaturation at 95°C for 5 sec, followed by 30-sec annealing at 60°C and 30-sec extension at 72°C). The primers used in each reaction were as follows (28): E-cadherin forward, 5′TGCCCAGAAAATGAAAAAGG-3, and reverse, 5′-GTGTATGTGGCAATGCGTTC-3′; N-cadherin forward, 5′-GAGAACTTTGCCGTTGAAGC-3′, and reverse, 5′-GTGTATGTGGCAATGCGTTC3′; vimentin forward, 5′-GAGAACTTTGCCGTTGAAGC-3′, and reverse, 5′-GCTTCCTGTAGGTGGCAATC-3′; GAPDH forward, 5′-ACCCAGAAGACTGTGGATGG-3′, and reverse, 5′-TCTAGACGGCAGGTCAGGTC-3′. Data are expressed as the mean ± SD of three independent experiments.

Cells (1×10⁵/ml) in 6-well plates were cultured until 60% confluence and then incubated with TNF-α (10 ng/ml, 20 ng/ml, 40 ng/ml) or IFN-γ (25 U/ml, 50 U/ml, 100 U/ml) for 36 h. Total cell extracts were prepared from the cell cultures using radioimmunoprecipitation assay buffer (Sigma-Aldrich) on ice. The extracts were centrifuged at 11,739 × g for 20 min, and the supernatant was subjected to a bicinchoninic acid assay for protein quantification. The samples were then boiled for 10 min at 100°C. Proteins were resolved on a 10% gradient sodium dodecyl sulfate-polyacrylamide gel and transferred to polyvinylidene difluoride membranes (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Subsequently, the membranes were blocked with 5% bovine serum albumin (BSA) in Tris-buffered saline/Tween 20 (TBST) for 2 h at room temperature, and then incubated with primary antibodies (dilution, 1:1,000 in 2.5% BSA/TBST). The following mouse monoclonal antibodies were used: Anti-E-cadherin (cat no. 610812, BD Biosciences); and anti-N-cadherin (cat no. sc-59987), anti-vimentin (cat no. sc-66002) and anti-β-actin (cat no. sc-47778) from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Next, the samples were incubated with the appropriate horseradish peroxidase-conjugated secondary antibody (dilution, 1:5,000 in TBST; Zhongshan Golden Bridge Biotechnology Co., Ltd., Beijing, China). The densitometry of the immunoblot analysis results was measured using the ImageJ software (version 1.4w3u; National Institutes of Health, Bethesda, MD, USA) and data from three independent experiments.

Statistical analyses were performed using SPSS software (version 18.0; SPSS Inc., Chicago, IL, USA). The results are represented as the mean ± SD. Differences between the groups were determined using one-way ANOVA. P<0.05 was considered to indicate a statistically significant difference.

In order to explore whether TNF-α and IFN-γ can contribute to the metastasis and invasion of PTC cells in vitro, a wound-healing assay was used to assess cell migration following stimulation with TNF-α and IFN-γ. Compared with the control groups, incubation of the three thyroid carcinoma cell lines (TPC-1, BCPAP and K1) with TNF-α (20 ng/ml) significantly enhanced cell motility by ~2-fold (P<0.01; Fig. 1). The migration ability of the cells under the stimulation of INF-γ (50 U/ml) was similar to this, the motility of K1 and TPC-1 was increased by ~2-fold and the migration ability of BCPAP was increased by ~1.5-fold compared with the control (P<0.01; Fig. 1). Similarly, significant changes in cell invasive ability were observed compared with the control (P<0.01; Fig. 2). Following incubation with TNF-α and IFN-γ, the number of invasive TPC-1 cells was increased by ~6-fold, in comparison to cells incubated without any cytokines. In addition, subsequent to incubation of BCPAP cells with TNF-α, the fraction of invasive cells increased by 13-fold compared with the control group, while it increased by 10-fold following incubation with IFN-γ. After incubation of K1 cells with TNF-α, the fraction of invasive cells increased by 10-fold compared with the control, while it increased by 6-fold following incubation with IFN-γ. These results suggest that TNF-α and IFN-γ are able to significantly promote the invasiveness of PTC cells.

TNF-α-induced EMT has been reported in numerous tumor types; however, EMT induced by IFN-γ alone has rarely been observed (11,22,23). RT-qPCR was used to measure the expression levels of epithelial cell marker, E-cadherin, and the mesenchymal cell markers, N-cadherin and vimentin, in epithelial cell lines, and their expression was used as a measure of EMT.

Epithelial cell lines treated with TNF-α and IFN-γ, alone or in combination, expressed lower levels of E-cadherin mRNA and higher levels of N-cadherin and vimentin mRNA compared with untreated cells (Fig. 3). In TNF-α-treated TPC-1 cells, RT-qPCR revealed that the mRNA expression of E-cadherin was decreased by ~60–70% at 12 and 24 h, while the mRNA expression of N-cadherin was increased by 2-fold at 12 h and 3-fold at 24 h; however, no evident increase was observed in the mRNA expression of vimentin. By contrast, when treated with IFN-γ, the mRNA expression of E-cadherin was decreased by ~20–40% at 12 h and ~70% at 24 h, while that of vimentin was increased by 1.5-fold at 12 and 24 h; however, no evident increase was observed in the mRNA expression of N-cadherin. In BCPAP cells, the observations were similar.

Incubation with low concentrations of TNF-α and IFN-γ was also found to exert a synergistic effect on E-cadherin in TCP-1 and BCPAP cells. TPC-1 cells incubated with a combination of 10 ng/ml TNF-α and 25 U/ml IFN-γ for 12 h expressed 67% more N-cadherin mRNA compared with cells incubated with 10 ng/ml TNF-α alone and 108% more N-cadherin mRNA compared with cells incubated with 25 U/ml IFN-γ alone (P<0.01). In addition, BCPAP cells incubated with a combination of 10 ng/ml TNF-α and 25 U/ml IFN-γ for 12 h presented a 6-fold increase in N-cadherin mRNA expression compared with cells incubated with 10 ng/ml TNF-α alone, and a 5-fold increase compared with cells incubated with 25 U/ml IFN-γ alone (P<0.01). Although TNF-α and IFN-γ were both able to induce downregulation of E-cadherin mRNA expression, their effect on N-cadherin and vimentin expression differed. TNF-α predominantly upregulated N-cadherin expression, while IFN-γ predominantly upregulated vimentin expression.

These results were further validated by immunoblot analysis (Fig. 4), which indicated that the expression of E-cadherin protein was decreased to varying levels in all three cell lines, when compared with the control groups. E-cadherin protein expression was reduced by 10–59% in TPC-1 cells, by 23–78% in BCPAP cells and by 28–78% in K1 cells. Consistent with the mRNA levels, TNF-α also predominantly induced N-cadherin protein, while IFN-γ predominantly induced vimentin, with detailed results shown in Fig. 4. However, when administered in higher concentrations (40 ng/ml TNF-α and 100 U/ml IFN-γ), the synergistic effect of these cytokines on EMT progression was less pronounced. Upon measuring the protein expression to evaluate the superposition effect of these two cytokines (10 ng/ml TNF-α and 25 U/ml IFN-γ), the EMT progression was unclear, potentially due to the excessively high concentration of cytokine used for an extended period (36 h), which may have induced apoptosis.