Two HNC cell lines originating from human oral cancer (MSKQLL1,SCC1483) were provided by Prof. Se-Heon Kim (Yonsei University, Korea). MSKQLL1 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEN): Nutrient Mixture F-12 (DMEM/F12; GIBCO, Carlsbad, CA, USA). whereas SCC1483 cells were cultured in Minimum Essential Medium (MEM; GIBCO, Carlsbad, CA, USA). The cells were grown in MEM with 10% fetal bovine serum and 1% penicillin-streptomycin, respectively. The cultured cells were maintained at 37℃ in a humidified atmosphere with 5% CO₂/95% air.

We used a micro-nozzle plasma jet system. The microplasma jet nozzle is made up of a plasma generation module and its cases. The schematic view of its nozzle is shown on Supplementary Fig. 1A (only online). The main components of plasma generation module are a Ni-Co alloy electrode, a glass insulator, and an electrode ring. The Ni-Co alloy electrode and glass insulator were fabricated simultaneously by micromachining technology. Supplementary Fig. 1B (only online) shows the fabrication process of Ni-Co alloy electrode. A Cr/Au film was deposited on a glass wafer as a seed layer for electroplating. The 105 holes were arrayed along concentric circles on the Ni-Co alloy electrode and glass insulator, and the diameter and the depth of holes were 300 µm and 100 µm, respectively. Cells were grown up to 90%– confluence and exposed to NTP (He, Ar, and N₂) 3 cm away from the nozzle for 1, 3, and 5 min. The gas flow rate was maintained at 8 L/min, and voltages were 15 KV and 20 kHz, respectively. We developed and produced NTP with a designed dielectric barrier discharge type to generate a homogenous NTP jet for biomedical research applications, as described previously.2122

Quantitative apoptotic cell death by plasma was detected using the Annexin V-propidium iodide (PI) apoptosis detection kit I according to the manufacturer's protocol (BD Biosciences, Bedford, MA, USA), as described previously.12 The cells were treated with various gas device (Ar, N₂, and He) NTP for 1, 3, and 5 min and then incubated further for 24 h. The cells were harvested, washed with cold phosphate-buffered saline (PBS), and stained with Annexin V-fluorescein isothiocyanate and PI at room temperature for 15 mins in the dark. The early and late apoptosis were quantified according to the manufacturer's instructions. Apoptosis was detected using a FACS Aria system (BD Biosciences), with the excitation and emission settings of 488 and 530 nm, respectively.

For wound healing assasy, inducible MSKQLL1 and SCC1483 cells were plated on 12-well plates and grown to confluency (>90%), followed by serum starvation for 24 h. Wounds were generated by using a sterile 200-mL pipette tip and washing with PBS. The cells were then exposed to various gases (N₂, Ar, and He) of NTP for 3 min. Each experiment was performed in triplicate. The wound on the captured image was automatically recognized and measured by Metamorph® NX image software (Molecular Devices, Sunnyvale, CA, USA) and eluate of crystal violet staining was measured under a light microscopy.

The invasion ability of each cancer cell line was evaluated using Transwell (24-well) chambers (pore size 8 µm, Costar, Cambridge, MA, USA), as described previously.7 Initially, type I collagen (8 µg/filter) was dissolved in 100 µL of medium and poured into the upper part of the polyethylene filter (pore size, 8 µm). The wells were coated overnight in a laminar flow hood. Then, 1×10⁴ cells (in 100 µL of culture medium) were added to the top of the filter in the upper transwell chamber. The chamber was incubated for 24 h in 5% CO₂ at 37℃. Finally, attached cells in the lower section (invading cells) were stained with hematoxylin and eosin (H&E) and counted in four representative fields by light microscopy (×200 magnification).

MMP-2/9 activity was examined using gelatin zymography as described previously.17 The cells were treated with He, Ar, or N₂ plasma for 5 min and incubated for an additional 24 h. The supernatant (100 µL) from each sample was mixed with 1 µL of 100 mM APMA (4-aminophenylmercuric acetate, Sigma-Aldrich, St. Louis, MO, USA) and the samples were activated for 1 h at 37℃ incubator. Next, each sample was placed in sample buffer (without mercaptoethanol) for 10 min and electrophoresed in sodium dodecyl sulfate (SDS) page gelatin gels at 125 V for 120 min at room temperature using a Novex Xcell II system (Thermo Scientific, Waltham, MA, USA). The gels were incubated in renaturation buffer for 60 min at room temperature, followed by incubation for 18 h in 100 mL of Novex zymogram developing buffer (Thermo Scientific) at 37℃ with orbital shaking. The gels were then stained for 3 h with Coomassie brilliant blue. After decolorization in 400 mL of methanol, 100 mL of acetic acid, and 500 mL of distilled water, images were taken using an image analyzer.

For Western blot, reduced protein samples were lysed by RIPA buffer containing 150 mM NaCl, 1.0% nonidet-P 40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris, pH 8.0, and protease inhibitor cocktail (Roche Applied Science, Vienna, Austria; pH 7.4), as described previously.12 The following antibodies were used for Western blot analysis: anti-focal adhesion kinase (FAK), -integrin, -Paxillin, and -α-tubulin (1:1000; Cell Signaling Technology, Danvers, MA, USA).

After culture on a microscope cover glass (Thermo Scientific), the cells were treated with He, Ar, and N₂ of NTP, and incubated for an additional 24 h. The cells were fixed with 4% formaldehyde and blocked in bovine serum albumin in 5% PBS for 45 min. Slides were then incubated with a polyclonal rabbit anti-p-FAK antibody (1:50; Cell Signaling Technology) for 2 h, washed with PBS and incubated with an Alexa 546-labeled goat anti-rabbit antibody (1:250; Molecular Probe, Eugene, Oregon, CA, USA) for 45 min. After rinsing in PBS, Hoechst 33258 (Molecular Probe) was added to slides for 15 min to counterstain nuclei. Slides were washed with PBS, mounted with Vectashield (Vector laboratories, Inc., Burlingame, CA, USA), and then visualized using a fluorescence microscope (Carl Zeiss, Oberkochen, Germany).

All data from three independent experiments are expressed as mean±SD. A p<0.05 was considered statistically significant (^*p<0.05; ^**p<0.01; ^***p<0.001).

The means for the different groups were compared using One-way analysis of variance, followed by post hoc Tukey's test, performed using SPSS 20.0 statistical software (SPSS Inc., Chicago, IL, USA).

To estimate various species produced from the plasma jet, we performed optical emission spectrum analyses. Optical emission spectroscopy was conducted from 280 nm to 920 nm wavelengths by optical emission spectroscope (SV 2100, K-MAC, Daejeon, Korea) (Fig. 1), and the optical emissions of three gases were compared: N₂, He, and Ar from a NTP jet at 5 L/min. The N₂ emission spectrum (Fig. 1A) was settled mostly by the presence of nitrogen (N) species, including N₂ second (290–410 nm), first positive systems (600–700 nm) and N₂⁺ first negative system (410–600 nm). Moreover, considerably reactive radicals were tied to oxygen, such as oxygen ions (O₂⁺) at 500–600 nm and weak atomic N at 747, 822, and 868 nm, were observed. The Ar and He NTP emission spectra (Fig. 1B and C) showed that the respective plasma species tied to argon and helium were superior. From Ar and He NTP in the emission spectra, we detected hydroxyl radicals and excited oxygen atoms. However, the intensities of oxygen atoms in Ar and He NTP were weaker than in N₂ NTP. These findings imply that N₂ NTP would be more effective for biomedical applications than He or Ar NTP.

It has been previously demonstrated that NTP induce apoptotic cancer cell death.12 Therefore, to estimate the effects of NTP on apoptotic cell death by different gas types, we measured annexin V-PI staining according to the processing time of each gas (Fig. 2). No significant SCC1483 and MSKQLL1 cell death was induced by 1 and 3 min NTP treatment with various gas sources (N₂, Ar, and He). However, significant cell death in both SCC1483 (Fig. 2A) and MSKQLL1 (Fig. 2B) cell lines was noted in all three gas plasmas after 5 min treatments.

Our previous studies showed that NTP treatement with the mixture of O₂ and He induces cellular morphological change and reduces tumor cell migration.3 To estimate the effect of NTP on cell migration in OSCC by gas type, we performed a wound-healing assay (Fig. 3): NTP treatment for 3 min was used, because NTP treatment for 5 min induced significant cancer cell death (Fig. 2). Thus, cells were exposed to N₂, He, and Ar NTP for 3 min and incubated for 24 h. Results showed that that NTP with all three gases significantly suppressed the migration of SCC1483 (Fig. 3A) and MSKQLL1 (Fig. 3B) cells across the denuded zone compared with control group. Furthermore, N₂ NTP significantly suppressed the migration of both cell lines compared with He and Ar NTP.

To elucidate the suppressive effect of NTP induced by each gas on cellular invasion, we performed an invasion assay (Fig. 4). Twenty-four hours after incubation, H&E-stained OSCC cells were placed on the undersurface of the membrane. NTP treatment reduced the number of invading cells compared with gasonly treatment. It should be noted that N₂ NTP among N₂, He, and Ar NTP showed the most deterrent effect on cell invasion (Fig. 4A and B). In order to understand the mechanism of suppression effect of NTP on tumor invasion in vitro, we performed the gelatin zymography. Fig. 4C shows visible changes in MMP-2/9 activity in NTP-treated SCC1483 and MSKQLL1 cells (Fig. 4C), with the most significant reduction by N₂ NTP among the gases.

FAK, integrin, and paxillin are well known proteins involved in cell migration and invasion.23 Thus, to elucidate the suppressive effect of NTP on migration, the expressions of T-FAK, p-FAK, integrin β3, t-paxillin, and p-paxillin were examined by western blot anlysis. As shown in Fig. 5A, the levels of p-FAK and integrin β3 in N₂ NTP treated cells were significantly lower than those in He and Ar NTP in both cell lines. Moreover, immunocytochemistry confirmed that intracellular distribution of p-FAK underwent the most significant alteration in N₂ NTP-treated SCC1483 and MSKQLL1 cells (Fig. 5B and C), respectively.