Low temperature plasma promoting fibroblast proliferation by activating the NF-κB pathway and increasing cyclinD1 expression

The potential applications of low temperature plasma (LTP) in wound healing have aroused the concern of many researchers. In this study, an argon atmospheric pressure plasma jet was applied to generate LTP for treatment of murine fibroblast cell (L929) cultured in vitro to investigate the effect of NF-κB pathway on fibroblast proliferation. The results showed that, compared with the control, L929 cells treated with plasma for less than 20 s had significant increases of proliferation; the productions of intracellular ROS, O2 − and NO increased with prolongation of LTP treatment time; NF-κB pathway was activated by LTP in a proper dose range, and the expression of cyclinD1 in LTP-treated cells increased with the same trend as cell proliferation. After RNA interference to block p65 expression, with the same treatment time, RNAi-treated cells proliferated more slowly and expressed less cyclinD1 than normal cells. Furthermore, pretreatment with N-acetyl-L-cysteine (NAC) markedly prevented the plasma-induced changes in cells. In conclusion, the proliferation of L929 cells induced by LTP was closely related to NF-κB signaling pathway, which might be activated by appropriate level of intracellular ROS. These novel findings can provide some theoretical reference of LTP inducing cell proliferation and promoting wound healing.

when the treatment time was prolonged. Thirdly, we detected the expressions of phosphorylated NF-κB p65 (phospho-p65), IκB and translocation of phospho-p65 from cytoplasm into nucleus with Western blotting and immunofluorescence (IF), respectively. It was found that NF-κB pathway was activated by LTP within a proper dose range. Through analyzing the expression of cyclinD1 extracted from LTP-treated cells, we discovered that the changes of cyclinD1 expression had the same trend with cell proliferation. However, pretreatment with N-acetyl-L-cysteine (NAC) markedly prevented the plasma-induced changes described above in cells. Finally, when the NF-κB pathway was blocked with RNA interference, RNAi-treated cells proliferated more slowly and expressed less cyclinD1 than normal cells with the same treatment time. These findings will provide some beneficial support of LTP inducing cell proliferation and promoting wound healing.

Results
APPJ device and its optical emission spectra. The plasma source in argon was generated by a co-axial double ring electrodes configuration as described elsewhere 12 . The schematic diagram of the APPJ device is shown in Fig. 1. A hollow quartz tube was used as the barrier dielectric and had inner and outer diameters of 0.2 and 0.4 cm, respectively. The powered electrode and the grounded electrode were 1 cm-wide copper strips wrapped around the tube, and the distance between them was fixed at 1.65 cm. The distance between the grounded electrode and the quartz tube nozzle was 1 cm. The flow rate of working gas was set as 0.5 L/min. An intermediate frequency (~39.5 kHz) sinusoidal resonant power supply (CTP-2000K, Suman Electronics Co. Ltd., Nanjing, China) with a peak voltage of 7 kV was applied to the two electrodes to ignite the plasma discharges. The applied voltage was recorded by a 4-channel digital oscilloscope (Tektronix TDS 4034, Beaverton, OR) through a high voltage probe (Tektronix P6015A). Measured by an alcohol thermometer, the temperature of the plasma jet at the tip of the effluent varied from 25 to 36 °C, similar to room temperature. The emission spectra of the plasma jet were collected by a spectrometer (Mechelle 5000, Andor Technology Ltd., Belfast, UK) together with an intensified charge coupled device (ICCD) (iStar334, Andor Technology Ltd.).
The APPJ emission spectra in the wavelength range of 300-900 nm were captured and the identified lines are shown in Fig. 2. It can be seen that the mainly lines emitted by excited atoms of the feed gas argon were between 680 and 900 nm, indicating that the discharge was dominated by electron impact ionization. Peaks corresponding to N 2 were measured between 330 and 400 nm and were presumably present due to mixing of the feed gas argon with the surrounding ambient air along the plasma jet. Atomic oxygen line at 673 nm and OH radical at 309 nm were detected, presumably as a result of O 2 and H 2 O dissociation. Atomic oxygen and OH radical are highly reactive radicals that could play important roles in the potential biomedical applications  were detected with coumarin boronic acid (CBA) probe and shown in Fig. 3A. H 2 O 2 was assayed with and without the catalase by observing the fluorescence of the product, 7-hydroxycoumarin (COH) 17 . It could be seen that H 2 O 2 was produced in both PBS and DMEM obviously with the prolongation of LTP treatment time and the concentration in DMEM was lower than that in PBS of each groups (p < 0.05). Yan et al. found that the components in DMEM such as cysteine and methioine were the main factors that promoted degradation of reactive species including H 2 O 2 18 . Besides, to quantify nitrite/nitrate produced in the PBS and DMEM by LTP, 200 μl of PBS or DMEM were exposed to LTP for 0, 5, 10, 15, 20, 25 and 30 s and the amount of nitrite/nitrate were quantified by UV absorption spectroscopy, as described in the Material and Methods section. The results in Fig. 3B,C showed that a  time-dependent increase of each compound concentration, and the concentration of nitrite was lower than that of the nitrate. Normally, DMEM contains a certain amount of nitrate. Moreover the reductive components of DMEM including cysteine and serum maintain nitrite/nitrate in a low level. So a less than 10 s LTP did not cause a large change of nitrite/nitrate concentration. With the treatment time was more than 15 s, the accumulated nitrite/nitrate were sufficient to withstand the buffering of DMEM or PBS, resulting for the rapid increase of nitrite/nitrate (p < 0.05).

Intracellular ROS, O 2 − , and NO production in LTP-treated L929 cells. A series of fluorescence
probes were applied to characterize the intracellular oxidative production of LTP-treated L929 cells. The results demonstrated that all these three kinds of intracellular RONS increased gradually when the treatment time was longer than 5 s. The results were shown in Fig. 4A   group (37.02 ± 0.37) %. The ANOVA tests showed that the differences in the distributions of cell cycle among 10 s, 15 s and 0 s groups were statistically significant (p < 0.05). Besides, pretreatment cells with N-acetyl-L-cysteine (NAC), a ROS scavenger, significantly prevented the plasma-induced cell proliferation. (Fig. 5C).

LTP treatment activated NF-κB pathway.
Previous studies showed that NF-κB pathway could be activated by ROS and other stimulating factors 14 . In addition, LTP reacted with cell culture medium to form RONS and other activated species 7,19 . To determine whether LTP could activate L929 cells in vitro, phospho-p65 and IκBα, key molecules involved in the NF-κB pathway were extracted from LTP-treated cells and determined with Western blotting. Besides, the transfer of phospho-p65 from cytoplasm into nucleus was regarded as a significant symbol of NF-κB pathway being activated. So we applied IF staining to analyze phospho-p65 translocation in nucleus. The results indicated that an appropriate dose of LTP could activate NF-κB pathway in L929 cells in vitro. It could be seen in Fig. 6A, compared with control groups, in the 5 s, 10 s and 15 s treatment groups, the expression of phospho-p65 increased gradually (p < 0.05). However, it decreased with 30 s treatment (p < 0.05). When the treatment time was longer than 10 s, IκBα expression decreased compared with control (p < 0.05). Figure 6B exhibited the fluorescence images of LTP-treated cells stained with fluorescent dyes, in which nucleus was blue and phospho-p65 was red. These results showed that more phospho-p65 emerged in the nucleus and cytoplasm at 15 s treatment compared with 0 s and 30 s. Furthermore, 5 mM NAC was used to eliminate LTP induced ROS in cell culture medium, and then phosphor-p65 and IκBα were analyzed with western blotting. The results showed that the expression of phosphor-p65 had no obvious change with the treatment time prolongation compared with total p65. No variations were found in the expression of IκBα in each groups. It was indicated that LTP-induced ROS in medium were closely related to active NF-κB pathway (Fig. 6C).

CyclinD1 expression in LTP-treated L929 cells.
As a cell cycle related protein, cyclinD1 plays an important role in G1/S phase, and promotes cells from G1 phase into S phase and commences the DNA replication. We detected cyclinD1 expression with Western blotting, and discovered that the expression of cyclinD1 was related to LTP treatment time (Fig. 7A). CyclinD1 increased significantly from 5 s to 20 s and reached the peak at 15 s. The expression of cyclinD1 had the similar trend with cell viability after LTP treatment. Besides, cyclinD1 mRNA were detected by quantitative RT-PCR at the level of transcription and the results were shown in Fig. 7B. It could be seen that mRNA transcription level of the 15 s group increased 1.76-folds compared to control (p < 0.05). The expression of cyclinD1 in NAC pre-treatment cells was displayed in Fig. 6C. It could be seen that there were no obvious variation about cyclinD1 expression and mRNA transcription (Fig. 7B).  Cell proliferation, cell cycle distribution and cyclinD1 expression in p65si-transfetced L929 cells treated with LTP. In order to clarify the relationship between NF-κB signal pathway and LTP-induced proliferation, we attempted to construct a blocked NF-κB pathway cell model by RNA interference. We applied specific small interfering RNA (siRNA) to knock down the expression of NF-κB p65 subunit. Figure 8A showed the level of p65 mRNA transcription and protein expression. An inhibition rate (66.87 ± 8.52) % indicated that the blocked-NF-κB signal pathway cell model was constructed successfully.
The same plasma treatment protocol as mentioned previously was used to investigate whether LTP could induce p65siRNA-treated L929 cell proliferation or not. For p65siRNA-transfected cells, the results (Fig. 5A) revealed that cell viability significantly increased at 15 s (p < 0.05). However, cell cycle distribution in Fig. 8B showed that there were no statistically significant differences among the groups (p > 0.05). The data of cell proliferation of both normal and p65siRNA-transfected L929 cells were analyzed. With the same treatment time, the viability of p65siRNA-transfected cells was significantly lower than that of the normal cells at 5 s, 10 s and 15 s (p < 0.05). Though we treated these two types of cells for the same time, the cell proliferation of p65siRNA-transfected cells was significantly inhibited. These findings demonstrated that NF-κB pathway played a critical role in LTP-induced cell proliferation in vitro.  (Fig. 8C) showed that with 15 s LTP treatment, p65siRNA-transfected cells had 1.2-folds increase of cyclinD1 mRNA transcription compared with the control (p < 0.05). However, for normal cells, the transcription of cyclinD1 mRNA increased 1.8 folds compared with the control (Fig. 8C). Compared with normal cells (Fig. 8C), p65siRNA-transfected cells had cyclinD1 expression dropped by 45.61% (p < 0.05) with the same treatment time (15 s). These results suggested that LTP-induced cyclinD1 expression was closely related to NF-κB pathway in L929 cells in vitro.

Discussion
Fibroblasts play a key role in the proliferative and remodeling phases of wound healing. The main functions of fibroblasts include synthesis of extracellular matrix (ECM) molecules, release of multiple proteases, such as interstitial collagenases, gelatinase, matrilysin, et al., secretion of cytokines and contraction of wound 20,21 . What is noteworthy is that cytokines synthesized by fibroblasts, including bFGF, TGF-β, IL-1, IL-6, TNF-α, endothelin-1 and so on, not only stimulated proliferation of fibroblasts themselves, but also affected other cells in a paracrine fashion [22][23][24] . Our previous studies revealed that the number of fibroblasts in the tissues around the LTP-treated wound in mice significantly increased 12 . E Sysolyatina et al. discovered that LTP promoted fibroblast proliferation and suggested this effect might be related to cell autophagy 25 . Paola Brun et al. provided evidence that helium generated plasma promoted proliferation and migration in liver and intestinal fibroblast-like primary cells mainly by increasing intracellular ROS levels and elicited the PPAR-c anti-inflammatory molecular pathway 26 . In this study, we found a similar proliferative effect of LTP on L929 cells. By focusing on the role of NF-κB signaling pathway in LTP-treated cells, LTP activated NF-κB pathway and it association with cell proliferation were revealed.
RONS include superoxide anion (O 2 − ), hydroxyl radical (OH • ), hydrogen peroxide (H 2 O 2 ), nitric oxide (NO), peroxynitrite (ONOO − ), and so on, which derived not only from cellular metabolism but from external environment. It is important to maintain a dynamic balance of RONS in organisms 27 . Previous studies had proved that there was a dose-response relationship between RONS and cell functions. A normal level of RONS played important roles in organism signal transductions, including activating transcription factors (NF-κB) 14 , regulating phosphorylation and dephosphorylation of target proteins (MAPK), and maintaining calcium signaling channel 28,29 . However, higher level of RONS damaged nucleic acids, proteins, fatty acids and other molecules and induced cell apoptosis and even caused necrosis [30][31][32] . According to current researches, LTP-induced RONS were generated, transferred or interacted in culture medium. Furthermore, these extracellular RONS could enter cells possibly by diffusing and inducing new species of intracellular production 7 . O 2 − , with strong oxidization and very short half-life, dismutated to hydrogen peroxide (H 2 O 2 ) and oxygen by either enzyme SOD or no-enzymatic reaction system. Protonation of O 2 − produced hydroperoxyl radical (H 2 O − ), which subsequently caused lipid peroxidation and damaged biological membranes 33,34 . NO had diversified effects on cytostatics, chemotaxis, and vasodilation during early stage of wound healing, and it also regulated the proliferation, differentiation and apoptosis of cells involved in wound healing 35 . However, NO had negative effects: excessive or disordered production would cause a series of adverse outcomes including free radical damage, chronic inflammation et al. 36,37 . Our study observed that the productions of intracellular ROS, O 2 − and NO increased with prolongation of LTP treatment time. It may be the reason why the fibroblast proliferation, NF-κB pathway-related proteins, and cell cycle-related protein increased firstly and then decreased. It was reported that RONS produced extracellularly by plasma may immediately move across the cell membrane by active transport across the membrane, transient opening of pores in the membrane, then interact with amino acids and proteins and finally lead to production of long-lived reactive hydroperoxides or activation of pathways that modify intracellular ROS concentration 38 . In our present study, the intracellular RONS had no significant change, whereas cell proliferation, NF-κB pathway activation occured with 5 s LTP treatment. It could be explained that the intracellular RONS might mainly activate signaling pathway and promote cell proliferation, which led to RONS consumption. Therefore, the intracellular RONS in 5 s group could not accumulate effectively and were no statistically significant compared with control. Furthermore, in present study, NAC was added into cell culture medium to eliminate LTP-evoked ROS. No obvious variations of cell proliferation, cell cycle distribution, phospho-p65 and cyclinD1 expression were observed compared to normal cells. This stategy may provide evidence that LTP-induced ROS play the critical role in cell proliferation.
It should be noted that there is no way to accurately obtain qualitative and quantitative detection results of intracellular free radicals. As a result, the common fluorescent probes such as DHE, DCFH-DA and DAF-FM DA 39 were used in this study. These probes can only reflect the trend of free radicals variance but not quantify the concentration of free radicals.
Our study also demonstrated that LTP could activate NF-κB signaling pathway. NF-κB, as a familiar transcription factor, can respond to environmental stimuli with specific changes in gene expression. A wide variety of genes are regulated by NF-κB, including encoding cytokines, chemokines, adhesion molecules, acute phase proteins, inducible effector enzymes, antimicrobial peptides 13,40 . NF-κB forms homodimer and heterodimer with such five proteins: RelA (p65), c-Rel, RelB, NF-κB1 (p50), and NF-κB2 (p52), among which p65/p50 complex is the most abundant form and shows a great transcriptional activation. IκBs are bound with the nuclear localization signals (NLS) on NF-κB subunits. As a result, NF-κB is localized to the cytoplasm in an inactive form in unstimulated cells. As the major IκB protein, IκBα has been studied extensively. When cells are stimulated by cytokines, growth factor, oxidative stress, and other stimulus mentioned before, IκB kinases (IKKs) mediate IκBs phosphorylation, and then the protein is degradated by the ubiquitin and proteasome systems. With that an unmasked NLS p65/ p50 heterodimers would be phosphorylated by IKK and others on p65 subunits 13,41 . Activated NF-κB translocates to nucleus from cytoplasma and binds at κB sites in the DNA. Our study showed that NF-κB signaling pathway was activated through the decrease of IκBα, increase of phospho-p65 and translocation of phospho-p65 into nucleus after 24 hours of LTP treatment, as shown in Fig. 6.
ScIEntIfIc RepoRts | 7: 11698 | DOI:10.1038/s41598-017-12043-w NF-κB can be activated by oxidant in many types of cells. The activation methods mainly include ROS activating NF-κB signaling pathway by IκBα phosphorylation at serine 32 or 36 and phospherylation IκBα on tyrosine 42 or PEST domain of serine residues 42 For instance, hydrogen peroxide activated IKKs 43 , and singlet oxygen activates IκBα through an intact tyrosine 42 44 . Additionally, the phospherylation or dephospherylation and oxidation of thiol residues caused by ROS plays a critical role in ROS-induced transformation of the NF-κB signaling pathway 45 . In the previous studies, NO regulated NF-κB activation to inhibit the production of many proinflammatory factors. In normal human peripheral blood mononuclear cells, NO promoted cytoplasmic NF-κB proteins to translocate into nucleus 22 . Other studies with vascular endothelial cells revealed that NO could inhibit NF-κB target genes transcription through either protecting IκBα from proteolytic degradtion or upregulating transcription of IκBα gene 46 . In this study, we detected the crucial biomakers IκBα and phospho-p65 extracted from LTP-treated L929 cells, and the results demonstrated that LTP could activate NF-κB pathway at an appropriate dose. Besides, pretreatment with NAC markedly prevented the plasma-induced variations of cell viability, cell DNA synthesis, NF-κB pathway activation and cyclinD1 expression in cells. We presume that the mechanism is closely related with LTP-induced RONS productions.
As a DNA-binding protein, NF-κB can promote cells from G1 phase to S phase by enhancing expression of cyclinD1 47 . CyclinD1 plays an essential role in G1/S phase checkpoint and DNA synthesis. Normally it associates with cyclin-dependent kinases CDK4 and CDK6, and phosphorylates the retinoblastoma protein (Rb) to release transcription factor E2F in mid-G1 phase 48 . E2F regulates transcription of genes such as CDK1, cyclinA, and cyclinE. Our study revealed that the changes of cyclinD1 expression were similar to the tend of cell proliferation mentioned above.
However, cyclinD1 gene transcription is regulated by not only NF-κB, but also STAT5, Ets1, c-Jun, Otc1 and so on. To ealuate the hypothesis that LTP-induced cyclinD1 increases in connection with NF-κB pathway, we knocked down p65 expression, and subsequently treated these cells with LTP. The results in Fig. 8 showed that cells which blocked NF-κB pathway proliferated slowly with treatment time prolonged and more cells were arrested in G1 phase. With the same dose of LTP, cyclinD1 expression in cells which block NF-κB pathway were significantly decreased than normal cells. So we consider that NF-κB-regulated cyclinD1 expression plays an important role in LTP-promoted L929 cell proliferation in vitro.

Methods
Cell culture. L929 cells were cultured in DMEM with high glucose (GE healthcare) complemented with 10% fetal bovine serum (Biological Industries, Israel), 2mM L-glutamine (Amresco, USA) and 1% antibiotics (100U/ ml penicillin and 100 μg/ml streptomycin, Gibco, USA). The cells were maintained in a humidified conditions at 37 °C with 5% carbon dioxide incubator. LTP treatment procedure. Before LTP treatment, L929 cell suspension was seeded on 96-well plate with 200 μl per well at a concentration of 10 5 cells/ml, and then kept in carbon dioxide incubator overnight with the same conditions described before. Besides, for treatment with NAC (Beyotime, China) and plasma, 5 mM of NAC was added to the wells 4 hours prior to exposure to plasma. According to a series of preliminary experiments, we determined LTP exposure time to be 5 s, 10 s, 15 s, 20 s, 25 s and 30 s. Cell viability assay. Cell viability was determined with MTT assay. After LTP-treated cells were cultured for 44 h, 20 μL MTT (Sigma, USA) was added into each well, and the cells were incubated for another 4 h. The culture medium was discarded, and formazan crystals were dissolved into dimethyl sulphoxide (Sigma, USA). The A570 absorbance values were determined on microplate reader (Tecan Spectra, Austria).
Flow Cytometry analysis of cell cycle. After LTP-treated L929 cells were cultured for 48 h, all cultured medium is collected, and cells are digested with 0.25% trypsin (HyClone) without EDTA for 3 min. The digested cells and collected medium were centrifuged. The cells were washed with cold PBS twice and then fixed in 70% cold ethanol at least for 4 h. After incubation, cells were stained with propidium iodide (containing RNase1, Sigma, USA)), and then incubated in the dark at room temperature for 30 min. The samples were analyzed for DNA content profile with flow cytometry (Becton Dickinson, USA). The percentage of cells in G0/G1, S, and G2/M phases were assessed by ModFit LT software (Becton Dickinson, USA).
Immunofluorescence. LTP-treated cells in 96 well-plate were washed three times with PBS, and then fixed by 4% paraformaldehyde for 10 min, blocked with 10% goat serum at room temperature for 1 h, and incubated with anti-phospho-p65 antibody (1:50, Abcam) overnight. The primary antibody was removed and washed three times with PBS, incubated with CY3-labeled secondary antibody (Beyotime, China) for 1 h. Nuclei were labeled with 1 μg/ml DAPI (Roche, Swiss) for 5 min and washed twice with PBS. Cells were visualized with inverted fluorescence microscope (Nikon, Japan).