Differential Epigenetic Effects of Atmospheric Cold Plasma on MCF-7 and MDA-MB-231 Breast Cancer Cells

Cold atmospheric plasma (plasma) has emerged as a novel tool for a cancer treatment option, having been successfully applied to a few types of cancer cells, as well as tissues. However, to date, no studies have been performed to examine the effect of plasma on epigenetic alterations, including CpG methylation. In this study, the effects of plasma on DNA methylation changes in breast cancer cells were examined by treating cultured MCF-7 and MDA-MB-231 cells, representing estrogen-positive and estrogen-negative cancer cells, respectively, with plasma. A pyrosequencing analysis of Alu indicated that a specific CpG site was induced to be hypomethylated from 23.4 to 20.3% (p < 0.05) by plasma treatment in the estrogen-negative MDA-MB-231 cells only. A genome-wide methylation analysis identified “cellular movement, connective tissue development and function, tissue development” and “cell-to-cell signaling and interaction, cell death and survival, cellular development” as the top networks. Of the two cell types, the MDA-MB-231 cells underwent a higher rate of apoptosis and a decreased proliferation rate upon plasma treatment. Taken together, these results indicate that plasma induces epigenetic and cellular changes in a cell type-specific manner, suggesting that a careful screening of target cells and tissues is necessary for the potential application of plasma as a cancer treatment option.


Introduction
Non-thermal atmospheric pressure plasma is ionized media that contains a mixture of active particles, including electrons, ions, free radicals, reactive molecules and photons [1,2]. Part of this mixture consists of reactive oxygen and nitrogen species, such as ozone, superoxide, hydroxyl radicals, singlet oxygen, atomic oxygen, nitric oxide, nitrogen dioxide, nitrite, and nitrates [3,4].
Plasma has recently emerged in multiple medical applications, having been shown to be highly effective in wound healing and blood coagulation, as well as in the treatment of various diseases, including cancer [5,6]. For example, in ovarian cancer, chronic chemo-resistant purchased from the Korean Culture Type Collection (KCTC, Korea). MCF-10A and MCF-12A cells were grown in MEBM supplemented with MEGM Single Quots and cholera toxin (Lonza, Basel, Switzerland). MCF-7, MDA-MB-231, HCT-15, and NCI-H1299 cells were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum. The cultured cells grown to 50% confluence in 60-mm culture dishes were exposed to plasma for 30 sec 10 times in all experiment except for ROS detection wherein plasma was treated for 600 sec., at a 2.24 kV strength under argon gas using a DBD-type of plasma-producing device manufactured at the Plasma Bioscience Research Center (Kwangwoon University, Korea). The cells were then cultured for an additional 24 hr before harvest.

Apoptosis and cell proliferation assay
Cells treated with plasma in a 60-mm dish were harvested using 0.05% trypsin-EDTA (Gibco-BRL, Carlsbad, CA) and washed with PBS. To analyze apoptosis, 1x10 5 washed cells were treated with 5 μl of FITC Annexin V and 5 μl of propidium iodide (PI) using an FITC Annexin Apoptosis Detection Kit (BD Technologies, Franklin Lakes, NJ). To analyze proliferation, the washed cells were fixed in ice-cold 70% ethanol overnight. The cells were then treated with 50 μg/ml RNaseA (Sigma-Aldrich, St. Louis, MO) for 1 hr and stained with 50 μg/ml PI (Sigma-Aldrich) in the dark for 30 min at room temperature. Samples were analyzed using a FACS Canto II flow cytometer (BD Technologies) and read with a 488-nm laser. The cell proliferation index was calculated using the following formula: proliferation index = (S+G2+M)/ (G0/G1+S+G2+M) x 100%, where each letter represents the number of cells at each cell cycle stage. To monitor cell proliferation using a Cell Counting Kit-8 (Dojindo, Japan), 2 × 10 3 cells were plated onto a 96-well plate and cultured for 6 days followed by staining of the cells with WST-8. For colony formation assay, cells were seeded into 60-mm cell culture dishes with medium at a density of 1.3 × 10 4 cells/dish. Following treatment with plasma, cells were cultured for two weeks, fixed with acetic acid/methanol (1:7), and stained with 1% crystal violet.

Pyrosequencing
Pyrosequencing is a sequencing-by-synthesis method that quantitatively monitors the realtime incorporation of nucleotides through the enzymatic conversion of released pyrophosphate into a proportional light signal. After bisulfite treatment and PCR, the degree of methylation at each CpG position in a sequence is determined from the ratio of T and C [26]. One microgram of genomic DNA sample was sodium bisulfite-converted using the EZ DNA Methylation Kit (Zymo Research). The resulting DNA was used to determine the methylation of human LINE1 and Alu using quantitative pyrosequencing, as described previously using the PyroMark ID (Qiagen, Valencia, CA) [27]. Non-CpG cytosines served as internal controls to verify efficient bisulfite DNA conversion. The degree of methylation was expressed as the percentage of methylated cytosines over the sum of methylated and unmethylated cytosines. Sequencing was performed on five samples independently treated with plasma, and assays were performed a minimum of two times per sample.

Genome-wide methylation assay
Chromosomal DNA from cells grown in a 60-mm culture dish was isolated using the AllPrep DNA/RNA/miRNA Universal Kit (Qiagen) with a final elution in 50 μl of distilled water. Fifty ng of chromosomal DNA was applied to the Illumina Infinium Human Methylation 450 Beadchip to detect genome-wide methylation (> 485,000 CpG sites). A methylation index (β) was outputted for each site, which is a continuous variable ranging between 0 and 1 that represented the ratio of the intensity of the methylated-probe signal to the total locus signal intensity. A β value of 0 corresponds to no methylation, while a value of 1 corresponds to 100% methylation at the specific CpG locus measured. These values were then used to calculate a ratio of relative methylation between control and plasma-treated cells, with higher values corresponding to greater levels of methylation in plasma-treated relative to non-treated cells. All array data were uploaded to the Gene Expression Omnibus (GEO) database, and can be accessed via their website (http://www.ncbi.nlm.nih.gov/geo/; Series accession number GSE65087).

Pathway analysis
Functional categorization and pathway construction for the gene pool obtained from the genes affected by plasma treatment were performed using the Ingenuity Pathway Analysis (IPA) software tool (Ingenuity Systems, Redwood City, CA). P values for individual networks were obtained by comparing the likelihood of obtaining the same number of transcripts or greater in a random gene set as were actually present in the input file (i.e., the set of genes differentially methylated in non-treated versus plasma-treated groups) using Fisher's exact test, based on the hypergeometric distribution. The highest confidence functional network was designated as the top network.

Real-time reverse transcription-polymerase chain reaction (RT-PCR)
Total RNA from cells grown in a 60-mm culture dish was isolated using the AllPrep DNA/ RNA/miRNA Universal Kit (Qiagen) with a final elution in 50 μl of distilled water. Reverse transcription was conducted using 2 μg of total RNA with a reverse transcription kit (Toyobo, Japan). Gene expression was measured by real-time quantitative RT-PCR analysis using a Kapa SYBR Fast qPCR Kit (Kapa Biosystems, Woburn, MA) on an ABI 7300 instrument (Applied Biosystems, Foster City, CA). One microliter of cDNA was used for PCR, which was performed in duplicate. The primers used for the RT-PCR are listed in S1 Table. RNA quantity was normalized to GAPDH content, and gene expression was quantified according to the 2 -ΔCt method.

ROS detection assay
Cells were cultured in 96-well plates at 4 × 10 3 cells/well for 24 hr and treated with cold plasma for 600 sec. ROS generation was detected with DCFH-DA (Sigma-Aldrich) following the manufacturer's instructions. Briefly, after plasma treatment, cells were incubated with 1 mM of DCFH-DA for 30 min at 37°C in the dark. ROS were measured by an Infinite 200 Pro fluorescence plate reader (Tecan, Switzerland) at excitation and emission wavelengths of 485 and 535 nm, respectively. Assays were performed in triplicate.

Statistical analysis
In the microarray analysis, observations with adjusted p-values equal to or greater than 0.05 were removed and thus precluded from further analysis. Following adjustment, remaining genes were defined as differentially methylated if they displayed an increased or decreased methylation level equal to or higher than 0.2 compared to control. A student's t-test was used to detect differences in the mean levels of expression for selected genes. Statistical analyses were conducted using SPSS for Windows, release 17.0 (SPSS Inc., Chicago, IL). A p < 0.05 was considered to be statistically significant. The heatmap data were obtained using the Multiexperiment Viewer software (http://www.tm4.org/). A pool of genes that fit the top IPA network was submitted to the program.

Plasma-induced global hypomethylation of Alu in MDA-MB-231 cells
To examine the effect of plasma on the induction of epigenetic changes in breast cancer cells, cultured MCF-7 and MDA-MB-231 cells, representing estrogen-positive and estrogen-negative cancer cells, respectively, as well as the normal breast cell lines MCF-10A and MCF-12A, were treated with plasma. Two repetitive elements, LINE1 and Alu, were selected to monitor the global methylation changes induced by plasma treatment (Fig 1A). The pyrosequencing analysis indicated that Alu was hypomethylated at a CpG site, decreasing from 23.4 to 20.3%, by plasma treatment in the estrogen-negative MDA-MB-231 cells (p < 0.05) (Fig 1B and 1C). Two other CpG sites examined showed no significant changes. Also, neither the estrogen- positive MCF-7 nor the normal cell lines showed any significant methylation changes at all the examined CpG sites. In contrast to Alu, LINE1 did not show any significant methylation change throughout the CpG sites in the MDA-MB-231 cell (S1 Fig). These results indicate that the plasma acts in a cell type-specific, as well as a CpG site-specific, manner.
Cell-to-cell signaling and cancer-related pathways are involved in the network of methylation-altered genes by plasma To address the effect of plasma on the epigenomic profiles of the MCF-7 and MDA-MB-231 cells, genome-wide methylation changes were monitored through microarray analysis. Genes fitting our criteria of methylation level changes (i.e., β-values higher than 0.2 after plasma treatment) were selected for further analysis. Among the selected genes, only RASA3 appeared in both cell types (Δβ: 0.23, MCF-7; -0.20, MDA-MB-231), although the involved CpG sites were located in different parts of the gene. In the MCF-7 cells, 318 and 56 CpG sites were hyper-and hypomethylated, respectively, in the plasma-treated cells (Fig 2). The CpGs appeared throughout the chromosome with 151 being found in the coding region, 121 in the promoter region, and 102 in the intergenic region ( Fig 2B). The 374 CpG sites in the MCF-7 cells were examined for functional inter-relatedness using the Ingenuity Pathway Analysis software tool. The results indicated the "cellular movement, connective tissue development and function, tissue development" pathway as the top network and the "cancer"-involved pathway as the second top network, implying their roles in tumorigenesis ( Fig 3A, S2 Fig, and Table 1). Of note, genes regulated by, or regulating, NF-κB featured most prominently in the network. The transcripts displaying the most significant levels of altered methylation within this network were IL13RA2 (hypermethylated, Δβ = 0.27) and PELI2 (hypomethylated, Δβ = -0.37). IL13RA2 expression has previously been established to be upregulated in lung-metastatic breast cancer cells [28] and in brain tumors [29]. PELI2 comprises one of several families of E3 ubiquitin ligases that play roles in regulating innate and adaptive immune receptor-signaling pathways [30].
In the MDA-MB-231 cells, 76 and 63 CpG sites were found to be hyper-and hypomethylated, respectively, following plasma treatment. A total of 65 CpGs were in the coding region, 21 in the promoter region, and 53 in the intergenic region. IPA analysis for the 139 genes identified the "cell-to-cell signaling and interaction, cell death and survival, cellular development" pathway as the top network ( Fig 3B, S2 Fig, and Table 2). The transcripts displaying the most significant levels of altered methylation within this network were HLA-DQB1 (hypermethylated, Δβ = 0.21) and OPCML (hypomethylated, Δβ = -0.28). Upregulation of HLA-DQB1 has previously been reported in human esophageal squamous cell carcinoma [31] and in canine mammary tumor [32]. OPCML has a tumor-suppressor function in mammary and ovarian cancer, and epigenetic inactivation of the gene induces oncogenic transformation of ovarian surface epithelial cells [33,34].

Plasma induced an increase in apoptosis of the breast cancer cells
Based on the observation that plasma affected the methylation status of many genes involved in the "cell cycle", "cell-to-cell signaling and interaction", or "apoptosis pathway", the apoptosis and proliferation of the breast cancer and normal cell lines were examined by FACS analysis after treatment with plasma. For apoptosis, an increase in apoptosis was observed in both cancer cell lines following treatment with plasma, but the increased rate was higher in the estrogen-negative MDA-MB-231 cells (from 7.67 to 13.8%) than in the estrogen-positive MCF-7 cells (from 10.7 to 13.5%) (Fig 4A). Meanwhile, no significant change was observed in the MCF-10A cell. The apoptotic effect of plasma was further examined in two more cancer cell lines other than breast cancer. The lung cancer cell line, NCI-H1299, showed a high increase of apoptosis (from 7.13 to 10.2%) while the colon cancer cell line, HCT-15, showed a less significant increase (from 8.53 to 9.16%) by plasma ( S3 Fig). The effect of plasma on cell proliferation was examined using a cell proliferation assay kit as well as through colony forming assay and the results indicated that the least number of cells were observed in the plasma-treated MDA-MB-231 cell, confirming that this cell line is most vulnerable to plasma regarding apoptosis (Fig 4B and S4 Fig). All these results indicate that plasma induces epigenetic and cell activity changes in a cell type-specific manner, inducing hypomethylation of Alu and increasing cellular apoptosis, especially in the MDA-MB-231 cells.

Expression confirmation of the differentially methylated genes
Expression of selected hypermethylated or hypomethylated genes was determined by real-time RT-PCR for the plasma-treated MCF-7 and MDA-MB-231 cells used in the genome-wide methylation assay to check for consistency between mRNA levels and DNA methylation. 45 and 10 CpGs appeared within 200 bp upstream of the transcription start site in the plasmatreated MCF-7 and MDA-MB-231 cell, respectively (Fig 2B). Nine hypermethylated genes (ESR1, PRR7, CD86, FDX1, CREB3, DHRS7B, EIF1YA, BCL2, and BDNF) and two hypomethylated genes (DNAJC8 and POTED) were selected, of which CpG sites were located in the promoter region. As shown in Fig 5, all the genes showed up-or downregulation in accordance with the methylation status except for DHRS7B and EIF1YA, which were upregulated while the CpGs were hypermethylated.  Table 1. Genes in the top network displaying differential methylation in MCF-7 cells exposed to cold plasma.

Discussion
Altered methylation of Alu has been implicated in various cancer types, which may cause genomic instability, eventually leading to cancer of the cell. Across the four cell types examined in this study, only one specific CpG site in the estrogen-negative cancer cell line, MDA-MB-231, showed a significant change in methylation levels, being hypomethylated by the plasma treatment. The hypomethylation observed in this study is not considered to drive the cell into a more cancerous state because the plasma-treated cell showed anti-proliferation activities, as evidenced by the cell proliferation and apoptosis assays and the IPA pathway analysis. Specifically, among the MCF-10A, MCF-12A, MCF-7, and MDA-MB-231 cell lines, the MDA-MB-231 cells showed a remarkable increase of apoptosis by the plasma treatment. Also, it is notable that a significant methylation change in Alu was only observed in the MDA-MB-231 cells. The possibility that the other cell types were not properly exposed to the plasma can be excluded, as genomewide methylation changes were observed through the methylation chip analysis. Assuring sufficient exposure to the target cells is a crucial step before further examining the effect of plasma. A few molecular markers and cellular activities, including the production of hydrogen peroxide [2] and induction of apoptosis [35], are currently proposed. Alu can be used as an epigenetic marker, at least in limited cancer cells, including MDA-MB-231 cells. It seems that the plasma treatment utilized in this study did not induce wide-range methylation changes in terms of the number of affected sites because only 121 and 21 CpGs in the promoter region in MCF-7 and MDA-MB-231 cells showed significant changes. Meanwhile, other physical sources, such as γ-rays, have been reported to alter 442 CpGs in the promoter region in human bronchial epithelial cells [36]. In a previous study, ROS have been reported to relay signals inducing methylation changes at the CpG sites in a number of genes, including BCL-2, BDNF, p16, RUNX3, and E-cadherin [20]. Especially, BCL-2 (Δβ: 0.23) and BDNF (Δβ: 0.21) also appeared on the list of genes that have shown a significant methylation change by plasma treatment in the MCF-7 cells. Expression of the two genes was inversely correlated with their methylation status in the plasma-treated group, judged by RT-PCR (Fig 5A). We also observed an increase of ROS level in the MCF-7 and MDA-MB-231 cell by plasma treatment (Fig 6). Taken together, these results confirm that plasma provokes changes in cellular activities via ROS and eventually regulates gene activity through alteration of CpG methylation status. BCL-2 is an anti-apoptotic protein, originally identified as an oncogene at the chromosomal breakpoint of t(14;18) involved in human follicular B cell lymphomas [37]. BDNF has a critical role in tumorigenesis, promoting proliferation, differentiation, angiogenesis and invasiveness in several tumor types, including breast cancer [38]. It is notable that these genes were hypermethylated by plasma, possibly suppressing their oncogenic activities.
It should be mentioned that plasma and ROS shared little in common, as the majority of the genes altered by methylation appeared unique to each stimulus. The IPA pathway analysis revealed the "TGF-β Signaling" and "Calcium-induced T Lymphocyte Apoptosis" pathways as the highest probable canonical pathways in MCF-7 and MDA-MB-231 cells, respectively. TGF1, HoxC9, HLA-DQB1, and PRKCQ are representative genes involved in the pathways, showing altered methylation by plasma. In contrast, previous studies identified ATMdependent signaling [39] and ETM pathway [40] as ROS-linked pathways implying that plasma can also act on the cell by routes other than those of ROS.
A key limitation of our study design is the inability to assess methylation levels with respect to diverse plasma-treatment conditions. Thus far, information regarding both the types of plasma sources and applied exposure conditions is sparse, as the use of plasma treatment is still under development. In particular, choosing the appropriate exposure time of plasma treatment is crucial in observing the effects of plasma, as changing the strength or duration of plasma treatment causes different results. For example, HeLa and adipose-derived stem cells exhibited increased expression of apoptotic markers, including cleaved caspase-3, PARP, and phospho-p53 in proportion to the duration of the plasma treatment [14]. ROS, one of the mediators of plasma, also induces opposing cellular activities depending on its concentration [41]. Whereas low concentrations of ROS generally stimulate proliferation, high concentrations result in cell death. Low concentrations of ROS activated phosphorylation of ERKs, whereas high concentrations of ROS activated phosphorylation of JNKs. MLK3 directly phosphorylates JNKs and may control activation of ERKs [41]. Also in our studies, two independent plasma treatment conditions with different treatment schemes (one for a one-time 600 sec exposure and the other for 30 sec exposures a total of 10 times) caused changes in different cellular activities, with the latter causing increased cellular apoptosis (data not shown).
In summary, we present global, as well as genome-wide, methylation profiles of plasmatreated breast cancer cells, which we compiled using the most comprehensive methylation measurement techniques currently available. Also, the effects of plasma on cellular proliferation and apoptosis were presented. Our findings support the hypothesis that epigenetic dysregulation of crucial cancer-relevant molecules, including those pertinent to tumor development and apoptosis, may be involved in the effects of plasma. Further investigations into the mechanism(s) leading to differential methylation observed in this study, as well as the association of methylation status with cellular activities, may provide additional insights into the epigenetic role of plasma, along with determining its treatment options for cancerous cells and tissues.