Delivery and expression of plasmid DNA into cells by a novel non-thermal plasma source

Abstract

Medical applications such as plasma assisted gene transfer is a minimally invasive approach that can substantially reduce potential discomfort of treated area. Atmospheric pressure plasma discharge is an effective approach to deliver plasmid DNA for in vitro and in vivo applications. We investigated plasma assisted delivery in vitro in mouse melanoma cells (B16F10) using a novel surface plasma device, which is operated in air. We evaluated the influence of applied voltage and distance between the surface device and cell monolayer. We found no significant effect on the viability of cells. Highest expression following delivery of a plasmid encoding green fluorescent protein was achieved with an applied voltage of 11.25 kV at a 2 mm distance and 5 second exposure time. To better understand the influence of oxidative damages and stress on cells after plasma delivery, a mRNA expression study was performed. Our results indicated that TNFα mRNA was significantly upregulated. The mRNA response may be attributed to the RONS generated by plasma; however, this mRNA upregulation was not adequate to be reflected in a coordinate protein upregulation. From the results reported here, it is clear that this novel plasma device could be used for plasmid delivery.

1. Introduction

Non-thermal plasma has been studied in a wide range of biological applications, for example bacterial decontamination, wound healing and ablation of cancer cells [1–6]. Non-thermal plasma can be operated at biological temperatures, which enables its use for biological treatments as well as in sterilization of heat sensitive objects. These plasmas can be produced by sources such as dielectric barrier discharge (DBD) and jets [7]. Plasma discharge depends on the feed gas (noble gases, oxygen, nitrogen, air) composition, which influences its chemical processes. Noble gases are often used to ignite plasma to reduce energy consumption. In contrast, air has an advantage due to accessibility and price. In addition, air gas plasma interacts with liquid biological samples producing reactive oxygen and nitrogen species (RONS) that can further modulate cellular pathways [8, 9]. Jets in particular have the advantage in delivery of plasma activated air into focused areas producing localized treatment.

In recent years, non-contact DNA delivery has been observed in diverse cell types employing multiple plasma devices, including noble gas or air jets and DBDs with various voltages, frequencies and irradiation exposure times [10–16]. Various cell types were exposed to plasma, mainly in medium or PBS for the protective buffering effect [10–13, 15–20]. Conversely, Edelblute et al. completely removed growth medium and added a small volume of plasmid DNA leaving the monolayer of cells exposed to plasma [14]. Plasma could be used to deliver plasmid DNA into cells with a minimal reduction of viability [14]. Moreover, plasma can successfully introduce plasmids into cells in vivo. Connolly et al. and Edelblute et al. observed higher expression of a plasmid encoding luciferase after plasma exposure in mice [21–24].

Plasma produces electric fields, UV radiation, charged particles and direct chemical reactions of neutral reactive species [25]. These effects may contribute to gene delivery. Recent studies have reported that the production of RONS are potentially involved in plasma delivery [14, 15, 18]. However, other factors generated by plasma may also contribute to gene delivery.

In this work, we present a novel plasma source capable of delivering a plasmid encoding green fluorescent protein (GFP) into a murine melanoma cell monolayer. The influence of the plasma source on viability and damage was investigated as well as induced gene expression was quantified.

2. Materials and methods

2.1. Plasma device

The experimental design of the reactor for generating plasma discharge was similar to the work of Malik et al. [26]. The electrodes were 50 μm thick aluminum strips. The scheme of an electrode setup, the electrode dimensions and a gap between them is shown in Fig. 1A. The electrode layer was attached to a glass slide (25 × 75 × 1 mm) and the slide was attached to the reactor with adjustable screws, which allowed changing of the distance of the electrodes from the sample. The reactor operated in the surrounding air at atmospheric pressure.

Fig. 1.

Plasma Device.

A) Drawing showing the dimensions of the plasma device used in the reported experiments; B) Voltage wave form for peak voltage 11.25 kV and 6.4 kV; C) Time integrated images of the plasma at the applied voltages of 6.4 kV and 11.25 kV.

A compact Pulsed Power Modulator MPC3000S-0P1 (Suematsu Electronics Co., Ltd., Japan) was used to deliver high voltage pulses of 11.25 or 6.4 kV for particular experiments. The power supply provided pulses of repetition frequency 500 Hz. A Tektronix P6015A voltage probe and a Tektronix TDS 2024C oscilloscope were used to measure the development of voltage. A typical waveform of the applied voltage signal is presented in Fig. 1B.

We used the device to produce discharge at two voltage settings, 11.25 kV and 6.4 kV, the images show top view of plasma generated at electrodes (Fig. 1C). Multiple photographic images were taken over time and integrated into a single photo (time integrated images). The time integrated images were taken while maintaining the identical exposure time for each experimental condition.

2.2. Cell culture conditions

Prior to seeding, glass slides were treated by poly-L-lysine solution (0.01%) to facilitate cell attachment. B16F10 melanoma cells (CRL-6475, American Type Culture Collection, Manassas, VA, USA) grown in a complete growth medium of McCoy’s 5A (Iwakata and Grace Modified) consisting 10% heat-inactivated FBS and 0.5% gentamicin were then seeded onto the slides. The initial concentration was 2.5×10⁵ cells/ml in medium and 50 μl was placed on each slide followed by 24 hours of incubation at a temperature of 37°C and 5% CO₂ concentration. This seeding density achieved a sub-confluent state of approximately 70% for the experiment.

2.3. Plasma treatment

Prior to plasma exposure, the entire volume of growth medium was aspirated. One hundred μg gWizGFP (Aldevron, Fargo, ND, USA), was carefully layered over the cells in 100 μl of phosphate buffered saline (PBS). Cells were exposed to the plasma at a 2 mm or 5 mm distance for 5 seconds (sec) for each experiment. Immediately after plasma irradiation, 5 ml of complete growth medium was layered over the cells, and the slides were placed in the incubator for an additional 24 hours.

We also investigated the influence of the electric field on plasma-induced GFP expression. The glass slide was flipped over so other processes such as exposure to reactive species could not reach the target cells. For these experiments, the distance between cell surface and plasma was maintained at 2 mm at a voltage of 11.25 kV.

2.4. Viability

Propidium iodide (PI) stains only dead cells, whereas Hoechst 33342 (Thermo Fisher Scientific, Waltham, MA, USA) is a nuclear stain for both live and dead cells. Twenty-four hours after plasma exposure and incubation in the dark at 37°C for 15 minutes, the medium was replaced with complete growth medium containing 10 μg/ml Hoechst and 4 μg/ml propidium iodide (Sigma-Aldrich, St. Louis, MO, USA). After incubation, cells were washed with PBS and cells were observed with a fluorescence microscope (Olympus IX-71, Center Valley, PA, USA) to determine their condition. A mercury lamp and DAPI and TRITC optical filters were used. Images were acquired homogenously through the glass slide utilizing a grid of three horizontal and five vertical images making a set of fifteen images. The areas of cell death were analyzed by ImageJ software [27]. The percentage of dead cells was calculated as the ratio of PI area to Hoechst area multiplied by 100.

2.5. GFP protein expression

The technique of evaluation of GFP positive cells was identical to the measurement of viability, except that DAPI and FITC filters were used. The area of GFP fluorescent cells was normalized to the surface of the treated area as determined by Hoechst staining for each image. The percentage of GFP positive cells was calculated as the ratio of GFP-stained area to Hoechst area multiplied by 100.

2.6. RNA isolation and cDNA quantification by real-time PCR

RNA was collected 24 hours after plasma delivery. Before RNA collection, the slide was carefully dipped in PBS in order to remove growth medium. RNA was extracted from the cells using silica membrane columns (RNeasy Mini Kit, Qiagen, Germantown, MD, USA). The RNA concentration in each sample was normalized after quantification using a NanoVue spectrophotometer (GE Healthcare, Chicago, IL, USA). Complementary DNA was reverse transcribed from RNA using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Waltham, MA, USA) following manufacturer’s instructions. Quantitative real-time PCR (qPCR) analysis was performed in 20 μl reactions containing PowerUp SYBR Green Master Mix (Applied Biosystems), 200 nmol of each primer, and 5 μl of 1:10 diluted cDNA. The PCR cycling parameters were five minutes at 95 °C, followed by 35 cycles of 95 °C for 15 sec and 62 °C for 60 sec (CFX96 Real Time PCR Detection System, Bio-Rad, Hercules, CA, USA). Primers were purchased from IDT DNA (Coralville, IA, USA); sequences are listed in Table 1. Each sample was analysed in triplicate and mean Ct values were calculated. The relative expression levels were calculated using the ddCt method [28]. Endogenous values were normalized to the levels of the reference genes GAPDH and GUS. Plasma treated cells with gWizGFP plasmid were normalized to non-treated cells exposed to gWizGFP plasmid only. GFP expression was normalized to detection of the CMV promoter sequence.

Table 1.

PCR targets and primers used in this study.

Gene	Primer	Sequence
Glutathione reductase, GSR	forward	CAC CGA GGA ACT GGA GAA TG
	reverse	ATC TGG AAT CAT GGT CGT GG
Superoxide dismutase 1, SOD1	forward	TGT GTC CAT TGA AGA TCG TGT G
	reverse	TTC CAG CAT TTC CAG TCT TTG
Superoxide dismutase 2, SOD2	forward	TGC TCT AAT CAG GAC CCA TTG
	reverse	CAT TCT CCC AGT TGA TTA CAT TCC
Superoxide dismutase 3, SOD3	forward	TGT CAC CAT GTC AAA TCC AGG
	reverse	ATC CAG ATC TCC AGC ACT TTG
Tumor Necrosis Factor, TNFα	forward	CCC TCC AGA AAA GAC ACC ATG
	reverse	GTC TGG GCC ATA GAA CTG ATG
Glyceraldehyde 3-phosphate dehydrogenase, GAPD	forward	TTC ACC ACC ATG GAG AAG GC
	reverse	GGC ATG GAC TGT GGT CAT GA
β-glucuronidase, GUS	forward	TCG CCG ACT TCA TGA CGA A
	reverse	GCT GTC TCT GGC GAG TGA AGA
Green fluorescent protein, GFP	forward	CTCTGTGCTATGGTGTTCAATG
	reverse	TGTCTTGTAGTTGCCGTCATC
Cytomegalovirus promoter region, CMV	forward	ATCATATGCCAAGTACGCCC
	reverse	TGAGTCAAACCGCTATCCAC

2.7. ELISA assay

Cells were lysed in Mammalian Protein Extract Buffer (GE Healthcare Life Sciences). Followed by protein quantification by the Pierce^™ BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) to normalize the protein concentration. TNFα protein levels were measured by ELISA (Mouse TNFαELISA Kit, Invitrogen, Carlsbad, CA, USA).

3. Results

3.1. Plasma GFP gene transfer

Gene transfer efficiency, viability and gene expression were evaluated after plasma discharge associated gene transfer to B16F10 cells. Initial experiments focused on viability. Variables examined were distance of plasma source from treated glass slide (2 and 5 mm) and on applied voltage (11.25 and 6.4 kV) using PI staining as an indicator for dead cells and Hoechst for nuclear staining. Cells were exposed to plasma for 5 sec for all experiments. One day after plasma exposure, we observed that there was a negligible change in viability, approximately 99% of cells not exposed to plasma were observed to be PI negative, the lowest viability was found using the condition of 5 sec, 2 mm and 11.25 kV resulting in approximately 98% of cells being PI negative. The difference in viability between treated samples and untreated sample was not significant (data not shown). We then examined the same voltage and distance conditions as well as the same time point after treatment to determine the plasmid delivery conditions producing the highest transfection efficiency. Delivery of gWizGFP plasmid to cells was compared to the inoculation of plasmid DNA alone. Figure 2A shows representative images of cells exposed to plasmid and plasma and cells inoculated with plasmid only. All cells were stained by Hoechst and propidium iodide and examined for GFP expression. It is clear from the merged images that GFP expression is present within cells. In Fig. 2B, we further evaluated the results following plasma treatment and observed significant elevation in GFP positive cells at two conditions; 2.4 and 0.2% for 11.25 kV and 6.4 kV respectively, both for a distance of 2 mm between cellular surface and plasma device. When the distance was 5 mm, the expression was negligible for both chosen voltages (5 mm, 11.25 kV-0.05%; 5 sec, 5 mm, 6.4 kV-0.02%). The efficiency of gene transfer was evaluated using 15 images obtained from the glass slide containing the treated cells. In Fig. 2C, the representative stitched image shows GFP expressing cells primarily under the high voltage (HV) electrode. The delivery conditions producing the highest expression were 5 sec, 2 mm and 11.25 kV

Fig 2.

Delivery of plasmid encoding GFP.

A) Microscopic visualization of cells expressing GFP after plasma delivery at 2mm, and 11.25 kV. Observed are was under the electrode. From left to right representative images were taken with DAPI, TRITC, FITC filters and final image is a merged image of all three and indicates GFP positive cells (green color), nuclear stained cells (blue color) and dead cells (red color). Magnification is 200 X and scale bar represents 200μm. B) Glass slides with B16F10 cells were treated using different plasma conditions. The percentage of GFP positive cells was calculated as the ratio of GFP cells (FITC channel) and Hoechst stained cells (DAPI channel). Statistical significance is indicated by askerisks determined using student’s t-test: *p<P<0.05, **p<P<0.01. The standard error of the mean (n≥3) is shown. C) After plasma delivery at 2 mm and 11.25 kV, this stitched image composed of 15 individual images acquired in FITC filter from one glass slide demonstrates the distribution of GFP positive cells across the slide. The percentage of GFP positive cells was calculated as the image.

3.2. The contribution of electric field to plasma assisted gene transfer

A potential contributor to gene delivery is the electric field. To evaluate this possibility, we explored the influence of a plasma generated electric field by comparing samples with plasmid inoculated only, plasmid innoculated and exposed to complete plasma (synergy of all generated processes, i.e. reactive species, charged particles, UV irradiation, electric field) and plasmid innoculated and exposed to plasma generated electric field protected from any other plasma generated processes to cells by reversing the slide (Fig 3). GFP expression was elevated for both conditions compared to plasmid DNA only control (2.4% and 0.5% for synergy of plasma generated processes and plasma generated electric field only, respectively).

Fig 3.

The influence of electric field to plasma assisted DNA delivery.

Statistical significance is indicated as askerisks determined by student’ s t-test: **p<P<0.01. The standard error of the mean is shown (n≥3).

3.3. ROS and damage induced genes influenced by plasma

Although we observed that the viability of cells was unaffected, we studied the influence of plasma exposure on the mRNA levels of five genes responding to oxidative stress [8]. Superoxide dismutases are major antioxidant enzymes and may be located throughout the cell (SOD1), in the mitochondrial matrix (SOD2) or in the plasma membrane (SOD3) [29]. Glutathione reductase (GSR) maintains reduced glutathione, important in ROS control [30]. Finally, proinflammatory cytokines such as TNFα respond to ROS exposure [31–33].

After 4 hours, the mRNA levels of TNFα increased by 5.5-fold compared to a non-treated control group (Fig. 4A). These levels returned to background 24 hours after the experiment (data not shown). However, TNFα protein was not found out by protein analysis in any sample (data not shown); protein levels remained below the detection limit of the ELISA. Gene expression was further studied on cells inoculated with gWizGFP plasmid followed by plasma exposure (Fig 4B). In these experiments, SOD2 mRNA levels were minimally but significantly elevated compared to a non-treated control group.

Fig. 4.

mRNA expression levels following exposure to plasma

A) Expression of mRNA levels influenced by ROS or potential cell damage by plasma discharge. B) Expression of genes influenced by ROS and damage influenced by plasma in B16F10 cells innoculated by gWizGFP plasmid. Statistical significance compared to naïve control cells is indicated as askerisks determined by student’s t-test: *p<P<0.05, **p<P<0.01. Shown is the standard error of the mean (n≥4).

4. Discussion

The objective of this study was to investigate the capacity of a novel surface plasma device to deliver plasmid DNA to cells. One advantage of plasma for this application is that the device is not in direct contact with treated surface. This study was carried out using gWizGFP plasmid for the ability to fluoresce, enabling straightforward evaluation. The highest gene expression after delivery with this plasma was observed using a 2 mm distance between the device and the cells, a plasma exposure time of 5 sec and an applied voltage of 11.25 kV. Under these conditions, damage to cells was negligible. Fig. 2B shows that 2.4% of cells in a monolayer on a glass slide expressed GFP after delivery. We observed that specific areas were expressing higher levels of GFP as was shown in the representative image (Fig 2A), where the plasmid DNA transfer to cell efficiency was 18% within an area of 2.2 ×1.6 mm (3.52 mm²) of microscopic field.

The highest efficiency of plasma mediated plasmid delivery observed in our study was comparable to that of the work of Sakai et al. [11]. In a comparison of plasmid DNA transfer efficiency using arc plasma, a plasma jet, DBD plasma and microplasma, Jinno, et al. observed efficiency from 2.6–5.5% within an area of 0.5×0.5 mm depending on the specific plasma source; however, the efficiency on the area comparable to our studied area (2.21 × 1.67 mm²) was 0.1–7.9% [16]. Xu et al. reported 70–90% efficiency for DNA-FITC oligomeric single stranded DNA, smaller than the plasmid DNA used in our experiments [15]. However, many parameters varied between these studies, including cell line, cellular division or distribution of cells and as well as the gas admixture, geometry of electrodes and discharge power. Taking this into consideration, the process of plasma mediated plasmid delivery was problematic to compare.

We detected a significantly elevated level of GFP protein expression four hours after plasma treatment, suggesting that the cells successfully transcribed the plasmid soon after transfer. The transgene expression kinetics may be similar to those observed after gene electrotransfer [34]. Increased levels of GFP mRNA were detected two hours after electrotransfer, peaking at six hours. Another reporter protein, luciferase, was detected four hours after electrotransfer.

One interesting observation was the localization of plasmid DNA delivery to cells under the HV electrode, which can be seen in Fig. 2C. A previous study demonstrated that electric field intensity is highest in the area of the HV electrode [35]. The structure of the electrode configuration enhanced the ignition of streamers from the edges of the HV electrode to the grounded electrode. Our observation supported the hypothesis that the induction of electric field by plasma can be a contributor to the plasmid delivery process. In addition, Jinno et al., suggested that the generated electric field does contribute to plasma mediated DNA delivery [18]. Hence, we studied the effect of electric field in more detail. We eliminated the plasma generated processes except for the electric field by simple mechanical protection of the cellular monolayer by glass. However, while there was a small increase in expression related to the electric field, the applied electric field alone did not significantly increase GFP expression (Fig. 3).

Edelblute et al. [14] and Xu et al. [15] proposed that the electric field is not involved in the process; each study concluded that the production of reactive species was the main process supporting plasma DNA delivery. RONS production by plasmas has tremendous interest in many fields of plasma interactions as important agents influencing many biological processes [8]. Malik et al. demonstrated that a discharge device similar to ours produced nitric oxide and ozone [35]. Our results support the concept that RONS together with electric field and other unknown plasma components are involved.

A crucial criterion for further development of a suitable system for gene delivery is viability. Generally, plasma can be used in cancer cell eradication both in vitro and in vivo [36]. Cancer cells in particular exhibit lower viability after plasma exposition [5, 37, 38]. We did not observe significant changes in cell viability. Thus, we postulate that the plasma device used in our study operated under the applied voltage conditions is not cytotoxic in accordance with several previous reports [11, 14–17, 24]. It is fundamental to establish proper plasma conditions that provide the desired effect with respect to cell damage and viability.

We performed mRNA expression study of genes or proteins influenced by oxidative damages and stress to better understand the cell conditions after plasma delivery. It is well accepted that plasma produces RONS which can influence gene expression and modulate signaling pathways [33, 39]. An increased RONS concentration can both contribute to disease development or have a role in the immune system as a signaling molecule.

In this study, we evaluated the mRNA expression levels of SOD1, SOD2, SOD3, GSR and TNFα. Our results indicated that TNFα mRNA was significantly upregulated, confirming previous studies [32, 33]. The mRNA response may be attributed to the RONS generated by plasma; however, this mRNA upregulation was not adequate to be reflected in a coordinate protein upregulation. Moreover, previous studies have suggested that cancer cells may be more susceptible to killing following exposure to RONS at the concentration induced by plasma [37, 40]. Nevertheless, the effect of oxidative species does not necessarily mean damage to cells. Oxidative species may help in the temporal disruption of membranes and thus molecules diffuse into the cell more effectively [41]. Vernier et al. showed that pretreatment of electroporated cells with hydrogen peroxide helped increase efficiency of plasmid transfer while hydrogen peroxide alone did not [42]. Thus, we hypothesize that the effect of plasma generated reactive species can support the process of plasma mediated delivery.

5. Conclusions

In the study reported here, we demonstrated that a novel plasma device could be used for plasmid delivery resulting in reporter protein expression. Under the plasma parameters used in this study, delivery could be performed without an adverse effect on cell viability. With this plasma device we found that the highest gene expression could be obtained with an exposure time of 5 sec and an applied voltage of 11.25 kV and a 2 mm distance between the device and cells. With these plasma conditions there was no significant effect on viability. We determined that the applied electric field generated by plasma did result in a small increase in expression that did not reach the level of significance, although RONS production contributed to a greater extent and resulted in a significant increase. In addition, we observed elevated expression of oxidative damage and stress mRNAs as a consequence of plasma treatment during plasmid delivery.

HIGHLIGHTS.

• Non-thermal plasma to deliver plasmid DNA

• Non-contact plasmid DNA delivery

• High viability approach

• Delivery of plasmid encoding green fluorescent protein

• Determination of delivery mechanism