Contactless delivery of plasmid encoding EGFP in vivo by high-intensity pulsed electromagnetic field

https://doi.org/10.1016/j.bioelechem.2021.107847 1567-5394/ 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). ⇑ Corresponding author. E-mail address: damijan.miklavcic@fe.uni-lj.si (D. Miklavčič). # Matej Kranjc and Simona Kranjc Brezar both are contributed equally to this paper. Matej Kranjc , Simona Kranjc Brezar , Gregor Serša , Damijan Miklavčič a,⇑


Introduction
Several medical applications for both diagnosis and therapy rely on the application of pulsed electromagnetic fields [1]. Most of the studies employing pulsed electromagnetic fields are focused on low-intensity magnetic fields, such as FDA approved healing nonunion or delayed bone fractures [2,3]. Transcranial magnetic stimulation (TMS) is the exception as it utilizes time-varying highintensity pulsed electromagnetic fields (HI-PEMF) to deliver highly localized brain stimulations by inducing an electric field in the human cortex [4,5]. This non-invasive and painless technique provides researchers and clinicians with a unique tool capable of stimulating both central and peripheral nervous systems [6,7]. In recent years, several studies demonstrated that HI-PEMF primarily used for TMS can also be applied for non-invasive permeabilization of the cell membrane [8][9][10]. One of the first studies on the possible effect of electromagnetic pulses on cell membrane permeability showed that time-varying magnetic field exposure of 2.2 T peak strength increases transmembrane molecular transport [11]. Pulsed electromagnetic fields were successfully applied in studies in vitro for reversible permeabilization of biological cells [8,12] and in studies in vivo, where the feasibility and antitumor effectiveness of magnetic pulses as a drug delivery system for cisplatin to murine melanoma subcutaneous tumors was evaluated [10]. Results showed that even a magnetic field below 1 T was sufficient to achieve membrane permeabilization of tumor cells and that the antitumor effect in treatment with pulsed electromagnetic field was related to increased drug uptake into tumor cells as demonstrated by the increased amount of platinum bound to the DNA. Mechanisms and pathways of increased molecular transmembrane transport induced by the HI-PEMF are however not known. The observed effect is similar to increased membrane permeabilization triggered by conventional electroporation, a process in which cells are exposed to high-intensity electric field pulses (hundreds of V/ cm) for a short duration (ms-ms) [13]. But in all of the abovementioned studies on HI-PEMF, the induced electric field value was at best 2.12 V/cm for in vitro studies [12] and 0.24 V/cm for in vivo studies [9], which is at least two orders of magnitude lower from what is usually reported as the electric field threshold leading to membrane electroporation.
Recently, HI-PEMF was applied for delivering siRNA molecules to silence enhanced green fluorescent protein (EGFP) in B16F10-EGFP tumors in vivo [9]. Since siRNA delivery is a promising gene therapy approach for inactivating oncogenes and tumor suppressor genes involved in cancer disease [14], the results obtained in the in vivo mouse model demonstrate the potential use of HI-PEMF for cancer therapy. Still, delivered siRNA molecules were 21 base pairs long, which is 200-times smaller compared to nucleic acids such as plasmid DNA that are delivered in gene therapies to various targets to generate therapeutic effect [15,16]. The simplest way of delivery is a direct injection of naked plasmid DNA into targeted tissue [17]. However, since the expression level after injection of naked plasmid DNA is generally limited, various physical approaches have been introduced to improve efficiency, such as electroporation, balistics, ultrasound, hydrodynamics and others [18,19]. Among them, electroporation (also referred to as electrotransfer) provides the most substantial change in the efficiency of plasmid-gene transfer [20]. Electroporation in vivo increases the transfection 100-10.000x in terms of gene expression [21,22], >100x in terms of serum protein levels [22,23], and 10-100x in terms of the number of transfected muscle fibers [21,24] compared to DNA injection alone. Furthermore, it strongly reduces intersubject variability [21,22] and was shown to be a powerful adjuvant in DNA vaccination [25]. Still, one of the drawbacks of electrotransfer that limits the efficiency of transfection is the tissue damage associated with the procedure, especially if electric pulse parameters are not appropriately chosen [26,27]. Another is the presence of electrochemical reactions in the electrode-electro lyte/tissue interface and alteration of the pH, potentially causing denaturation of the plasmid DNA [28][29][30]. The HI-PEMF approach of plasmid DNA delivery could have an advantage over conventional electroporation since HI-PEMF is non-invasive, contactless, and painless. Therefore, we explored HI-PEMF contactless delivery technique for electrotransfer of bigger molecules, i.e., plasmid DNA, by evaluating the transfection efficiency of plasmid DNA encoding EGFP (pEGFP-N1) in mouse muscle, skin, and tumor in vivo. In addition, a plasmid DNA encoding shRNA against melanoma cell adhesion molecule (MCAM), as a therapeutic approach in gene electrotransfer (GET) mediated by HI-PEMF, in the melanoma B16F10 tumor model was used to compare its antitumor effectiveness to conventional GET using contact electrodes.

Experimental design
Mice were randomly divided into 4 experimental groups: a group subject to single intramuscular or intradermal injection of endotoxin-free water combined with HI-PEMF treatment (HI-PEMF), a group subject to single intramuscular or intradermal injection of pEGFP-N1 (EGFP), a group subject to single intramuscular or intradermal injection of pEGFP-N1 combined with either HI-PEMF treatment (HI-PEMF + EGFP) or conventional electroporation (EP/GET + EGFP). Changes in the EGFP expression in muscles (left and right hind leg) and skin (left and right flank) of 72 mice were evaluated through the skin utilizing non-invasive stereomicroscopy fluorescence imaging at different time points Specifically, 24 animals were used for long term expression of EGFP and histology (up to 230 days after the treatment) and for the other short term expression and histology 24 animals per each time point (2 days and 5 days). At these time points after the treatment, ani-mals were humanely sacrificed, the muscle tissues were carefully removed and imaged for precise evaluation of transfection.
Antitumor response to pMCAM was evaluated on 62 mice that were randomly divided into 9 experimental groups: an untreated control group (Control, n = 6), tumors exposed to conventional electric pulses (EP group, n = 6) or to application of high intensity of pulsed electromagnetic fields (HI-PEMF group, n = 6) alone, tumors treated with intratumorally injection of plasmid DNA encoding shRNA against MCAM (pMCAM group, n = 6) or intratumorally injection of control plasmid (pControl group, n = 6), tumors exposed to treatment with conventional electroporation and intratumoral injection of pMCAM (GET pMCAM group, n = 8) or intratumoral injection of pControl (GET pControl, n = 8) and tumors exposed to treatment with HI-PEMF and intratumoral injection of pMCAM (HI-PEMF pMCAM group, n = 8) or intratumoral injection of pControl (HI-PEMF pControl, n = 8). Each therapy was performed three times (on day 0, 2 and 4).

Plasmid DNA (pEGFP-N1, pMCAM, pControl)
In the experiments for optimization of transfection protocol by HI-PEMF in different tissues (tumor, muscle or skin) a plasmid DNA encoding EGFP (Clontech Laboratories, Inc., Mountain View, CA, USA) was used. Further, to evaluate the HI-PEMF as a potential delivery method of a therapeutic gene into the tumors two plasmids with constitutive CMV promoter were used, therapeutic plasmid DNA encoding shRNA against MCAM (pMCAM; pENTR/U6 CD146) [31] and control plasmid DNA, encoding shRNA with no homology to any gene in the mouse genome as a negative control (pControl; pENTR/U6 pControl) [32]. Plasmids were isolated using a Qiagen Ò Endo-Free Plasmid Mega Kit (Qiagen, Hilden, Germany) and dissolved in sterile endotoxin-free water to a final concentration of 2 mg/ml or 4 mg/ml. The tibialis cranialis muscle and skin area (diameter 1 cm) on the flank of SKH1-Elite mice were used in the experiments with pEGFP-N1. In addition, B16F10 melanoma tumors were implanted subcutaneously into the C57Bl/6 mice in the right flank of the mice by inoculation of a suspension 1 Â 10 6 B16F10 melanoma cells prepared in 100 ml of PBS for experiments silencing MCAM with HI-PEMF and conventional gene electrotransfer (GET). The optimization procedures were performed in BALB/c and C57Bl/6 female mice, which are described in Supplemetary data.

Mice, tumor models
2.4. In vivo delivery of plasmid DNA (pEGFP-N1 or pMCAM, pControl) The skin area over the tibialis cranialis muscle on the left and right hind leg, over the tumor and the skin on the flank of mice were depilated (Reckitt Benckiser Health Care Ltd). After the animals were anesthetized, 15 ml (30 mg) of pEGFP-N1 was injected intramuscularly through the skin (medial anterior position) or intradermally in the skin. In the combined treatment schedule, the muscle or skin was exposed to the treatment with either HI-PEMF or conventional electroporation 10 s after the injection. In the case of tumors as target tissue, 25 ml (50 mg) plasmid DNA (pMCAM, pControl) was slowly injected intratumorally followed by HI-PEMF or conventional electroporation applied after 10 min.

Gene transfer using High-Intensity pulsed electromagnetic field
HI-PEMF was delivered by a custom-made magnetic field pulse generator connected to an applicator consisting of a round coil with 68 turns. The generator supplied the applicator with bipolar electric pulses that generated a time-varying HI-PEMF in the vicinity of the coil (Fig. 1). The applicator was positioned over the treated tissue so that the treated tissue was located at the periphery of the coil just below the middle of the windings where the induced electric field is the highest. To improve induced electric field in the tissue, we added conductive gel (G006 ECO, FIAB, Vicchio, Italy) between the applicator and tissue in order to decrease opposing electric field established in the tissue due to charge accumulation on air-tissue boundary [33]. The pulse sequences of electric current delivered to the applicator consisted of 400 bipolar pulses at the repetition frequency of 33 Hz (the application of HI-PEMF lasted a total of 12 s). The optimization procedures of HI-PEMF are described in Supplementary material.

Gene electrotransfer
For gene electrotransfer, different electric pulse parameters and electrodes were applied. The electric pulse parameters for muscle gene electrotransfer were chosen based on the study of Tevz et al, where efficient and long-term gene expression in muscle was demonstrated [27]. Briefly, the muscle was placed between plate electrodes 4 mm apart immediately after the plasmid DNA intramuscular injection and exposed to square wave electric pulses: one high voltage (HV, 600 V/cm, 100 ms long) followed by four low voltage (LV, 80 V/cm, 100 ms long) at a frequency of 1 Hz [27]. In the skin, GET was performed immediately after the intradermal injection of plasmid DNA using 24 electric pulses (570 V, 100 ms long) at 5 kHz, that are known to provide gene expression in the superficial skin layers and induced a local response [34]. The electric pulses were delivered through a noninvasive multi-electrode array consisting of 7 spring-loaded pins arranged in a hexagonal mesh and spaced 3.5 mm apart from each other. For conventional GET in the muscle and skin, a CLINIPORATOR TM electric pulse generator (IGEA s.r.l., Carpi, Italy) was used. In tumors, GET was performed 10 min after intratumoral injection of plasmid DNA. Tumors were placed between plate electrodes 6 mm apart and exposed to eight square-wave electric pulses (600 V/cm, 5 ms, 1 Hz; b Tech, Leroy, France). The electric pulse parameters were chosen based on the studies Kranjc Brezar et al [35], where the expression of the transgene at protein and mRNA level, as well as the antitumor effectiveness of the transgene were determined. Good contact between the electrodes and the skin over the muscle, over the tumor or the skin on the back of the animal, was ensured by the use of conductive gel (G006 ECO, FIAB, Vicchio, Italy).

Assessment of the EGFP expression
EGFP expression in muscles and skin was evaluated utilizing non-invasive stereomicroscope fluorescence imaging using Carl Zeiss SteREO Lumar V12 fluorescence stereomicroscope equipped with a NeoLumar S 0.8Â objective and an AxioCam MRc5 digital camera (all from Carl Zeiss, Jena, Germany). This equipment allowed the observation of the EGFP expression in the same animal for several days. The animals were anesthetized with isoflurane inhalation anesthesia and placed on a custom-designed holder. Aiming to achieve the best conditions for in vivo imaging in the treated tissue, the targeted area was shaved and depilated. The exposure time was set at 2 s with no binning. The EGFP fluorescence from the treated tissue was quantitatively evaluated at the measured time points. From the bright-field images acquired at 12.5Â magnification, the target tissue was located and manually gated to determine the region of interest. For the images taken with the GFP filter, a suitable threshold was applied, and the fluorescence intensity in the whole area of the target tissue, i.e., the region of interest determined from bright-field images, was determined with image analysis software ImageJ/Fiji (National Institutes of Health, Bethesda, MD, USA) [36]. Besides, the expression of EGFP in the upper layer of skin was determined by noninvasive imaging confocal laser scanning microscope equipped with a water immersion objective W Plan-Apochromat 20 Â 1.0 DIC and GaAsP wavelength detector 485-700 nm (Examiner. Z1, Zeiss LSM 800, Zeiss, Germany) and an AxioCam 506 color digital camera 2, 7 and 14 days after the transfection using HI-PEMF. A laser was used for excitation at 488 nm (green signal) and a signal was acquired for each slice/time-point. First, the z-stack was acquired with a spacing of 1 lm between slices to obtain a 3D image of the skin in the treated area, starting at the the top and finished in total 152 lm of skin thickness.

Histological assessment of transfected tissue
To evaluate tissue damage in muscle tissue caused by the treatment, muscles were removed at different time intervals after the procedure to assess early tissue damage (2 days), as well as late tissue damage (6 months). The harvested muscles were cross sectioned in the middle (horizontal direction) and fixed in Zn-fixativ for 24 h, afterward placed in 70% of ethanol for 24 h and embedded in paraffin blocks that were cut into tissue sections of 10 lm. Ten consecutive sections per muscle were cut. Tissue sections were differentially stained with hematoxylin-eosin dye. The extent of histological change was determined as a fraction of the area with inflammatory mononuclear cell infiltration concerning the whole muscle area. Besides, muscles were inspected for the presence of necrosis. Images of muscle sections (5 fields of view/section) were captured by a digital color camera (Olympus, Hamburg, Germany) and analyzed with CellSens Dimension software. The field of view did not include the site of injection, but it is at the site of treatment and expression.

Assessment of antitumor response to pMCAM
Three orthogonal diameters (a, b, c) of tumors were measured by Vernier caliper every second day and the tumor volume was calculated according to the formula V = a Â b Â c Â p/6. From the tumor volumes, arithmetic means and standard error of the mean for each group were calculated. The antitumor effect was assessed by tumor doubling time (DT), determined as the time in which the tumor doubled the volume from the initial day of the experiment. In addition, the tumor growth delay (GD) was calculated as the difference in tumor DT of the therapeutic and control group. The growth of tumors was followed until the tumor volume reached 400 mm 3 (data used growth curves). The animals were assigned as cured if they were tumor-free until 100 days after the treatment. Thereafter, cured animals were challenged with a secondary subcutaneous injection of the B16F10 melanoma cells as described above in the left flank. The outgrowth of tumors was followed till 100 days after the injection of tumor cells and the animals were considered as resistant to secondary challenge if no tumor growth at day 100 was observed. Animal weight loss was monitored as a sign of systemic toxicity of the treatments.

Statistical analysis
All data were tested for a normal distribution with the Shapiro-Wilk test (a = 0.05). A one-way analysis of variance (ANOVA) followed by a Holm-Sidak test or unpaired, nonparametric Mann-Whitney test was used to evaluate the differences between the experimental groups. A P-value of <0.05 was considered significant. SigmaPlot software (Systat Software, Chicago, IL, USA) was used for statistical analysis and graphical presentation.

Assessment of the EGFP expression in tibialis cranialis muscle and skin
To test whether HI-PEMF enables gene transfer to tibialis cranialis muscle and skin, the HI-PEMF was performed immediately (10 s) after intramuscular or intradermal injection of plasmid pEGFP-N1 in SKH1 hairless mice. The HI-PEMF treatment was compared to conventional electroporation using high voltage square wave electric pulses delivered via plate electrodes to muscle or via a multi-electrode array to the skin. The transfection efficiency was noninvasively assessed by fluorescence stereomicroscope imaging at different time points. One third of the animals were humanely sacrificed 5 days after the treatment for a precise evaluation of transfection. Altogether, the expression of EGFP after HI-PEMF application was observed in both tibialis cranialis muscle (Fig. 2) and skin (Figs. 3 and 4). As expected, no expression of EGFP was observed in any of the control groups: plasmid DNA (EGFP) injection only or application of HI-PEMF only. The fluorescent intensity, transfected area and dynamic of expression obtained in the muscle (Fig. 2) after the combined treatment of HI-PEMF and injection of pEGFP-N1 (HI-PEMF + EGFP) were comparable to the treatment with conventional electroporation (EP + EGFP). Whereas in the skin, higher fluorescence intensity and larger areas were determined with HI-PEMF compared to conventional electroporation (Fig. 3B,C). The expression of EGFP was obtained in deeper skin layers, with a maximal peak at day 7, thereafter started to decline reaching the plato (lasted to 80 days) and afterwards between 80 and 90 days started to blur in most cases or even disappear in some cases (Fig. 3). Furthermore, the expression of EGFP was determined also in the upper layer of skin (Fig. 4), which started to blur after one and disappear after two weeks. In both, muscle and in the skin, the highest fluorescence intensity was detected in areas that were closest to the surface of the applicator, i.e., where the induced electric field was the highest.

Histological analysis after gene electrotransfer in muscle tibialis cranialis
As can be observed in Fig. 5, 48 h after the application of either high intensity pulsed electromagnetic field (HI-PEMF), conventional electroporation (EP), injection of pEGFP-N1 plasmid (pEGFP), injection of pEGFP-N1 and application of HI-PEMF (HI-PEMF + pEGFP) or injection of pEGFP-N1 and application of conventional electroporation (EP + pEGFP) the inflammatory infiltrate was diffusely distributed in the endomysium around the muscle fibers. Lymphocytes and plasma cells dominating among inflammatory cells. Combined treatment of plasmid injection and application of conventional electric pulses resulted in a significant increase in inflammatory mononuclear cells up to 38% of muscle area. Interestingly, combined treatment of plasmid injection with the application of HI-PEMF induced the infiltration of inflammatory mononuclear cells up to 15% of muscle area. There was no inflammatory immune cell infiltration after HI-PEMF application or conventional application using electric pulses and contact electrodes alone.

Antitumor effect of silencing MCAM mediated by HI-PEMF
As a model for evaluating and comparing the antitumor effect of silencing MCAM mediated by HI-PEMF to conventional electroporation, a B16F10 melanoma was used that has previously shown the therapeutic potential of MCAM silencing [9]. Silencing MCAM using HI-PEMF delayed tumor growth (up to 5 days compared to control) with no tumor cures observed (Table 1, Fig. 6,). The antitumor effect was also observed after electrotransfer of nontherapeutic plasmid mediated by HI-PEMF but was less evident compared to the therapeutic plasmid, resulted in a 3.6 days shorter tumor growth delay. The most pronounced antitumor effect of therapeutic and nontherapeutic plasmid DNA was observed after using conventional electroporation for GET (GET pMCAM, GET pControl). Silencing MCAM using conventional GET (GET pMCAM) significantly delayed tumor growth compared to all other groups (control, monotherapies, GET pControl, and groups where electrotransfer of a therapeutic and nontherapeutic gene was mediated by HI-PEMF) (Table 1, Fig. 6). In addition, silencing of MCAM using conventional GET therapy cured 37,5% of tumors and 66.7% of these mice were resistant to secondary challenge (Tumor growth curves of mice that were not resistant to secondary challenge are available in Supplementary material). A similar but significantly lower antitumor effect was determined after electrotransfer of plasmid DNA encoding of a non-therapeutic gene by conventional electroporation (GET pControl), which resulted in 6.6 days shorter tumor growth delay compared to the group treatead with EP only (EP) and 25% of tumor cures, but none of them were resistant to secondary challenge.

Discussion
In our study we demonstrated the use of contactless high intensity pulsed electromagnetic field (HI-PEMF) treatment as a feasible noninvasive approach to achieve in vivo transfection by enabling the transport of larger molecules such as plasmid DNA encoding EGFP into muscle and skin. Furthermore, the antitumor effectiveness using HI-PEMF for electrotransfer of therapeutic plasmid, i.e., silencing MCAM was demonstrated. HI-PEMF treatment may thus represent an important tool in research and clinical applications for noninvasive gene/nucleic acid delivery in vivo since HI-PEMF is easy to perform, is non-invasive, contactless, and painless.
Generally two target tissues, muscle and skin, were shown as interesting in gene manipulation [27,34,[37][38][39][40]. The specific biological characteristics of skeletal muscle, i.e., high capacity of protein synthesis, postmitotic nature, the longevity of muscle fibers, the presence of muscle precursor cells enable long-lasting transgene expression either locally or systemically, thus makes it as promising approach in gene therapy [17,21,23,40]. Muscle gene transfection has been tested in the treatment of muscular disorder, local secretion of angiogenic and neurotrophic factors and systemic secretion of therapeutic proteins such as erythropoietin, coagulation factors and anti-inflammatory cytokines [38,39]. Another tissue attractive for gene therapy and vaccination is skin.
Specifically, the skin complexity, different layers consisted of different cell types, enable the long term expression in transfected muscle cells (rarely divide), short term in transfected keratinocytes (fast dividing cell), also in deeper layers of skin [34,37,[41][42][43], which could altogether affect on the localization and expression of transgene.
We have shown previouly that small molecules can be delivered into cells with HI-PEMF in vitro [8] and in vivo [9,10]. In this study, we obtained a long term expression of EGFP in the muscle and skin after HI-PEMF and conventional electroporation, in some mice even up to 230 days (Fig. 2a) and up to 190 days (Fig. 3a), respectively. The results are in agreement with the studies where long-term expression up to 18 months using conventional electroporation was reported [21,27,62]. Besides, Tevz et al determined 25% of the transfected area of the whole muscle after electrotransfer of 30 lg of plasmid DNA encoding EGFP at day 7 posttransfection. Similarly in our study around 30% of the transfected muscle area was obtained at day 5 post-transfection (Fig. 2d). Furthermore, HI-PEMF was tested for transfection in skin tissue. The obtained transfected area in the skin using HI-PEMF was significantly higher, up to 3-fold compared to conventional electroporation (Fig. 3d). Targeting different types of cells with different  turnover in skin layers resulted in different duration of transgene expression; i.e. long-term in muscle cells and short-term in keratinocytes [41,63]. In a previous study using multi-electrode array electroporation it was reported that by using different pulse parameters gene expression in the skin can be controlled; the high-voltage electric pulses for superficial expression of the transgene which enabled a local response; and the low-voltage electric pulses in the deeper skin layers (muscle layers Panninculus carnosus below the dermis), which enable the prolonged gene expression and higher transgene production, possibly with systemic distribution [34]. Interestingly, in our study we observed transfection of both, in the upper layers (keratinocytes) and the deeper layers (including muscle) of the skin using HI-PEMF. Visually detected transfection started to blur after 80-90 days, thus we presume that the transfection of muscle Panniculus carnosus was observed, which has longer turn over compared to cells in the epidermis and dermis. Thus, HI-PEMF as gene delivery method could be used in vaccination, where local immune response of the skin cell is expected, as well for the muscle transfection used for delivery of molecules with systemic action. Therefore, in the translation into the clinics, it will be important to evaluate the level and duration of transgene expression after using HI-PEMF. As a proof of principle, the HI-PEMF as a delivery method was further tested in antitumor effectiveness of plasmid encoding MCAM (therapeutic plasmid) in B16F10 immunologically responsive melanoma and compared to conventional gene electrotransfer. Recently it has been shown that conventional gene electrotransfer of pMCAM in B16F10 melanoma, significantly delayed tumor growth (up to 13 days compared to control) and cured 36% of mice, of which 60% were resistant to secondary challenge [35]. Similarly, the antitumor effect was confirmed with GET of nontherapeutic control plasmid (pControl); the tumor growth delays up to 10 days, 31% of cured tumor of which 20% were resistant to secondary challenge. The results from that and other studies indicated the induction of immune response after GET of therapeutic and nontherapeutic plasmids [32,35,[64][65][66]. In the same tumor model,   Marrero et al [65] obtained up to 71% of animals that were resistant to the secondary challenge after GET of control pVaX1 plasmid, while in the studies of Savarin et al [32] and Kranjc Brezar et al [35], they determined up to 100% of animals that were resistant to the secondary challenge after GET of pControl. Similar results were obtained also in our study after conventional gene electrotransfer of pMCAM or pControl. In the case of GET pMCAM 37.5% of tumors were cured and 66.7% of them were resistant to secondary challenge. A less pronounced effect was observed after GET pControl treatment resulting in 25% of tumor cures and no resistance to secondary challenge ( Table 1). The antitumor effect after using HI-PEMF to deliver therapeutic and nontherapeutic plasmids in melanoma was less evident. The growth of tumors after HI-PEMF pMCAM was delayed significantly less (for 8.9 days) compared to conventional gene electrotransfer and no tumor cures were observed (Table 1, Fig. 6) and insignificantly less compared to GET pControl. Nevertheless, tumor growth was still significantly delayed (up to 5.6 days) compared to all other pertinent control groups (Control, either of therapeutic or nontherapeutic plasmid injection only, HI-PEMF, HI-PEMF pControl). It was reported in several studies that electroporation as a delivery method of plasmid DNA or other molecules induces local tissue damage and lymphocyte infiltration [27]. The infiltration of mononuclear cells after the injection of plasmid DNA or GET combined with the injection of plasmid DNA determined 48 h after the treatment reached up to 30% and 60% of muscle area, respectively. Similarly, we observed a significant increase of 38% in inflammatory mononuclear cells of muscle area after conventional gene electrotransfer of pEGFP compared to control. However, the delivery of pEGFP using HI-PEMF induced significantly less infiltration of inflammatory mononuclear cells compared to conventional gene electrotransfer (Fig. 5). In controlling of systemic action of expressed proteins, HI-PEMF induced no tissue damage and less mononuclear cell infiltration compared to conventional electroporation.
Although nonviral delivery of genetic material into cells using conventional electroporation holds great promise for basic research and clinics, it has some limitations, i.e., low transfection efficiency compared to other methods [67][68][69][70], presence of electrochemical reactions, alteration of pH [28][29][30] and difficulties in con-trolling the delivery due to an incomplete understanding of its mechanisms [71][72][73][74]. In the gene electrotransfer using conventional electroporation of cells and tissues, it is postulated that plasmid DNA can enter through pores only partially. Mainly clathrinmediated or Rac-1-dependent endocytosis processes are then responsible for the transport of plasmid DNA across the membrane and cytoplasm to the nucleus [73][74][75][76]. Such processes demand optimized electric pulse parameters, i.e., longer, high voltage or a combination of short, high voltage and long, low voltage pulses [27,71,[76][77][78][79][80][81][82]. Furthermore, the alterations in the pH of the medium during electroporation may cause denaturation of the plasmid DNA or reduces pDNA uptake and thus plays an important role in DNA electrotransfer [28,29]. Additionally, the temperature affects the cell membrane fluidity and consequently, its electroporation [83]. It was shown that gene electrotransfer could be thermally assisted. Specifically, heating applied before electroporation enabled a similar level of transfection efficiency in skin using an approximately 30% lower electric field [84]. In our study, we observed an increase in the temperature of the magnetic coil (to approximately 40°C, data not shown), and we assume that some transfer of heat could have led to a thermal effect in the tissue during electroporation, thus affecting the transfection efficiency in the skin but this is not a case in muscle and tumor due to slow thermal diffusion in tissue. Besides, the heterogeneity of tissue, i.e., the cell type organization and extracellular matrix composition, has a substantial impact on transfection efficiency. It was shown that modification of the tissue extracellular matrix improved the transfection efficiency, due to a better plasmid DNA distribution within the tissue [79,85]. In summary, we are aware of all these factors that might have affected the transfection efficiency in the present study and intend to address this possibility in a further study, both by optimizing the conditions of HI-PEMF and improving the equipment, i.e., the coil and magnetic pulse generator. Our long-term goal is to develop the devices and applicators for noninvasive noncontact electroporation and mechanistic knowledge that will allow further optimization of the device, applicators and gene transfer protocols and by this development of gene therapies, which will be safer and also more accessible.

Conclusion
In conclusion, our results show that a high-intensity pulsed electromagnetic fields (HI-PEMF) at a time-varying magnetic field below 1 T are capable of a facilitating delivery of large molecules of plasmid DNA (pEGFP-N1) in different tissues (muscle, skin and tumors). Furthermore, the antitumor effectiveness using HI-PEMF for electrotransfer of therapeutic plasmid pMCAM was demonstrated. Due to the potential and contactless application of HI-PEMF, this approach represents for gene therapy and DNA vaccination a viable alternative to conventional electroporation. Further studies are warranted to improve the equipment, HI-PEMF parameters and to optimize the protocols for gene transfection.   6. Growth curves of B16F10 melanoma tumors after treatment with gene electrotransfer of plasmid DNA encoded shRNA against MCAM (pMCAM) mediated by HI-PEMF or conventional electroporation (GET). Each therapy was performed three times (on day 0, day 2 and day 4). The data represent the arithmetic mean and standard error of the mean, n = 6-8; *P < 0.05 significant to HI-PEMF pMCAM; **P < 0.05 significant to HI-PEMF pControl; One-way ANOVA; black circlescontrol; black squares -EP; black triangles -HI-PEMF; black upside down triangles -pMCAM; black diamonds -pControl; yellow circles -GET pMCAM (2 CR); blue squares -HI-PEMF pMCAM; green triangles -GET pControl (1 CR); violet upside down triangles -HI-PEMF pControl.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.