Introduction

Skin injury leading to chronic wounds is of high interest due to the increasing number of patients suffering from this symptom [1]. Generally, wounds, being the result of damage to the skin, are highly susceptible to pathogenic microorganisms [2]. Widespread wound infections occur after surgeries or the use of several medical instruments such as cannula, drips, and needles [3]. Giacometti et al, [3] suggested that Staphylococcus aureus, Pseudomonas aeruginosa, and Escherichia coli are the most common bacteria detected in the wounds. To inhibit bacterial growth, the most popular therapies are based on the application of antibiotics [4], which might lead to antibiotic resistance in the environment. In addition to killing microorganisms responsible for wound infections, there is also a high need to activate the normal skin cells in order to enhance the wound healing efficiency. For that reason, alternative wound healing therapies have been searched and extended. Very recently, special attention has been paid to the use of cold atmospheric pressure plasma (CAPP) in dermatology, thanks to the ability of these ionized-gas sources is to improve the wound healing process [5,6,7,8,9,10,11,12,13,14,15,16,17,18].

Over the past few years, the interest of the controlled applications of CAPPs has rapidly increased. They include not only the utilization of CAPPs in dermatology [5,6,7,8,9,10,11,12,13,14,15,16,17,18], but also application of CAPP in the synthesis of nanomaterials [19,20,21], the germination of seeds [22], inhibiting the proliferation of cancers cell lines [23, 24], and so on. This is possible because CAPP operation results in the generation of different reactive oxygen species (ROS, i.e., O3, H2O2, O·, \({\text{O}}_{2}^{ - }\), OH·) and reactive nitrogen species (RNS, i.e., NO, NO2), in addition to the emission of UV radiation and heat [7,8,9, 18, 25, 26].

The use of CAPPs in dermatology is related to their ability to inactivate pathogenic microorganisms as well as to enhance cell angiogenesis, to stimulate cell proliferation and migration, and to activate receptors on the cell surface. Angiogenesis, which is the formation of new blood vessels, is a major part of the wound healing process [27]. In order to increase cell motility, proliferation, and angiogenesis in wound healing without triggering the inflammation or apoptosis processes, the concentration of the above-mentioned reactive species needs to be tightly controlled [15]. For that reason, the effect of various CAPP treatments on skin cell lines such as fibroblasts [18, 28,29,30] and keratinocytes [18, 29,30,31,32,33,34] has been examined. These studies have established that CAPP-based devices are effective tools for stimulating the wound healing process [18, 29,30,31,32,33,34,35,36]. Interestingly, the type and amount of generated ROS and RNS, as well as the emitted UV radiation and heat, which facilitated the wound healing processes, were found to be highly dependent on the plasma source.

To date, a few reaction-discharge systems have been developed for further applications comprise medicine and dermatology. These reaction-discharge systems include dielectric barrier discharges (DBDs) [14, 15], corona discharges (CDs) [37], microwave-induced plasmas (MIPs) [11, 12], and plasma jets named kINPen [5, 7,8,9,10, 17] or COST [38, 39], in addition to the non-thermal atmospheric pressure plasma jets studied and utilized by e.g. Rad et al. [40], Gidon et al. [41, 42], Darmawati et al. [43], Cheng et al. [44], Busco et al. [45]. To be applied in biological systems, the reaction-discharge systems should provide a low gas temperature, i.e., less than 40 °C, along with the controlled production of reactive chemical compounds such as ROS and RNS, and the emission of UV radiation [9].

Different computational models based on the Monte Carlo simulation, fluid dynamics (CFD), and 2D cylindrically symmetric hydrodynamics have been developed to better understand various properties (e.g. the effect of the supplied power, the temperature distribution, and the different species density distribution) of the kINPen plasma jet device [9]. Additionally, Gidon and co-workers [41] reported a predictive modelling approach for adequately describing the system dynamics of a He atmospheric pressure plasma jet using a model predictive control strategy [41].

The main aim of the present study was to develop a portable and efficient He-DBD-based reaction-discharge system of a controlled gas temperature for the selected biological activities, including the activation of normal human skin cell lines by stimulating their proliferation, migration, and angiogenesis. Compared to others studies of atmospheric pressure plasma jets [5, 7,8,9,10, 17, 38,39,40,41,42,43,44,45], the presented He-DBD-based reaction-discharge system is characterized by a unique in its class DBD generator that provides an ability to modulate either frequency or duty cycle, and it is cheaper and simpler compared to the classical dc or ac generators commercially used. The design of experiments (DOE) approach followed by the response surface methodology (RSM) were applied for the multivariate optimization of the operating conditions of the above mentioned system and the selection of settings under which a rotational temperature of the OH molecules (Trot(OH)), reflecting the gas temperature (Tg) of the He-DBD plasma, was less than 37 °C (310 K). Additionally, the selected spectroscopic parameters, i.e., the excitation (Texc), vibrational (Tvib), and rotational (Trot) temperatures, and the electron number density (ne), of the He-DBD were determined when the developed reaction-discharge system was working under optimal operating conditions. Optical emission spectrometry (OES) was used to identify the type of ROS and RNS produced in the gaseous phase of He-DBD under defined operating conditions. The concentrations of the selected ROS and RNS in the liquid phase treated by He-DBD were assessed using spectrophotometric methods. Finally, the effect of the He-DBD plasma generated under the optimal operating conditions were assessed on the proliferation, migration, and angiogenesis of the skin cell lines keratinocytes (HaCaT) and fibroblasts (MSU-1.1), and on human skin microvascular endothelial cells (HSkMEC.2).

Experimental Procedures

Construction of the Portable He-DBD-Based Reaction-Discharge System

In order to provide a better wound healing efficiency according to the proliferation, migration, and angiogenesis of the normal skin cell lines, the portable He-DBD-based reaction discharge system was developed and optimized. As shown in Fig. 1, the main part of this system was a reaction-discharge head, made of an E-57 epoxy resin, containing a quartz tube (ID = 10 mm) onto which two ring-like tungsten electrodes (ID = 2.0 mm) were mounted. The above-mentioned parts were covered with a CORIAN® insulator and inserted into a ceramic tube. The edge of the quartz tube was 15 mm below the edge of the ceramic tube. Stable-in-time and homogenous glow He-DBD was sustained in the quartz tube by supplying a high voltage (HV) potential up to 15 kV to both electrodes from a pulse modulated alternating current (40 kHz) power generator (Dora Electronics Equipment, Poland). Helium 6.0 (99.9999%, Linde, Poland) was used as the discharge gas due to its high thermal conductivity of 190.6 × 10−3 W m−1K−1 at 400 K. The voltage was modulated with a square wave at frequencies changing from 2.1 to 2.8 kHz and with the duty cycles changing from 25 to 75%. The actual discharge currents were measured using a Rogowski coil current probe (Dora Electronics Equipment, Poland), while a HV probe (1:1000), built-in in the power generator, was applied for the voltage measurements. Finally, the discharge current and the voltage signals were processed by a two-channel 60 MHz SDS1104X oscilloscope (Siglent, China). To control the flow rate of He within 4.0–11.0 L min−1, a 1355G Sho-Rate low flow glass tube flowmeter (Brooks Instruments, USA) was used.

Fig. 1
figure 1

A schematic representation of the portable He-DBD-based reaction-discharge system

Response Surface Methodology: The Optimization of the Portable He-DBD-Based Reaction-Discharge System

Since the portable He-DBD-based reaction-discharge system was planned to treat the normal human skin cells, its Tg should be less than 37 °C (310 K). Hence, it was necessary to optimize the operating parameters of the DBD-based reaction-discharge system to gain such a Tg value. Accordingly, the RSM was used to model the Tg of the system versus the selected operating parameters and to find on the basis of the established response surface regression model the optimal settings that could provide the desired Tg value. To have insight into the kinetic energy of the gaseous phase of the He-DBD, the Trot(OH), as determined based on the emission spectra of He-DBD, was used to accurately estimate the Tg [46, 47]. A Box-Behnken experimental design (BBD) was used to model the system versus the following operating parameters: A—the frequency used to modulate the sinusoidal voltage (in kHz), B—the duty cycle (in %), and C—the He flow rate (in L min−1). The RSM experimental design comprised 15 randomized runs at the 3 different levels of operating parameters and included the 3 central points. The levels of the operating parameters were selected based on the observations of the stable operation of homogenous He-DBD and were as follows: A—2.10, 2.45, and 2.80 kHz; B—25%, 50%, and 75%; and C—4.0, 7.5, and 11.0 L min−1. All runs were carried out in one block according to the run order given in Table 1.

Table 1 The Box-Behnken experimental design with the actual and coded (in brackets) values of the operating conditions of the He-DBD-based reaction-discharge system along with the standard and run orders and the values of the response

To reproduce the variability in the system, the response of the system was modeled with a complete quadratic function, including the main linear and quadratic effects of the operating parameters on the Trot(OH), and their two-way interactions [48]. In addition, to reduce the dimensionality of the final response surface model, the insignificant terms were eliminated using a stepwise-selection-of-terms algorithm [48]. In this case, the terms were added during the selection procedure to maintain a hierarchical model of the system. The suitability of the established model was assessed using the ANOVA test and the established p-values for each source and the determination coefficient for the whole response surface regression model (R2). The model was also tested for the lack-of-fit that provided the evidence of its adequacy and efficacy [48]. Finally, the quality of the model and its reliability were checked by examining the residuals. In this case, the normal probability plot and a scatterplot of the residuals versus the run order were looked through for any heteroscedasticity.

Determination of the Optical Temperatures and the Electron Number Density

The emission spectra of He-DBD were acquired to characterize the developed and studied CAPP source. Radiation emitted by the He-DBD was observed in the range from 200 to 900 nm, alongside the plasma tip. It was directly collimated using a UV quartz lens on the entrance slit of a Shamrock SR500i high-resolution imaging spectrograph (Andor, UK), equipped with two holographic gratings (1800 and 1200 lines mm−1) and a CCD camera (1024 × 255 pixels, Andor, UK). The acquisition time was 10 s. Such an observation layout was reasonable because of the application of the He-DBD plasma tip in the vertical position for the wound healing. All the emission spectra of the He-DBD system were acquired under the operating conditions given in more details in Table 1 [each parameter setting is defined, including the frequency of modulation (in kHz), the duty cycle (in %), the He flow rate (in L min−1)]. The unresolved emission spectra of the OH molecule in the range of 300–320 nm, belonging to the OH (A-X) system, the (0–0) vibrational-rotational transition with the band head at 308.9 nm, was chosen for the determination of the Trot(OH). Similar procedure was also applied in the case of Trot(N2) and the (0–2) band of N2(C-B) with the bandhead at 380.2 nm was selected for that purpose. The emission spectra of N2 were collected in the range within 376–380 nm. The Trot(OH) and Trot(N2) values were determined by fitting the recorded OH (A-X) and N2(C-B) emission spectra with the spectra simulated at various temperatures using the Lifbase software (2.2 version, SRI International, USA) and Specair (2.1 version, Spectrafit, France), respectively. This temperature was assumed to be a very good estimator of the Tg of He-DBD. In addition, other optical temperatures such as the Texc, the Tvib, and the Trot, and the ne were assessed for He-DBD operated under the optimal conditions. The Texc of H atoms was determined using the two-line method, using 656.20 nm (Hα line) and 486.13 nm (Hβ line) emission lines [49].The Tvib of the OH molecules was evaluated based on two OH (A-X) molecular bands, i.e., the (1–0) and (0–0) transitions. In turn, experimentally measured OH molecular bands in the range within 280-320 nm were fitted with those simulated using the Lifbase software at different values of the Tvib [50]. A similar procedure was also applied to determine the Trot of the \({\text{N}}_{2}^{ + }\) molecules. In this case, the emission spectra of the \({\text{N}}_{2}^{ + }\) (B-X) system, the (0–0) transition with the band head at 391.4 nm, were simulated in the range of 388–392 nm at various values of the Trot. Then, these simulated emission spectra were compared with the experimental data. The precision (expressed as the relative standard deviation, RSD) of the temperature measurements was better than 10%. The ne was determined on the basis of the Stark broadening (ΔλS) of the Hβ line at 486.13 nm, using the formula given by Gigosos et al., i.e., (ΔλS/4.800) = (ne/1017)0.68116 [50].The experimentally measured profiles of the Hβ line were fitted using the Voight algorithm. Only the Lorentz part (ΔλL) of the Voight algorithm was applied to calculate the ΔλS. Additionally, the Van der Walls broadening (ΔλvdW) of the Hβ line profile was also used in the final calculations of the ΔλS, i.e., ΔλS = ΔλL − ΔλvdW.

Elucidation of the Reactions and Processes Running in the Gaseous Phase of He-DBD and the Liquid Phase Treated by He-DBD

Several experimental techniques were applied to reveal the possible reactions and processes occurring in the gaseous phase of the He-DBD plasma as well as in the liquid treated by this plasma. These analyses were undertaken when the He-DBD-based reaction-discharge system was operated under optimal conditions, found on the basis of the validated response surface regression model.

Accordingly, the ROS and RNS produced in the gaseous phase of He-DBD were identified on the basis of its emission spectra measured with OES. These spectra were recorded from the same plasma region within 200–900 nm, using the same instrument as was described in the “Determination of the Optical Temperatures and the Electron Number Density” section.

The spectrophotometric methods were used for the quantitative determination of the selected ROS and RNS, i.e., hydrogen peroxide (H2O2), and nitrite ions (\({\text{NO}}_{2}^{ - }\)), produced in a physiological saline solution (0.85% m/v NaCl) treated by He-DBD, which was operated under the optimal operating conditions, i.e. the He flow rate of 10.6 L min−1, the frequency of modulation of 2.1 kHz, and the duty cycle of 74%. The glass vials were put onto the magnetic stirring and vigorously mixed during the treatment. The distance between the tip of the He-DBD plasma and the surface of the treated liquid was 15 mm.

The concentration of H2O2 was measured based on the production of the peroxovanadium cation (\({\text{VO}}_{3}^{2 + }\)) [49]. The above-mentioned orange \({\text{VO}}_{3}^{2 + }\) cation, showing a maximum absorbance at 450 nm, was obtained as a result of the reaction between the H2O2 molecules with ammonium metavanadate (NH4VO3) in the medium of 0.58 mol L−1 H2SO4. Hence, 2.5 mL of the 0.85% (m/v) NaCl solution (Sigma-Aldrich, Germany), treated for 300, 600, or 1200 s by He-DBD, were mixed with 2.50 mL of the NH4NO3/H2SO4 stock solution, and diluted to 25.0 mL using re-distilled water. The suggested plasma treatment time was chosen to provide the best accuracy in these types of measurements. The concentrations of NH4VO3 (Avantor Performance Materials, Poland) and H2SO4 (Avantor Performance Materials, Poland) in the above-mentioned stock solution were 6.2 mmol L−1 and 0.058 mol L−1, respectively. The concentration of H2O2 in the analyzed samples was calculated based on the absorbance value, recorded at 450 nm using a double-beam UV/Vis spectrophotometer (Specord 210 Plus, Analityk Jena, Germany). The spectrophotometer was zeroed using a respective procedural blank sample.

The concentration of \({\text{NO}}_{2}^{ - }\) was determined using the Rivanol method [51, 52]. In this method, a red-purple diazonium salt, showing a maximum absorbance at 595 nm, was produced as the product of the reaction between Rivanol (2-ethoxy-6,9-diamino-acridinium lactate, C18H21N3O4) and the \({\text{NO}}_{2}^{ - }\) ions. To determine the \({\text{NO}}_{2}^{ - }\) concentration in the analyzed samples, 5.0 mL of the 0.85% (m/v) NaCl solution, treated for 300, 600, or 1200 s by He-DBD, were mixed with 1.25 mL of a 0.1% Rivanol solution (Prolab, Poland), acidified with 0.1 mol L−1 HCl (Avantor Performance Materials, Poland), and topped with re-distilled water to 10 mL. The suggested plasma treatment time was chosen to provide the best accuracy in these types of measurements. The concentration of the \({\text{NO}}_{2}^{ - }\) ions in the analyzed samples was calculated based on the absorbance, recorded at 595 nm using a double-beam UV/Vis spectrophotometer (Specord 210 Plus, Analityk Jena, Germany). The spectrophotometer was zeroed using a respective procedural blank sample.

The pH (the total concentration of the H3O+ ions) of the 0.85% (m/v) NaCl solution was estimated using a CPC-505 pH meter (Elmetron, Poland) before and immediately after their 300, 600-, and 1200-s treatments with He-DBD.

Cells and Culture Conditions

The HaCaT cell line coming from the spontaneously immortalized human keratinocytes [53] was used to test the effect of the He-DBD plasma on biological activity in reference to the proliferation, migration, and angiogenesis of normal human cells. In addition, the MSU-1.1 cell line, obtained by the v-myc oncogene transformation of the foreskin fibroblasts [54] was used as representative of dermal skin cells. Finally, the human microvascular endothelial skin cell line, HSkMEC.2 [55], was used for the study of pseudotubes formation. The cell lines were cultured at 37 ºC in an incubator under the 5% CO2 and 95% air atmosphere. The cells were maintained in Opti-MEM with GlutaMAX (Thermo Fisher Scientific Inc., USA) medium supplemented with 3% fetal bovine serum (FBS, Gibco, UK), 100 U mL−1 penicillin and 100 µg mL−1 streptomycin (Sigma-Aldrich, Germany). They were routinely passaged using a 0.05% trypsin/0.02% EDTA (w/v) solution.

He-DBD-Based Treatment of Human Skin Cell Lines

The human skin cell lines HaCaT, MSU-1.1, and HSkMEC.2 were detached and suspended in 0.85% (m/v) NaCl to a final concentration of 1 × 106 cells mL−1, and treated with the He-DBD plasma operated under the optimal conditions, however, maintaining a 15 mm distance between the tip of the He-DBD plasma and the treated biological materials, i.e., cells that were suspended in an Eppendorf tube. The normal human cells were treated with the He-DBD plasma for a defined time, i.e., from 0 s (control), 10, 30, 60, 90, 180, 270, and 360 s, and then seeded onto the 96-flat bottom well plates in triplicates. After treatment cells were suspended in culture medium and used in the following in vitro experiments.

Proliferation Assay

The cell proliferation was investigated using the standard MTT test (Sigma-Aldrich, Germany). Briefly, 2 × 103 cells were treated with He-DBD for a defined time (0–360 s) and then seeded onto the 96-flat bottom well plates in triplicates [56]. The test was carried out for up to 7 days. It was unequivocally evidenced in this test that the cells treated longer than 90 s stopped proliferating and their apoptosis was observed.

Wound Healing Assay

For the scratch wound assay He-DBD-treated cell lines HSkMEC.2, HaCaT, and MSU-1.1 were cultured separately in 48-well plates in medium supplemented with serum and used in the experiments after the cells reached 100% of confluence. Next, scratches lines were made with 200 μL tips. PBS was then used to wash cells three times and cells scratched out were removed. The remaining adherent cells were cultured with serum-free medium in an incubator with 5% CO2 at 37 °C. Pictures were collected at 0, 16, 24, 48, 72, and 84 h, and photographed under an inverted light microscope to measure the scratch distance. The closure area was calculated using ImageJ software and presented as relative wound closure (RWC, %) as previously described [57].

Angiogenic Assay

The angiogenic properties of the HSkMEC.2 cells treated with He-DBD for the defined time were evaluated by seeding the cells on the surface of growth factor reduced-Matrigel matrix (Corning, USA) and examined by following the formation of pseudovessels. The endothelial cells (1.5 × 104 cells/well) were seeded in duplicates on the Matrigel-coated 96-well plates and cultured in the standard conditions for up to 24 h. Photographs were taken after 6 h of culture. Angiogenesis was evaluated by several parameters: the mean mesh size, the number of nodes, and the total tube length. The results were calculated using the ImageJ Angiogenesis Analyzer Software (version 1.52o) and presented as mean ± SD.

In all in vitro tests all the graphs as well as statistical analysis were prepared using GraphPad Prism software (version 7.04, license number 885438).

Results and Discussion

Optimization of the Portable He-DBD-Based Reaction Discharge System

A Box-Behnken experimental design followed by the RSM was used to model the response, i.e., the Tg, being estimated by the Trot(OH), of the He-DBD system versus selected operating parameters and find such optimal settings that could provide the Tg of 37 °C (310 K). To establish the response surface model, the He-DBD-based reaction-discharge system was run at different operating conditions (as given in Table 1 for a certain run) and the emission spectra of the He-DBD plasma were acquired to determine the Trot(OH). At the beginning, it was verified that the variability in the mean response for all 15 runs was higher than the variability of the repeated measurements of the response for each run. In addition, neither correlation nor trend was observed in the scatter plot of the response versus the run order. Therefore, it was concluded that the variability in the response was due to the changes of the operating parameters within the arbitrarily selected ranges. In this way, it was not necessary to stabilize the variance in the response through its transformation [47, 58].

To establish the response surface model, the Trot(OH) values were fitted with a full quadratic polynomial response surface model. The equation of this model was as follows: Trot(OH) [°C] = − 500.0 + 11.7 × A + 11.0 × B + 38.8 × C − 0.1 × A2 − 1.2 × BC. It should be noted, that frequency of modulation (A) in model was coded, i.e., values of frequency: 2.10, 2.45, and 2.80 kHz correspond to respective values: 25%, 50% and 75% in model. The accuracy of this model was tested by the ANOVA and lack-of-fit tests. The results of both tests at α to enter equal of 0.25 [47] are given in Table 2.

Table 2 The outcomes of the ANOVA and lack-of-fit tests for the response surface model found with the aid of the stepwise-selection-of-terms algorithm (α = 0.25) for the He-DBD-based reaction-discharge system

As can be seen, the response surface model was statistically significant (p = 0.056) and the following linear, square and 2-way interaction terms were significant, i.e., B (p = 0.136), C (p = 0.047), A2 (p = 0.194), and BC (p = 0.044), which all were lower than α = 0.25. The statistically insignificant term A (p = 0.293 > 0.25) was included in the model, considering the hierarchy of the terms. The R2 was relatively high (65%), indicating that the gathered responses were well fitted with the model. In addition, the lack-of-fit test was statistically insignificant (p = 0.454 > 0.25), showing that there was no reason to reject the model nor evidence that it did not fit the experimental data well. This goodness-of-fit was also proved by the distribution of residuals. The normal probability plot showed that the values of the residuals were approximately normally distributed. A random pattern of the residuals was observed in their scatterplot versus the run order; they were uniformly distributed on both sides of the zero axis, except for one outlier (the observation no. 11 with the highest Trot(OH) value of 580 °C, (853 K). Both scatterplots additionally confirmed that the response surface model was correct and well fitted the experimental values of the Trot(OH) of the studied He-DBD reaction-discharge system.

Hence, the model was used to find the operating parameters of the He-DBD plasma so that its Trot(OH) ≅ Tg was 37 °C (310 K). It was established that the target value could be attained with the desirability of 0.999 when operating He-DBD at a frequency of modulation of 2.14 kHz, a duty cycle of 74%, and a He flow rate of 10.6 L min−1. Using the above-mentioned settings, the Trot(OH) value fitted with the model was less than 37 °C (310 K). To confirm the correctness of this selection of these operating conditions, the emission spectra of He-DBD, as operated at the above-mentioned operating parameters, were acquired and the Trot(OH) was evaluated. It was just 37 ± 10  °C ( 310 ± 10 K). This measure confirmed the usefulness of the RSM approach to quickly and reliably established the operating conditions of the He-DBD-based reaction-discharge system to operate it at the Tg of 37 °C (310 K). The selected operating parameters, based on the established response surface model, were used to operate the system in all further experiments.

Spectroscopic Characteristic of the Developed He-DBD-Based Reaction-Discharge System

In order to describe the investigated here He-DBD plasma source, the optical temperatures, i.e., the Texc, the Tvib, and the Trot as well as the ne, were additionally assessed. The evaluated spectroscopic parameters of He-DBD, working under the optimal operating conditions, were shown in Table 3. As can been seen, the Texc (~ 3300 K) was comparable with the Tvib (~ 3000 K) and was much higher than the Trot. It was also observed that the Trot(\({\text{N}}_{2}^{ + }\)) was about twice higher than the Trot(OH) and the Trot(N2). Probably, the excited states of the \({\text{N}}_{2}^{ + }\) molecules, i.e., \({\text{N}}_{2}^{ + }\)(B), were produced by the interaction of ground state N2 molecules (N2(X)) with the metastable states of the He atoms (Hem), i.e., He (3S) and/or He (1S), as a result of the Penning ionization process, i.e., N2(X) + Hem = \({\text{N}}_{2}^{ + }\)(B) + He + e [59]. The threshold energies of the He(3S) and He(1S) states (19.8 and 20.9 eV, respectively) are much higher than the ionization-excitation energy of the \({\text{N}}_{2}^{ + }\)(B) state, i.e., 18.7 eV, and thus the additional energy from the He metastable states are converted to the rotational energy of the \({\text{N}}_{2}^{ + }\) molecules during the Penning ionization process. It should be noticed that the Trot(OH) was equilibrated to the Trot(N2), i.e., Trot(OH) ≈ Trot(N2), and hence close to the Tg of He-DBD, likely due to an increase in the frequency of the molecule collisions in the developed CAPP source [60]. The determined values of the Trot of the \({\text{N}}_{2}^{ + }\) and OH molecules were consistent with those reported in previous works related to various DBDs jets generated in the He atmosphere [59, 61]. Finally, the relations between the determined optical temperatures were as follows: Texc ≈ Tvib ≫ Trot(\({\text{N}}_{2}^{ + }\)) > Trot(OH) ≈ Trot(N2), indicating that developed He-DBD was far from the thermodynamic equilibrium. The ne determined for He-DBD was estimated to be ~1.5 ×1014 cm−3.

Table 3 The spectroscopic parameters determined for the He-DBD plasma operated under the optimal operating conditions in the developed reaction-discharge system

Identification of the Reactive Species in the Gaseous Phase of He-DBD and in the Liquid Phase Treated by the Plasma

Before the quantitative determination of the selected ROS and RNS in the liquid phase treated by He-DBD operated under the optimal conditions, initially, the emission spectrum of the plasma was acquired and analyzed to identify the possible reactive species in the gaseous phase of He-DBD (Fig. 2). As displayed in Fig. 2, the N2, \({\text{N}}_{2}^{ + }\), NH, NO, and OH species were easily excited in He-DBD. Its emission spectrum was dominated by the emission bands of the N2 molecules belonging to the C3Πu-B3Πg system and the emission bands of the OH radical belonging to the A2Σ − X2Π system. The most intense band heads of the above-mentioned systems were observed for the following transitions: (0–0) in case of OH at 309.4 nm as well as (1–0), (0–0) and (0–1) in case of N2 at 315.9 nm, 337.1 nm and 357.7 nm, respectively. Additionally, few other emission bands were identified, including the (0–0), (0–1), and (0–2) transitions for the NO molecules with the band heads at 226.9 nm, 237.0 nm, and 247.9 nm, respectively, belonging to the γ-system (A2Σ+ − X2Π), the (0–0) transition for the NH (A3Π − X3Σ) system with the band head at 336.0 nm, as well as the (0–0), and (0–1) transitions for the \({\text{N}}_{2}^{ + }\) molecules with the band heads at 391.4 nm and 427.8 nm, respectively, belonging to the \({\text{B}}^{2} \Sigma_{\text{u}}^{ + } - X^{2} \Sigma_{\text{g}}^{ + }\) system. The O I lines at 777.2 nm and 844.6 nm with the excitation energy ~ 11 eV, and the H lines of the Balmer series at 486.1 nm, and 656.2 nm with the excitation energies within 12–13 eV were also observed. It should be noted that the most prominent He I lines were not identified in the spectra of the examined He-DBD plasma, probably due to the high excitation threshold energy levels of the He atoms, i.e., ~ 22–23 eV.

Fig. 2
figure 2

The emission spectrum of the He-DBD plasma operated under the optimal conditions

Since it was found that N2, \({\text{N}}_{2}^{ + }\), NO, and OH were the predominant reactive species produced in the gaseous phase of He-DBD, the selected ROS and RNS were likely to be produced in the liquid phase treated by the He-DBD plasma as a result of the plasma-liquid interactions. Hence, the concentrations of the H2O2 molecules, and the \({\text{NO}}_{2}^{ - }\) ions, in addition to the pH were determined in the 0.85% (m/v) NaCl solutions treated by He-DBD under optimal conditions. Although the subsequent treatment of the human skin cell lines with He-DBD was performed only for 90 s (see below), the treatment times of 300, 600, and 1200 s were used in this examination due to the technical constraints. It was established that the concentrations of all three measured species increased with the increased He-DBD treatment time (Fig. 3). The correlation between the concentration of the H2O2 molecules and the He-DBD operation time might be associated with the H2O oxidative processes, as was suggested by Locke et al. [62]. In the case of the \({\text{NO}}_{2}^{ - }\) ions, their fabrication in the He-DBD-treated solution was possibly associated with the synthesis of the HNO2 molecules [52]. The decreasing pH might be related to the irradiation of the solutions with the positive ions, e.g., H2O+, as was suggested previously by Tochikubo et al, [63] Our results, hence, were in line with those of other studies that also reported the increase in the concentrations of the studied reactive species following the CAPP treatment [15, 64].

Fig. 3
figure 3

The H2O2 and \({\text{NO}}_{2}^{ - }\) concentration along with pH of a 0.85% (m/v) NaCl solution following the 300-, 600-, and 1200-s He-DBD treatment

Evaluation of the Effect of the Portable He-DBD-Based Reaction-Discharge System on the Wound Healing

The effect of the He-DBD plasma on the wound healing was examined using the developed portable reaction-discharge system operated under the optimal conditions. The treatment of the normal human skin cell lines with He-DBD for 0 and 90 s was found to be non-toxic. The longer treatment lengths were established to induce a strong cytotoxic effect (data not shown). Therefore, all the functional tests presented in this study were performed with the studied cells that were treated with He-DBD for 0–90 s.

Figure 4 shows the augmentation of the proliferation of the HSkMEC.2 cells activated with He-DBD for 10 s, the MSU-1.1 cells activated with He-DBD for 60 s and the HaCaT cells activated for 10–90 s on Day 5 as compared to the untreated cells (no He-DBD activation). These results showed that He-DBD might stimulate the proliferation of the normal human skin cells, which are involved in the wound healing process. Similar results were obtained by other research groups, who observed the increased proliferation of the cells treated with He-DBD over the short time periods [17, 65]. In comparison to the epidermal keratinocytes and dermal fibroblasts, which are relatively robust towards the CAPP treatments, our results showed that the endothelial cells were more sensitive to the direct He-DBD treatment. Longer than 10 s time plasma exposure reduced skin endothelial cells proliferation and may be unfavorable for their angiogenic activity. Moreover, in this study for the first time has been documented proliferation of human non-malignant endothelial cells of skin-origin treated with CAPP.

Fig. 4
figure 4

The proliferation of the skin endothelial cells (HSkMEC.2), fibroblasts (MSU-1.1), and keratinocytes (HaCaT) treated with the He-DBD plasma sustained in the developed reaction-discharge system for 10–90 s from 0 to 7 days. As a control, the cells not treated with He-DBD (0 s) were used. Data are presented as mean ± SD of three analyses done in triplicate. The statistical comparison was performed using the Mann–Whitney test (*p < 0.05)

For proper wound closure, migration of cells involved in wound healing is necessary. In vitro cell migration was assessed by scratch test and proved heightened migratory potential of skin vessel endothelial cells, keratinocytes, and fibroblasts treated with He-DBD compared to control (Fig. 5). Similar results were obtained in previously published papers where keratinocytes and fibroblasts were used [66, 67]. In our study the in vitro tests were extended and combine all population of skin-derived cells such as keratinocytes, fibroblasts, and endothelial cells to assess their motility after direct He-DBD treatment. There was a significant increase in cell motility and scratch area closure for short treatment times (10 s) in all cell lines. Obtained results approves that He-DBD treatment might also influence the wound healing process by accelerating migration ability of skin-origin cells involved in regenerative effect in the wound site.

Fig. 5
figure 5

The scratch assays with the skin endothelial cells (HSkMEC.2), the fibroblasts (MSU-1.1), and the keratinocytes (HaCaT) were performed to monitor the artificial in vitro wound healing. All cell types were grown to confluence after the exposure to He-DBD (0–90 s). Subsequently, the monolayers were scratched, and the rate of the migration was measured after 16 and 24 h, or 24, 48, 72, and 84 h by quantifying the relative wound closure (RWC %). There was a significant increase in the cell motility for the short treatment times in all cell lines. Data are presented as mean ± SD of three analyses done in triplicate. Statistical analysis was performed using Mann–Whitney test and comparison to untreated controls (*p < 0.05)

Proper wound revascularization has beneficial effect to accelerate wound healing. Pseudo-vessel tube formation in vitro was induced when the cells were treated with He-DBD for 30 s compared to the untreated control (0 s), whereas the HSkMEC.2 cells formed reduced numbers of tubes after 60 s of the He-DBD treatment and did not form any vascular sprouts after the 90-s exposure to He-DBD (Fig. 6).

Fig. 6
figure 6

The tube formation by the human endothelial skin cells (HSkMEC.2) after the direct He-DBD treatment. The HSkMEC.2 cells were directly treated with He-DBD for 0, 10, 30, 60, or 90 s and the number of tubes was determined using the Matrigel® tube formation assay after 6 h. a The images showing the tube formation by the HSkMEC.2 cells after various times of He-DBD treatment b the evaluated mean mesh size, the total length, and the number of nodes of the HSkMEC.2 cells as determined by the ImageJ software. Data are presented as mean ± SD of three analyses done in duplicate. Statistical analysis was performed using Mann–Whitney test and comparison to untreated controls (*p < 0.05)

All the results were in accordance with the observations published by Arndt and co-workers showing that the exposure of the human endothelial cells to the plasma for the short periods stimulated their angiogenesis [68]. For that reason, the developed portable He-DBD-based reaction-discharge system might be an interesting and viable alternative to other CAPP-based devices described so far in the literature, distinguishing due to its simplicity and versatility. Thus, in vitro studies documented that short-term CAPP treatment (10 s) improve keratinocytes and fibroblasts migration, enhanced pro-angiogenic activities and accelerate closure of gaps in scratch test. This observation suggests that He-DBD treatment may be applied in vivo to enhance cutaneous wound healing.

Conclusions

As a remedy to the high demand for new, fast, and efficient therapies related to the activation of the skin cells responsible for the wound healing, the portable and safe He-DBD-based reaction-discharge system was developed and dedicated to the activation of the normal human skin cells. Because of the variety of the operating parameters that might affect the gas temperature, the DoE followed by the RSM were applied to reliably find the optimal conditions under which the gas temperature in the developed portable He-DBD-based reaction-discharge system would be close to 37 °C (310 K). Using this multivariate optimization approach it was established and verified that the developed reaction-discharge system should be operated at the frequency of modulation of 2.14 kHz, the duty cycle of 74%, and the He flow rate of 10.6 L min−1. The examination of the reactive species formed in the gaseous phase of He-DBD as well as in the He-DBD-treated liquid was performed in order to find out which reactive forms were produced during the CAPP operation. It was shown that our portable He-DBD-based reaction-discharge system successfully stimulated the skin cell (including keratinocytes, fibroblasts, and endothelial cells) proliferation and migration, as well as the angiogenesis of the endothelial cells. It was observed that by changing the He-DBD treatment time, the biological effect could be varied. In addition, it is worth noting that the results of the present work on the optimization of the developed He-DBD-based reaction-discharge system for the activation of the normal human skin cells are competitive to the well-described and studied atmospheric pressure plasma jets. Future studies must reveal which reactive species produced during the CAPP operation are finally responsible for the determined activation of the skin cells biology and the CAPP treatment in the in vivo model.