Antimicrobial effects of microwave‐induced plasma torch (MiniMIP) treatment on Candida albicans biofilms

Summary The susceptibility of Candida albicans biofilms to a non‐thermal plasma treatment has been investigated in terms of growth, survival and cell viability by a series of in vitro experiments. For different time periods, the C. albicans strain SC5314 was treated with a microwave‐induced plasma torch (MiniMIP). The MiniMIP treatment had a strong effect (reduction factor (RF) = 2.97 after 50 s treatment) at a distance of 3 cm between the nozzle and the superior regions of the biofilms. In addition, a viability reduction of 77% after a 20 s plasma treatment and a metabolism reduction of 90% after a 40 s plasma treatment time were observed for C. albicans. After such a treatment, the biofilms revealed an altered morphology of their cells by atomic force microscopy (AFM). Additionally, fluorescence microscopy and confocal laser scanning microscopy (CLSM) analyses of plasma‐treated biofilms showed that an inactivation of cells mainly appeared on the bottom side of the biofilms. Thus, the plasma inactivation of the overgrown surface reveals a new possibility to combat biofilms.


Introduction
Non-thermal plasmas (NTPs) are also known as nonequilibrium plasmas, honouring the fact that the cooling of ions and uncharged particles is more effective than an energy transfer of energetically excited electrons to the latter particles. Thus, the gaseous environment is not heated up, which is contrary to thermal plasmas (Fridman, Friedman et al., 2008). Today, NTPs play an increasingly important role in both medicine and industry. They combine the advantage of non-thermal operation and high antimicrobial activity (Surowsky, Schl€ uter et al., 2015). Hence, NTPs are used in today's medicine for many different applications such as wound healing (Shekhter, Serezhenkov et al., 2005;Ghaffari, Jalili et al., 2007), cell detachment as well as reattachment (Kieft, Darios et al., 2005;Kieft, Kurdi et al., 2006) and biological decontamination (Laroussi, Alexeff et al., 2000;Laroussi, Mendis et al., 2003). Due to their capability of biological decontamination, NTPs have gained interest in many industries. In the dairy and food industry, biofouling caused by microbial biofilms is a serious problem, which leads to a considerable loss of resources. This is caused not only by the microbial contamination of end-products such as meat, fruits and vegetables (Kumar and Anand, 1998), but also by increased corrosion rates at processing surfaces and increased fluid frictional resistance and heat flow across the surface (Criado, Suarez et al., 1994).
The predominant microbial life cycle implies attachment to solid surfaces and the formation of three-dimensional, multicellular aggregates called biofilms, which are embedded in a self-produced, extracellular matrix (ECM; Steenackers, Parijs et al., 2016, Serra, Hidalgo-Bastida et al., 2017. Especially, food-borne pathogens and spoilage microorganisms like Candida albicans prefer biofilm formation on stainless steel, aluminium, glass, polytetrafluoroethylene (PTFE) seals and polyamide (PA) material, which are typically found in food-processing environments (Herald and Zottola, 1988;Mafu, Roy et al., 1990;Notermanns, Dormans et al., 1991). Candida spp., for example, can often be isolated from conveyor tracks of the food and beverage industry (Loureiro and Malfeito-Ferreira, 2003;Brugnoni, Lozano et al., 2007).
Economic losses in the food and healthcare sector as well as a more sensitive public awareness put food safety into the spotlight. Consequently, European countries implemented the Good Manufacturing Practice (GMP) and Hazard Analysis Critical Control Point (HACCP) for food industries. As a consequence, in 2016 the European Food Safety Authority was able to identify 49,950 foodborne outbreaks resulting in illness, 3869 hospitalizations and 20 deaths (European Centre for Disease Prevention and Control & European Food Safety Authority, 2017). Notably, 99,392 outbreaks resulting in illness were reported in the United States, leading to 2625 hospitalizations and 115 deaths (Centers for Disease Control and Prevention, 2016). Conventional methods like the mechanical removal of biofilms by high-pressure cleaners or brushing and wiping indeed lead to a loss of biomass, but not to an efficient extent. Due to intense and brief-acting shear forces, the majority of the biomass is removed but a thin biofilm most likely remains which further grows much denser with an increased resistance (Liu and Tay, 2002). Additionally, in older mature biofilms, significant components of the biofilm often remain, causing the biofilm to grow again (Jang, Rusconi et al., 2017). The treatment of biofilms with biocides is just as problematic as the biofilms show increased resistance to these biocides compared to their planktonic counterparts. This is mainly due to the extracellular matrix which prevents deep penetration of the biofilm with the biocides (Bridier, Dubois-Brissonnet et al., 2011). This leads to low concentrations in the deeper layers of the biofilm, which in turn leads to increased horizontal gene transfer of resistance genes (Jutkina, Marathe et al., 2018). Thus, there is a strong need for novel strategies combating biofilms in the food industry. These range from the treatment of biofilms with special oils (Kerekes, Vid acs et al., 2015) and enzymes (Meireles, Borges et al., 2016) to bacteriophages. The treatment of microorganisms with NTPs is a constantly growing field in which new insights into the effects are constantly being gained (Sladek, Filoche et al., 2007;Koban, Holtfreter et al., 2011;Xu, Tu et al., 2011;Alkawareek, Algwari et al., 2012;Ermolaeva, Sysolyatina et al., 2015;Flynn, Higginbotham et al., 2015). Results obtained in the here presented study suggest that the microwaveinduced plasma torch (MiniMIP) is a powerful tool for microbial decontamination. In order to meet the industrial requirements, it is of particular importance to find standardized parameters for the plasma source. Despite antimicrobial effects of a different plasma source on C. albicans having been shown (Handorf, Weihe et al., 2018), no studies exist, which investigated the effects of the MiniMIP on C. albicans biofilms up to this date.

Results
The path of the development from a new plasma source concept to its effective use in a specific application also includes the investigation of a potential antimicrobial impact on surface-bound biofilms and their basic physical adaptations to it. Consequently, the presented work summarizes test series to determine the antimicrobial effect of the MiniMIP plasma source on eukaryotic biofilms. Due to its ubiquitous presence in medical and food sectors, C. albicans has been chosen as a model microbe, because its biofilm formation has been intensely studied and it is contaminant of medical as well as industrial importance (Kabir, Hussain et al., 2012;Morata and Loira, 2017). In particular, C. albicans strain SC5314 is known for its rapid vertical growth and could be compared with Saccharomyces cerevisiae, a yeast of great importance in the food and beverage industry because of its high sugar consumption and the fermentation of juices into alcoholic end-products (Battey, Duffy et al., 2002;Walker and Stewart, 2016;Lorenzini, Simonato et al., 2019). This study investigated the viability of the cells [revealed via the fluorescence assay (2.5)] and cellular metabolism [XTT assay (2.6)] and was complemented by the determination of the posttreatment viability (CFU 2.4). Additionally, we used fluorescence microscopy (2.7), CLSM (2.8) and AFM (2.9) to obtain optical evidence of the plasma influence on the cells. Finally, OES (2.10) was used to give a general overview of the chemical composition of the plasma gas.
Effects of plasma treatment on the proliferation of the cells A reduction factor (RF) was calculated as the difference between the log 10 (CFU) of an untreated control and the log 10 found for the samples after the treatment. The controls hosted in average 10 6 cells. The RFs quantify the inhibitory effect of the plasma treatment. The treatment with the MiniMIP revealed a RF of 2.97 after a 50 s treatment (Fig. 1A). The RF for the 10 s treatment with the MiniMIP shown to intersect the x-axis was not statistically significant. A continuous increase in RF could be detected up to a 40 s plasma treatment. Longer plasma treatment times did not lead to a significant increase in RF.

Effects of plasma treatment on the viability of the cells
The fluorescence LIVE/DEAD assay, which was used to detect the viability of the cells after the plasma treatment, showed a declined G/R ratio from 2.6 to 0.58 after a 20 s MiniMIP plasma treatment (Fig. 1B), which corresponds to a 77% reduction of the G/R ratio. The maximum reduction observed for a MiniMIP treatment was reached after 20 s. Longer treatment times revealed no further changes in the viability.
Effect of plasma treatment on the metabolism of the cells The XTT assay, which was used to determine the metabolic activity of the cells after plasma treatment, showed a reduction in the absorption from 2.09 to 0.20 after a 40 s MiniMIP plasma treatment (Fig. 1C). It represented a reduction of 90%. No further decrease in the cell metabolism could be measured for longer treatment times.

Fluorescence microscopic confirmation of plasma treatment effects
Fluorescence microscopy indicated a massive impact of the plasma treatment on cells which were located on the bottom of the biofilms already after a 20-30 s treatment (Fig. 2C,D). The biofilm degenerated from its exterior, and the inactivation of cells amplified into the centre of the biofilm with increasing treatment times ( Fig. 2E-G). After 60 s treatment time, almost the entire biofilm was affected by the treatment (Fig. 2G).

CLSM confirmed cell inactivation processes predominantly on the bottom of the biofilms
The influence of a plasma treatment on the three-dimensional structure of a C. albicans biofilm was studied by CLSM (Fig. 3). The right side of the figure shows an inferior view of the 3D biofilm structure, which reveals highly influenced regions already after a 10 s treatment. Although the propagation of inactivated cells in the biofilm was more pronounced in the inferior regions, an inactivation was already noticeable in more peripheral layers of the biofilm, which became obvious in a 3Dmodel ( Fig. 3) as well as in orthogonal views of the top layers. With increasing plasma treatment time, almost the entire bottom of the biofilm was inactivated and clear effects were visible on the top layers of the biofilms. Furthermore, increasing separations and holes in the biofilm could be detected with longer treatment times.

AFM confirmed alterations in the cell morphology of the biofilm after plasma treatment
With the aid of AFM, more profound insights in the cell morphological alteration of the biofilm after plasma treatment have been obtained. Due to the method of cantilever visualization, AFM can only visualize the morphological changes of the cells on the surface of the biofilm. The control cells appeared vital and commonly shaped (Fig. 4A). In contrast, cells treated with plasma for 60 s appeared more spherical and partially ruptured with distinct sites fractures. In addition, they were smaller than the control cells (Fig. 4C). OES demonstrated the chemical composition of the plasma gas OES detects molecular and atomic emission bands of electromagnetic radiation, which provide information about the plasma composition. Unfortunately, differences in the peak height do not simultaneously indicate differences in the quantity of the molecules. Nevertheless, qualitative statements could be made based on the molecule spectra. Molecular absorption bands for nitrogen and hydroxyl groups were obtained, as well as spectral lines of atomic oxygen and argon in the measurements of the plasma gas. However, additional bands in the range of 520-620 nm were visible in the effluent of the MiniMIP compared to the kINPen09, a well-studied plasma device with regard to its plasma gas composition (Fig. 5).
Each of these plasma sources have specific application areas for which they are suited best. Plasma needles, for example, have already been used in dentistry for root canal treatments (Sladek, Stoffels et al., 2004;Goree, Liu et al., 2006). Radiofrequency plasma jet (RFPJ) like the kINPen09 or the new version kINPen MED has already been used for chronic wound healing in medicine (Lademann, Ulrich et al., 2013;Bekeschus, Schmidt et al., 2016). DBD is a promising tool for the microbial decontamination of water (Baroch and Saito, 2011) and has already been applied for the treatment of surfaces (Oehmigen, H€ ahnel et al., 2010;Baroch and Saito, 2011;Banaschik, Luke s et al., 2015), for example textiles (M€ uller, Zahn et al., 2010;Simor, Creyghton et al., 2010). In comparison, the decontamination of biofilms with cold atmospheric pressure plasmas is an emerging field of research with a series of promising results (Machala, Chladekova et al., 2010;Ehlbeck, Schnabel et al., 2011;Misra, Tiwari et al., 2011;Scholtz, Pazlarova et al., 2015;Liguori, Cochis et al., 2017).
So far, microwave-induced plasmas (MIPs) are mainly used in spectroscopy for the analysis of gas components (Broekaert and Engel, 2006), surface modifications (Jia, Kuraseko et al., 2008) or the processing of biogas (Tippayawong, Chaiya et al., 2015). Not much is known about microwave plasmas which are already used in medical technology apart from SteriPlas (Adtec, Hounslow, UK). The latter technology is based on the microwave plasma technology of the Max Planck Institute for extraterrestrial Physics in Germany. During an 8 year period, chronic wounds of 379 patients were treated with this plasma source and significant reductions in the bacterial count were detected (Isbary, Morfill et al., 2010). Based on the RFs obtained in our experiments, the Mini-MIP shown in this study appeared as a versatile tool to combat microbial biofilms in the food and beverage industry.
In general, a distinction has to be made between the type of treatment and the type of plasma generation/ ignition (Niemira, 2012). In this work, we used a microwave-driven plasma source (type of plasma generation) and an indirect treatment of the biofilms (type of treatment). For a comparison, the most reasonable way is to work with plasma sources with a comparable power output and the same working gas. If two plasma sources have to be compared in their antimicrobial effects, they should be used in a standardized assay (Mann, Schnabel et al., 2015;Sarangapani, Patange et al., 2018). For instance, Ehlbeck et al. (2008) showed the treatment of contaminated PET bottles with microwave-induced plasma. In this case, the type of treatment and the type of plasma generation were the same and they reached reduction factors of up to 7. Contrary, comparison of the  reduction factors is difficult, as the plasma used in that work had a much higher power and was operated with compressed air instead of argon gas.
The trend for microwave plasmas in microbial decontamination tends more in the direction of plasma-processed air (PPA) or plasma-treated water (PTW) which subsequently affect the microorganisms (Schnabel, Andrasch et al., 2014;Thirumdas, Kothakota et al., 2018;Schnabel, Handorf et al., 2019). Especially in terms of industrial manufacturing, PPA and PTW offer advantages over a conventional plasma treatment. For instance, not all areas which need to be treated are easily accessible for plasma devices. In this context, washing and flushing processes with PTW or PPA can offer a decisive advantage.
Indirect treatment of microorganisms where the effluent has been brought into a short distance over the biofilms is published for plasma jets or DBDs (Maisch, Shimizu et al., 2012;Khan, Lee et al., 2016;Handorf, Weihe et al., 2018). If only the achieved reduction factors are considered for the same treatment time regardless of the type of plasma generation and type of application, RFs of 0.6 for the kINPen09, 2.9 for the hollow electrode dielectric discharge (HDBD) and 2.3 for the volume dielectric discharge (VDBD) were obtained after 1 min treatment time (Koban, Matthes et al., 2010). In relation to these results, the MiniMIP showed the strongest inactivation with a reduction factor of 2.92 after 1 min treatment time (Fig. 1A).
In the present study, three different methods (CFU, fluorescence and XTT assay) were used to investigate the viability of C. albicans cells organized in a biofilm, after a plasma treatment. However, these methods do not reveal the same effects on the cells. The CFU shows the ability of the cells to proliferate after treatment. The fluorescence assay indicates membrane damage of the cells, and the XTT assay monitors the metabolic activity of the cells after treatment. The results of the methods cannot be directly compared. Not only based on the current study, the increasing importance of viable but nonculturable (VBNC) cells becomes more and more obvious but should be discussed more nuanced, and it is essential that the results of several different methods are considered together. Concretely, together they give a comprehensive overview which makes it possible to exclude a VBNC status as far as possible (Ramamurthy, Ghosh et al., 2014;Saprykina, Bolgova et al., 2016;Bolgova, Saprykina et al., 2017). The results (Fig. 1) showed that already 20 s after treatment the cells had significant membrane damages and their metabolic activity was strongly reduced. Their ability to proliferate was also severely restricted.
In terms of three-dimensional effects, RFPJ treatments showed centralized spots of dead cells while the majority of biofilm cells were not affected by plasma treatment (Handorf, Weihe et al., 2018). This occurred mainly due to the relatively fine and centralized effluent of plasma jets. In contrast, a very broad effect has been demonstrated in the treatment of biofilms with the MiniMIP, which rather spreads from the marginal areas to the centre and affected the complete biofilm already after 20-40 s (Fig. 2C-E). Most publications showed that the biofilm structures with increasing treatment times were either inactivated on their surfaces or completely inactivated during the course of the treatment (Pei, Lu et al., 2012;Traba and Liang, 2015;Delben, Zago et al., 2016). This could have been a result of the higher power of the plasma sources or the longer treatment times.
It is very likely that the applied plasma treatment times of 10-60 s mirror the dynamic range. In the indirect treatments, the formed reactive oxygen species (ROS) and reactive nitrogen species (RNS) played a major role in inactivation Xu, Shen et al., 2015;Ziuzina, Boehm et al., 2015). There may have been an increase in the concentration of ROS/RNS in the liquid residues of the biofilm that have accumulated on the bottom of the biofilm. In combination with water, RNS such as nitrates and nitrites are to be expected when treating with the MiniMIP. Consequently, PTW was generated at the bottom layers of the biofilm and predominantly led to an inactivation in that area. This was also indicated in treatments with the kINPen09 in the same experimental set-up (Handorf, Weihe et al., 2018).
The AFM images revealed differences in the cell morphology between a treatment with the MiniMIP or the kINPen09 (Handorf, Weihe et al., 2018). It is most likely caused by the device-dependent reaction pathways, which lead to ROS/RNS (Yusupov, Neyts et al., 2012;Gilmore, Flynn et al., 2018). However, more in-depth investigations of the MiniMIP plasma gas constituents have not yet been completed. The PLexc microwave plasma comes closest to the gas physics of the MiniMIP (Pipa, Andrasch et al., 2012). Investigations of the gas physics of this plasma source have shown that RNS is mainly produced during the latter process (Schnabel, Handorf et al., 2019). This knowledge was additionally supported by the temperature-dependent dissociation rates of the different gas molecules (Drost, 1980). Investigations have already shown that the temperature in the plasma discharge area within the MiniMIP is above 2000°C, where ROS were already dissociated and mainly RNS still exist (Baeva, B€ osel et al., 2012). Therefore, the OES spectra in this paper serve as an overview of the gas molecules.
Notably, plasma treatment could overcome various limitations of conventional antimycotic drugs. For instance, it is well-known that reactive species of plasma gases are able to overcome the barrier of the ECM and inhibit cells in the biofilm or even completely degrade the ECM (Delben, Zago et al., 2016;Modic, McLeod et al., 2017;Gilmore, Flynn et al., 2018). C. albicans exploits a range of resistance mechanisms to conventional antimicrobial treatment strategies. Its resistance is primarily due to the activity of efflux pumps, the production of an ECM and the presence of recalcitrant persister cells during biofilm growth (Nobile and Johnson, 2015). The efflux pumps of C. albicans comprise two major classes: the ATP-binding cassette transporter superfamily and the major facilitator class (Anderson, 2005;Cowen, 2008). The ECM represents a mechanical barrier to drugs and thus leads to higher drug resistance of C. albicans cells embedded in biofilms (Baillie and Douglas, 2000;Al-Fattani and Douglas, 2006). Persister cells are a small subset of metabolically dormant yeast cells in biofilms that are extremely resistant to antimycotics (LaFleur, Kumamoto et al., 2006). Comparative studies between plasma treatment of C. albicans biofilms and treatment with antimycotics or chemical disinfectants showed a stronger reduction in the CFU during plasma treatment and indicated no correlation between plasma effects and efflux pumps of the pathogen (Koban, Matthes et al., 2010). Furthermore, it could be shown that pre-treatment with plasma even significantly increases the effect of antimycotics on the pathogens (Sun, Yu et al., 2012).
The results of our work offer new promising fields of application for MIPs. The results of the MiniMIP compared to already well-investigated plasma sources showed a stronger reduction of C. albicans biofilms in a shorter treatment time despite a larger distance of the plasma source to the surface of the biofilm (Handorf, Weihe et al., 2018). The MiniMIP is therefore a promising new tool in the field of plasma-based antimicrobial decontamination of biofilms. Although the investigation of the effect of MiniMIP on prokaryotic and eukaryotic biofilms is still in its infancy, the results shown here are highly motivating and represent a solid basis for further investigations and applications of this innovative plasma source.

Conclusion
Our results show the influence of plasma treatment of C. albicans biofilms with the microwave-induced plasma source MiniMIP for the first time. A stronger influence of the MiniMIP on the biofilms could be shown within a shorter plasma treatment time and greater distance compared to already commercially available and well-studied plasma sources, for example the RFPJ kINPen09. The composition of the plasma gases can be of decisive importance for the different results shown in the CFU, the fluorescence and the XTT assay. Furthermore, an influence mainly on the bottom side of the biofilms could be shown with the aid of fluorescence microscopy and CLSM. Diffusion processes, water channels and plasma flow dynamics might be crucial for this phenomenon. Based on the highly reproducible and new findings, which are generated by the selection of the plasma source and its settings to a certain problem, it would be a decisive step for the industrial use of MIPs in several value chains. For instance, the removal of biofilms from an overgrown surface is of great interest in the food industry. If the inactivation of the bottom side of the biofilms shown in this work also led to its surface detachment, it could be a crucial advantage for the application in the industry.

Fungal strain and growth conditions
Because of its primarily vertical growth, C. albicans SC5314, which is a commonly used strain in laboratory experiments, was grown on Sabouraud agar 4% glucose (Roth, Karlsruhe, Germany) for 24 h at 37°C. Grown colonies were suspended in 10 ml phosphate-buffered saline (PBS; pH 7.2, according to S€ orensen) to an OD 600 of 0.375-0.385. Afterwards, 1 ml of the suspension was pipetted in 9 ml RPMI medium without bicarbonate (Merck, Darmstadt, Germany). From this inoculum, 200 µl was pipetted into each well of a 96-well plate and subsequently incubated at 37°C and 80 rpm on a rotary shaker for 90 min to achieve homogeneous oxygen distribution in the biofilm resulting in improved biofilm growth compared to a static grown biofilm. Subsequently, the medium was removed and each well was washed with 200 µl PBS and refilled with 200 µl RPMI medium to remove non-adhered cells. The plate was incubated at 37°C and 80 rpm on a rotary shaker for 24 h. 96-well coated polystyrol plates (Sarstedt, N€ umbrecht, Germany) were used for biofilm cultivation. For scanning probe measurements (SPM), 12-well plates with 12 mm coverslips (Sarstedt, N€ umbrecht, German) were used. This protocol was kindly provided by the research group of Christiane Yumi Koga-Ito, Institute of Science and Technology -UNESP, Department of Oral Sciences and Diagnosis (Borges, Lima et al., 2018).

Plasma source
The MiniMIP worked at a frequency of 2.45 GHz at atmospheric pressure with a forward power in the range of 20-200 W. The microwave discharge was induced in a ceramic tube, which has an inner radius of 0.75 mm and an outer radius of 1.5 mm. The length of the ceramic tube is approximately 31 mm (Fig. 6). The ceramic tube leads completely into the inside of the housing, where it is encased by an aluminium tube. The ignition of the plasma takes place at the front edge of the aluminium tube, and plasma propagation is driven by the surrounding microwave field and the gas flow along the longitudinal axis of the ceramic tube leading to a small plasma plume of about 10 mm in length and 3 mm in diameter outside the device (Baeva, B€ osel et al., 2012). In this work, the MiniMIP was operated with a forward power of 40 W and a reverse power of 20 W at a gas flow of 5 slm pure argon gas.
Plasma treatment of C. albicans SC5314 biofilms After the incubation, mature biofilms (2.1) were washed with PBS (pH 7.2 according to S€ orensen) followed by a complete removal of the liquid. Throughout the treatment, the plasma source was attached to a xyz table, holding a constant distance between the source and the biofilm surface. For the treatment, the MiniMIP was horizontally guided into the centre of the biofilm during the treatment and then brought into the desired distance to the biofilm via the computer control software. During the horizontal movement between the treatments of individual well, the plasma source was vertically positioned at the distance of approximately 30 cm (10 times of the treatment distance). A distance of 3 cm was chosen from the beginning of the outer part of the ceramic tube to the surface of the biofilms during the treatment. Each biofilm underwent one of six different treatment times. To avoid dehydration effects, all sample groups were treated in a row. For the colony-forming units (CFU), the fluorescence LIVE/DEAD assays and the XTT assays, 200 µl PBS (pH 7.2 according to S€ orensen) per well was added. Subsequently, the biofilms were mechanically removed from the overgrown surface by repeated pipetting of 200 µl PBS and the resulting cell suspensions were collected. To ensure the transfer of the entire biofilm, this step was repeated three times in total which resulted in a final suspension volume of 600 µl.

Analytical determinations
First, the influence of the temperature on a biofilm was determined. For temperature measurements, thermal images of the biofilms were acquired directly after plasma treatment with a FLIR thermal imaging camera (FLIR Systems, Frankfurt am Main, Germany) at distance of 20 cm. The temperature had no significant influence on the biofilms (data not shown). The viability of a C. albicans biofilm after a plasma treatment was analysed by counting the viable number of microorganisms via the CFU method. Therefore, a 1:10 serial dilution of the samples with maximum recovery diluent (MRD, 0.85% NaCl, 1% tryptone) was performed. Controls and the samples were diluted 1:10 000 and 1:1000 respectively. Each dilution step was plated on Sabouraud agar by pipetting 10 µl per dilution onto the plate and spread out using the tilting technique. Subsequently, the plates were incubated at 37°C for 24 h. The colonies of the dilution levels were counted, and the CFU/ml was calculated as follows: (1) 10 x = the dilution factor for the lowest dilution; v = the volume of diluted cell suspension per plate in ml; ∑c y = the total number of colonies on all (n y ) plates of the lowest evaluated dilution level 10 -x ; ∑c y+1 = the total number of colonies on all (n y + 1) plates of the next highest dilution level evaluated 10 -(x+1) (Bast, 2001). After the calculation (1), the reduction factor (RF) was determined as follows: RF ¼ MV k log 10 ÀMV p log 10 (2) MV k log 10 = the mean value of the CFU/ml of the reference group. MV p log10 = the mean value of the CFU/ml of the treated specimens.
For the final illustration of the data points, the weighted mean value of the total population was used with the formula:x The weight p i is calculated according to the following formula: where k = arbitrary constant. This ensures that the weight of the mean values is not included in the calculation as values. The weighted mean value has the advantage compared to the arithmetic mean value that it is more resistant to aberrations (Gr€ anicher, 1994). The propagation of error was calculated for each treatment group. This finally resulted in four different error propagations for each treatment time from which the weighted error was calculated and used as error bars in the illustration (Gr€ anicher, 1994). The experiments were repeated fourfold with n = 6.

Fluorescence LIVE/DEAD assay
The LIVE/DEAD BacLight TM Bacterial Viability Kit (Thermo Scientific, Waltham, USA) was prepared according to product instructions. Subsequently, 0.9 µl of the mixture was added to 300 µl of the sample solution (2.3) followed by an incubation on a rotary shaker in the dark at room temperature for 20 min. A fluorescence microplate reader (Varioskan Flashâ, Thermo Scientific, Waltham, USA) was used to determine the fluorescence of each well of a 96-well plate with an excitation wavelength of 470 nm and an emission wavelength of 530 nm (G, green) or 630 nm (R, red). Conclusively, a ratio G/R was calculated by dividing the fluorescence intensity value of green fluorescence by the value of red fluorescence.

XTT assay
A colorimetric assay was used to determine the cell viability after plasma treatment (XTT Cell Proliferation Assay Kit; AppliChem, St. Louis, MO, USA). Therefore, XTT was applied to reveal the cell viability as a function of redox potential, which arises from a trans-plasma membrane electron transport (Scudiero, Shoemaker et al., 1988). The sterile activation solution, which contains N-methyl dibenzopyrazine methyl sulphate (PMS) as an intermediate electron carrier, and the XTT solution were mixed 1:50. For each well, these mixtures were added at a ratio of 1:3 to the sample solution (2.3). The 96-well plate was incubated at 37°C with continuous horizontal shaking (80 rpm) in the dark for 2 h. After the incubation time, 96-well plates were scanned at a wavelength of 470 nm using the Varioskan Flash â device.
The obtained values were blank-corrected using XTT and activation solution mix without sample. The experiments were repeated fourfold with n = 6.

Fluorescence microscopy
Black 96-well plates with a glass bottom (PerkinElmer, Hamburg, Germany) were used for fluorescence microscopy. To avoid dehydration of the biofilms, plasma-treated biofilms were resuspended in 300 µl of 0.85% NaCl after treatment. The LIVE/DEAD BacLight TM Bacterial Viability Kit was used as previously described (2.5). Epifluorescence images were acquired using Operetta CLS High-Content Imager (PerkinElmer, Hamburg, Germany) using a 59 objective (air, NA = 0.16, Zeiss, Oberkochen, Germany). Depending on the experiment, several fields of view were recorded and combined in the software. SYTO TM 9 was excited by a 475 nm (110 mW) LED, and the fluorescence was collected with a 525 AE 25 nm band-pass filter. Propidium iodide was excited by a 550 nm (170 mW) LED, and the emission light was collected with a 610 AE 40 nm band-pass filter. A laser autofocus (785 nm) was available for all measurements. The images were displayed using Harmony 4.6 software.

Confocal laser scanning microscopy (CLSM)
Biofilms were cultivated (2.1), plasma-treated (2.3) and LIVE/DEAD (2.5)-stained as previously described. The supernatant was removed after the staining and the washing procedure. Subsequently, the biofilms were analysed using a Zeiss LSM 510 microscope (Carl Zeiss, Jena, Germany) equipped with a 109 objective (air, NA = 0.1). The filter and detector settings were adapted to the fluorescent dyes SYTO TM 9 and propidium iodide. The dyes were excited by an argon laser at 488 nm and the emission was collected at 505-530 nm (band-pass filter) and 650 nm (long-pass filter) respectively. Three-dimensional images were acquired using the ZEN 2009 software (Carl Zeiss) with an area of 1272.2 µm 9 1272.2 µm and z-stack sections of 5.5 µm.

Atomic force microscopy
For topographic atomic force microscopy (AFM), coverslips were placed on a gel-like mass Gelrite TM (Duchefa, Haarlem, Netherlands), which avoids unwanted adherence on the bottom side of the coverslip. Because of its rapid curing process, 50 ml Gelrite was autoclaved and directly used hereafter. For the preparation of the liquid mass, a 12-well plate was used, which was filled with 200 µl per well. While it cooled down, coverslips were placed at the surface of the hardening mass. The biofilms were cultivated as described above (2.1), except that 1 ml of the C. albicans-RPMI mix was pipetted to each well until the coverslips were completely topped with the mix. Subsequently, the 12-well plates were incubated at 37°C and 80 rpm on a rotary shaker for 90 min. Hereafter, the biofilms underwent additional washing steps with 0.85% NaCl. The biofilms were stored overnight in 37°C to ensure the growth of enough biofilm mass. The cultivation medium was removed, and the biofilm-overgrown coverslips were repeatedly washed with 1 ml 0.85% NaCl the following day.
For the experiments, samples were treated 30 s and 60 s with the MiniMIP. Dehydration of the coverslips before AFM analysis was avoided by using a humidity chamber. The AFM measurements were carried out on a DI CP II SPM (Veeco, Plainview, USA), which was mounted on a vibration-free object table (TS-150,  Table Stable, Zwillikon, Switzerland). The set-up was mounted on an optical bench encased by an additional acoustic protection. The AFM was equipped with a linearized piezo scanner, on which the coverslips were mounted with a metal sample holder with leading tabs. The samples were measured using cantilevers with nominal spring constant of k = 0.1-0.6 N 9 m 2 in contact mode, a frequency of 0.4 Hz and set point = 8 N/m 2 with a picture size of 20 µm 2 . Pictures were edited with Gwyddion (Czech Metrology Institute, Brno, Czech Republic).

Optical emission spectroscopy (OES)
For stable OES measurements, the plasma device was left in operation for 30 min until condensation water escaped from the gas pipes and the effluent was optimally adjusted. An optical fibre with an internal diameter of 400 µm was connected to the USB compact spectrometer AvaSpec 2048 (Apeldoorn, Netherlands), which was connected to a laptop. For better focusing, the optical fibre was clamped within a holding bracket. For measurement correction, a black cap was placed on the tip of the fibre optic cable and the dark current was determined. Afterwards, the cap was removed, the fibre optic cable was placed in front of the effluent, and the spectra were measured. The measurement was done for the MiniMIP and the kINPen09, an already commercially used radiofrequency plasma source, for comparison of the OES. The data were read out using MATLAB â (MathWorks â , Natick, MA, USA) and evaluated and displayed visually as diagrams using Excel.