In bacterial biofilms, bacteria arrange themselves in three-dimensional structures, which adhere to the surface. Within a biofilm, bacteria interact with each other and an extracellular polymeric matrix provides the community with protection from hostile environment and exposure of antimicrobial agents (Hoiby et al., 2010; Zhou et al., 2015). The extracellular matrix is an important factor in antibiotic resistance. It acts primarily in slowing down the penetration of antibiotics to the cells inside the biofilm, thus reducing the effective dose of antibiotics delivered to the cells. The decreased rate of diffusion of antibiotics into the biofilm enables cellular expression of genes mediating resistance to antibiotics to be activated in the deeper layers of the biofilm (Donlan and Costerton, 2002; Hoiby et al., 2010). Slow penetration of antibiotics inside the matrix material gradually lowers the growth rate of biofilm-associated bacterial cells, which slows down the intake of antibiotics. In addition, the cells killed by the antibiotic in the upper layers of the biofilm “dilute” antibiotics on a per cell basis (Donlan and Costerton, 2002; Bagge et al., 2004; Mai-Prochnow et al., 2008). Biofilms structures also naturally limit the penetration of nutrients to the inner cell layers, and the lack of nutrients results in persistent cells that have higher tolerance to antibiotics (Costerton et al., 1999; Donlan and Costerton, 2002). It has also been proposed that biofilms cannot be completely eradicated due to the survival of persistent cells which accumulate mutations leading to antibiotic resistance (Lewis, 2007). Bacterial cells in biofilms also show increased resistance to the host immune system (Leid et al., 2005; Cerca et al., 2007). Bacterial biofilms are thus recognized as an important cause of chronic infections, including biofilm formation on medical devices, in wounds, and in immunocompromised patients (Costerton et al., 1999; Bjarnsholt et al., 2008). Common bacteria, such as Staphylococcus epidermidis, Escherichia coli, and Pseudomonas aeruginosa, that are generally opportunistic or pathogenic, can cause severe chronic infections in immunocompromised individuals (Kaper et al., 2004; de Bentzmann and Plesiat, 2011). The inappropriate use of antibiotics can cause bacteria to develop resistance, rendering infections by antibiotic resistant bacteria which is a major emerging threat (Ventola, 2015a). According to WHO, the global threat of antibiotic resistance is likely to usher a post-antibiotics era (Kostakioti et al., 2013; WHO, 2014). Thus, in addition to managing the usage of antibiotics, there is a significant need of novel approaches for treatment of bacterial infections (Ventola, 2015b). When it comes to bacterial biofilms, antibiotics or other known antimicrobial agents are not strong enough to eradicate them completely (Hoiby et al., 2010; Hoiby, 2011). In this context, the use of a combined therapeutic strategy is recommended to obtain a strong bactericidal effect against bacterial biofilms.

Plasma has been defined as the fourth state of matter, in addition to solid, liquid, and gas (Tendero et al., 2006). The plasma state can be described as an ionized gas, where ionization is obtained by adding energy to the gas. In case of medical applications, the ionizing energy most commonly comes from electricity (Tendero et al., 2006). For plasma to be used in medicine, it has to be generated at atmospheric pressure and the gas temperature must be suitable for treatment of living tissue, with minimal damage to surrounding healthy tissue (Laroussi, 2009). Plasma used in medicine is therefore often called cold atmospheric plasma (CAP; Graves, 2014). In medical treatments, plasma exposure is indirect or remote, where the plasma itself does not come into contact with the tissue, rather, it is an afterglow of the ionized gas (Kong et al., 2009; Bárdos and Baránková, 2010). These plasma sources are called remote plasma jets. In these devices the operating gas flows between two coaxial electrodes, an inner electrode usually powered with radiofrequency (RF) and an outer one grounded. One or both electrodes can be covered with a dielectric barrier (DBD), resulting in an ionized gas jet out of the equipment's nozzle (Bárdos and Baránková, 2010). CAP has received considerable attention in microbiology due to its high bactericidal activity (Maisch et al., 2012; Mai-Prochnow et al., 2014). The main killing mechanism of CAP is the generation of reactive oxygen and nitrogen species (RONS), which have an adverse effect on the cellular biochemistry: damaged proteins and nucleic acids (Joshi et al., 2010; de Geyter and Morent, 2012). In addition, CAP has been shown to have low toxicity toward healthy human cells (Joshi et al., 2010; de Geyter and Morent, 2012). The current CAP-based treatments are not efficient enough in eradicating persistent bacterial biofilms, and we therefore looked for a combined therapeutic strategy which could complement it. Vitamin C, which is a common food additive and can be safely used in medical treatment, is known to be a ROS-generating agent, and can kill some bacterial species (Vilchèze et al., 2013). We investigated if vitamin C pre-treatment could enhance the antibacterial efficacy of CAP at different exposure times on a 48 h bacterial biofilm. The study has been conducted on a non-pathogenic Gram-positive model bacterium Bacillus subtilis, and three facultative pathogens: Gram-negative E. coli and P. aeruginosa, and Gram-positive S. epidermidis. The non-pathogenic B. subtilis was very sensitive to CAP alone, but for the three pathogens with CAP-resistant biofilms, the pre-treatment with vitamin C dramatically increased the bactericidal effect.

CAP has emerged as a useful tool for treating local microbial infections and decontamination of various biomedical surfaces (Cahill et al., 2014; Flynn et al., 2015). Previous studies have demonstrated complete eradication of planktonic bacteria after a couple of minutes of CAP treatment (Yu et al., 2006; Flynn et al., 2015). Several studies have also tested the efficacy of cold plasma against the bacterial biofilm (Ermolaeva et al., 2011; Flynn et al., 2015). Their results suggested that bacterial biofilms are more resistant to CAP, and there is a need to increase the duration of exposure for complete decontamination. However, there are no studies regarding the long term cold plasma exposure. Biofilms are considered to be up to 1,000 times more resistant than the planktonic state of bacteria, and it accounts for 60% of human infections, with a high probability of becoming resistant to antimicrobials (Mai-Prochnow et al., 2015). In this study we evaluated the anti-biofilm activity of long term CAP exposure (5–60 min) against four different bacterial strains, one non-pathogenic and three facultative pathogens. Furthermore, we also evaluated whether vitamin C pre-treatment could be used to shorten the duration of the CAP treatment while achieving the same effect.

The response to CAP exposure for both the Gram-negative and Gram-positive bacteria seems to be species dependent. As shown in the Figure 1 and Table 1, non-pathogenic B. subtilis is highly sensitive to CAP exposure, dropping to under 2% survival after 5 min. E. coli and S. epidermidis were remarkably resistant to CAP treatment. These results indicated that biofilms of E. coli, P. aeruginosa, and S. epidermidis tend to be more resistant to CAP compared to B. subtilis, and that different biofilm matrix compositions, which are species-specific, can play a significant role in CAP resistance. The initial viability drop occurred after 5 min of treatment, but then saturation ensued, and the biofilms exhibited as much as 10–15% live bacteria even after 60 min of treatment. P. aeruginosa was almost completely eradicated after a 60-min exposure, but its inactivation curve was different. The biofilm survived quite well up to 30 min of treatment, and then the viability dropped radically from 30 to 60 min.

Such “two-slope” inactivation curves were reported previously for biofilm inactivation (Joaquin et al., 2009). The first killing phase corresponds to fast destruction of the top layers of the biofilm, and is followed by the slower destruction of the bottom layers, which are covered by dead cells and exopolysaccharide biofilm matrix. It was also reported that the CAP treatment triggers the bacterial cell entry into a viable-but-non-cultivable (VBNC) survival state during the initial drop in the inactivation curve, followed by the second phase in which the cells are actually killed. The VBNC state has been reported for many Gram-negative bacteria, which undergo morphological changes and decrease in size (Joaquin et al., 2009).

Since the CFU method does not differentiate the dead cells from the ones in the VBNC state, it is important to combine it with other methods for confirmation. Our live/dead fluorescent staining images (Figure 2) confirm the 30-min results of the CFU counting, with B. subtilis being most sensitive, and the three facultative pathogens comparably more resistant. Taken together, the CFU counts and live/dead staining suggest that a large portion of P. aeruginosa cells might be in the VBNC state up to 30 min of exposure, and are finally killed after 60 min.

For all the tested facultative pathogens, a 5 min CAP treatment (which is a feasible duration for medical treatments) did not efficiently eliminate the biofilm. The survival after 5 min of CAP treatment was 25% for E. coli, 65% for S. epidermidis, and 100% for P. aeruginosa (Figure 1). Since the bactericidal effect of CAP is based on oxidative stress due to generation of ROS, we investigated whether the efficiency of CAP could be enhanced by pre-treatment with other ROS-generating agents known to be harmless to humans. Vitamin C, an important dietary supplement for humans, also has been demonstrated to enhance the antimicrobial activity of several antimicrobial agents (Cursino et al., 2005; Khameneh et al., 2016). Therefore, we pre-treated the biofilms for 15 min with 5 mM vitamin C, also known to generate ROS in most bacterial cells (Vilchèze et al., 2013). The result was very encouraging. Vitamin C alone lead to some loss of viability, but most importantly, when followed by a 5 min CAP treatment, the viability of biofilms was reduced to only 1.9% for E. coli, 17.6% for S. epidermidis, and 10.7% for P. aeruginosa (Figure 4). This was a dramatic improvement (three- to five-fold) over the 5 min CAP treatment alone. Our results indicate that in combination with vitamin C, a very short CAP exposure can be efficient against resilient biofilms of bacterial pathogens. The powerful synergistic effect of CAP combined with vitamin C is clearly visible in the Figure 5, where the treatment is affecting the membrane integrity. Vitamin C treated cells were more susceptible to collapsing during sample processing and as a result more altered S. epidermidis cells were observed followed by complete disintegration of cellular structure after subsequent exposure to cold plasma. The pretreatment with vitamin C might show great impact on current CAP based treatment strategy such as wound and root canal treatment by enhancing the bactericidal effect. In vivo studies are needed to evaluate the effect of proposed combined treatment on real multispecies biofilm consortium.

Overall, our results suggest that a brief pre-treatment with vitamin C, followed by a very short CAP treatment could be a viable route for treating various types of infectious lesions.

SP, SH, and VM designed and carried out the experiments. SP, SH, VM, and IM analyzed the data and wrote the manuscript. FW and PA made substantial contributions to conception and design of the work and analysis and interpretation of data, and they critically revised the manuscript.

This work was funded by grants from the Chalmers University of Technology and VINNOVA to IM and FW, and ÅForsk to IM.