Here we report that treatment with low level DCs can effectively reduce the viability of B. subtilis cells. The effects of DCs and pre-treated media on the viability, morphology and gene expression of B. subtilis were studied. There was less killing of biofilm cells by incubating in the pre-treated media than when the current was directly applied, especially for biofilms treated with 250 μA/cm2 (Figure 4). This finding suggests that the movement of ions or some transient species might be important for the killing of biofilm cells.
In contrast to the biofilm samples, planktonic cells were much more susceptible to DCs. However, planktonic cells exposed to current and to pre-treated media showed similar reduction in cell viability. It is possible that the presence of the biofilm matrix could reduce the effects of current-generated ions. The majority of the planktonic cells are not likely to be in direct contact with the electrode surface, especially given the vertical positioning of the electrodes (the turbidity in the cuvette appeared to be homogeneous). In contrast, biofilms are formed on the surface of the electrodes, positioned vertically, and held there by EPS. When a current is applied directly, biofilm cells are in direct contact with the metal cations released, possibly for the entire period of treatment as the ions were generated from the working electrode and diffused through the biofilm matrix. In the pre-treated LB medium, metal cations may have been converted to more inert forms relatively rapidly through reactions with water, oxygen, or hydroxide. In addition, biofilms treated with pre-treated LB media were not exposed to current directly; this may lead to a decreased exposure to metal cations, which were released from the anodic electrode. This can probably explain why treatments of biofilms with applied currents were more effective than using the pre-treated media prepared with the same level and duration of DC, especially at 250 μA/cm2. Precipitation of metal complex may also explain the additional killing by treating planktonic cells with 25 and 83 μA/cm2 DC compared to pre-treated media. At 250 μA/cm2, however, applied DC was less effective than pre-treated media. This is probably due to the changes in electrochemistry, which may generate metal complex that are more effective than ions moving in an electric field as existed for treatments with DC. The exact nature of these reactions remains to be determined.
During electrochemical reactions involving stainless steel as the working electrode, a multitude of ions and other chemical species can be formed depending on the voltage and current levels and composition of the medium. In particular, the chemical species formed of five key elements are of particular interest with regards to cell viability include iron, chromium, chlorine, oxygen and hydrogen (pH). Fe2+ ions can be generated during electrochemical reactions with stainless steel or graphite as an electrode (Dickinson and Lewandowski 1998). This effect may be intensified by the presence of biofilms on the stainless steel due to an increase in the resistance of the system, leading to an increased voltage when current is held constant (Dickinson and Lewandowski 1998). Ferrous ion can react with hydrogen peroxide via the Fenton reaction, resulting in the production of ferric ion, hydroxide ion, and the hydroxyl radical (Segura et al. 2008). This reaction has been reported to kill bacteria through further formation of the superoxide radicals (Andrews et al. 2003). In B. subtilis, oxidative stress due to H2O2 causes several genes to be up-regulated based on the response by the per regulon (Chen et al. 1995; Selinger et al. 2000). The induction of katA by 83 μA/cm2 and of the hemBCDLX operon by 83 μA/cm2 suggests that oxidative stress due to hydrogen peroxide may have been present. The decreased cell viability in biofilms treated with current may be in part due to oxidative stress as a result of the products of the Fenton reaction.
The second-most abundant metal in stainless steel is chromium, at amounts of up to 20% in 304L. Chromium ions, specifically Cr(VI) in chromate and dichromate, are highly toxic to bacterial cells (Garbisu et al. 1998). The presence and concentration of Cr(VI) in our system during treatment is unknown. B. subtilis 168 has a metabolic pathway by which it can reduce Cr(VI) to the much less toxic Cr(III) that functions when chromate ions are present in concentrations of up to 0.5 mM (Garbisu et al. 1998). However, genes for chromate reduction (ywrAB, ycnD) did not show significant changes in expression under our experimental conditions. It has been reported that the presence of heavy metals, such as zinc, cadmium, and copper, can inhibit chromate reduction by B. subtilis (Garbisu et al. 1997). Genes related to zinc, cadmium, and copper toxicity (copAB) were induced in the presence of 250 μA/cm2 current in our study. This finding suggests that ions of some heavy metals may be present in our system when using stainless steel as electrodes. Chromium reduction can also occur by chemical processes in solution, and can be enhanced or inhibited by other chemical species in the medium. Most significantly, the presence of Fe2+ enables the reduction of Cr(VI) to Cr(III), at a ratio of 3 Fe2+ to 1 Cr6+, possibly forming Fe/Cr complexes (Buerge and Hug 1997). However, the presence of organic ligands can modify this reaction; ligands specific for Fe2+ inhibit the reaction, while those for Fe3+ enhance it (Buerge and Hug 1998). In summary, the interactions of chromium within the system are complex, and killing via hexavalent chromium cannot be ruled out. However, the significant killing of B. subtilis using graphite electrodes suggests that the Cr(VI) ions are not indispensible for the killing effects of DC.
If metal cations are responsible for a loss of cell viability, we would expect to see genes that are related to metal tolerance to be up-regulated. Indeed, nine metal resistance genes were induced or up-regulated such as arsBCR, appBCF and zosA at 83 μA/cm2, and copAB at 250 μA/cm2. The arsBCR operon is responsible for the transport of arsenate, arsenite, and antimonite (Sato and Kobayashi 1998). These molecules bear little resemblance to divalent iron or hexavalent chromium compounds. It is interesting to note that arsenic is in the same group as phosphorous. It is possible that up-regulation of this operon may be related to the phosphate starvation.
In the absence of metal ions in solution as charge carriers, chloride ions in solution can react with hydroxyl ions to form hypochlorite, which is well known to be toxic to cells (Shirtliff et al. 2005). However, the experiments with graphite electrodes in M56 medium that did not contain chlorine showed that the metal ions are likely to be the dominating factors responsible for killing B. subtilis under our experimental conditions.
The bioelectric effect reported previously (Costerton et al. 1994) suggests that electric currents have a synergistic effect with antibiotics to improve the overall efficacy of killing biofilm cells. Surprisingly, in this study we observed that when ampicillin was added to the solution with current, the amount of killing was not significantly altered versus treatment with current alone. In the case of biofilms grown on graphite electrodes and treated in chlorine-free M56 buffer with 50 μg/mL ampicillin and 83 μA/cm2 current there was even a slight decrease in killing. It is well documented that iron can interfere with the action of antibiotics, including ampicillin (Ghauch et al. 2009), through a variety of mechanisms including chelation of ferric cations by antibiotics (Ghauch et al. 2009; Nanavaty et al. 1998). It is possible that the presence of iron and other metal cations is inhibiting ampicillin activity through chelation mechanisms under our experimental condition. Such interaction may be dependent on the nature of antibiotics since some other antibiotics do show synergy with electric currents in killing biofilm cells (Costerton et al. 1994). It is also important to note that in this study we employed a shorter treatment time (15 min) than that in the study by Costerton and co-workers (24 h, Costerton et al. 1994). To obtain a deeper insight into the mechanisms at the molecular level, it will also be important to follow the kinetics of viability and gene expression over time.
In summary, we conducted a detailed study of the effects of weak DC on viability, gene expression and morphology of B. subtilis. The data suggest that the ions and oxidative species generated by electrochemical reactions have significant influence on bacterial gene expression and viability. Further testing with additional conditions and different antibiotics as well as study with mutants of key genes will help unveil the mechanism of bioelectric effects.