INTRODUCTION
In 2011, the World Health Organization (WHO) stated that in Europe alone, approximately 4.5 million patients are affected by health care-associated infections (HCAIs) each year, resulting in 16 million extra days of hospital stay, at an estimated cost of €7 billion, with a mortality rate of 37,000 deaths (
1). The inanimate environment and “high-touch” surfaces have been verified as common reservoirs of bacteria causing HCAIs (
2,
3). The onset of a HCAI usually occurs approximately 48 to 72 h or more after hospital admission, but the risk increases significantly by 50% to 75% if the prior occupants of the ward had a HCAI (
4).
Within the hospital environment, contaminated surfaces have been demonstrated to play an important role in the transmission of microorganisms causing health care-associated infections (
5). The bacterial infections associated with primary surface colonization include methicillin-resistant
Staphylococcus aureus (MRSA), vancomycin-resistant enterococci (VRE), and extended-spectrum-beta-lactamase (ESBL)-producing Gram-negative organisms, such as
Escherichia coli and
Acinetobacter baumannii, which prevail in the hospital environment for extended periods, i.e., months, in viable form. Contaminated objects include hospital bed rails and bed linen, mattresses, patients' gowns and clothing, curtains, overbed tables, and stethoscopes (
6–13). These pathogens may survive on dry surfaces for extended periods and thus facilitate transmission between patients and health care workers (
14). Primary transmission onto surfaces originates from hands, patients, hospital water systems, and airborne sources (
15–19). Infection prevention and control practices to prevent HCAIs include the use of biodecontamination. However, current sterilization and disinfection methods have critical limitations in terms of efficacy, environmental impact, clinical downtime, and economic cost. In addition, more-aggressive decontamination approaches, such as the use of hydrogen peroxide gas and ultraviolet (UV) radiation, pose logistical difficulties, as both require the evacuation of patients and health care staff for a number of hours (
20,
21). Therefore, new approaches that would combine safety and efficiency in terms of minimal disruption in clinical areas are needed. One such method being evaluated involves cold atmospheric pressure plasma (CAPP). CAPP has numerous chemical and physical properties which can affect microbicidal outcomes. Depending on the plasma-generating mechanism (e.g., plasma jet, dielectric barrier discharge, etc.), CAPP systems are sources of positive and negative ions, reactive atoms and molecules (e.g., atomic oxygen, ozone, superoxide, and oxides of nitrogen), intense electric fields, and UV radiation. In many cases, CAPP sources produce a “cocktail” of all of the physicochemical properties listed above at the same time, in various proportions and densities. Positive and negative ions can lead to electrostatic disruption of bacterial cell walls. Oxidative atoms and compounds (e.g., atomic oxygen and ozone) can physically etch the cell wall and interfere with transport within the cell. Furthermore, such reactive compounds can induce DNA double and single breakage. Sufficiently intense electric fields can result in electroporation, whereas UV radiation (particularly sub-260-nm-wavelength UV) is well known to induce damage to DNA and intracellular proteins (
22).
The biomedical and clinical applications of CAPP have been evaluated in various areas, such as dermatology and wound treatment (
23–25), bone regeneration, implant treatments (
26,
27), and dental procedures, including bleaching and root canal disinfection (
28–30). However, CAPP has an innate antibacterial activity, making it an interesting decontamination technique and a possible solution for environmental decontamination, particularly in the clinical environment. In this study, we describe an
in vitro evaluation of a CAPP single-jet system for the decontamination of materials commonly found in the clinical environment.
MATERIALS AND METHODS
Bacterial strains and growth conditions.
Two Gram-positive organisms (MRSA and VRE) and two Gram-negative organisms (E. coli and A. baumannii) were chosen for this study. MRSA strain 43300 and ESBL-positive E. coli strain CL2 are clinical strains from our collection, the VRE clinical strain was provided by the Beaumont Hospital Microbiology Department, and the A. baumannii 19606 reference strain was sourced from the American Type Culture Collection (ATCC).
Bacteria were stored at −20°C on cryovial preservation beads (Microbank; Pro-Lab Diagnostics, Merseyside, United Kingdom). MRSA and A. baumannii strains were revived on Columbia blood agar (CBA) (Oxoid Ltd., Basingstoke, United Kingdom) plates, the E. coli strain was revived on Mueller-Hinton (MH) (Fluka, Sigma-Aldrich, Ireland Ltd.) agar plates, and the VRE strain was revived on Trypticase soy broth (TSB) (Oxoid Ltd., Basingstoke, United Kingdom) agar plates before each experiment. Overnight (16 to 18 h) bacterial cultures were grown aerobically at 37°C, with rotation, in TSB supplemented with 5% NaCl, for MRSA and VRE only or brain heart infusion (BHI) broth for A. baumannii or MH broth for E. coli strains.
Test surface preparation.
The test surfaces used in this study were 5-cm2 sections of marmoleum flooring (Forbo Flooring, Dublin, Ireland) and polyurethane mattress (Meditec Medical, Dublin, Ireland) commonly used in hospitals and provided by Beaumont Hospital, Dublin, polypropylene (GoodFellow Cambridge Ltd., United Kingdom), powder-coated mild steel (Watermark Engineering, Ireland), and stainless steel. To decontaminate before use, the soft surfaces, i.e., marmoleum and mattress, were placed in a 1% Virkon solution (Sparks Lab Supplies, Dublin, Ireland) for 30 min, rinsed three times in distilled water, and dried in the laminar flow cabinet for 1 h. The solid surfaces, i.e., polypropylene, powder-coated mild steel, and stainless steel, were soaked and wiped with 70% ethanol and left to dry in a laminar flow cabinet. All surfaces were then placed into petri dishes and placed under UV light for 30 min.
Preparation of the bacterial inoculums.
A volume of 25 ml of the appropriate broth was inoculated with one isolated colony from an overnight culture plate. Fresh overnight cultures were used for each assessment. Overnight cultures were centrifuged for 10 min at 15,500 × g (11,000 rpm) (Eppendorf centrifuge 5804R) and washed three times with sterile phosphate-buffered saline (PBS). The bacterial concentration was adjusted to a 3 to 4 McFarland standard (approximately 8 to 9 log10 CFU per ml) in 3 ml of sterile PBS, from which 50 μl was taken to inoculate each of the test surfaces.
CAPP single-jet system experimental design.
The CAPP single-jet system, shown in
Fig. 1, consists of a hollow, cylindrical polyether ether ketone (PEEK) body with a grounded stainless steel conical nozzle. A high-voltage (HV) stainless steel pin electrode runs through the axis of the PEEK cylinder, which is sealed at the end opposite to the nozzle. A sinusoidal high voltage is applied to the center pin at a frequency of 8 kHz and an amplitude of approximately 2.5 kV. Compressed air is forced through an orifice perpendicular to the jet axis at a flow rate of 12 standard liters per min (slm).
CAPP single-jet treatment.
The artificially inoculated test surfaces were exposed to the plasma jet plume for 30 s, 60 s, and 90 s, operating at approximately 25 W and 12 liters/min flow rate. The plume temperature did not exceed 45°C. The distance between the plume and the test surface was 1 cm (
31). All experiments were carried out at least three times in duplicate. The plasma system was maintained within a fume hood installed with an ozone detector.
Bacterial recovery and enumeration.
The entire areas of both test and control (nontreated) surfaces were swabbed using flocked eSwabs (Copan, Italy). Swabs were placed into Falcon round-bottom tubes (BD Bioscience, United Kingdom) with 3 ml of PBS, briefly subjected to a vortex procedure, and cultured onto CBA plates for MRSA and A. baumannii, ESBL Brilliance agar plates (Oxoid Ltd., Basingstoke, United Kingdom) for ESBL-positive E. coli, and VRE Brilliance agar plates (Oxoid Ltd., Basingstoke, United Kingdom) for VRE for bacterial enumeration. One-in-10 serial dilutions were performed when needed to determine a total viable count (TVC), i.e., the number of CFU/ml of one sample (30 to 300 countable colonies on the plate).
AFM.
Atomic force microscopy (AFM) images were completed in ambient air with a Dimension 3100 AFM microscope controlled by a Nanoscope IIIa controller (Digital Instruments, Santa Barbara, CA), operated in tapping mode, using standard silicon cantilevers (Budget Sensors, Bulgaria) with a 7-nm radius of curvature and a spring constant of 42 N/m (nominal values) to assess the physical effects of the plasma on the bacterial cells. Samples were prepared as described above and plasma treated, and AFM was performed. Multiple areas (approximately 10 areas per surface) were imaged to ensure good representation of the total surface inoculated. Images were then examined and edited using WxSM software (Nanotec Electronica S.L., Madrid, Spain) to generate phase and profile data (
32). Gwyddion software was also used to perform data analysis on the AFM scans (
www.gwyddion.net). The original two-dimensional (2D) scans obtained from the AFM were corrected by removing the polynomial background; this results in an accurate zero value on the surface, therefore verifying the exact height distribution of the cells on the surface.
Rt analysis was also carried out.
Rt is defined as the maximum peak-to-peak-valley height. This statistically analyzes the absolute value of the difference between the highest and lowest peaks indicative of the roughness and height of the cells as they are distributed across the surface. To further evaluate the AFM images, height distribution data analysis was also performed. This provides an overall comparison of the root mean square analysis of the cell on the surface, which is the quadratic mean, a statistical measure of the magnitude of a quantity of various points.
Statistical analysis.
Statistical data analysis was carried out using GraphPad Prism 5.00 software. The means of the log (CFU/ml) values from comparisons between recovered control and plasma-treated samples over 30 s, 60 s, and 90 s were determined by one-way analysis of variance (ANOVA).
RESULTS
Bactericidal effect of CAPP single jet on A. baumannii, ESBL-producing E. coli, MRSA, and VRE inoculated on various surfaces.
The bactericidal effect of the plasma on
A. baumannii, ESBL-producing
E. coli, MRSA, and VRE inoculated onto marmoleum, mattress, polypropylene, powder-coated mild steel, and stainless steel is summarized in
Fig. 2. For all the microorganisms and surfaces tested, the effect of the CAPP single jet was dependent upon the length of exposure to the plasma, with the maximum log reduction achieved at 90 s. For each set of data, a clear trend was observed over time correlating with the duration of exposure time and effect. There were, however, different effects noted depending upon the types of surface material.
Following exposure to the CAPP single jet, the highest log (CFU/ml) reductions compared to the recovered inoculum for A. baumannii were observed on the soft surfaces of mattress and marmoleum: 3.18 ± 1.26 and 3.12 ± 0.57, respectively. On stainless steel and polypropylene, there were log reductions of 2.97 ± 0.27 and 2.73 ± 0.27, respectively, followed by 1.66 ± 0.50 on powder-coated mild steel.
For ESBL-producing E. coli, the CAPP single jet was more effective after shorter exposure times, with a complete killing after 90 s for all surfaces except on powder-coated mild steel. Following a 60-s exposure time, high log reductions of 3.40 ± 0.20 on stainless steel and 2.78 ± 0.93 on the marmoleum were observed. Similarly, a 60-s exposure reduced the log (CFU/ml) numbers by 3.40 ± 0.20 on the polypropylene and by 2.44 ± 0.43 on the mattress. Ninety-second treatments of the powder-coated mild steel reduced the numbers of ESBL-producing E. coli by log 2.71 ± 0.24.
For MRSA, the best results were achieved on polypropylene, with a log reduction of approximately 5.87 ± 0.6, and log reductions of 4.08 ± 0.32, 3.95 ± 0.89, 3.82 ± 0.15, and 3.42 ± 0.90 were achieved on mattress, stainless steel, marmoleum flooring, and powder-coated mild steel, respectively, after 90 s.
The effects of the plasma on VRE following 90-s treatments resulted in the best log reduction on marmoleum flooring of approximately 5.19 ± 0.86, followed by log reductions of 5.01 ± 0.35, 4.02 ± 0.45, 2.80 ± 0.56, and 2.21 ± 0.08 on polypropylene, mattress, stainless steel, and powder-coated mild steel, respectively.
The bacterial log reduction as an outcome of the effect of the CAPP was confirmed to be statistically significant for all microorganisms inoculated on all surfaces (P < 0.05 following one-way ANOVA).
Atomic force microscopy imaging of the bactericidal effect of the CAPP single jet.
Atomic force microscopy imaging of all microorganisms inoculated on powder-coated mild steel before and after 90-s exposure to CAPP is shown on
Fig. 3. Powder-coated mild steel was chosen as the model surface to image as some of the other surfaces cannot be imaged using AFM due to forces exerted between the surface and the cantilever. Micrographs A and B illustrate the 2D topography of the applied cells, while micrographs C and D illustrate the 3D topography of the applied cells before and after CAPP treatment, respectively. Each micrograph represents an area of 5 μm, edge to edge, and is representative of multiple experiments (
n = 10). Plots E and F represent the surface topography in
Rt measurements and height distributions, respectively, of untreated and CAPP-treated cells on powder-coated mild steel.
A. baumannii cells before treatment (A and C) were observed as cellular aggregates indicative of pellicle formation, a morphological characteristic of biofilm-forming
A. baumannii 19606 (
33–36) whereby the secretion of exopolysaccharide causes the cells to clump together. This characteristic is considered to extend the survival of the organism in the environment. Following 90 s of exposure of
A. baumannii to CAPP (micrographs B and D), a significant disruption of the cell aggregates can be observed, with single cells showing disruption of the cell wall and leakage of cellular content. In panels E and F, the changes in the surface topography, registered in
Rt measurements and height distribution, respectively, can be seen. The noise in the measurement of the treated cells is indicative of the severe etching effect caused by the plasma corresponding to surface damage of the cells. Cell disruption is verified by a reduction in the
Rt value and the median height distributions from the untreated cells (151.8 nm and 153.2 nm) to the treated cells (118.7 nm and 118.9 nm).
ESBL-producing E. coli AFM micrographs show smooth and individual cells for the untreated control in panels A and C. However, severe cellular disruption can be seen, with only cell debris left on the surface, and no residual intact cells, following 90-s exposures to CAPP (B and D). In panels E and F, the surface topography in Rt and height distribution measurements show a considerable reduction in the Rt value and cell height of 302 nm and 312 nm compared to the untreated-cell values of 139.6 nm and 131.2 nm, consistent with the physical disruption of the bacterial cells.
AFM imaging of the MRSA cells inoculated on powder-coated mild steel shows smooth and morphologically intact cells, with no disruption visible on the 2D and 3D micrographs, respectively (panels A and C). Following 90 s of CAPP treatment, cellular distortions can be seen in panels B and D, with obvious cellular debris present and very few intact cells remaining. The surface topography and roughness assessed in Rt measurements (E) and height distributions (F) show increases in the Rt and height measurements from 262.0 nm and 260.0 nm to 414.0 nm and 415.4 nm, possibly due to a buildup of cell debris on the surface following CAPP treatment, indicative of the physical disruption of the cells by the air plasma jet.
Finally, untreated VRE cells are seen to be intact and with a slightly oval shape, which is characteristic of Enterococcus spp., on the corresponding 2D and 3D micrographs of panels A and C, respectively. Cellular malformations arose in treated VRE cells after 90 s of exposure to CAPP, with cells appearing distorted and what could possibly be intracellular material leaching out of damaged cells (micrographs B and D). In panels E and F, the surface roughness expressed in Rt values and median of height for the untreated and treated cells were 290.5 nm and 291.0 nm and 294.3 nm and 193.9 nm, respectively.
DISCUSSION
The present study aimed to evaluate the antimicrobial effect of a CAPP single-jet prototype on bacteria of clinical significance, including MRSA, VRE, ESBL-producing E. coli, and A. baumannii, on inanimate surfaces commonly found in the clinical setting.
A recent review on environmental contamination has highlighted the importance of this source as a primary mode of transmission of HCAI. Current effective decontamination methods pose logistical difficulties and limitations, but the results presented here suggest that the use of CAPP is a promising tool for environmental biodecontamination, achieving a >log 5 reduction for some bacteria on certain materials after 90 s. Although other studies have been performed on biomedical device materials, skin models, and pagers and in solution, this is the first study performed on materials of surfaces of clinical relevance (
37–40).
Previous studies on the antimicrobial effects of plasma involved different treatment exposure times mainly due to the physical state of the bacteria and demonstrated shorter times for planktonic cells in solution (
41) and longer times for cells dried on test surfaces and in biofilms. In this study, the optimum antimicrobial activity of the air plasma was observed after 90 s, producing reductions of log 3 to 5 for MRSA, log 2 to 5 for VRE, log 2 to 3 for
E. coli, and log 1.7 to 3 for
A. baumannii, all of which were air-dried on each test surface. Maisch et al. (
42) evaluated the efficacy of a CAPP device on MRSA- and
E. coli-contaminated porcine skin and showed that longer exposure times were required to achieve log reductions similar to those seen in our study, i.e., 6 min for a log 3 reduction and 8 min for a log 5 reduction of both strains, but those are relatively prolonged periods in the busy clinical environment for surface decontamination. Similarly, the efficacy of a plasma microjet in killing
S. aureus and
Enterococcus faecalis inoculated on agar found that treatment exposure times of 4 to 5 min were required to achieve a log 4 reduction. For biofilms, the treatment times increase significantly, in some cases taking as long as 30 min to achieve a log 3 reduction (
43). Recently, remote plasma exposure of MRSA strains, in biofilm form, has proven to be effective also. However, the treatment in that case required up to 1.5 h to inactivate the biofilm completely (
44). Few studies have assessed the effects of CAPP on
A. baumannii, but one found that this bacterium was more resistant to plasma than
S. aureus and other Gram-negative organisms (
45).
The isolation of Gram-negative bacteria from the environment poses challenges as they may enter a viable but nonculturable state, and this may partly explain why they are isolated less frequently than Gram-positive bacteria. Although they are still capable of causing infection, recovering them from the environment is difficult in this state (
46). This was also reflected in our results as there was an evident decline in log numbers between the applied inoculum and the recovered control after air-drying. Morphological cellular effects following plasma exposure were observed for Gram-negative bacteria as seen in
Fig. 3 for treated
A. baumannii and ESBL-producing
E. coli in the 2D and 3D AFM images (
Fig. 3B and
D) compared to the untreated controls (
Fig. 3A and
C). CAPPs produce numerous reactive ions, including reactive oxygen species (ROCs), reactive nitrogen species (RONs), and UV, which, as originally suggested by Laroussi et al. (
47), chemically and physically alter various bacterial, fungal cells, tissues, and surfaces. These species not only affect bacterial cells on a surface level but also intracellularly cause a cascade of effects leading to cell wall disruption, cytoplasm leakage, lipid peroxidation, and DNA damage (
48–50). For both MRSA and VRE, cellular disruption and physiological changes were observed whereby, following 90 s of treatment, few intact cells remained, with visible cellular debris observed (
Fig. 3, MRSA and VRE panels B and D).
Montie et al. (
50) suggested that leakage of the cytoplasm occurs due to initial “etching” or physical damage of the bacterial cell wall and that, once the cell wall has been compromised, the reactive oxygen species then filter through into the cell, causing oxidative damage and eventually leading to cell death. The rates at which this occurs differ between Gram-positive and Gram-negative bacteria chiefly due to metabolic and biochemical pathway differences, in addition to the differences in the amounts of peptidoglycan present in the cell walls. Another publication by Yusupov et al. (
48) verified the disruption of important C-N, C-O, and C-C bonds in peptidoglycan by O
3 and O
2 molecules and O atoms following plasma treatment. As the thicknesses of the peptidoglycan layer differ between Gram-positive bacteria (20 to 30 nm) and Gram-negative bacteria (6 to 7 nm), it can be speculated that the effects of the plasma on the cell wall may be more pronounced in Gram-negative bacteria. In the present study, both ESBL-producing
E. coli and
A. baumannii showed more-severe physical damage and in some cases total cell disruption as seen in AFM micrographs of ESBL-producing
E. coli (
Fig. 3B and
D), where only cell debris can be seen following treatment of
E. coli cells for 90 s. Similar effects were seen for
A. baumannii (
Fig. 3,
A. baumannii panels B and D). Height distribution measurements produced a “noisy” graph that may be indicative of significant etching of the cell walls (
Fig. 3,
A. baumannii panel E).
The data presented in this study have verified the efficacy of CAPPs for use as a biodecontaminating agent in clinical environments. The air plasma source used shows significant bactericidal effects on both Gram-positive and Gram-negative organisms, with a maximum log reduction of approximately >log 5 after 90 s. In addition, the design and configuration of the plasma jet used here produce and deliver reactive species in a controlled manner. This suggests that the use of such a system could greatly enhance infection control procedures currently existing in the clinical setting.
In conclusion, we have shown that CAPP significantly reduces bacterial numbers on a range of surfaces commonly found in the clinical environment within 90 s. Further work is required to develop a prototype that could be used in the clinical environment and to evaluate this against spore-forming bacteria such as Clostridium difficile and mixtures of bacteria with protein and other substances that mimic contamination in a clinical setting. If efficacy is confirmed, CAPP would represent an important and valuable alternative to surface decontamination in health care facilities.