Correspondence: Michelle Maclean, The Robertson Trust Laboratory for Electronic Sterilisation Technologies, Department of Electronic and Electrical Engineering, University of Strathclyde, Royal College Building, 204 George Street, Glasgow, Scotland, G1 1XW, UK. Tel.: +44 0 141 548 2891; fax: +44 0 141 552 5398; e-mail: firstname.lastname@example.org
This study was conducted to investigate the bactericidal effects of visible light on methicillin-sensitive and methicillin-resistant Staphylococcus aureus (MRSA), and subsequently identify the wavelength sensitivity of S. aureus, in order to establish the wavelengths inducing maximum inactivation. Staphylococcus aureus, including MRSA strains, were shown to be inactivated by exposure to high-intensity visible light, and, more specifically, through a series of studies using a xenon broadband white-light source in conjunction with a selection of optical filters, it was found that inactivation of S. aureus occurs upon exposure to blue light of wavelengths between 400 and 420 nm, with maximum inactivation occurring at 405±5 nm. This visible-light inactivation was achieved without the addition of exogenous photosensitisers. The significant safety benefit of these blue-light wavelengths over UV light, in addition to their ability to inactivate medically important microorganisms such as MRSA, emphasises the potential of exploiting these non-UV wavelengths for disinfection applications.
The increasing problem of microbial antibiotic resistance has generated interest in alternative methods of inactivating problematic organisms such as methicillin-resistant Staphylococcus aureus (MRSA). One area of interest involves the use of light-based treatment technologies. Much research has been carried out involving the photodynamic inactivation (PDI) of S. aureus and MRSA using light, particularly light of visible wavelengths, and exogenous photosensitisers such as phenothiazinium dyes, including toluidine blue O (Wilson & Yianni, 1995; Wainwright et al., 1998), methylene blue (Wainwright et al., 1998; Zeina et al., 2001), and porphyrin derivatives such as hematoporphyrin (Bertoloni et al., 2000). The induction of intracellular porphyrin production through pretreatment with δ-aminolaevulinic acid followed by light exposure has also provided a successful method for staphylococcal PDI (Nitzan & Kauffman, 1999; Nitzan et al., 2004).
The observation that visible light alone has bactericidal properties has been previously documented for bacteria, most notably, the acne-associated bacterium Propionibacterium acnes. Research into P. acnes has found that irradiation of this organism with blue light leads to photosensitisation of intracellular porphyrins, stimulation of which leads to the production of reactive species, predominantly singlet delta oxygen (1O2), and consequently, cell death (Papageorgiou et al., 2000; Ashkenazi et al., 2003; Hamblin & Hasan, 2004). Other bacteria that have been found to be susceptible to inactivation solely through visible-light exposure include Helicobacter pylori and some oral black-pigmented bacteria (Feuerstein et al., 2004; Ganz et al., 2005; Soukos et al., 2005).
The present study was carried out to investigate and fully characterise the bactericidal effect of visible light on S. aureus. Firstly, the bactericidal effect of high-intensity visible light (400 nm and longer) on S. aureus and MRSA was investigated. Subsequently, to identify the region of the visible spectrum that induces staphylococcal inactivation, the visible wavelength sensitivity of S. aureus was investigated through a series of optical filter studies. The results demonstrate that blue light between 400 and 420 nm, but optimally 405 nm (±5 nm), has bactericidal effects against S. aureus.
Materials and methods
The bacteria used in this study were as follows: S. aureus NCTC 4135, Escherichia coli NCTC 9001 (National Collection of Type Cultures, Collindale, UK), MRSA LMG 15975 (The Belgian Co-ordinated Collections of Micro-organisms, Gent, Belgium) and MRSA 16a (a clinical wound isolate obtained from the Royal Infirmary, Glasgow). Test strains were inoculated into 100 mL of nutrient broth (Oxoid, UK) and cultivated at 37 °C under rotary conditions (at 125 r.p.m.). After an 18-h incubation period, the broth was centrifuged at 3939 g for 10 min and the resultant pellet resuspended in 100 mL phosphate-buffered saline (PBS) (Oxoid, UK). This suspension was then diluted in PBS to give a population density of c. 2.0 × 105 CFU mL−1 for experimental use.
Visible light source
A xenon broadband white-light source (Hamamatsu Photonics UK Ltd), together with a 400 nm longwave pass (L-P) filter, was used for the visible-light exposure of bacterial suspensions and the relevant emission spectra are shown in Fig. 1. The use of this filter allowed transmission of wavelengths longer than 400 nm (i.e. visible light) and thus eliminated UV-light inactivation.
High-intensity exposure experiments
The experimental set-up was as follows: a 2 mL volume of bacterial suspension, with a population density of 2.0 × 105 CFU mL−1, was transferred to one well of a 12-well multidish (Nunc, Denmark), which also contained a 7 mm × 2 mm magnetic follower. The 400 nm L-P filter was placed on top of the well. The dish was then positioned directly under the light source on a magnetic stirrer. This, in conjunction with the magnetic follower, permitted continuous mechanical agitation of the sample during light exposure. For the experiments detailed, all parameters were maintained constant: the sample volume used in each experiment was 2 mL and the irradiance of the xenon lamp through the 400 nm L-P filter was 350–400 mW cm−2 (measured using a radiant power meter and detector; L.O.T.-Oriel Ltd, UK). A separate sample well containing 2 mL bacterial suspension was used for each exposure time, and after each exposure, samples were plated onto nutrient agar (NA) (Oxoid, UK) (see ‘Bacterial Plating and Enumeration’) and incubated at 37 °C for 24 h. Control samples were also set-up; these were subjected to identical conditions but were not exposed to high-intensity visible light, but left in normal laboratory lighting conditions.
Filter study for identification of inactivation wavelengths
The experimental set-up for the filter studies on S. aureus NCTC 4135 was identical to that described above for the high-intensity exposure experiments, with the 400 nm L-P filter being substituted with, firstly, a selection of commercially available L-P and shortwave pass (S-P) filters (L.O.T.-Oriel Ltd), which would identify the causative wavelength range, and secondly, a selection of narrow bandpass (B-P) filters ranging from 400 to 500 nm (L.O.T.-Oriel Ltd; Ealing Catalog. Inc.), which would identify the causative bandwidth to within 10 nm.
For comparison of the germicidal efficiencies for each narrow bandwidth, the output intensity of the lamp was amended for each B-P filter so that the same irradiance (3.27 mW cm−2) was transmitted onto each bacterial sample for each filter. This 3.27 mW cm−2 irradiance was transmitted through each of the filters and illuminated each bacterial sample for an equal time period of 2 h. From these values, the absolute dose, also termed energy density, in Joules per square centimetre [irradiance (W cm−2) × time (s)] for each 10 nm bandwidth being applied to the S. aureus suspensions was calculated as 23.5 J cm−2 (3.27 mW cm−2 for 2 h).
Bacterial plating and enumeration
In order to obtain accurate viable cell counts, several standard plating methods were used in this study. The spiral and spread plate methods, using 50 and 100 μL sample volumes, respectively, were prepared on NA using a WASP 2 spiral plater (Don Whitley Scientific Ltd, UK). For samples with anticipated low CFU counts, pour plates, using a 1-mL sample volume, were prepared manually using molten NA. Spiral plates were enumerated using the supplied counting grid and tables. Spread and pour plates were counted manually with the aid of a colony counter (Stuart Scientific, UK). The resultant counts from each of these methods were then converted into viable CFU counts mL−1 of sample.
With regard to the replication and recording of experimental results, in the high-intensity exposure experiments each data point on the graphs represents the results from two independent experiments, with a minimum of triplicate samples being taken for each experiment. These results are documented as mean values with SDs being included. Significant differences in the light-treatment results were calculated at the 95% confidence interval using anova (one way) with Minitab software Release 15. Data on the germicidal efficiency of different wavelengths, as discussed later, were obtained from replicated CFU count data and calculated as mean log10 reduction per unit dose.
High-intensity exposure experiments
Figure 2 shows the results for the exposure of S. aureus suspensions to high-intensity light of wavelengths >400 nm. It can be seen that the light had a significant bactericidal effect on S. aureus, with a 5-log10 reduction being achieved after a dose of 630 J cm−1 (350 mW cm−2× 30 min). The exposed E. coli suspensions demonstrated negligible inactivation over a 30-min exposure time, although exposures of 45 and 60 min did demonstrate significant differences compared with the associated control samples, indicating that a more prolonged exposure may induce further inactivation of the exposed E. coli population. In Fig. 3, it can be seen that when the MRSA strains were light-exposed, inactivation results that were similar to those obtained with S. aureus NCTC 4135 were observed (Fig. 2). The population densities of all control samples stayed constant throughout this series of experiments (Figs 3 and 4). Temperature was also monitored throughout the light-exposure experiments and was found to not rise above 37 °C, eliminating any thermal inactivation effects.
Identification of inactivation wavelength range
As a first step to identify the wavelength range within the visible spectrum inducing staphylococcal inactivation, different visible-wavelength ranges were selected using S-P and L-P filters using similar irradiances. Figure 4 shows the effects of the different wavelength ranges on the rate of inactivation of S. aureus NCTC 4135 cells in PBS suspensions. Exposure to wavelengths of 500 nm and below induced the most rapid inactivation rate, and this was likely the result of the inclusion of UV wavelengths, which are well known to have a strong germicidal effect. Wavelengths longer than 400 nm also caused total inactivation. When longer wavelengths of 500 nm and above were investigated, no inactivation was observed. This confirmed that the visible wavelengths inducing staphylococcal inactivation were in the wavelength region of 400–500 nm.
Identification of the inactivation bandwidth
In order to identify the narrow bandwidth of visible light between 400 and 500 nm inducing staphylococcal inactivation, B-P filters were used. The S. aureus suspensions exposed to each narrow 10 nm bandwidth between 400 and 500 nm received an absolute dose of 23.5 J cm−2, and significant log10 reductions were achieved through exposure to 400–420 nm bandwidths, as shown in Table 1. Here it can be seen that the maximum log10 reduction of S. aureus cells resulted from exposure to 405±5 nm wavelength light. Exposure to bandwidths of 430–500 nm did not cause significant inactivation of the bacteria.
Table 1. Log10 reduction and germicidal efficiency (η) values for the inactivation of Staphylococcus aureus NCTC 4135 following exposure to 10 nm bandwidths of light from 400 to 420 nm, each for a dose of 23.5 J cm−2
The inactivation capability at each wavelength can be quantified as the germicidal efficiency, defined as the log10 reduction of a bacterial population by inactivation per unit dose in Joule per square centimetre (Wang et al., 2005). Thus, germicidal efficiency, η=log10(N/N0) J cm−2.
Numerical data for the germicidal efficiencies achieved through light exposure to 10 nm bandwidths of light from 400 to 430 nm (with centre wavelength increments of 5 nm) are listed in Table 1. The results show that the germicidal efficiency peak of 0.102 log10 J cm−2 was at 405 nm, but the 400 nm light also demonstrated good germicidal activity for S. aureus, with a value of 0.064 log10 J cm−2. Bandwidths between 430 and 500 nm demonstrated no significant germicidal efficiency against S. aureus suspensions.
Investigations using a high-intensity xenon lamp, in conjunction with a selection of commercially available L-P, S-P and B-P filters, have demonstrated the sensitivity of S. aureus to visible light, and also identified the bactericidal wavelengths inducing maximum visible-light inactivation to within a 10 nm bandwidth. The results have highlighted that inactivation is evident using 400–420-nm-wavelength blue light, with the most effective bactericidal activity at 405±5 nm. Wavelengths of longer than 430 nm were found to induce no effect on the viability of S. aureus cells. The occurrence of the peak at 405 nm suggests that an inactivation process is at its optimum within the S. aureus cells at this specific wavelength.
The identification of this narrow band of inactivation wavelengths highlights that the vast majority of the illuminating wavelengths emitted by the broadband xenon lamp (>400 nm) were superfluous to the inactivation process. Therefore, in Figs 2 and 3, although c. 630 J cm−2 was required for a 5-log10 reduction of S. aureus and MRSA strains, this was the total irradiance, only a small fraction of which was responsible for the inactivation. Although this value is useful for comparison with other studies that have used broadband light sources, the experiments using the narrowband filters provide more meaningful values of absolute dose: 23.5 J cm−2 of 405 nm light resulting in a 2.4 log10 reduction of S. aureus.
The use of these dose values can also be used to explain why throughout previous studies on the PDI of S. aureus using exogenous photosensitisers, inactivation in the absence of photosensitisers or δ-ALA-induced porphyrins has either been dismissed or not been discussed. Bertoloni et al. (2000) and Zeina et al. (2001) both used broadband light sources emitting visible light 400–700 nm (as used in this study; Figs 2 and 3) for illumination of the bacterial samples but only applied maximum doses of 3.6 and 20.16 J cm−2, respectively, compared with the much greater 630 J cm−2 dose applied in the present study (Figs 2 and 3).
The finding that exposure to high-intensity visible light, at intensities that resulted in a 5 log10 reduction in S. aureus and MRSA populations, had a negligible effect of E. coli reflects results found in previous PDI studies involving photosensitisers. In one study using pretreatment with δ-ALA, which used a white-light source for illumination (as in the present study), the inactivation of E. coli required at least 10-times higher doses to achieve similar activation levels to that for S. aureus (Nitzan & Kauffman, 1999).
The sigmoidal shape of the inactivation curve for visible-light inactivation reveals an initial period of low biocidal activity that may indicate a requirement for a build-up of energy, reactive molecules or cellular damage, which must occur before induction of bacterial inactivation is initiated. It is likely that the mechanism of inactivation is quite different to that of continuous UV-C exposure, which induces DNA damage, primarily as a result of UV absorption by the DNA bases in the wavelength region 240–280 nm (Blatchley & Peel, 1991), or near-UV light, inactivation by which has been accredited to sublethal damage of DNA repair systems (Tyrrell & Peak, 1978). Visible-light inactivation, on the other hand, as established for other bacteria such as P. acnes, H. pylori and some black-pigmented bacteria (Ashkenazi et al., 2003; Feuerstein et al., 2004; Ganz et al., 2005; Soukos et al., 2005), has been accredited to the photo-stimulation of endogenous intracellular porphyrin molecules. These studies identified the stimulating wavelengths to be visible light in the wavelength region 400–500 nm, and more specifically 400–420 nm in the cases of P. acnes and H. pylori (Ashkenazi et al., 2003; Elman et al., 2003; Ganz et al., 2005). Because of the similarity in causative wavelengths, it is hypothesised that the inactivation mechanism occurring with the S. aureus and MRSA strains in the present study is also the result of photo-stimulation of intracellular porphyrins, which results in the production of reactive species, predominantly singlet delta oxygen (1O2), and consequently, cell death.
Papageorgiou et al. (2000) investigated the effect of blue light on P. acnes and their findings are in general agreement with those established in this study. They found that the sensitivity of P. acnes to visible light was at a maximum in the blue region of 415 nm. Papageorgiou et al. (2000) state that 415 nm corresponds to the absorption maximum of the specific porphyrin molecules produced by P. acnes. The maximum of 405 nm determined in the present study may indicate that porphyrin molecules that have different absorption maxima are present within S. aureus bacteria or are produced by them. The recent work of Guffey & Wilborn (2006a, b), however, reports inactivation of S. aureus at both 405 and 470 nm. In their study, low populations of S. aureus plated onto agar surfaces were exposed to doses of magnitude similar to those used in the present study. For 405 nm inactivation, they measured an approximate single log10 reduction in bacterial concentration for a dose of 15 J cm−2 compared with the reduction of 2.4 log10 for a dose of 23 J cm−2 found in this work. At 470 nm, Guffey & Wilborn (2006b) measured a log10 reduction of around 0.4–0.5 for a dose of 15 J cm−2; our results, on the other hand, indicate that no inactivation of S. aureus occurs at 470 nm.
The germicidal efficiency of the 405-nm-wavelength light is much lower than that for UV wavelengths; however, this disadvantage may be more than outweighed for some applications by the greater safety afforded at this wavelength. While it is well established that UV irradiation provides a method of inactivating pathogenic microorganisms, it is not apposite to expose humans to UV radiation because of the well-recognised risks of eye damage and skin cancer. Significantly, this study has shown that these blue wavelengths within the visible-light spectrum are capable of inactivating S. aureus, including MRSA, while posing a negligible threat to human health (ACGIH, 2007).
The first author would like to thank The Engineering and Physical Sciences Research Council (EPSRC) for their support through a Doctoral Training Grant (awarded in 2002/2003). All authors would like to thank The Robertson Trust for their funding support.