|Home | About | Journals | Submit | Contact Us | Français|
Anaerobic cultures of Shewanella oneidensis MR-1 reduced toxic Ag(I), forming nanoparticles of elemental Ag(0), as confirmed by X-ray diffraction analyses. The addition of 1 to 50 μM Ag(I) had a limited impact on growth, while 100 μM Ag(I) reduced both the doubling time and cell yields. At this higher Ag(I) concentration transmission electron microscopy showed the accumulation of elemental silver particles within the cell, while at lower concentrations the metal was exclusively reduced and precipitated outside the cell wall. Whole organism metabolite fingerprinting, using the method of Fourier transform infrared spectroscopy analysis of cells grown in a range of silver concentrations, confirmed that there were significant physiological changes at 100 μM silver. Principal component-discriminant function analysis scores and loading plots highlighted changes in certain functional groups, notably, lipids, amides I and II, and nucleic acids, as being discriminatory. Molecular analyses confirmed a dramatic drop in cellular yields of both the phospholipid fatty acids and their precursor molecules at high concentrations of silver, suggesting that the structural integrity of the cellular membrane was compromised at high silver concentrations, which was a result of intracellular accumulation of the toxic metal.
Silver is an element that has been used widely in industrial processes as diverse as photographic processing, catalysis, mirror production, electroplating, alkaline battery production, and jewelry making (18). It has been known for some time that silver ions and silver-based compounds can be highly toxic to microorganisms, and with increasing concern about pathogenic “superbugs” with high resistance to conventional antibiotics, silver is attracting much interest as a potential biocide (11, 18, 36, 42). Silver has no known physiological functions and can exist in several oxidation states, although it is most commonly encountered in its elemental [Ag(0)] and monovalent [Ag(I)] forms. Although use of nanoscale elemental Ag(0) as a biocide has been increasing, for example, in wound dressings and as an antimicrobial coating on consumer products, little is known about its mode of toxicity. This is despite the surprising ability of actively growing Fe(III)-reducing bacteria such as Geobacter sulfurreducens to precipitate nanoscale Ag(0) particles within and around the cell surface via reduction of Ag(I) (18). Ionic Ag(I), in contrast, has been the focus of more studies on the mode of metal toxicity. Previous research showed that silver ions have antimicrobial activities against a wide diversity of bacteria (19). They have been shown to disrupt the respiratory chain of Escherichia coli (3) and inhibit the exchange of phosphate and its uptake (34). Ag(I) has also been linked to copper metabolism in E. coli, potentially competing with copper binding sites on the cell surface and subsequent copper transport into the cell (8). However, the toxicity of silver is not limited to prokaryotes, as long-term exposure in humans can cause argyria, impaired night vision, and abdominal pain (31, 32, 36). The detailed mechanism of toxicity in prokaryotes or eukaryotes remains to be identified, although it has been proposed that silver ions react with cellular proteins via SH groups (16), leading to the disruption of cellular metabolism.
Microbial cells have evolved an extremely diverse range of mechanisms to survive high concentrations of toxic metals. The mechanisms invoked include biosorption, bioaccumulation, special efflux systems, alteration of solubility and toxicity via reduction or oxidation, extracellular complexation or precipitation of metals, and lack of specific metal transport systems (1). For example, for silver ions the bacterial cell wall can be an efficient permeability barrier to block the uptake of metal (21), with additional complexation in the periplasm by specific silver-binding proteins (35). Redox transformations also offer the potential to detoxify Ag(I) ions, e.g., through the reduction to insoluble elemental Ag(0) (30). In addition, the energy-dependent efflux of toxic Ag(I) is perhaps the best-studied resistance mechanism for silver, mediated via ATPases and chemiosmotic cation/protons antiporters (9).
Shewanella spp., Gram-negative, dissimilatory metal-reducing bacteria, can use a wide variety of terminal electron acceptors for growth (23, 39), including high oxidation state metals such as Fe(III), Mn(IV), Cr(VI), U(VI), and Au(III) (5, 17, 26, 28, 41). Shewanella species also have the potential to reduce Ag(I), given their similar activities against Au(III), and the reduction of Ag(I) to form nanoscale deposits of Ag(0) within the cell has been documented for other Fe(III)-reducing bacteria (18). This metabolic versatility offers considerable potential for bioremediation applications, for example, via reduction of U(VI) to insoluble U(IV) (5, 17, 26, 28, 41), and the recovery of precious metals such as silver and gold via reductive precipitation. It also offers an interesting model organism to study the metabolism of toxic metals such as silver, including the physiological impact of ionic Ag(I) and nanoscale Ag(0).
This paper describes interactions of Shewanella oneidensis MR-1 with various concentrations of Ag(I), including demonstrations of the reduction and deposition of silver nanoparticles under anaerobic conditions. A range of techniques, including X-ray diffraction (XRD) and analytical transmission electron microscopy (TEM), were used to investigate the nature and cellular localization of the precipitates, while Fourier transform infrared (FT-IR) spectroscopy metabolic profiling techniques were used to identify the impact of toxic metal accumulation on the cell. The disruption of membrane integrity was implied by these investigations and confirmed by fatty acid methyl ester (FAME) analysis, which showed a dramatic decrease in the quantities of membrane lipid components.
S. oneidensis strain MR-1 was obtained from the Manchester University Geomicrobiology Group culture collection and stored at −80°C in 50% glycerol prior to use. All cultures used were grown in a fully defined minimal medium, based on the recipe of Myers and Nealson (27): 1.2 g/liter (NH4)2SO4, 1 g/liter K2HPO4, 0.4 g/liter KH2PO4, 0.17 g/liter NaHCO3, 0.25 g/liter MgSO4·7H2O, 0.072 g/liter CaCl2·2H2O, and trace elements (1 liter of medium contained 1 mg FeSO4·7H2O, 0.2 mg MnCO3, 1 mg CoCl2·6H2O, 0.28 mg ZnCl2, 0.8 mg Na2MoO4·2H2O, 3.5 mg H3BO3, 1 mg NiCl2·6H2O, 0.05 mg CuSO4·5H2O, 0.2 mg Na2SeO4, and 22 mg Na2EDTA). Amino acid solution (20 mg/liter l-arginine hydrochloride, 20 mg/liter l-glutamate, 20 mg/liter l-serine) was also added. Sodium dl-lactate (100 mM) was used as the carbon source and electron donor, and fumarate (20 mM) was the electron acceptor. The medium was sparged with nitrogen, and the pH was adjusted to 7.6 before sterilization. All chemicals were purchased from Sigma unless noted otherwise.
Experiments were conducted in a Coy anaerobic chamber under an atmosphere of 5% hydrogen/95% nitrogen (vol/vol), unless otherwise noted. Anaerobic cultures were grown in 80 ml defined medium in 100-ml serum bottles (flushed with N2 for 10 min and sealed with a butyl rubber stopper before use). In all experiments six biological replicates were prepared. A silver nitrate (100 mM) stock solution was then filter sterilized and added to give concentrations ranging from 1 to 400 μM in the defined medium. Inocula were grown overnight anaerobically to early exponential phase at 30°C (approximately 40 h). A 10-ml aliquot of inoculum was then added to the 80 ml of medium in sterile 100-ml serum bottles, and the experimental cultures were incubated at 30°C in the dark. Samples were taken in the anaerobic chamber for protein analyses. FT-IR and high-performance liquid chromatography samples were harvested by centrifugation (13,000 × g for 10 min) and washed with sterile 0.9% NaCl solution, the wet weight of the cell pellet was determined, and the cells were frozen at −80°C until further analysis.
At regular time points, 1 ml of the broth was sampled and its optical density at 600 nm measured. The total protein concentrations were measured using a bicinchoninic acid/copper sulfate kit (BCA; Sigma, United Kingdom). Aliquots (800 μl) were centrifuged at 12,000 × g for 5 min, and the cell pellet was resuspended in 100 to 800 μl of double-distilled water. Fifty-microliter samples were mixed with 950 μl BCA reagent and then incubated at room temperature overnight. The protein concentrations of “unknown” test samples were calculated from A562 measurements compared against a standard curve prepared with known concentrations of bovine serum albumin.
Ag-containing mineral phases were prepared for XRD analysis in 1-liter bottles of anaerobic medium. S. oneidensis MR-1 cultures were grown in 1 liter of medium containing a final concentration of 100 μM AgCl at 30°C without agitation for 24 h. Cells and precipitate were harvested by centrifugation (13,500 × g for 5 min), washed twice in distilled water, dried in an anaerobic cabinet, and analyzed using a Philips PW 1730 diffractometer with Cu K α1 radiation. Precipitates formed over 24 h in the absence of S. oneidensis cells were also prepared for analysis.
The S. oneidensis MR-1 cells and biogenic silver particles were also observed by TEM. The cells from anaerobic cultures challenged with silver were collected by centrifugation (13,500 × g for 5 min) and then prefixed overnight in 2.5% glutaraldehyde in HEPES buffer (pH 7) at 4°C. The cell pellet was then enrobed in 2% Nobel agar and the small agar block immersed in 100 mM HEPES buffer. The cells were then dehydrated at room temperature with 25, 50, 75, and 95% ethanol for 10 min each and 100% ethanol for 15 min. The specimens were then embedded in LR white resin (London Resin Co., London, United Kingdom). Increasing concentrations of LR white resin in ethanol were infiltrated into the agar block with one 15-min wash with an LR white resin-ethanol (50:50) mixture and then two times for 2 h with 100% LR white treatments. The agar blocks were cured in LR white resin for 1 h at 60°C, and specimens of 60 to 80 nm were cut using an ultramicrotome and mounted on a carbon-coated copper TEM grid. The sections were stained with 2% uranyl acetate in distilled water for 10 min and then with a 0.2% (wt/vol) lead citrate solution (18) for 5 min prior to TEM analysis.
The metabolic profiles of cells grown aerobically and anaerobically were recorded by FT-IR spectroscopy periodically. Samples (5 ml) were collected and centrifuged at 4°C at 12,000 × g for 10 min, the supernatant was removed, and the cell pellet was washed twice with sterile 0.9% NaCl solution and centrifuged again prior to storage at −80°C. After all the samples had been collected, they were defrosted and resuspended in sterile 0.9% NaCl solution to 0.8 g/liter biomass (dry weight).
All FT-IR analyses were conducted using an Equinox 55 infrared spectrometer equipped with a high-throughput motorized microplate module (HTS-XT; Bruker Optics, Coventry, United Kingdom). The motorized module of this instrument introduces the plate into the airtight optics of the instrument, in which tubes of desiccant are contained to provide a moisture-free environment (10). A deuterated triglycine sulfate (DTGS) detector was employed for transmission measurements of the samples to be acquired. A 96-well Si sample plate was washed thoroughly with 2-propanol and deionized water and allowed to dry at room temperature prior to use. Twenty microliters (16 mg of biomass) of each bacterial sample was evenly applied in triplicate onto the plate. The plate was consequently dried at 60°C in an oven for 10 min and loaded into the FT-IR motorized module. Spectra were collected over the wavelength range of 4,000 to 600 cm−1 under the control of a computer programmed with Opus 4, operated under Microsoft Windows 2000. Spectra were acquired at a resolution of 4 cm−1, and 64 spectra were coadded and averaged to improve the signal-to-noise ratio. The collection time for each spectrum was approximately 1 min.
The ASCII data were imported into Matlab version 6 (The MathWorks). The FT-IR spectra were normalized such that the smallest recorded absorbance was set to 0 and the highest was set to 1 for each spectrum. Baseline correction, extended multiplicative scatter correction (EMSC) (20), and row normalization preprocessing steps were applied before exporting the data to Pychem v3.0.3 for further analysis (12). An unsupervised data reduction method, principal components analysis (PCA), was employed to reduce the dimensionality of the FT-IR data from 1,764 absorbance measurements at different wave numbers to 10 principal components (PCs), representing 99% of the total variance in the FT-IR data. The PCs were used as inputs for discriminant function analysis (DFA), a supervised technique that discriminates groups using a priori knowledge of class membership. The algorithm works to maximize between-group variance and minimize within-group variance. In this study, knowledge of the sampling time was passed to the algorithm, and all models were cross-validated by 3-fold cross-validation, using each biological repeat measurement for model testing in each validation (29).
The key bands contributing to DFA classifications were recovered by inspection of PC-DFA loadings (13, 14). These variables were used to reform DFA, and the data were considered validated if the replotted PC-DFA was the same as in the previous PC-DFA. The most discriminatory bands for DFA classification were recovered, and DFA were reperformed for classification by using 11 or 15 of the 1,764 most contributed bands.
Total lipid extracts (TLEs) were obtained from the harvested microbes using a modified Bligh and Dyer extraction procedure (2). The TLEs were separated into three fractions by Bond-Elut column chromatography (bonded phase NH2, 50 mg Biotage isolute) (15). The neutral lipid fractions were obtained by eluting with dichloromethane/isopropanol, (2:1, vol/vol), the acid fractions by eluting with 2% acetic acid in diethyl ether, and the phospholipid fractions with methanol. Ten microliters (0.1 mg ml−1) of internal standard (tetracosane D50; Isotec) was added to the phospholipids and free fatty acid fractions. The phospholipids were saponified using 95% methanolic NaOH (0.5 M) and heating at 70°C for 1 h. The acid and the saponified phospholipid (phospholipid fatty acid acid [PLFA]) fractions were derivatized with BF3 in methanol (Sigma-Aldrich; heated at 70°C for 6 min) and O-bis(trimethylsilyl) trifluoroacetamide (Sigma-Aldrich; heated at 70°C for 6 min) to convert the acid moieties into their corresponding methyl esters and the alcohol moieties into their trimethylsilyl ethers. The acid and PLFA fractions were analyzed using gas chromatography-flame ionization detection (GC-FID) and gas chromatography-mass spectrometry (GC-MS). Laboratory blanks were analyzed simultaneously, and glass Bond-Elut columns were used to avoid contamination (33). No significant contamination was observed.
Samples were analyzed using an Agilent 7890A GC with FID and an Agilent 7683B autosampler and programmable temperature vaporization (PTV) inlet. The samples were dissolved in hexane and injected using a pulsed splitless injection (1 μl; inlet pressure of 25 lb/in2 for 0.25 min) and separated on an HP-5 capillary column (5% diphenyl-dimethylpolysiloxane; length, 30 m; inner diameter [ID], 320 μm; film thickness, 0.25 μm; J+W Scientific). The samples were run at constant pressure (12.675 lb/in2) with helium as the carrier gas. The oven temperature was programmed from 70°C to 130°C at a rate of 20°C min−1 and then to 300°C at 4°C min−1, where it was kept for 15 min.
GC-MS was performed using a similar type of GC apparatus, inlet, and autosampler as described above, which was interfaced to an Agilent 5975C MSD mass spectrometer operating with electron ionization at 70 eV and scanning from m/z 50 to 600 at 2.7 scans second−1. The heated interface temperature was set at 280°C with the mass source temperature at 230°C and the MS quadrapole at 150°C. Analyses were performed using an HP-5 MS capillary column (5% diphenyl-dimethylpolysiloxane; length, 30 m; ID, 250 μm; film thickness, 0.25 μm; J+W Scientific). Compounds were identified by comparison of spectra with those reported in the literature. Quantitative data were determined by comparison of individual peak areas and comparison with a known concentration of internal standard.
The anaerobic growth rates and biomass yields of S. oneidensis MR-1 were measured in the presence of various concentrations (1 to 400 μM) of Ag(I), as Ag(I) nitrate, added to fumarate-containing medium. Growth was noted up to 100 μM added Ag(I), and there were pronounced color changes at the higher concentrations of silver, from very pale pink in control cultures to intense dark brown at 100 μM Ag(I) (Fig. (Fig.1).1). Before inoculation of the cultures, the medium contained a white precipitate after the addition of high concentrations of silver (data not shown). However, despite the potential precipitation of Ag(I), for example as AgCl, quantifiable concentrations of Ag(I) remained in solution to elicit a toxic response. When 1, 10, or 100 μM Ag(I) was added to the growth medium, after 24 h of incubation in the absence of an inoculum, 0.5, 4.8, and 25.7 μM Ag(I) remained in solution, respectively. There was no color change of silver-containing media in control experiments that lacked added cells.
The growth profiles of S. oneidensis MR-1 grown under anaerobic conditions with silver added over a range of concentrations are shown in Fig. Fig.2.2. The presence of precipitates in the presence of added silver made it impossible to monitor growth through changes in turbidity. Growth was monitored instead by quantifying total cellular protein in the cultures, and protein yields noted in control cultures without added Ag(I) were similar to those reported previously (120 mg/liter after 40 h of growth) using similar media (38), with a doubling time of approximately 7.5 h, again consistent with previous data (38). Also, our data based on quantification of cell numbers through protein measurements correlated well with those from CFU counts (data not shown); the doubling times in the presence of lower concentrations of added Ag(I) were largely similar to those in control replicates (6.4 h with 1 μM and 5.8 h with 10 μM silver), although at 100 μM added Ag(I) the doubling time was extended significantly to 24 h, indicating metal toxicity. This was backed up by measurements of the total protein yield, which was reduced from 120 mg/liter cellular protein in the controls to 72 mg/liter in the presence of 100 μM added Ag(I), while no growth was recorded at 400 μM added Ag(I). There was a stimulation in growth noted at 1 and 10 μM added Ag(I).
After the addition of Ag(I) to the minimal medium, a white precipitate formed, which was collected from the medium supplemented with 100 μM Ag(I) by centrifugation and analyzed by XRD. The precipitate was identified as a mix of AgCl and Ag3PO4 (Fig. (Fig.3A).3A). After inoculating the growth medium with cells of S. oneidensis MR-1, the white precipitate changed color to brown/black (see above). XRD analysis of the precipitate, collected after growing the cells for 24 h in growth medium supplemented with 100 μM Ag(I), showed the presence of elemental Ag(0) (Fig. (Fig.3B).3B). This indicated that the Ag(I), present as AgCl and Ag3PO4, was reduced to elemental Ag.
TEM analysis of thin sections showed the accumulation of electron-dense precipitates of an approximate diameter of 10 nm around the cell after 24 h of incubation in the presence of 10 μM Ag(I), presumably due to the reduction of Ag(I) to Ag(0) at the surface of the cell (Fig. (Fig.4B).4B). At higher magnifications it was apparent that the nanoparticles were precipitated on the outside of the outer membranes of the Gram-negative cells (data not shown here). These precipitates were absent when Ag(I) was not added to the growth medium.
In comparison, the TEM images of S. oneidensis cells grown in anaerobic medium supplemented with 100 μM Ag(I) contained larger (20- to 50-nm-diameter) electron-dense particles, presumably Ag(0), and these were predominantly precipitated within the cytoplasm (Fig. (Fig.4C).4C). In addition, there were significant morphological changes in the cells containing the nanoparticles, including condensation of the electron-dense components of the cytoplasm within the cell and their detachment from the inner membrane/cell wall.
All infrared absorption spectra of S. oneidensis MR-1 samples treated with and without silver were typical of those of Gram-negative bacteria (Fig. (Fig.5).5). The biological characteristic bands, including the acyl vibration bands from fatty acids and lipids, amide I and II protein vibration bands, nucleic acid vibration bands, and phospholipids (22), were recognizable and are highlighted in Fig. Fig.55.
Changes for certain characteristic bands between the cell samples treated with or without silver could be observed in these spectra. The amide I and II bands (1,530 cm−1 to 1,654 cm−1) decreased with increasing silver concentration, as did the fatty acid bands (3,050 cm−1 to 2,800 cm−1) (Fig. (Fig.5).5). By contrast, the bands from carboxylic groups elongated slightly with increasing silver concentration. To reveal more information from these FT-IR spectra, multivariate statistical techniques were used to extract information from these metabolomics data.
The PCA score plot of all of the FT-IR data clearly indicated that S. oneidnsis MR-1 cells grown to log phase under conditions of 400 μM Ag(I) grouped separately from the rest of the samples (data not shown), which was hardly surprising since the cells failed to grow at this high concentration. Therefore, reanalysis of the cells growing under 0 to 100 μM Ag(I) concentrations was conducted using PC-DFA (Fig. (Fig.6A),6A), and this revealed a trend in the first and second PC-DFA scores which was correlated with the increasing concentration of Ag(I). The cells grown in the no-Ag(I) control and samples treated with 1 to 50 μM Ag(I) were separated using either score, while the cells grown in 100 μM Ag(I) were significantly separated from the other cells in the first PC-DFA score. This indicated that the physiology and metabolism of the cells had significantly changed when the Ag(I) concentration was increased to 100 μM.
In order to investigate this finding, the spectral loadings for the first two PC-DFA variables were plotted (Fig. (Fig.6B),6B), and this indicated that the following three regions increased in intensity in samples treated with 100 μM silver: 1,716 to 1,672 cm−1, 1,587 to 1,577 cm−1, and 4,000 to 3,650 cm−1. By contrast, regions from 2,943 to 2,906 cm−1, 2,962 to 2,951 cm−1, and 1,005 to 872 cm−1 decreased in the high-dose treatment group and were higher in cells exposed to 0 to 50 μM Ag(I).
The absorption bands at 1,716 to 1,672 were due to C=O stretching of carboxyl groups, and changes between 1,587 and 1,577 cm−1 formed an asymmetrical C(=O)2 stretching band also from the carboxylate anion group (COO—). These regions both increased, and it is interesting that this is typical of the complexation of the carboxylate anion functional group by coordination with metal cations (7). The absorption informations from 2,962 to 2,951 cm−1 and 2,943 to 2,904 cm−1 revealed decreasing changes in fatty acids. The band centers at 2,924 cm−1 were assigned to the stretching vibrations of CH vibrations. Therefore, binding of silver nanoparticles in this research may have a significant impact on lipid metabolism.
GC-FID and GC-MS analyses of the PLFA fractions purified from S. oneidensis MR-1 cells grown with 10 and 100 μM Ag(I) showed distribution patterns similar to the control [no Ag(I) added]. These profiles were dominated by a homologous series of C14 to C18 fatty acids, with the C16:0 fatty acid being the most abundant component (Fig. (Fig.7).7). The ratios of unsaturated to saturated fatty acids (as defined in Table Table1)1) showed comparable values for all PLFA fractions, with values ranging from 0.64 to 0.77 (Table (Table1).1). By contrast, the total amounts of PLFAs present were reduced substantially from 158 μg/g of dry weight biomass for the S. oneidensis cultures grown with no Ag(I) added to 2 to 21 μg/g of dry weight biomass for the S. oneidensis cells grown with 10 and 100 μM Ag(I) added, respectively (Table (Table11).
Analysis of the free fatty acid fractions, which are precursors for PLFA synthesis, revealed similar distribution patterns to those noted for the PLFA fractions, with dominance by a homologous series of C14 to C18 fatty acids, with C16:0 again being the most abundant component (Fig. (Fig.8A).8A). However, in contrast to the PLFA fractions, in addition to a substantial reduction in the total amounts of fatty acids present with the addition of Ag(I) (from 20.1 μg/g of dry weight biomass to 1.3 to 8.9 μg/g of dry weight biomass), the ratio of unsaturated fatty acids to saturated fatty acids also changed (Table (Table1).1). While the fatty acid fraction of S. oneidensis cells grown with no Ag(I) added was characterized by the presence of substantial amounts of unsaturated fatty acids, predominantly C16:1 fatty acid (Fig. (Fig.8A),8A), the presence of these compounds dropped substantially when the cells were grown in the presence of 10 and 100 μM Ag(I) (Fig. (Fig.8B).8B). This difference was reflected in the ratio of unsaturated to saturated fatty acids, showing a shift from 0.71 for the culture with no Ag(I) added to 0.18 to 0.19 for cultures grown with Ag(I) added (Table (Table11).
It is generally considered that toxic metals are inhibitory to a wide range of microorganisms, impacting microbial growth through concerted actions against diverse metabolic pathways (4). In these experiments, the growth of S. oneidensis MR-1 cells was delayed by the addition of 100 μM Ag(I) and completely inhibited by the addition of 400 μM Ag(I), although there was a slight stimulation of growth rate and cell yields at lower concentrations (10 μM and 1 μM) of the metal. The mechanism of growth stimulation at low silver concentrations remains to be investigated, although it could be linked to the addition of nitrate as a counterion [Ag(I) was supplied as a nitrate salt].
Upon the addition of a 10% inoculum of S. oneidensis MR-1 cells grown to late log phase into medium containing Ag(I), there was a rapid change in color of the medium from cloudy/white to black/brown (Fig. (Fig.1).1). The color change was complete within 1 h. XRD analysis (Fig. (Fig.2)2) indicated the reduction of Ag(I) to Ag(0), with at least some of the Ag(I) present as insoluble AgCl and Ag3PO4. Although the concentrations of soluble Ag(I) were not quantified, it is conceivable that reduction of aqueous Ag(I) ions could have resulted in detoxification of the metal in these cultures. Analyses of the site of Ag(0) precipitation within cell cultures by using TEM gave useful insights into the mechanism of toxicity of silver in these experiments. When added at a 10 μM concentration, the Ag(I) was reduced to form insoluble Ag(0) particles on the surface of the cell, consistent with the involvement of outer membrane c-type cytochromes (24, 25) or secreted flavins (40) in the reduction of extracellular Fe(III) and Mn(IV) minerals. It should be noted that Ag(I) reduction by outer membrane c-type cytochromes has been proposed for other Fe(III)-reducing bacteria (18). At higher concentrations [100 μM Ag(I)], the metal accumulated within the cell as nanoparticles of Ag(0), suggesting that the metal was able to enter the cell either as reduced nanoparticles formed on the cell surface or as Ag(I) ions that were subsequently reduced and precipitated within the cytoplasm. Previous studies suggested that silver can inactivate cellular proteins and DNA replication (37), leading to morphological changes in Gram-negative E. coli and Gram-positive Staphylococcus aureus cells consistent with those noted in the S. oneidensis cultures at 100 μM Ag(I) in this study. These include shrinkage of the inner membrane from the cell wall, detachment of the outer membrane, and condensation of DNA molecules within the center of the cell (6).
The FT-IR data suggested that there were metabolic differences between cells grown with different Ag(I) concentrations in minimal medium under anaerobic conditions. For 1 to 50 μM Ag(I), the FT-IR spectra were grouped relatively closely with the control samples, indicating that the metabolic fingerprints were comparable to those of the control cells (Fig. (Fig.6A).6A). A trend that was related to increasing Ag(I) was clearly observed, indicating a common metabolic shift that changed with increasing Ag(I) from 1 to 50 μM. By contrast, the spectra of cells treated with 100 μM Ag(I) were grouped completely separately (Fig. (Fig.6A),6A), suggesting that the higher concentration of silver had a significant impact on the cellular metabolism. Computational analysis of the FT-IR spectra was used to identify biochemical changes in composition of the bacteria at different concentrations of added Ag(I). PC-DFA revealed that the lipids, free amine groups in the proteins, and the complexation of phosphate and the carboxylate anions in the cells were all affected by the addition of Ag(I). In this study, the carboxylate anion in the cells may have been affected by a high concentration of Ag(I), as the asymmetrical stretching bands centered at 1,583 cm−1 showed the strongest change in the samples treated with 100 μM silver only (Fig. (Fig.6B).6B). The CH2 vibrations (2,962 to 2,951 and 2,943 to 2,906 cm−1) from lipids within the cell showed pronounced signals in cells treated with 1 to 50 μM Ag(I), with a reduction in intensity in this region of the FT-IR spectra when 100 μM Ag(I) was added (Fig. (Fig.6B).6B). This indicated that the higher concentration of silver may have an impact on the lipid components of the cell, including the cellular membranes.
In testing these findings further, GC-FID and GC-MS analyses of both membrane phospholipids and free fatty acids indicated substantial differences between the cells incubated with and without added Ag(I). In particular the total amounts of membrane (PLFA) and free fatty acids (determined as μg/g of dry weight) dropped substantially with the addition of Ag(I) (Table (Table1).1). Although the total PLFA and free fatty acid yields decreased with added Ag(I), no substantial differences in the membrane acid distribution patterns could be observed, for example, as shown by the ratios of unsaturated to saturated fatty acids (Table (Table1).1). This indicated that S. oneidensis may lose its membrane integrity rather than produce different types of these lipids when grown in higher concentrations of added Ag(I). Although these data collectively provide a model for silver toxicity based on decreased membrane integrity and Ag(I) uptake at higher metal concentrations, some results remain challenging to explain. For example, it is unclear why there were higher concentrations of PLFA and free fatty acids at 100 μM Ag(I) than with 10 μM Ag(I) and also why there were enhanced yields of biomass at 100 μM Ag(I). It is clear that more research is required to understand the detailed biochemical response of S. oneidensis to a range of Ag(I) concentrations and also to determine if intermediate concentrations of the metal could be beneficial to anaerobic growth, for example, through use as a terminal electron acceptor.
The results presented in this study clearly show that low micromolar concentrations of Ag(I) have low toxicity for cells of S. oneidensis MR-1, and at 10 μM Ag(I) the metal is precipitated outside the cell, presumably at the site of reduction. At higher concentrations we have provided evidence that the Ag(I) can penetrate the cell and potentially impact on several areas of metabolism, most notably lipid metabolism and membrane integrity. Collectively, these data suggest that silver, which is being used more frequently as a broad-specificity antimicrobial, may not be an efficient biocide against all Gram-negative bacteria, especially when supplied as nanoscale Ag(0) particles analogous to those precipitated on the surface of S. oneidensis. The toxicity of these nanoscale Ag(0) particles is most likely due to oxidation to the more bioavailable Ag(I), and so it is reasonable to suggest that organisms with the capacity to reduce any Ag(I) formed, including Shewanella species, may have enhanced resistance to these biocides. This clearly warrants further work. Finally, we also suggest on the basis of our data that FT-IR spectroscopy may be a useful tool for assessing the metabolic impact of toxic metals and other biocides on microbial cells.
This research was supported by BBSRC grant BBS/B/03718 and a Manchester-CSC Ph.D. scholarship to H. Wang.
Katherine Hollywood is acknowledged for helping with FT-IR data acquisition.
Published ahead of print on 11 December 2009.