Antimicrobial nanotechnology: its potential for the effective management of microbial drug resistance and implications for research needs in microbial nanotoxicology

Abstract

The development of antibiotics revolutionized human health, providing a simple cure for once dreaded diseases such as tuberculosis. However, widespread production, use, and mis-use of antibiotics have contributed to the next-generation concern for global public health: the emergence of multiple drug-resistant (MDR) infectious organisms (a.k.a. “superbugs”). Recently, nanotechnology, specifically the use of nanomaterials (NMs) with antimicrobial activity, has been presented as a new defense against MDR infectious organisms. We discuss the potential for NMs to either circumvent microbial resistance or induce its development in light of our current state of knowledge, finding that this question points to a need for fundamental research targeting the molecular mechanisms causing antimicrobial activity in NMs. In the context of current microbial nanotoxicology studies, particularly reductionist laboratory studies, we offer suggestions and considerations for future research, using an illustrative example from our work with silver nanoparticles.

---

Environmental impact

Multidrug resistant (MDR) pathogenic microbes, which present a growing crisis for public health, have arisen via processes both at the host cellular level as well as environmental systems. Antimicrobial nanotechnology may be a part of next-generation defenses against MDR organisms. We discuss the potential for antimicrobial nanomaterials (NMs) to curb the proliferation of MDR pathogens while avoiding the generation of new NM-resistant organisms. While we find NMs to have promise for the treatment of MDR pathogens, we also find indications that NMs could encourage the development of antibiotic resistance in the environment. We conclude that more information is needed concerning the mechanisms behind the antimicrobial activity of nanomaterials and their potential for influencing the development of resistance in environmental systems.

Introduction

Prior to the development and administration of antibiotics in medicine, the prospect of early death or debilitation from infectious diseases was a fact of life for the world's population.¹ While antibiotics appeared to potentially be a long-term solution for this issue, the emergence of multi-drug-resistant (MDR) microorganisms raises the specter of a future in which we are again lacking pharmaceutical defenses against infection.² So far, microorganisms have been able to keep pace with or arguably overtake human innovation in antibiotic design, developing resistance even to multiple synthetic antibiotics. Strains of Mycobacterium tuberculosis,³ methicillin-resistant Staphylococcus aureus (MRSA),⁴ and several other strains carrying the enzyme New Delhi metallobetalactamase-1 (NDM-1)³ occur for which even last-resort antibiotics may be, and in many cases are, ineffective.

In light of this pressing issue, there has been much interest in possible alternative antimicrobial therapies, including the use of nanotechnology. Nanotechnology is defined as any technology that utilizes nanomaterials (NMs), which are materials with dimensions from 1–100 nm. NMs have frequently been found to have novel physical and chemical properties with respect to their larger-sized (bulk) counterparts, such as greater chemical reactivity, even when normalized for the increase in surface-to-volume ratio, and useful optical features (e.g., strong visible fluorescence).⁵ Such properties are being exploited for a wide array of new products, making nanotechnology an industry that is growing exponentially.

Several NMs are already used as broad-scale antimicrobial agents in consumer products, for example nanoparticles of metallic silver that are impregnated into socks and other garments.⁶ In human medicine, there has been a strong interest in harnessing NMs for targeted drug-delivery.^7,8 Most recently, a new class of NMs has been introduced that has been specifically touted as a weapon against MDR bacteria.⁹ These NMs consist of macromolecular antimicrobial polymers (MAPs), which selectively lyse microbial membranes.^10,11 Their recent modification rendering them biodegradable further enhances their promise in human medicine.⁹ The potential advantage of nano-MAPs is that because they directly destroy bacterial membranes without targeting a very specific step in their metabolic pathways, as do most antibiotics, there may be less opportunity for mutation or other alteration to impart resistance. This may be responsible for the suppressed resistance to nano-MAPs that has been observed.^12,13 Also, because they are non-metallic, there is less possibility of co-selection, as it has been commonly observed that metal and antibiotic resistant elements are carried on the same mobile genetic elements.¹⁴

In this review, we discuss the potential for NMs and associated nanotechnologies to help provide alternatives to traditional antibiotics. Can NMs circumvent microbial drug resistance mechanisms, or will they be added to the pile of antibiotic substances “defeated” by microbial ingenuity? We consider what is currently known about the mechanisms causing NM toxicity to microorganisms, and what this might indicate about the potential for drug resistance issues. Subsequently, we suggest future research directions to address the phenomenon of microbial resistance, using an illustrative example from our own work on silver nanoparticles.

Microbial interactions with NMs: mechanisms of antimicrobial activity and potential for development of microbial resistance

NMs functionalized with antibiotics

NMs functionalized with molecular antibiotics include polymer nanoparticles in which an antibiotic is dispersed or encapsulated,¹⁵ liposomes (vesicles with an aqueous center),^16,17 dendrimers (hyperbranched polymers),¹⁸ and inorganic nanoparticles with antibiotic molecules attached to the surfaces.¹⁹ Such antibiotic-functionalized NMs have been found to be effective in killing MDR isolates of pathogenic bacterial species.²⁰ This is thought to be the case for two primary reasons. First, such functionalized NMs can have improved drug-delivery characteristics as opposed to the molecular antibiotic alone. For example, vancomycin-resistant S. aureus was treated in vitro with vancomycin conjugated to folic acid-tagged chitosan nanoparticles, and minimum inhibitory concentration (MIC) values were reduced to less than 2 μg mL⁻¹, as compared to unconjugated vancomycin (MIC ∼ 50 μg mL⁻¹).²¹ Interestingly, when the chitosan nanoparticles were not tagged with folic acid, they were not effective against the bacteria. It was suggested that a “Trojan Horse” mechanism was taking place, in which the bacterial cells were “deceived” into consuming the nanoparticles because they were coated with a food source. Second, combinations of NM and traditional antibiotics have been shown to impart interesting synergistic effects. A recent study by Brown and co-workers¹⁹ concerned Ag and Au nanoparticles functionalized with ampicillin. An in vitro toxicity study was performed with these functionalized nanoparticles and MDR strains of Pseudomonas aeruginosa and Enterobacter aerogenes as well as MRSA. Ag nanoparticles conjugated with ampicillin had a faster kill rate than Ag nanoparticles alone. Au–ampicillin was toxic to bacteria at a dose of 4 μg mL⁻¹ but Au nanoparticles were not toxic by themselves. Mechanisms behind this and other observations of synergy between NMs and traditional antibiotics are still not clear.²²

The prospects for conjugates or combinations of NMs and traditional antibiotics to circumvent microbial resistance mechanisms give cause for cautious optimism. Particularly, NM–antibiotic conjugates may not be subject to the most common drug resistance mechanisms, such as blocking entry of a molecular antibiotic or efflux pumping out of the cell. However, as traditional antibiotics generally have one specific target (e.g., disabling a certain type of enzyme), we believe that there is still potential for resistance to develop, either by changing the target, breaking down the antibiotic, or upregulation. Furthermore, co-resistance could play a role if genes enabling resistance to antimicrobial NMs are physically connected on the same genetic element and thus “co-selected” when either gene presents an advantage in the face of an antimicrobial. Still, if drug delivery is highly effective, then resistance may be prevented as the offending organism would receive a concentrated dose of antibiotic rather than the sublethal doses that can allow for resistant populations to develop.²³ The synergistic effects between some NMs and traditional antibiotics may also lead to less drug resistance; for example, if the mechanism of killing is multifaceted or less targeted, it may be more difficult for microbes to develop resistance as opposed to when the drug is laser-focused.

NMs with intrinsic antimicrobial properties

Several NMs possess either confirmed or suspected antimicrobial activity. These materials include inorganic nanoparticles of metal (silver, copper) or metal oxides, fullerenes, nanotubes, and MAPs.^9,24–27 While there is some overlap with mechanisms of traditional antibiotic action, there are also distinctions that offer hope that NMs will not select for resistance in the way that antibiotics do. Traditional antibiotics target five main aspects of the cellular machinery:^28,29 (1) cell wall synthesis (e.g., penicillin); (2) folic acid production (e.g., sulfonamides); (3) ribosomal function (e.g., tetracyclines and macrolides); (4) DNA replication (e.g., quinolones); and (5) cell membrane damage (e.g., polymixins and other peptide antibiotics). Bacteria typically fight back by either modifying these targets or the pathways for their synthesis, altering or degrading the antibiotic, or pumping the antibiotic out of the cell (i.e., efflux). The main challenge is that all of these resistance mechanisms are encoded by antibiotic resistance genes (ARGs), which are stable molecules encoded in the DNA and once established can spread from mother to daughter cells or between neighboring bacterial cells by horizontal gene transfer. If ARGs present an advantage for survival (i.e., in the presence of an antibiotic), bacteria will tend to continue to carry them. Even in the absence of antibiotic selective pressure, they do not always incur an energetic cost of carriage and can persist.³⁰ Furthermore, one antibiotic may select for resistance to another antibiotic because the ARGs are physically present on the same element and thus selected together (co-selection).

The sections below detail the known antimicrobial actions of NMs and how they may or may not trigger or select for genetic mechanisms of resistance.

(1) Cell membrane damage. NMs are capable of interacting with bacterial membranes and incurring damage to this vital protective barrier.^31–34 Polymixin antibiotics similarly affect bacterial membranes. Polymixins are comprised of a hydrophobic tail with a positively charged cyclic peptide, which enables it to specifically recognize, interact with and disrupt Gram negative bacterial membranes. Resistance to polymixins exists through modifications of the “lipid A” target, which inhibits binding of the antibiotic.³⁵ Therefore, it is conceivable that bacteria could develop similar means of resistance to antimicrobial NMs through membrane modification. However, the action of NMs against bacterial membranes is generally non-specific, and is not known to involve recognition of a specific membrane component, as is the case for polymixins. Likewise, even in the case of MAPs, resistance mechanisms have been remarkably absent.^12,13 It has been proposed that even some NMs could still selectively bind and inactivate certain surface biomolecules (e.g., thiol groups in proteins could bind to silver nanoparticles), which could enhance chances of developing resistance. After interacting with the bacterial membrane, secondary destructive reactions may also occur, such as generation of reactive oxygen species (ROS) by the nanoparticles (see next section). Intrinsically, as such a mechanism does not entail a specific biochemical pathway, it is likely to be more difficult for bacteria to gain resistance, albeit not impossible. Bacteria could potentially increase the rate of cell membrane synthesis, secrete extracellular polysaccharides to limit ROS from reaching the cell membrane, or generally protect against collateral membrane damage incurred by any of the above mechanisms.

(2) Generation of reactive oxygen species (ROS). Various NMs have been shown to produce reactive oxygen species (ROS), which are well-known to cause catastrophic damage in biological systems by chemically degrading a wide range of organic compounds including DNA, RNA, and proteins.^26,36,37 Metal ions, such as the silver ions that can be released from silver nanoparticles, potentially can catalyze the formation of ROS as well.³⁸

Low concentrations of ROS are widely occurring in nature and correspondingly defenses against them are well-developed in many species of bacteria. Also, both plants and animals can generate ROS to attack unwanted bacteria.³⁹ Bacteria can fight back against ROS by producing enzymes, such as superoxide dismutase, that neutralize ROS radicals. Upregulation of extracellular polysaccharides can also present a “shield” by which to exhaust the ROS on relatively inert cell components. Two examples of well-studied systems for handling oxidative stress are the SoxRS system (responding to superoxide) and the OxyR system (responding to hydrogen peroxide). In E. coli, OxyR and SoxR sense the presence of oxidants and then induce various genes against oxidative stress that are involved in removing oxidants, repairing damaged cell components, and maintaining reducing conditions in the cell. Whereas OxyR responds primarily to H₂O₂ and nitrosylating agents, SoxR is known to respond primarily to superoxide and nitric oxide.^40,41 The ROS response has also been studied and involves induction of a suite of genes to rapidly repair bacterial DNA damage in last resort situations. Intriguingly, the ROS response, which tends to induce a high rate of DNA mutations, has also been linked to the generation and horizontal transfer of antibiotic resistant mutations.^42–45 Bacteria exposed to nanoparticles have demonstrated signs of oxidative stress, which may have been caused by ROS.⁴⁶ However, the extent to which the NM-generated ROS enter the cell to trigger more extensive damage is unknown.

(3) Release of toxic metals. Many of the NMs known to be toxic to bacteria are colloidal nanoparticles of silver or metal oxides. Toxicity of dissolved metals to bacteria is widely known, thus calling to question whether antimicrobial action of metallic NMs is really due to the metal ions, and not necessarily intrinsic to the NMs. It has been demonstrated that metallic NMs can dissolve, releasing toxic metal ions.^26,47–50 However, it is not known whether toxic metal ion release is the primary mechanism by which all such nanoparticles are toxic.

Multiple bacterial species have demonstrated adaptation to high levels of toxic ionic metals, including mechanisms for efflux of the metals as well as sequestration.^14,51 If the mechanism behind the toxicity of some NMs lies in their release of metal ions, this has implications for the potential development of microbial resistance to NMs. One example of concern is the ability of bacteria to resist the toxicity of ionic silver,⁵² particularly as silver nanoparticles are being used commercially for their antimicrobial properties.

Another general means by which bacteria can be resistant to the antimicrobial activity of NMs of any type is via the formation of a biofilm. A biofilm contains bacterial cells within a matrix of extracellular polysaccharides, lipids, and proteins, and can be far more resistant to traditional antibiotics than cells in planktonic form, and hence difficult to eradicate from an infected host. It appears that NMs may be susceptible to the same challenge. Recent studies examining the interaction of Ag nanoparticles with E. coli in both planktonic and biofilm form have demonstrated a higher survival rate of the bacteria in the biofilm.⁵³ This resistance has been linked to factors such as the limited diffusion of Ag nanoparticles in the film, and aggregation of the nanoparticles on the biofilm surface.⁵³ Conversely, tailored NMs, such as magnesium fluoride nanoparticles, have been developed to prevent biofilm formation.⁵⁴

It is important to note that NMs are not entirely new to bacteria. Just as bacteria had evolved over millions of years in the presence of natural antibiotic substances, they have also evolved in the presence of natural NMs such as clay particles, metal oxides, and metal sulfides.⁵⁵ Bacteria have also been found to manufacture NMs in nature. Two notable examples of these NMs are bacterial magnetosomes⁵⁶ and ferrihydrite nanoparticles.⁵⁷ Hence, while NMs are not currently in widespread use to combat disease, it is reasonable to postulate that bacteria do very likely harbor biochemical defense mechanisms that could be brought to bear upon any engineered NM.

Nevertheless, it is equally important to consider that engineered NMs are distinct from natural NMs in many ways. In human medicine, engineered NMs may ideally be designed to strategically attack various aspects of the microbial machinery. For example, it has been suggested that one possible strategy for more effective antibiotics is to develop materials that attack more than one bacterial biochemical process or system simultaneously.⁵⁸ It is plausible that many NMs may be toxic due to more than one factor. For example, they may poison bacteria both via their release of metal ions as well as the generation of ROS.

Also, as mentioned above, NMs have the potential to be tailored to precisely target pathogenic microbes. Already there are numerous examples of targeted biolabeling applications for NMs, for example, by functionalizing the NM to have an affinity for a specific tissue or organism.^59,60 Targeted delivery offers multiple advantages that could help alleviate drug resistance issues. First, it can potentially mean being very specific and efficient in delivery of antimicrobial action. Both are important in that ideally only the target pathogen should be affected by the antibiotic (as opposed to a wider population, including potentially beneficial microflora). Also, high-efficiency delivery reduces the chances that a population would be subjected to sublethal doses which would allow for resistance development.⁶¹

Impact of NM exposure upon microbial resistance to other compounds (e.g. antibiotics)

Exposure of microbial populations to substances that induce cellular stress can lead to selection pressures that favor the development of microbes resistant to traditional molecular antibiotics. There are two major mechanisms by which this can occur. First, there is the phenomenon of cross-resistance, in which a population develops a resistance mechanism for one substance (e.g., heavy metal ions) that also effectively combats other substances (e.g., a molecular antibiotic). Second, if horizontal gene transfer between microbes is induced by a stressor substance (e.g., an antimicrobial NM), the phenomenon of co-resistance can occur, in which a mobile genetic element (e.g., a plasmid encoding resistance genes for multiple antimicrobial substances, such as the NM and antibiotics) can be transferred between microbes. In both cases, the end result is clear—exposure to one type of stressor can result in selection of resistance to both that stressor as well as another antimicrobial substance. NM have the potential to be that stressor, as the potential for NM with antimicrobial activity to be released into the environment is high. Use of antimicrobial NM is on the rise in both consumer products and industry. One example includes the plethora of products coated with antimicrobial silver nanoparticles (textiles, cookware, etc.). Use of products with potential incidental antimicrobial activity is also increasing, such as sunscreen creams that contain zinc oxide or titanium dioxide nanoparticles.

A recent study in PNAS indicates that there is cause for concern.⁶² In this study, horizontal gene transfer of ARGs was stimulated between two bacteria of different genera (E. coli and Salmonella) by the addition of nanoalumina, which is beginning to be used as a high-surface area alternative to bulk alumina in drinking water treatment plants. Horizontal gene transfer occurred via conjugation and at a rate up to 200-fold greater than that of controls. Microbes exposed to the alumina nanoparticles showed signs of oxidative stress and displayed cell membrane damage. These results have serious implications for how antimicrobial NM could impart unintended consequences in terms of development of resistance to existing antimicrobial pharmaceuticals.

Antimicrobial NMs and resistance in microorganisms: implications for research directions in microbial nanotoxicology

In considering the potential for microbes to develop resistance to antimicrobial NMs, as well as the effect of these NMs upon resistance to other drugs, we believe that a number of directions for future research are evident. First, there is a need to determine whether microbes are capable of developing resistance to antimicrobial NMs. To date, we are not aware of peer-reviewed work in this area. Along these lines, it is also important to continue exploring the effects of nanomaterials upon resistance to other compounds, as initiated with the work of Qiu et al.⁶²

Second, we believe that work focusing upon the mechanisms behind nanomaterial toxicity is also critical, as the ability of a microbial population to develop resistance to NMs or have co-selection phenomena occur depends upon the means by which the NMs interfere with microbial functions.

A considerable amount of excellent work focusing upon antimicrobial nanomaterial–microbe interactions has been performed to date. The expertise accumulated through such work is very likely critical to achieving our need for a mechanistic understanding of nanomaterial toxicity, and should be harnessed. The natural question to ask is then how current studies of microbial nanotoxicology offer mechanistic insights, and to consider how they may be refined and focused.

Arguably the most widely used experimental model for examining NM–microbial toxicology is the reductionist approach of exposing microbes (often a pure culture of one bacterial strain) in a set growth medium to a series of NMs, generally with systematic variation of NM physicochemical parameters (e.g. shape, size, coating).^63–66 (We acknowledge that there are notable studies with complex microbial systems but elect not to discuss them here as in our opinion a reductionist approach may be initially a simpler path to understanding toxicological mechanisms). Such studies can be very powerful, but we feel that very careful attention has to be paid to designing a study such that it results in conclusive data about mechanism. There are many issues that can complicate interpretation of results.

We offer suggestions for refining this approach, using a small study we performed as an illustrative example. In our study we examined the effect of surface coating upon the toxicity of silver nanoparticles to a wild-type strain of the Gram-negative bacterial species Pseudomonas aeruginosa (P. aeruginosa is omnipresent in the natural environment, but is also of concern as an opportunistic pathogen; MDR strains pose a problem for patients in hospital settings). Three types of colloidal silver nanoparticles were tested, all coated with polyvinylpyrrolidone (PVP) of three different chain lengths (described according to the average molecular weight cutoffs (MWCOs) of the polymer in Daltons): 10k, 55k, and 360k. P. aeruginosa was grown in sterile defined media containing different concentrations of these nanoparticles in a standard 24-hour minimum inhibitory concentration (MIC) assay, which determines a concentration of NM at which no growth is observed after 24 hours. We offer here some of the salient points of what we learned from this study in the context of other in vitro bacteria–NM studies, hopefully offering some insight for the design of future investigations into microbe–NM interactions of potential medical interest. For further information regarding experimental details, please refer to the Materials and Methods section. We summarize here what we discovered through our work.

(1) Varying single material parameters (size, morphology, surface coating, etc.) still has the potential to simultaneously alter several features of the physicochemical behavior of the NM being tested. In the case of our experiment, varying the length of the PVP polymers coating the silver nanoparticles could vary multiple physicochemical behaviors such as: the kinetics of silver ion release from the nanoparticles, the kinetics of reactive oxygen species (ROS) generation and transport to bacterial cell surfaces, and steric repulsion between cell surfaces and the coatings. Changes in any of these behaviors (or a combination thereof) could be the mechanism behind differences we observed in MIC measurements using Ag nanoparticles with different PVP coatings. To compensate for this fact, we made auxiliary measurements examining factors such as ionic silver release, which will be further described later in this text.

In general terms, the now classic nanotoxicology experiment in which microbes are exposed to a variety of NMs with controlled variations in physicochemical parameters may not be the most straightforward approach to deducing toxicological mechanisms.

(2) The physicochemical state of the NM being tested may change throughout the experiment, and this needs to be monitored if possible. If such changes are not taken into account, the interpretation of the toxicological data may not be meaningful. It is entirely possible for the mechanism behind the toxicity of the NM to vary during the course of the experiment.

Measurement of the physicochemical state of a given NM during an experiment (e.g. its size, degree of dissolution, or aggregation state) is not straightforward. For example, in our experiment, we found that the MICs for the silver nanoparticles were in the 32–128 ppb Ag range. Other studies have shown silver nanoparticle toxicity in significantly higher concentration ranges.^67–69 We suggest that this may have been due to the defined, simple media (no biomolecules in solution to significantly alter reactivity), the highly dispersed nature of the nanoparticles, and even potentially the amount of cells with which the nanoparticle–media solutions were inoculated.

At this low concentration of nanoparticles, many methods for monitoring the state of the nanoparticles during bacterial growth (e.g., elemental analysis or optical spectroscopy) could not be applied. We therefore had to perform tests monitoring the aggregation state and dissolution of the silver nanoparticles at much higher concentrations and extrapolate these results to our actual system, which is not entirely optimal. Not only do chemical and physical phenomena vary with concentration, but the possible effects of the bacteria upon the NMs (e.g., production of extracellular polysaccharides that could cause aggregation) could not be directly studied in this circumstance.

(3) The NM being studied may not be entirely “ideal” and researchers must account for this in their interpretation of toxicology data. For example, if a NM tested is purchased commercially, it may have some unaccounted chemical impurities that could impact toxicological results. We circumvented this issue by utilizing home-synthesized nanoparticles and subjecting them to extensive characterization.

It is rare that even in the case of a series of home-synthesized NMs that one can easily vary a single material parameter without varying anything else about the NM. For example, it is common when varying the size of a NM that its morphology will change as well. An illustration of this is in studies on varying sizes of iron oxide nanocrystals^70,71 in which 7 nm nanocrystals have a hexagonal platy morphology, while 40 nm nanocrystals have a rhombohedral morphology. Therefore while a researcher will be focusing upon differences in toxicology with respect to size, he or she needs to equally consider the possible impact of a change in morphology.

In our study, we had 10k PVP-coated nanoparticles and the 55k PVP-coated nanoparticles uniformly sized at 28 nm (of silver component measured by TEM). We then synthesized 360k PVP-coated nanoparticles. Unfortunately, these nanoparticles were only 17 nm. The variation of two major physicochemical parameters, namely size (hence total surface area) and the polymer chain length of the coating further complicated interpretation.

(4) Biological variables must be considered, such as the physiological state of bacteria, species variations amongst bacteria, and growth/exposure conditions, particularly when assessing the broader implications of a given NM–microbe experiment. Experiments throughout the literature vary the physiological state and mode of exposure of the bacteria to the NM in question. In our study, partially because we were following a previously published standard MIC protocol, the bacteria tested were in stationary state when initially exposed to the NM. In other studies, the bacteria have been exposed in log phase.^72,73 Parameters concerning the mode of exposure can be of importance as well. In our study, the bacteria are exposed to silver nanoparticles for the entirety of the 24-hour test period, while in other experiments, exposures are briefer.^27,74

We have noticed that, as expected, bacterial growth and physiology can vary depending upon conditions such as the composition of the growth media, and this certainly has implications for the interaction of the bacteria with NMs. For example, in the absence of EDTA, we noted that P. aeruginosa would secrete optically detectable amounts of pyoverdin, a siderophore, which could certainly affect the surface chemistry of a given NM.

(5) There is room for more sophistication in measurements of the biological responses of microbes to NM exposure. Our study, as with many studies, was limited to a simplistic measure of biological response—optical density measurements of cell growth. Note that at the low concentrations of silver nanoparticles that we were using, an optical signal from the nanoparticles themselves was not detectable. Unfortunately, that meant that even when we saw a change in cell density, we could not be certain as to what was happening to cause that change. Were bacteria dying, or was their growth simply stalled? Were the nanoparticles inducing oxidative stress in the bacteria? We are in this case rather blind. Notable progress is being made in providing more sophisticated biological analyses, such as using GFP-labeled indicator strains of bacteria.⁷⁵ Further careful biological analyses are needed, such as gene expression analysis combining reverse transcription with quantitative polymerase chain reaction (RT-qPCR) or microarrays, which can measure expression levels of large numbers of genes simultaneously. Results from our study are summarized in Table 1. For experimental details, please refer to the materials and methods section.

Table 1 Summary of toxicity study results and characterization

Measurement	10k PVP Ag nanoparticles	55k PVP Ag nanoparticles	360k PVP Ag nanoparticles
a Total silver includes both ionic and nanoparticulate silver. b Inferred from dynamic light scattering study performed in media, notably in absence of a live culture. c Optical density of cultures grown with the same mass of PVP polymer as the sole carbon source in the media, measured after 24 hours of growth. d Ionic Ag is functionally defined as any silver that could pass through a 5k MWCO ultrafilter.
MIC (total^a Ag)	32 ppb	64 ppb	64 ppb
Size (nm) by TEM	27.9 (±6.4) nm	28 (±13) nm	17.8 (±7.6) nm
Hydrodynamic diameter in media	25.3 (±0.09) nm	39.2 (±1.2) nm	25.8 (±0.89) nm
Aggregation state throughout experiment^b	Dispersed	Dispersed	Dispersed
Zeta potential (in water)	−29.38 (±16.75) mV	−27.56 (±11.60) mV	−23.3 (±8.75) mV
TOC	0.15 ppm/ppm Ag	0.06 ppm/ppm Ag	0.8 ppm/ppm Ag
Coating metabolized by P. aeruginosa?^c	Yes, readily (OD600 = 1.08 @ 24 h)	Yes, slowly (OD600 = 0.21 @ 24 h)	Very slowly (OD600 = 0.025 @ 24 h)
% Ionic Ag before incubation^d	1.83 (±0.24)	0.62 (±0.09)	0.99 (±0.09)
% Ionic Ag after incubation^d	5.15 (±1.46)	1.24 (±0.20)	1.01 (±0.05)

---

What can be concluded from this study about the causes of differences in MIC between these nanoparticles with varying coatings, and the mechanisms involved in nanoparticle toxicity? We can with some confidence state that the greater toxicity of the 10k MWCO PVP-coated silver nanoparticles (hence referred to as 10k-PVPnanoAg) is not likely to be due to differences in aggregation state, given the DLS data. The size of the naked silver nanoparticle did not appear to be affecting toxicity, as the 10k-PVPnanoAg has the same silver particle diameter as the 55k-PVPnanoAg (as measured by TEM), yet the 10k-PVPnanoAg are more toxic. While we were unable to conduct zeta potential measurements successfully in the media used, no significant difference was detected between the zeta potentials of the nanoparticles in water. Given the similarity of the three types of coatings, we believe it is likely that their surface charge behavior in the media would not be significantly different. Such results could suggest that the chain length of the coating polymer used was the main parameter influencing toxicity. However, this observation does not provide much information about the mechanisms behind the toxicity of the nanoparticles. As mentioned above the differences in coating chain lengths could alter a number of variables in the system that can affect toxicity, such as distance between the bacterial cell surface and the nanoparticle and diffusion of water to the silver nanoparticle surface (hence affecting factors such as dissolution kinetics).

As we were interested in the mechanisms behind the antibiotic activity of the silver nanoparticles, we made auxiliary measurements as much as was feasible for the system. For example, to account for the effect of coatings upon the release of ionic silver, we incubated solutions of silver nanoparticles in our growth media, and used ultrafiltration to separate out nanoparticles from ionic silver (defined as anything that could pass through a 5k MWCO ultrafilter). We measured the amount of ionic silver present before and after 24 hours of incubation. It should be noted that in order for our experiments to be well above detection limits for silver, we had to conduct them in the ppm range. Also, these dissolution experiments could not account for biological effects, as total Ag concentrations in ppm ranges had been found to be toxic to P. aeruginosa in our system. Therefore, dissolution data obtained from these experiments may only suggest what happened in our actual system. 10k-PVPnanoAg had a higher concentration of ionic silver present before incubation than 55k-PVPnanoAg and 360k-PVPnanoAg. It is unclear whether this higher amount was due to dissolution-related phenomena or a higher concentration of ionic Ag remaining from the synthesis. After incubation the percentage of ionic silver increased for the 10k-PVPnanoAg and to a lesser extent for the 55k-PVPnanoAg sample; there was no significant change in the percentage of ionic silver present for 360k-PVPnanoAg. It may seem that the plausible explanation for the greater toxicity of the 10k-PVPnanoAg particles was due to a greater release of ionic silver. However, it should be noted that at the MIC concentration of total silver, 32 ppb, the amount of ionic silver present is 2.71 ppb, which is lower than the MIC in our AgNO₃ control (4 ppb). This does not rule out the possibility that release of ionic Ag plays a role in degree of toxicity. We did not take detailed data on the rate of ionic Ag release, which could make a difference. Also, this may be a matter of how the ionic Ag is delivered to the cells. If the nanoAg is attaching to cell surfaces in some fashion, the localized concentration of ionic Ag may be far higher than that measured for bulk solution. While we did examine bacterial cells with SEM as well as whole-mount and thin-section TEM, we were not able to observe any obvious association of nanoAg with the cells; this may have been due to the low concentrations of nanoAg we applied.

In addition to examining the dissolution behavior of the Ag nanoparticles, we also checked to see if the PVP coating polymers were in any fashion toxic to P. aeruginosa. On the contrary we discovered that PVP can be metabolized. In follow-up experiments we found that P. aeruginosa could readily grow even when PVP was the sole carbon source provided. After a 24 hour incubation, growth of P. aeruginosa on 10k PVP was especially pronounced relative to 55k PVP, and 360k PVP as per optical density measurements. The difference in the degree to which the bacteria could consume the shorter chain length PVP was remarkable. While this points to an intriguing factor in considerations of toxicity, again, we could not conclusively determine whether this influenced our results.

From this study, we can conclude that even what seems to be a relatively non-remarkable change in surface chemistry (chain length of polymer coating), could indeed cause a difference in toxicity. While one might argue that the lack of a difference in the MIC values for 55k PVP-coated nanoparticles and 360k PVP-coated nanoparticles might argue against this supposition, it should be noted that the dimensions of the silver component in 360k PVP-coated nanoparticles were significantly smaller than that of the 55k PVP-coated nanoparticles. This size difference could mean that more reactive surface area was available in the 360k-PVPnanoAg which could obscure potentially toxicity-lowering effects of the longer-chain 360k PVP polymer coating.

While a number of mechanisms behind the observed toxicity could be proposed, it is not possible to definitively pinpoint the exact cause from the experiments we performed. Certainly one set of analyses we had not deeply explored was that of the biological response—e.g. expression of metal resistance genes, or changes in metabolism. This is left for future studies.

Conclusion

Nanomaterials show promise as a powerful weapon in the continuing race against MDR organisms. They first present a different approach to damaging microbial cell functions. Rather than focusing upon particular biochemical processes, as is the case with traditional antibiotics, they are likely to disrupt multiple cellular processes in a less specific fashion. This multifaceted approach may make it more difficult for microbes to develop resistance. Second, NMs offer the possibility of more efficient and targeted delivery of antibiotic agents. Better drug delivery can lower the likelihood of sublethal dosing of antibiotics as well as broad spectrum microbial exposures, which in turn could reduce the development of resistance to NMs.

Predicting the potential for NMs as antimicrobial agents as well as being able to effectively engineer NMs to be effective antibiotics require a deeper understanding of at least two main areas of study. The first concerns the specific mechanisms by which NMs have antimicrobial activity. Elucidating these mechanisms requires careful study to deconvolute the many variables affecting antimicrobial activity. As we demonstrate with an example from our own research, even what appears on the surface to be a carefully designed NM–bacteria experiment can still be in some ways poorly controlled, leading to inconclusive results. The second is to understand how NMs may affect the genetic transfer of antibiotic resistance amongst organisms, much of which can happen in environmental systems (e.g., wastewater treatment plants).

What we do currently know about the mechanisms behind the antimicrobial activity of NMs still indicates that the development of resistance is possible. Beyond the cell-host system, a recent study indicates that NMs have the potential to induce horizontal gene transfer in environmental systems, which could lead to increased resistance.

We conclude that while NMs do clearly display promise as antibiotic agents effective even against MDR organisms, further information is greatly needed. We recommend that caution be exercised in the use of NMs for antimicrobial purposes, as there is still the potential for increased MD-resistance, particularly if NMs are not judiciously applied.

Materials and methods

Preparation of PVP-coated silver nanoparticle suspensions

Polyvinylpyrrolidone (PVP)-coated silver nanoparticles were prepared by using the polyol method.⁷⁶ In brief, 20 g of PVP with a molecular weight cutoff (MWCO) of 10 kDa, 55 kDa or 360 kDa were dissolved in 75 mL of ethylene glycol at room temperature under magnetic stirring. Subsequently 1.5 g of AgNO₃ were added to this solution. Once the AgNO₃ was completely dissolved, the solution was heated to 120 °C at a rate of 1 °C min⁻¹. Heating and stirring were maintained for 24 hours, after which the reaction mixture was allowed to cool to room temperature. To retrieve the nanoparticles, the reaction mixture was diluted with Nanopure water (volume ratio 1 : 10) and then centrifuged at 15 000 rpm for 30 minutes. The supernatant was then decanted and the nanoparticles dispersed in Nanopure water again. This process was repeated three times. The resulting suspension of nanoparticles was then stored in the dark at 4 °C.

Characterization of PVP-coated silver nanoparticles

After synthesis, the 10k PVP-, 55k PVP-, and 360k PVP-coated silver nanoparticles were characterized with transmission electron microscopy (TEM), UV-visible absorption spectroscopy, and dynamic light scattering (DLS). Primary particle size and morphology were characterized using a Tecnai G² Twin TEM (at 200 kV) equipped with a CCD camera system (FEI, Hillsboro, OR, USA). TEM samples were prepared by pipetting 10–20 μL of the PVP-coated silver nanoparticle suspensions onto TEM grids with Formvar/carbon support films (Ted Pella Inc. Redding, CA, USA) which were then allowed to air-dry at room temperature. Images were analyzed using the ImageJ image processing program (National Institutes of Health, USA) to measure the average primary particle size and to assess the uniformity of the particles' size and shape. Mean particle sizes for the 10k PVP-, 55k PVP-, and 360k PVP-coated silver were estimated from measurements on >500 particles for each sample.

A variety of analyses were performed upon nanoparticles suspended in both Nanopure water and PAM9 media (growth media used for experiments described below), except for zeta potential measurements which were performed in water. UV-visible spectroscopy measurements of the PVP-coated silver nanoparticles suspensions were obtained using a Cary 500 scan UV-VIS-NIR spectrophotometer (Varian, CA, USA). Estimates of hydrodynamic diameter (Z-average) and zeta potential were obtained using a CGS 3 (ALV-GmbH, Germany) equipped with a helium–neon laser (632.8 nm). The concentration of PVP was estimated by measuring total organic carbon (TOC) using a Shimadzu 5050A TOC analyzer (Columbia, MD, USA). Total silver concentrations of the PVP-coated silver nanoparticle suspensions were measured using inductively coupled plasma atomic emission spectrometry (ICP-AES) with a Spectro ARCOS SOP (Spectro Analytical Instruments, Germany) preceded by nitric acid digestion. Final silver concentration of the PVP-coated silver nanoparticle suspension was adjusted by measured recovery in a silver standard that underwent nitric acid digestion followed by ICP-AES quantification. Samples and calibration standards were prepared in a matrix of 2% nitric acid to minimize adsorption to sample tubing or internal instrument surfaces.

Bacterial strain and media

Wild-type Pseudomonas aeruginosa strain PA01 HS was obtained from Professor Nancy Love (University of Michigan Ann Arbor). The media used was PAM9 minimal media consisting of the following in g L⁻¹: Na₂HPO₄ (6), NaH₂PO₄ (3), NH₄Cl (1), CH₃COOH (0.94), NaCl (0.5), MgSO₄·7H₂O (0.25); in mg L⁻¹: CaCl₂·2H₂O (39), KH₂PO₄ (22), FeSO₄·7H₂O (3.8), Na₂EDTA·2H₂O (5), ZnCl₂ (0.25), MnSO₄·H₂O (0.18); in μg L⁻¹: CuCl₂·2H₂O (32), NaMoO₄·2H₂O (6), H₃BO₃ (3), CoCl₂·6H₂O (1). All chemicals used in the media were reagent grade or higher. The final pH of the media was adjusted to 7.0 with NaOH. After autoclaving, the media was sterile filtered (0.2 μm filter) to remove precipitates that could interfere with particle size analyses and optical measurements.

Bacterial growth study

Cells of PA01 HS were grown in PAM9 overnight at 37 °C into lag phase and diluted to an OD₆₀₀ of 0.1 with fresh PAM9. To test the effects of PVP-coated silver nanoparticles upon cell growth, culture media were prepared with 10k PVP-, 55k PVP-, and 360k PVP-coated silver nanoparticle suspensions as additives. In order to compare the toxicity of nanoparticulate silver with equivalent amounts of ionic silver, culture media was also prepared with AgNO₃ as an additive. To examine the effects of the PVP coatings upon cultures, culture media was prepared with PVP polymer added, including preparations in which PVP was the only carbon source provided to the bacteria. Finally, to determine whether any remaining reaction byproducts had detectable effects on cell density, silver nanoparticle suspensions were subjected to ultrafiltration with Amicon filters (Millipore, MA, USA) to remove the nanoparticles, and the filtrates were used as additives to culture media. All culture media with additives were sterile-filtered through 0.2 μm syringe filters (Acrodisc with Supor membranes, Pall Co., NY, USA).

The PVP-coated silver nanoparticle media was 2-fold serially diluted from its highest to lowest concentrations in a 96-well sterile culture plate (Corning, USA) to obtain total silver concentrations ranging from 1 ppb to 128 ppb. Media with other additives were diluted in the same fashion. PAM9 media was added to all wells to obtain a final volume of 100 μL in each well. At every concentration, 10 μL of diluted cell culture were added to half of the wells. The same volume of sterile culture was added to the remaining wells, which were abiotic controls. For every plate used, a set of additive-free control cultures was grown.

Plates were incubated at 25 °C for 24 hours in the dark. Growth was monitored after 24 hours by measuring absorbance at 600 nm in a SpectraMax 190 plate reader (Molecular Devices, CA, USA). All growth experiments were repeated three times. To determine minimum inhibitory concentrations (MICs) for the silver nanoparticles and ionic silver, a local regression (LOESS) model was used.

Dissolution of PVP-coated silver nanoparticles during incubation

We conducted measurements to estimate the release of silver ions by silver nanoparticles, as this could affect toxicity. PVP-coated silver nanoparticles suspended in PAM9 media were incubated under darkness at 25 °C for 24 hours. Prior to incubation, these suspensions were analyzed with ICP-AES for total silver concentration (silver nanoparticles + ionic silver). They were also subjected to centrifugal ultrafiltration with 5k MWCO Amicon filters (Millipore, MA, USA) to remove nanoparticulate silver, and the remaining filtrate was analyzed for the concentration of ionic silver. After incubation, the suspensions were subjected to the same processing and analyses for total and ionic silver. All dissolution experiments were performed in triplicate.

In order to ensure detection of silver concentrations via ICP-AES, higher concentrations of silver nanoparticles were used for dissolution experiments (2, 5, and 7 ppm total Ag) than those used in culture experiments (1–128 ppb). As total silver concentrations ≥64 ppb were found to be toxic to P. aeruginosa PA01 HS, these dissolution experiments were performed abiotically. Changes in ionic silver are therefore reported as percentages of the initial total silver measured.

Acknowledgements

A grant from the National Science Foundation (NSF) and the Environmental Protection Agency (EPA) under NSF Cooperative Agreement EF-0830093, entitled Center for the Environmental Implications of Nanotechnology (CEINT), provided major financial support for this study. D.M.A. acknowledges support from National Science Foundation grant DEB-0610373 (Postdoctoral Fellowship) during part of this work. This manuscript was completed during her current service at NSF. This work was also supported by National Science Foundation CBET CAREER award #0852942 and the Virginia Tech Institute for Critical Technology and Applied Science (ICTAS). The authors are appreciative of important assistance from Stephen McCartney in the Nanoscale Characterization and Fabrication Laboratory at Virginia Tech, as well as assistance from Athena Tilley of the Virginia Tech Soil Testing Laboratory. We also thank Professor Jeffrey Kuhn (Virginia Tech) for the use of his laboratory and plate reader. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NSF or the EPA. This work has not been subjected to EPA review and no official endorsement should be inferred.

References

1. J. Carlet, P. Collignon, D. Goldmann, H. Goossens, I. C. Gyssens, S. Harbarth, V. Jarlier, S. B. Levy, B. N'Doye, D. Pittet, R. Richtmann, W. H. Seto, J. W. M. van der Meer and A. Voss, Lancet, 2011, 378, 369–371.
2. J. Carlet, V. Jarlier, S. Harbarth, A. Voss, H. Goossens and D. Pittet, Antimicrob. Resist. Infect. Contr., 2012, 1, 11.
3. Z. F. Udwadia, R. A. Amale, K. K. Ajbani and C. Rodrigues, Clin. Infect. Dis., 2012, 54, 579–581.
4. D. P. Calfee, Curr. Opin. Infect. Dis., 2012, 25, 385–394.
5. D. M. Aruguete, J. Liu and M. F. Hochella, in Environmental and Human Health Consequences of Manufactured Nanoparticles, ed. J. Lead and E. Smith, John Wiley and Sons, Chichester, UK, 2009, ch. 3, pp. 79–108.
6. T. M. Benn and P. Westerhoff, Environ. Sci. Technol., 2008, 42, 4133–4139.
7. C. Seeney, J. O. Ojwang, R. D. Weiss and J. Klostergaard, Nanomedicine, 2012, 7, 289–299.
8. R. A. Petros and J. M. DeSimone, Nat. Rev. Drug Discovery, 2010, 9, 615–627.
9. F. Nederberg, Y. Zhang, J. P. K. Tan, K. J. Xu, H. Y. Wang, C. Yang, S. J. Gao, X. D. Guo, K. Fukushima, L. J. Li, J. L. Hedrick and Y. Y. Yang, Nat. Chem., 2011, 3, 409–414.
10. R. E. W. Hancock and H. G. Sahl, Nat. Biotechnol., 2006, 24, 1551–1557.
11. I. S. Radzishevsky, S. Rotem, D. Bourdetsky, S. Navon-Venezia, Y. Carmeli and A. Mor, Nat. Biotechnol., 2007, 25, 657–659.
12. K. A. Brogden, Nat. Rev. Microbiol., 2005, 3, 238–250.
13. R. I. Lehrer and T. Ganz, Curr. Opin. Immunol., 1999, 11, 23–27.
14. C. Baker-Austin, M. S. Wright, R. Stepanauskas and J. V. McArthur, Trends Microbiol., 2006, 14, 176–182.
15. G. Mohammadi, A. Nokhodchi, M. Barzegar-Jalali, F. Lotfipour, K. Adibkia, N. Ehyaei and H. Valizadeh, Colloids Surf., B, 2011, 88, 39–44.
16. D. Pornpattananangkul, L. Zhang, S. Olson, S. Aryal, M. Obonyo, K. Vecchio, C. M. Huang and L. F. Zhang, J. Am. Chem. Soc., 2011, 133, 4132–4139.
17. C. M. Huang, C. H. Chen, D. Pornpattananangkul, L. Zhang, M. Chan, M. F. Hsieh and L. F. Zhang, Biomaterials, 2011, 32, 214–221.
18. M. K. Mishra, K. Kotta, M. Hali, S. Wykes, H. C. Gerard, A. P. Hudson, J. A. Whittum-Hudson and R. M. Kannan, Nanomedicine, 2011, 7, 935–944.
19. A. N. Brown, K. Smith, T. A. Samuels, J. R. Lu, S. O. Obare and M. E. Scott, Appl. Environ. Microbiol., 2012, 78, 2768–2774.
20. A. J. Huh and Y. J. Kwon, J. Controlled Release, 2011, 156, 128–145.
21. S. P. Chakraborty, S. K. Sahu, S. K. Mahapatra, S. Santra, M. Bal, S. Roy and P. Pramanik, Nanotechnology, 2010, 21, 105103.
22. A. M. Allahverdiyev, K. V. Kon, E. S. Abamor, M. Bagirova and M. Rafailovich, Expert Rev. Anti-Infect. Ther., 2011, 9, 1035–1052.
23. J. Davies, G. B. Spiegelman and G. Yim, Curr. Opin. Microbiol., 2006, 9, 445–453.
24. D. Y. Lyon, D. A. Brown and P. J. J. Alvarez, Water Sci. Technol., 2008, 57, 1533–1538.
25. J. R. Morones, J. L. Elechiguerra, A. Camacho, K. Holt, J. B. Kouri, J. T. Ramirez and M. J. Yacaman, Nanotechnology, 2005, 16, 2346–2353.
26. L. K. Adams, D. Y. Lyon and P. J. J. Alvarez, Water Res., 2006, 40, 3527–3532.
27. A. Thill, O. Zeyons, O. Spalla, F. Chauvat, J. Rose, M. Auffan and A. M. Flank, Environ. Sci. Technol., 2006, 40, 6151–6156.
28. T. M. Bakheet and A. J. Doig, BMC Bioinf., 2010, 11, 195–204.
29. M. A. Kohanski, D. J. Dwyer and J. J. Collins, Nat. Rev. Microbiol., 2010, 8, 423–435.
30. D. E. Rozen, L. Mcgee, B. R. Levin and K. P. Klugman, Antimicrob. Agents Chemother., 2007, 51, 412–416.
31. L. L. Zhang, Y. H. Jiang, Y. L. Ding, M. Povey and D. York, J. Nanopart. Res., 2007, 9, 479–489.
32. A. L. Neal, Ecotoxicology, 2008, 17, 362–371.
33. Q. L. Li, S. Mahendra, D. Y. Lyon, L. Brunet, M. V. Liga, D. Li and P. J. J. Alvarez, Water Res., 2008, 42, 4591–4602.
34. S. Kang, M. Pinault, L. D. Pfefferle and M. Elimelech, Langmuir, 2007, 23, 8670–8673.
35. A. X. Tran, M. E. Lester, C. M. Stead, C. R. H. Raetz, D. J. Maskell, S. C. McGrath, R. J. Cotter and M. S. Trent, J. Biol. Chem., 2005, 280, 28186–28194.
36. O. Choi and Z. Q. Hu, Environ. Sci. Technol., 2008, 42, 4583–4588.
37. Y. Li, W. Zhang, J. F. Niu and Y. S. Chen, ACS Nano, 2012, 6, 5164–5173.
38. M. J. Sweet and I. Singleton, Adv. Appl. Microbiol., 2011, 77, 115–133.
39. S. O. Kotchoni and E. W. Gachomo, J. Biosci., 2006, 31, 389–404.
40. M. S. Koo, J. H. Lee, S. Y. Rah, W. S. Yeo, J. W. Lee, K. L. Lee, Y. S. Koh, S. O. Kang and J. H. Roe, EMBO J., 2003, 22, 2614–2622.
41. E. Cabiscol, J. Tamarit and J. Ros, Int. Microbiol., 2010, 3, 3–8.
42. P. Ysern, B. Clerch, M. Castano, I. Gibert, J. Barbe and M. Llagostera, Mutagenesis, 1990, 5, 63–66.
43. K. Drlica and M. Malik, Curr. Top. Med. Chem., 2003, 3, 249–282.
44. J. W. Beaber, B. Hochhut and M. K. Waldor, Nature, 2004, 427, 72–74.
45. L. R. Mesak, V. Miao and J. Davies, Antimicrob. Agents Chemother., 2008, 52, 3394–3397.
46. T. Xia, M. Kovochich, M. Liong, L. Madler, B. Gilbert, H. B. Shi, J. I. Yeh, J. I. Zink and A. E. Nel, ACS Nano, 2008, 2, 2121–2134.
47. X. Y. Yang, A. P. Gondikas, S. M. Marinakos, M. Auffan, J. Liu, H. Hsu-Kim and J. N. Meyer, Environ. Sci. Technol., 2012, 46, 1119–1127.
48. E. Navarro, F. Piccapietra, B. Wagner, F. Marconi, R. Kaegi, N. Odzak, L. Sigg and R. Behra, Environ. Sci. Technol., 2008, 42, 8959–8964.
49. G. Q. Liu, D. M. Wang, J. M. Wang and C. Mendoza, Sci. Total Environ., 2011, 409, 2852–2857.
50. R. Ganesh, J. Smeraldi, T. Hosseini, L. Khatib, B. H. Olson and D. Rosso, Environ. Sci. Technol., 2010, 44, 7808–7813.
51. G. M. Teitzel and M. R. Parsek, Appl. Environ. Microbiol., 2003, 69, 2313–2320.
52. S. Silver, FEMS Microbiol. Rev., 2003, 27, 341–353.
53. O. Y. Choi, C. P. Yu, G. E. Fernandez and Z. Q. Hu, Water Res., 2010, 44, 6095–6103.
54. J. Lellouche, A. Friedman, J. P. Lellouche, A. Gedanken and E. Banin, Nanomedicine, 2012, 8, 702–711.
55. M. F. Hochella, Jr., D. Aruguete, B. Kim and A. S. Madden, in Nature's Nanostructures, ed. A. S. Barnard and H. Guo, Pan Stanford Publishing, Singapore, 2012, ch. 1, pp. 1–42.
56. A. Kornig and D. Faivre, in Nature's Nanostructures, ed. A. S. Barnard and H. Guo, Pan Stanford Publishing, Singapore, 2012, ch. 9, pp. 249–271.
57. H. Konishi, H. Xu, and H. Guo, in Nature's Nanostructures, ed. A. S. Barnard and H. Guo, Pan Stanford Publishing, Singapore, 2012, ch. 3, pp. 75–113.
58. M. A. Kohanski, D. J. Dwyer, B. Hayete, C. A. Lawrence and J. J. Collins, Cell, 2007, 130, 797–810.
59. M. F. Frasco and N. Chaniotakis, Anal. Bioanal. Chem., 2010, 396, 229–240.
60. P. C. Ray, S. A. Khan, A. K. Singh, D. Senapati and Z. Fan, Chem. Soc. Rev., 2012, 41, 3193–3209.
61. F. Baquero, Drug Resist. Updates, 2001, 4, 93–105.
62. Z. G. Qiu, Y. M. Yu, Z. L. Chen, M. Jin, D. Yang, Z. G. Zhao, J. F. Wang, Z. Q. Shen, X. W. Wang, D. Qian, A. H. Huang, B. C. Zhang and J. W. Li, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 4944–4949.
63. G. A. Martinez-Castanon, N. Nino-Martinez, F. Martinez-Gutierrez, J. R. Martinez-Mendoza and F. Ruiz, J. Nanopart. Res., 2008, 10, 1343–1348.
64. S. Pal, Y. K. Tak and J. M. Song, Appl. Environ. Microbiol., 2007, 73, 1712–1720.
65. P. Wick, P. Manser, L. K. Limbach, U. Dettlaff-Weglikowska, F. Krumeich, S. Roth, W. J. Stark and A. Bruinink, Toxicol. Lett., 2007, 168, 121–131.
66. H. Yang, C. Liu, D. F. Yang, H. S. Zhang and Z. G. Xi, J. Appl. Toxicol., 2009, 29, 69–78.
67. J. R. Rupelia, A. K. Chatterjee, S. P. Duttagupta and S. Mukherji, Acta Biomater., 2008, 4, 707–716.
68. L. Kvitek, A. Panacek, J. Soukupova, M. Kolar, R. Vecerova, R. Prucek, M. Holecova and R. Zboril, J. Phys. Chem. C, 2008, 112, 5825–5834.
69. L. Bo, W. Yang, M. Chen, J. Gao and Q. Xue, Chem. Biodiversity, 2009, 6, 111–116.
70. T. Echigo, D. M. Aruguete, M. Murayama and M. F. Hochella, Jr., Geochim. Cosmochim. Acta, 2012, 90, 149–162.
71. A. S. Madden and M. F. Hochella, Jr., Geochim. Cosmochim. Acta, 2005, 69, 389–398.
72. G. Applerot, A. Lipovsky, R. Dror, N. Perkas, Y. Nitzan, R. Lubart and A. Gedanken, Adv. Funct. Mater., 2009, 19, 842–852.
73. D. A. Pelletier, A. K. Suresh, G. A. Holton, C. K. McKeown, W. Wang, B. Gu, N. P. Mortensen, D. P. Allison, D. C. Joy, M. R. Allison, S. D. Brown, T. J. Phelps and M. J. Doktycz, Appl. Environ. Microbiol., 2010, 76, 7981–7989.
74. M. Heinlaan, A. Ivask, I. Blinova, H. Dubourguier and A. Kahru, Chemosphere, 2008, 71, 1308–1316.
75. G. Rosenblum, C. L. Chen, J. Kaur, X. N. Cui, Y. E. Goldman and B. S. Cooperman, Nucleic Acids Res., 2012, 40, e88,  DOI:10.1093/nar/gks232.
76. P. Y. Silvert, R. Herrera-Urbina, N. Duvauchelle, V. Vijayakrishnan and K. T. Elhsissen, J. Mater. Chem., 1996, 6, 573–577.

---