Microbial Cytotoxicity of Carbon-Based Nanomaterials: Implications for

Feb 25, 2009 - Microbial Cytotoxicity of Carbon-Based Nanomaterials: Implications for River Water and Wastewater Effluent. Seoktae Kang, Meagan S. Mau...
0 downloads 0 Views 3MB Size
Download PDF
Environ. Sci. Technol. 2009, 43, 2648–2653

Microbial Cytotoxicity of Carbon-Based Nanomaterials: Implications for River Water and Wastewater Effluent SEOKTAE KANG, MEAGAN S. MAUTER, AND MENACHEM ELIMELECH* Department of Chemical Engineering, Environmental Engineering Program, Yale University, New Haven, Connecticut 06520-8286

Received November 7, 2008. Revised manuscript received February 2, 2009. Accepted February 3, 2009.

This study evaluates the cytotoxicity of four carbon-based nanomaterials (CBNs)ssingle-walled carbon nanotubes (SWNTs), multiwalled carbon nanotubes (MWNTs), aqueous phase C60 nanoparticles (aq-nC60), and colloidal graphitesin gram negative and gram positive bacteria. The potential impacts of CBNs on microorganisms in natural and engineered aquatic systems are also evaluated. SWNTs inactivate the highest percentage of cells in monocultures of Escherichia coli, Pseudomonas aeruginosa, Bacillus subtilis, and Staphylococcus epidermis, as well as in the diverse microbial communities of river water and wastewater effluent. Bacterial cytotoxicity displays time dependence, with longer exposure times accentuating toxicity in monocultures with initial tolerance for SWNTs. In Bacillus subtilis, an additional 3.5 h of incubation produced a five fold increase in toxicity. Elevated concentration of NOM reduces the attachment of bacteria on SWNT aggregates by 50%, but does not mitigate toxicity toward attached cells. CBN toxicity in bacterial monocultures was a poor predictor of microbial inactivation in chemically and biologically complex environmental samples.

Introduction Microbial communities serve diverse functions in natural and engineered aquatic systemssas primary producers, facilitators of nutrient recycling, and agents of pollutant degradation. Some microbial ecosystems depend upon keystone species for ecological function, while in others, productivity is correlated to biocomplexity in the microbial community (1). Environmental impact assessment of novel nanomaterials will include screening for potential alterations to keystone species distribution and community richness in microbial systems. Carbon-based nanomaterials (CBNs), including singlewalled carbon nanotubes (SWNTs), multiwalled carbon nanotubes (MWNTs), aqueous phase C60 nanoparticles (aqnC60), and colloidal graphite, will interact with microbial communities in natural and engineered aquatic systems. Probable exposure routes include disposal of consumer products containing CBNs (2), direct application of CBNs in water treatment and environmental remediation applications (3, 4), and accidental release of nanomaterials into the * Corresponding author e-mail: [email protected]; phone: (203) 432-2789. 2648

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 7, 2009

environment. Existing studies compare the relative toxicity of CBNs in eukaryotic cells (5-7), but the implications and mechanisms of cytotoxicity are significantly different in bacterial systems. Assessing the cytotoxic response of bacteria to CBNs with a broad set of physicochemical properties may inform a deeper understanding of toxicity in microbial systems. CBN cytotoxicity in aquatic environments is a complex function of nanomaterial physicochemical properties, transport behavior, solution chemistry, and system ecology. Most existing literature on CBN microbial toxicity addresses these factors separately. Several publications report on the bactericidal activity of CBNs in monoculture studies on model gram positive or gram negative bacteria (8-10). Cytotoxicity is contingent upon direct contact with the cell membrane (10), though conclusive characterization of mechanistic pathways for cytotoxicity is difficult since many of these studies omit physicochemical analysis of the CBN. More recent work seeks to close this gap by systematically isolating physicochemical properties underlying MWNT bacterial cytotoxicity (11). These results link antimicrobial activity to MWNT aqueous phase properties that enhance cell-nanomaterial contact opportunities, including dispersivity in solution, functionalization, and short nanotube length. Functionalization of carbon nanotubes (CNTs) with carboxylic acid moieties, for instance, promotes aqueous phase dispersion by increasing the electrostatic repulsion of the carbon sidewalls and overcoming van der Waals attractive forces between the CNTs (11). Further effort will be necessary to decouple the physical conformation of the nanotubes (i.e., bundling) from the effects of nanomaterial surface chemistry (i.e., functionalization). Toxicity in natural and engineered aquatic systems is inseparable from environmental variables, including natural organic matter, ionic strength, and aggregation state. Dissolved organic matter (DOM) debundles MWNTs and induces steric and electrostatic stabilization of CBNs at environmentally relevant solution chemistries (12-14). High ionic strength promotes CBN aggregation by shielding electrostatic repulsive forces between similarly charged particles (15, 16). These modifications to the CBN aggregation state have secondary effects on the chemical and electronic properties of the nanomaterials. For instance, aggregation of fullerenes into crystalline nanostructures (i.e., aq-nC60) reduces ROS production by increasing the probability of self-quenching and triplet-triplet annihilation reactions (17, 18). This decrease in ROS production provides a mechanistic explanation for the inverse relationship between aq-nC60 toxicity and ionic strength (19). The environmental implications of toxicity will also depend upon the ecological composition and structure of complex microbial communities. Monoculture studies report varying degrees of toxicity across bacterial species (7, 20), but only a few experiments have probed the effects of carbonbased manufactured nanoparticles on diverse microbial communities. While low-resolution genetic analysis using denaturing gradient gel electrophoresis (DGGE) did not identify any changes to microbial community structure in soil (21) or anaerobic biosolids (22) after C60 exposure, it is impossible to extrapolate these results to other CBNs. Future analysis of environmental toxicity in microbial communities should also probe the functionality, stability, and performance of CBN-impacted microbial communities—characteristics not measured by DGGE. The present study contributes to an understanding of CBN microbial cytotoxicity in natural and engineered aquatic 10.1021/es8031506 CCC: $40.75

 2009 American Chemical Society

Published on Web 02/25/2009

FIGURE 1. SEM and TEM (insets) micrographs of carbon-based nanomaterials: (a) graphite, (b) Aq-nC60, (c) MWNTs, and (d) SWNTs. Samples were dispersed in ethanol (∼ 0.1 µg/mL) or in DI water (Aq-nC60), then dried in a vacuum chamber overnight on silicon wafers (SEM) or copper grids (TEM). environments. Well-characterized CBNs are evaluated for bacterial cytotoxicity in monocultures of gram negative and gram positive bacteria and in the diverse microbial communities of river water and wastewater effluent. SWNTs are the most toxic carbon-based manufactured nanomaterial in each investigated system, and toxicity is not significantly affected by the presence of NOM in experiments with short SWNT exposure times. The cytotoxicity of CBNs in natural and engineered aquatic environments is poorly correlated to results from monoculture studies, highlighting the shortcomings of risk assessment methods that extrapolate toxicity in heterogeneous environmental systems from toxicity in model organisms.

Materials and Methods Carbon-Based Nanomaterial Preparation. We used commercially available SWNTs, MWNTs, fullerenes (C60), and colloidal graphite. SWNTs and MWNTs were purchased from Stanford Materials (Aliso Viejo, CA) and NanoTechLabs Inc. (Yadkinville, NC), respectively, and purified with hydrochloric acid. The purification procedures and characteristic properties of SWNTs and MWNTs were explained in detail in our previous studies (3, 11). Aq-nC60 nanoparticles were prepared by stirring 1 g of commercially synthesized C60 powder (MER Corporation, Tucson, AZ) in 1 L of DI water for 5 weeks. Colloidal graphite samples originated from Electron Microscopy Sciences (Hatfield, PA). Characterization of Carbonaceous Nanomaterials. CBN characterization was performed as described in our previous work (3, 10, 11, 23). We verified manufacturer reports on the purity and physicochemical properties of each sample via microscopic, thermal, and spectroscopic techniques. SEM imaging on an XL30 microscope (FEI, USA) provided data on

sample dimensions and bulk aggregation state. The colloidal graphite, MWNT, and SWNT samples were dispersed in 0.1 µg/mL ethanol sonicated for 10 min in a bath sonicator, and deposited onto a silicon wafer for imaging. The aq-nC60 sample was diluted with DI water and deposited onto a silicon wafer to preserve the original aggregate structure. The distribution of sample diameters was analyzed using TEM imaging on a Tecnai F20 (Philips Electron Optics, Eindhoven, Netherlands). To perform TEM imaging, we dispersed the samples in ethanol and dried them on 200mesh copper grids coated with carbon-Formvar (Electron Microscopy Sciences). Images were captured using a 200 kV accelerating voltage. Thermo-gravimetric analysis (TGA) (SETSYS 16/18) measured sample purities. A minimum of 20 mg of each CBN was heated from 200 to 1000 °C at a heating rate of 10 °C/ min. Elemental compositions of CBNs were further analyzed via X-ray photoelectron spectroscopy (XPS; Kratos Analytical, Axis Ultra, NY). Raman spectra were collected at an excitation wavelength of 532 nm (Jobin Yvon, Japan), equipped with a confocal microscope (Olympus Corporation, Japan). In carbon-based nanostructures, the G-band (∼1580 cm-1) to D-band (∼1350 cm-1) ratio provides a qualitative indicator for comparing the bonding structure and amorphous carbon content of the samples. A low G/D ratio is characteristic of samples with extensive structural defects and a high amorphous carbon content (24, 25). A Raman emission peak around 1480 cm-1 is indicative of pristine C60 structures (26). Preparation of Bacterial Monocultures. Escherichia coli, Pseudomonas aeruginosa, Bacillus subtilis, and Staphylococcus epidermidis served as model organisms for assessing bacterial toxicity. Bacteria were grown in LB medium at their VOL. 43, NO. 7, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2649

optimum growth temperatures, either 37 °C (E. coli, P. aeruginosa, S. epidermidis) or 30 °C (B. subtilis) (27), and harvested during midexponential growth phase. Cells were washed twice and resuspended in saline solution (0.9% or 0.154 M NaCl) before exposure to carbonaceous nanomaterials. River Water and Wastewater Samples. River water (Mill River, New Haven, CT) and secondary wastewater effluent (rotating biological contactor treatment system; Wallingford, CT) were collected using nontransparent bottles and immediately placed on ice for transport back to the laboratory. The samples were allowed to settle for 30 min at room temperature prior to exposure to the CBNs in the toxicity assay. Important characteristics of the river water and wastewater effluent are summarized in Table S1. Bacterial Toxicity Assay. Past studies link CBN toxicity to direct contact with bacterial cells (10, 23). The previously described CBN-coated filter method facilitates large contact areas for direct quantification of bacterial cytotoxicity (10, 11, 23). CBN-coated filter preparation included (i) dispersing 4 mg of nanomaterial sample in 10 mL of dimethyl sulfoxide (DMSO) and sonicating for 15 min in a bath sonicator, (ii) filtering the suspension through a 5 µm PVDF membrane (Millipore), (iii) washing the filter with 100 mL of ethanol to remove residual DMSO, and (iv) rinsing with 200 mL of deionized water to remove any remaining ethanol. To perform the cytotoxicity assay, 2 × 106 bacterial cells were deposited onto a CBN-coated filter, and the filter was incubated in an isotonic solution of 0.9% (0.154 M) NaCl for 1 h in the dark at 37 or 30 °C (B. subtilis). This procedure was repeated for each model microorganism or environmental sample and each CBN-coated filter. A bare 0.45 µm PVDF membrane filter served as the negative control. After incubation, bacterial membrane integrity was quantified using a fluorescence-based, nucleic acid assay applicable for both gram negative and gram positive bacteria. Cells were stained with propidium iodide (PI) for 15 min, and counter-stained with 4′-6-diamidino-2-phenylindole (DAPI) or SYTO-9 for 5 min in the dark. The cells on the filter surface were then imaged under an epifluorescence microscope (Olympus Corporation, Japan). Ten representative images from different locations on each filter were captured for subsequent data analysis using direct counting methods. The percentage of inactivated cells was determined by the ratio of cells stained with PI to those stained with DAPI (or SYTO-9) plus PI. All experiments were conducted in triplicate. Bacterial Toxicity in the Presence of Suwannee RiverNatural Organic Matter. We assessed bacterial toxicity in the presence of Suwannee River-natural organic matter (SRNOM) using the suspended aggregate method described in earlier publications (10, 23). Previous studies confirmed equivalent levels of toxicity using the suspended aggregate and coated filter assessment methods (10, 23). The suspended system consisted of 5 mg/L SWNTs in 0.9% NaCl solution with and without 20 mg/L SR-NOM. All experiments were performed in triplicate.

Results and Discussion Physicochemical Properties of CBNs. Past work correlates physicochemical properties of MWNTs to varying rates of E. coli cytotoxicity and emphasizes the need for careful documentation of sample characteristics when reporting nanomaterial toxicity (11). Relevant properties include nanomaterial dimensions, purity, structural conformation, and elemental composition. In this work, we extend standard characterization methods for CNTs to colloidal graphite and aq-nC60. A summary of the physicochemical properties of the CBN samples is provided in Table 1. SEM and TEM micrographs (Figure 1) of the CBN samples confirmed the manufacturer’s dimension specifications. The diameters of our synthesized aq-nC60 aggregates were widely 2650

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 7, 2009

TABLE 1. Physicochemical Properties of Carbon-Based Nanomaterials physical properties

chemical properties elemental compositiond

diametera (nm)

lengtha (µm)

graphite 1.17 ( 0.34 Aq-nC60 0.36 ( 0.21 MWNT 17.4 ( 6.1 77 ( 31 SWNT 1.2 ( 0.3 17.8 ( 5.6

RMb (%)

G/Dc

0.1 3.57 0.1 1.13 2.1 2.47 6.0 31.03

carbon oxygen (%) (%) 96.1 99.7 99.1 99.3

3.9 0.3 0.9 0.7

a Average diameters and lengths of CBNs were determined from the Raman spectra breathing mode (SWNTs), SEM, and/ or TEM images. b Residual mass (%) after thermo-gravimetric analysis (TGA). c The ratio of Raman G-band (∼1580 cm-1) to D-band (∼1350 cm-1) peak heights at λlaser ) 532 nm (Figure S1). d From X-ray photoelectron spectroscopy (XPS) analysis.

distributed from tens of nanometers to several micrometers, which is typical for aq-nC60 aggregates prepared by stirring in DI water (8, 28). Thermo-gravimetric analysis (TGA) (Figure S1) provides metrics for sample purity, including the percentage of amorphous carbon and residual metal catalyst in the samples. The stability of the aq-nC60, MWNTs, and SWNTs at 300 °C indicates very low amorphous carbon content in these samples. Colloidal graphite had a much higher mass loss around 300 °C, suggesting that this sample was a mixture of ordered graphite nanoparticles and amorphous carbon, as well as modified with oxygen-containing functional groups (Table 1). This is also confirmed by the lower G/D ratio measured by Raman spectroscopy (Figure S2) compared to highly pristine graphite (29). The low residual mass after fully combusting the carbon nanomaterials at 1000 °C confirms the removal of catalytic metals from the CNT samples. Catalytic metal residues from nanotube synthesis were initially suggested as an explanation for observed toxicity in human keratinocyte cells (6), but subsequent work (10, 11, 23, 30) suggests that low concentrations of residual (iron) catalyst in the CNT samples is uncorrelated to bacterial toxicity. Past research suggests that functional groups on the surface modify CBN cytotoxicity by enhancing dispersion in polar solvents (11). Low oxygen content in the XPS analysis of elemental composition (Table 1) confirms that the samples are unfunctionalized. CBN Cytotoxicity in Gram Negative and Gram Positive Bacteria. Fluorescence-based cytotoxicity assays confirm that CBNs are not inert; each CBN induced bacterial cytotoxicity when compared to the control system (Figure 2). SWNTs display the highest cytotoxicity toward all types of bacteria tested in the study (representative images in Figure S3). Exposure to aq-nC60 also induced pronounced cytotoxicity in E. coli and P. aeruginosa. The antibacterial activity of MWNTs and colloidal graphite was moderate, though we observed a wide distribution in the degree of MWNT toxicity between bacterial species. Mechanistic explanations for this variation in the tolerance of specific microbial populations to CBN exposure is a topic of continuing research. The considerable variation in the cytotoxicity of a single manufactured nanomaterial toward different bacterial species is presented in Figure 2. Compared to the other monocultures tested in the present study, B. subtilis experienced little cytotoxicity after one hour of incubation with any of the four CBNs. This finding is consistent with prior observations that B. subtilis is less susceptible to C60 and aq-nC60 than other bacterial species (19, 31). A number of conjectures and hypotheses have been developed to explain the elevated tolerance of B. subtilis. Fang et al. (32) observed a significant

FIGURE 2. Summary of fluorescence-based toxicity assays following bacterial contact with bare (control), graphite-, aq-nC60-, MWNT-, or SWNT-coated filters. Cell suspensions were passed though the filters and incubated in 0.9% NaCl at 37 and 30 °C (B. subtilis) for 1 h. Error bars represent standard error (n ) 10). increase in the levels of iso- and anteiso-branched fatty acids and monounsaturated fatty acids in the cell membrane after aq-nC60 exposure, a physiological change that may enhance membrane fluidity. Mohanty et al. (9) conjecture that teichoic acid in gram positive cell walls improves tolerance toward CBNs. The robust peptidoglycan layer in gram positive B. subtilis (33) enhances the structural rigidity and resistance of the cell wall and may contribute to increased resistance. Many of the proposed explanations for the enhanced resistance of B. subtilis are consistent with our hypothesis that direct cell-CBN contact is required for membrane perturbation and bacterial cytotoxicity (10, 23). Additional work to investigate the role of cell wall structure, the protective effect of outer membrane surface features, or the unique defense and repair mechanisms of specific bacterial strains would aid in understanding CBN-cell envelope interactions. Longer Contact Time Increases Cytotoxicity. To more thoroughly characterize the resistance of B. subtilis toward CBNs, we investigated the effect of incubation time on toxicity rates. Longer incubation time is known to increase the cytotoxicity of CBNs in E. coli (10). We confirmed similar time-dependent bactericidal behavior in gram positive B. subtilis exposed to SWNTs (Figure 3). After one hour of incubation, cell inactivation on the filter surface was not statistically different from that of the control sample with no CBNs present. After four hours of incubation, however, cell inactivation increased to 55%. The diminishing resistance to CBNs after long exposure times implies that even robust bacteria in the natural environment may experience cytotoxic effects from manufactured nanomaterials. Endospore formation in B. subtilis is a common physiological response to environmental stressors, and bacterial endospores are expected to display high resistance to hypothesized mechanisms of antimicrobial activity, namely cell membrane perturbation, direct oxidation, or secondary oxidation of the lipid bilayer by reactive oxygen species (19, 23, 34). Although the time dependence of SWNT antimicrobial activity suggests that endospore formation is not responsible for the increased resistance of B. subtilis, additional experiments were performed to confirm that endospore formation is not the primary defense mechanism of this species (Figure S4 and related information described in the SI). The susceptibility of B. subtilis to thermal shock following incubation on the SWNT filter implies that a cellular attribute other than endospore formation is responsible for the increased tolerance of B. subtilis. Future work to investigate CBN cytotoxicity toward endospores will be

FIGURE 3. Effect of incubation time for the inactivation of B. subtilis cells on SWNT-coated filter. Cells were deposited on bare PVDF membrane (control) or SWNT-coated filter, and incubated in 0.9% (0.154 M) NaCl solution at 30 °C in the dark. Error bars represent one standard deviation.

FIGURE 4. Influence of natural organic matter (SR-NOM) on the inactivation of E. coli cells and their attachment to SWNT aggregates. Relative cell number represents the ratio of the number of cells on the SWNT aggregates when NOM was present compared to no NOM (the latter represents 100%). Error bars represent one standard deviation. important for characterizing potential environmental and drinking water applications of CBNs. Natural Organic Matter Does Not Influence Bacterial Cytotoxicity of SWNTs. NOM is ubiquitous in natural and engineered aquatic environments. Adsorbed NOM affects CBN surface charge, aggregation behavior, and mobility in aquatic environments (12, 13, 15, 16, 28). Surface coatings on CBNs will also modify interactions with microorganisms and inorganic solids. Although a number of groups report dispersion and stabilization of MWNTs and aq-nC60 by NOM (12, 13, 35), we do not observe SWNT debundling in SRNOM concentrations between 10 and 200 mg/L (data not shown). For cytotoxicity experiments in the presence of NOM, we used the suspended aggregate method described in our previous work (10, 23) to mimic interactions among microorganisms, NOM, and CBNs in natural aquatic systems. Despite evidence that NOM modifies the physicochemical properties of MWNTs (12, 15), the presence of SR-NOM in the SWNT-suspended system did not have a statistically significant impact on the percentage of inactivated cells when compared to the control experiment without SR-NOM (Figure 4). Bacterial inactivation by SWNTs increased by 12% in the presence of SR-NOM, but a comparable increase of 16.5% in the control sample (SR-NOM without SWNTs) indicates that the toxicity attributed to interactions with SWNT was VOL. 43, NO. 7, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2651

FIGURE 5. Inactivation of microorganisms in river water and wastewater treatment plant (WWTP) samples. Each 100 mL sample was filtered through control or CBN-coated filters and incubated for 1 h in 0.9% (0.154 M) NaCl solution at 37 °C. The average of in vitro inactivation rates for the monoculture studies (summarized in Figure 2) is provided as a visual reference. Error bars represent one standard deviation. unaffected. Although cytotoxicity was not mitigated by NOM adsorption, we did observe a 55% decrease in the number of bacteria attached to the SWNT aggregates in the presence of SR-NOM (Figure 4 and SI Figure S5). The observed reduction in bacterial attachment may be attributed to increased steric and electrostatic repulsive forces following NOM adsorption to SWNT aggregates (15, 28). Thus, we suspect that the toxicity of SWNTs in aquatic environments will be partially mitigated by NOM coating of SWNTs. CBN Toxicity in Engineered and Natural Aquatic Environments. Carbon-based nanomaterials will interact with microbial communities when discharged to natural or engineered aquatic systems such as river water and wastewater effluent. The dispersion and ecotoxicity of CBNs will be influenced by the physicochemical properties of the manufactured nanomaterial and solution chemistry parameters, including pH, ionic strength, and dissolved organic matter concentration. Existing literature does not report on the short-term ecotoxicity of CBNs in natural or engineered aquatic systems. This study compares the relative toxicity of CBNs toward the heterogeneous microbial communities of aquatic systems (Figure 5). Averages of inactivation values from the laboratory monoculture experiments are also included in Figure 5 for reference, though it is important to acknowledge the significant differences in biological and chemical conditions between monocultures and environmental samples (Table S1). (a) Wastewater Effluent. Wastewater treatment plants support bacterial communities rich in genetic and metabolic diversity. Industrial or household release of CBNs into receiving wastewaters may perturb the community distribution, density, or function. Although previous research suggests the impact of nC60 aggregates on soil microbial communities is minimal (21), the degree of CBN cytotoxicity has not been analyzed in aquatic systems. Wastewater effluent characteristics, including cell concentration, pH, conductivity, DOC, UV absorbance at 254 nm, and major cation concentrations, are presented in Table S1. The higher conductivity and divalent cation concentrations (e.g., Mg2+ and Ca2+) in the wastewater effluent sample make it likely that the nanomaterials are more aggregated than in river water (15, 28). Elevated concentrations of suspended and dissolved organic matter in the wastewater sample could complicate toxicity analysis as we discussed 2652

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 7, 2009

earlier with NOM. Fluorescence microscope and SEM images provided in Figure S6 reveal a thick coating of colloidal organic particulates on the CBN-coated filter surface. This coating hindered direct contact between the bacteria and the CBNs and complicated interpretation of fluorescence-based toxicity assays. To ensure contact between the cells and the CBNcoated filter during toxicity experiments, we diluted the wastewater effluent sample four-fold with wastewater effluent that was filtered through a 0.22 µm membrane. Despite macromolecule and particulate matter coatings on the CBN filters, microbial inactivation paralleled cytotoxicity trends from monoculture experiments in isotonic solution. SWNTs continue to induce the highest rates of microbial inactivation, while aq-nC60 and graphite showed more moderate toxicity. MWNTs are less toxic toward wastewater samples than predicted by monoculture experiments, suggesting that these simplified experiments may not accurately predict nanomaterial cytotoxicity in heterogeneous water samples with diverse microbial communities. (b) River Water. Microbes are the foundation of ecosystem productivity in aquatic environments. Significant differences in microbial composition and solution chemistry between natural and engineered aquatic systems necessitate the independent evaluation of CBN toxicity in river water. For instance, the specific UV absorbance (i.e., UV absorbance at 254 nm normalized to the TOC concentration) of the river water sample is three times greater than that of the wastewater effluent sample, indicating a higher concentration of NOMlike organic matter in river water. As with previous samples, CBNs induce cytotoxicity in river water microbial communities a short time after introduction into the aquatic system. Cell inactivation results from the river water sample correlate poorly with the average from the monoculture toxicity experiments, except in the case of SWNTs (Figure 5). Aq-nC60 induced toxicity in monoculture experiments at a significantly higher rate than observed in environmental samples. This variation may be explained by the strong dependence of nC60 aggregation on environmental conditions such as ionic strength and dissolved organic matter (16, 28). We suggest that indicator tests for CBNs in the aquatic environment include bacteria native to impacted waters and experimental conditions that reflect the aquatic chemistry of the environment. Environmental Implications. CBNs may enter the aquatic environment through accidental release or intentional application. In river water and wastewater ecosystems, the physicochemical properties and biological activity of CBNs will be modified by adsorbed NOM, polysaccharides, or biomacromolecules. The present study suggests that NOM coatings reduce the deposition and attachment of bacteria on the surface of the nanomaterial, but do not mitigate SWNT toxicity toward attached cells. Although some bacteria appear resistant to the toxicological effects of CBNs, extended contact time reduces cell viability and membrane integrity. The strong dependence of CBN toxicity on physicochemical properties, aqueous solution chemistry, and microbial species necessitates the standardization of protocols for analyzing the potential impacts of CBNs in natural and engineered aquatic environments.

Acknowledgments We acknowledge the support of the National Science Foundation under Research Grants BES-0646247 and BES0504258. M.S.M. acknowledges generous support from the NSF Graduate Research Fellowship Program (GRFP) and the EPA Science to Achieve Results (STAR) Graduate Fellowship Program. We thank Dr. K. L. Chen at Johns Hopkins University for the preparation of aq-nC60 and TEM images, and Dr. S. Chae at Duke University for his kind measurement of XPS of CBNs.

Supporting Information Available Summary of river water and wastewater characteristics (Table S1). TGA analysis data for carbon-based nanomaterial samples (Figure S1). Raman spectra for carbon-based nanomaterial samples (Figure S2). Representative images of bacterial cells on SWNT-coated filter (Figure S3). Growth curves of B. subtilis cells and endospores (Figure S4). Representative images of E. coli cells on SWNT aggregates (Figure S5). Representative fluorescence and SEM images of microorganisms in wastewater effluent in contact with a SWNT-coated filter (Figure S6). This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Smith, V. H. Microbial diversity-productivity relationships in aquatic ecosystems. FEMS Microbiol. Ecol. 2007, 62, 181–186. (2) Mueller, N. C.; Nowack, B. Exposure modeling of engineered nanoparticles in the environment. Environ. Sci. Technol. 2008, 42, 4447–4453. (3) Brady-Estevez, A. S.; Kang, S.; Elimelech, M. A single-walledcarbon-nanotube filter for removal of viral and bacterial pathogens. Small 2008, 4, 481–484. (4) Mauter, M. S.; Elimelech, M. Environmental applications of carbon-based nanomaterials. Environ. Sci. Technol. 2008, 42, 5843–5859. (5) Tian, F. R.; Cui, D. X.; Schwarz, H.; Estrada, G. G.; Kobayashi, H. Cytotoxicity of single-wall carbon nanotubes on human fibroblasts. Toxicol. in Vitro 2006, 20, 1202–1212. (6) Shvedova, A. A.; Castranova, V.; Kisin, E. R.; Schwegler-Berry, D.; Murray, A. R.; Gandelsman, V. Z.; Maynard, A.; Baron, P. Exposure to carbon nanotube material: Assessment of nanotube cytotoxicity using human keratinocyte cells. J. Toxicol. Environ. Health A 2003, 66, 1909–1926. (7) Neal, A. L. What can be inferred from bacterium-nanoparticle interactions about the potential consequences of environmental exposure to nanoparticles? Ecotoxicology 2008, 17, 362–371. (8) Lyon, D. Y.; Adams, L. K.; Falkner, J. C.; Alvarez, P. J. J. Antibacterial activity of fullerene water suspensions: Effects of preparation method and particle size. Environ. Sci. Technol. 2006, 40, 4360–4366. (9) Mohanty, B.; Verma, A. K.; Claesson, P.; Bohidar, H. B. Physical and anti-microbial characteristics of carbon nanoparticles prepared from lamp soot. Nanotechnology 2007, 18, 1–9. (10) Kang, S.; Pinault, M.; Pfefferle, L. D.; Elimelech, M. Single-walled carbon nanotubes exhibit strong antimicrobial activity. Langmuir 2007, 23, 8670–8673. (11) Kang, S.; Mauter, M. S.; Elimelech, M. Physicochemical determinants of multiwalled carbon nanotube bacterial cytotoxicity. Environ. Sci. Technol. 2008, 42, 7528–7534. (12) Hyung, H.; Fortner, J. D.; Hughes, J. B.; Kim, J. H. Natural organic matter stabilizes carbon nanotubes in the aqueous phase. Environ. Sci. Technol. 2007, 41, 179–184. (13) Hyung, H.; Kim, J. H. Natural organic matter (NOM) adsorption to multi-walled carbon nanotubes: Effect of NOM characteristics and water quality parameters. Environ. Sci. Technol. 2008, 42, 4416–4421. (14) Wang, X. L.; Lu, J. L.; Xing, B. S. Sorption of organic contaminants by carbon nanotubes: Influence of adsorbed organic matter. Environ. Sci. Technol. 2008, 42, 3207–3212. (15) Saleh, N. B.; Pfefferle, L. D.; Elimelech, M. Aggregation kinetics of multiwalled carbon nanotubes in aquatic systems: measurements and environmental implications. Environ. Sci. Technol. 2008, 42, 7963–7969. (16) Chen, K. L.; Elimelech, M. Influence of humic acid on the aggregation kinetics of fullerene (C-60) nanoparticles in monovalent and divalent electrolyte solutions. J. Colloid Interface Sci. 2007, 309, 126–134.

(17) Lee, J.; Yamakoshi, Y.; Hughes, J. B.; Kim, J. H. Mechanism of C-60 photoreactivity in water: Fate of triplet state and radical anion and production of reactive oxygen species. Environ. Sci. Technol. 2008, 42, 3459–3464. (18) Hotze, E. M.; Labille, J.; Alvarez, P.; Wiesner, M. R. Mechanisms of photochemistry and reactive oxygen production by fullerene suspensions in water. Environ. Sci. Technol. 2008, 42, 4175– 4180. (19) Lyon, D. Y.; Fortner, J. D.; Sayes, C. M.; Colvin, V. L.; Hughes, J. B. Bacterial cell association and antimicrobial activity of a C-60 water suspension. Environ. Toxicol. Chem. 2005, 24, 2757– 2762. (20) Handy, R. D.; von der Kammer, F.; Lead, J. R.; Hassellov, M.; Owen, R.; Crane, M. The ecotoxicology and chemistry of manufactured nanoparticles. Ecotoxicology 2008, 17, 287–314. (21) Tong, Z. H.; Bischoff, M.; Nies, L.; Applegate, B.; Turco, R. F. Impact of fullerene (C-60) on a soil microbial community. Environ. Sci. Technol. 2007, 41, 2985–2991. (22) Nyberg, L.; Turco, R. F.; Nies, L. Assessing the impact of nanomaterials on anaerobic microbial communities. Environ. Sci. Technol. 2008, 42, 1938–1943. (23) Kang, S.; Herzberg, M.; Rodrigues, D. F.; Elimelech, M. Antibacterial effects of carbon nanotubes: Size does matter. Langmuir 2008, 24, 6409–6413. (24) Kim, U. J.; Furtado, C. A.; Liu, X. M.; Chen, G. G.; Eklund, P. C. Raman and IR spectroscopy of chemically processed singlewalled carbon nanotubes. J. Am. Chem. Soc. 2005, 127, 15437– 15445. (25) Pimenta, M. A.; Dresselhaus, G.; Dresselhaus, M. S.; Cancado, L. G.; Jorio, A.; Saito, R. Studying disorder in graphite-based systems by Raman spectroscopy. Phys. Chem. Chem. Phys. 2007, 9, 1276–1291. (26) Dresselhaus, M. S.; Dresselhaus, G.; Eklund, P. C. Raman scattering in fullerenes. J. Raman Spectrosc. 1996, 27, 351–371. (27) Budde, I.; Steil, L.; Scharf, C.; Volker, U.; Bremer, E. Adaptation of Bacillus subtilis to growth at low temperature: a combined transcriptomic and proteomic appraisal. Microbiol-Sgm 2006, 152, 831–853. (28) Chen, K. L.; Elimelech, M. Interaction of fullerene (C60) nanoparticles with humic acid and alginate coated silica surfaces: Measurements, mechanisms, and environmental implications. Environ. Sci. Technol. 2008, 42 (20), 7607–7614. (29) Tuinstra, F.; Koenig, J. L. Raman Spectrum of Graphite. J. Chem. Phys. 1970, 53, 1126–&. (30) Shvedova, A. A.; Kisin, E. R.; Mercer, R.; Murray, A. R.; Johnson, V. J.; Potapovich, A. I.; Tyurina, Y. Y.; Gorelik, O.; Arepalli, S.; Schwegler-Berry, D.; Hubbs, A. F.; Antonini, J.; Evans, D. E.; Ku, B. K.; Ramsey, D.; Maynard, A.; Kagan, V. E.; Castranova, V.; Baron, P. Unusual inflammatory and fibrogenic pulmonary responses to single-walled carbon nanotubes in mice. Am. J. Physiol.-Lung C 2005, 289, L698–L708. (31) Lyon, D. Y.; Brunet, L.; Hinkal, G. W.; Wiesner, M. R.; Alvarez, P. J. J. Antibacterial activity of fullerene water suspensions (nC(60)) is not due to ROS-mediated damage. Nano Lett. 2008, 8, 1539–1543. (32) Fang, J. S.; Lyon, D. Y.; Wiesner, M. R.; Dong, J. P.; Alvarez, P. J. J. Effect of a fullerene water suspension on bacterial phospholipids and membrane phase behavior. Environ. Sci. Technol. 2007, 41, 2636–2642. (33) Vollmer, W.; Blanot, D.; de Pedro, M. A. Peptidoglycan structure and architecture. FEMS Microbiol. Rev. 2008, 32, 149–167. (34) Fortner, J. D.; Lyon, D. Y.; Sayes, C. M.; Boyd, A. M.; Falkner, J. C.; Hotze, E. M.; Alemany, L. B.; Tao, Y. J.; Guo, W.; Ausman, K. D.; Colvin, V. L.; Hughes, J. B. C-60 in water: Nanocrystal formation and microbial response. Environ. Sci. Technol. 2005, 39, 4307–4316. (35) Xie, B.; Xu, Z. H.; Guo, W. H.; Li, Q. L. Impact of natural organic matter on the physicochemical properties of aqueous C-60 nanoparticles. Environ. Sci. Technol. 2008, 42, 2853–2859.

ES8031506

VOL. 43, NO. 7, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2653