Stability, Bioavailability, and Bacterial Toxicity of ZnO and Iron-Doped

Dec 6, 2010 - Department of Civil & Environmental Engineering, California NanoSystems Institute, and Molecular Screening Shared Resource, University o...
8 downloads 15 Views 1MB Size
Download PDF
Environ. Sci. Technol. 2011, 45, 755–761

Stability, Bioavailability, and Bacterial Toxicity of ZnO and Iron-Doped ZnO Nanoparticles in Aquatic Media M I N G H U A L I , †,‡ S U M A N P O K H R E L , § ¨ D L E R , ‡,§ X U E J I N , †,‡ L U T Z M A R O B E R T D A M O I S E A U X , ‡,| A N D E R I C M . V . H O E K * ,†,‡ Department of Civil & Environmental Engineering, California NanoSystems Institute, and Molecular Screening Shared Resource, University of CaliforniasLos Angeles (UCLA), Los Angeles, California, 90095, United States, and Foundation Institute of Materials Science (IWT), Department of Production Engineering, University of Bremen, Germany

Received July 5, 2010. Revised manuscript received November 4, 2010. Accepted November 15, 2010.

The stability and bioavailability of nanoparticles is governed by the interfacial properties that nanoparticles acquire when immersed in a particular aquatic media as well as the type of organism or cell under consideration. Herein, high-throughput screening (HTS) was used to elucidate ZnO nanoparticle stability, bioavailability, and antibacterial mechanisms as a function of iron doping level (in the ZnO nanoparticles), aquatic chemistry, and bacterial cell type. ζ-Potential and aggregation state of dispersed ZnO nanoparticles was strongly influenced by iron doping in addition to electrolyte composition and dissolved organic matter; however, bacterial inactivation by ZnO nanoparticles was most significantly influenced by Zn2+ ions dissolution, cell type, and organic matter. Nanoparticle IC50 values determined for Bacillus subtilis and Escherichia coli were on the order of 0.3-0.5 and 15-43 mg/L (as Zn2+), while the IC50 for Zn2+ tolerant Pseudomonas putida was always >500 mg/ L. Tannic acid decreased toxicity of ZnO nanoparticles more than humic, fulvic, and alginic acid, because it complexed the most free Zn2+ ions, thereby reducing their bioavailability. These results underscore the complexities and challenges regulators face in assessing potential environmental impacts of nanotechnology; however, the high-throughput and combinatorial methods employed promise to rapidly expand the knowledge base needed to develop an appropriate risk assessment framework.

1. Introduction With the rapidly increasing production of nanomaterials worldwide, the potential for their release into the environment and subsequent impacts to ecological and human health have become great concerns. Timely evaluation of * Corresponding author address: University of CaliforniasLos Angeles, Department of Civil and Environmental Engineering, 5732 Boelter Hall, P.O. Box 951593, Los Angeles, CA, 90095-1593; Tel: (310) 206-3735; Fax: (310) 206-2222; E-mail: [email protected]. † Department of Civil & Environmental Engineering, UCLA. ‡ California NanoSystems Institute, UCLA. § University of Bremen. | Molecular Screening Shared Resource, UCLA. 10.1021/es102266g

 2011 American Chemical Society

Published on Web 12/06/2010

the potential environmental impacts of nanotechnology will enable regulators to assess potential risks, provide industry with information needed to develop safer nanomaterials, and improve public trust of nanotechnology (1). Current knowledge about environmental impacts of nanotechnology is limited, particularly for aquatic microorganisms (2). Biotopes of aquatic microorganisms may be particularly vulnerable to direct contact with nanoparticles if they enter the ecosystem through industrial effluent, domestic wastewater, accidental spills, or direct runoff from urban areas (3). Moreover, many nanomaterials are intentionally designed to inactivate the bacteriasespecially metal and metal oxide nanoparticles (4). It is possible that the high specific surface area, more reactive surfaces, and the novel properties associated with these nanomaterials could enhance the transport, persistence, bioavailability, and toxicity of nanoparticles. The “nano effect” refers to any unique behavior and resulting hazardous properties due to the nanosize of a nanomaterial (5). Therefore, there is an urgent need for the information on the potential impacts on aquatic microorganisms before large amounts of these nanomaterials enter the environment. Zinc oxide (ZnO) nanoparticles are used in a wide range of applications such as cosmetics, paints, plastics additives, ceramics, and semiconductors (6). Although ZnO is generally recognized as a safe material, recent published studies consistently demonstrate that ZnO nanoparticles produce toxic responses in mammalian cells, bacteria, copepods, crustaceans, and fish (7-12). However, one recent study suggests the toxicity of ZnO nanoparticles to mammalian cells is related to dissolution of Zn2+ ions, and both stability and toxicity in mammalian cell culture media can be tuned by iron doping (8). The mammalian cell types selected and the specificity of interfacial properties acquired by ZnO nanoparticles suspended in biological media, which dictate nanoparticle stability (aggregation, dissolution, sedimentation) and bioavailability, prevent confident extrapolation of these results to environmental systems (13, 14). Considering the complexity and variability of natural water chemistry in both location and time, the implications of laboratory data using deionized water (DI) or nutrient-rich media for real environmental systems is uncertain. The transport, fate, and bioavailability of nanoparticles are highly dependent on the media that they are exposed to (15). Relatively subtle changes in water chemistry (e.g., pH, ionic strength, types of ions and organic matter) can significantly alter interfacial morphology, hydrophobicity, and charge, which dictate the stability (aggregation, sedimentation, deposition), reactivity (dissolution and reprecipitation), and, consequently, the bioavailability and potential toxicity of nanoparticles (13, 14). Therefore, it is absolutely necessary to assess the impacts of water chemistry on the behavior of nanoparticles in a combinatorial manner. Indeed, differential results observed in many published nanotoxicology studies may be largely due to the fundamental dissimilarity of testing media and cell types employed. Intuitively, the toxic responses caused by the same nanoparticles can vary dramatically among different species and across trophic levels. Nevertheless, very few systematic studies are available in the literature where the nanoparticle properties, aquatic chemistry, and cell types are systematically varied. This is partially due to the amount of time and labor required to conduct such studies using conventional toxicological approaches. However, high-throughput screening (HTS) and high-content screening (HCS) tools enable VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

755

thousands of samples to be screened using several different physical, chemical, and biological assays in a single day (16). The research presented herein employed a HTS bacterial viability assay to simultaneously evaluate the influences of ZnO nanoparticle iron doping and aquatic chemistry on nanoparticle stability, bioavailability, and toxicity for three different bacterial strains. The aims of the study were (1) to determine if there is a nano effect associated with ZnO nanoparticle toxicity or if it is solely due to released Zn2+, (2) to determine whether the iron doping levels influence Zn2+ dissolution and bacterial toxicity in environmentally relevant aquatic media, (3) to evaluate impacts of aquatic inorganic and organic chemistry on nanoparticle stability and bioavailability, and (4) to compare and contrast the antibacterial activity of ZnO nanoparticles for different bacteria cells.

2. Experimental Section 2.1. Bacteria Cells. One Gram-positive bacteria, Bacillus subtilis, and two Gram-negative bacteria, Pseudomonas putida and Escherichia coli, were selected to contrast the toxic responses among different bacterial cell types. P. putida is known to have high tolerance to ionic Zn2+ (17, 18). The use of P. putida can also help reveal ZnO toxicity due to Zn2+ ion release as reported recently (7, 19). Two synthetic fresh water matrices and five different organic macromolecules were evaluated to illustrate the water chemistry influence. Six different ZnO nanoparticles with various iron doping levels were tested for their stability and antimicrobial activities to elucidate the potential influence due to structure modification of the nanoparticles. 2.2. Nanoparticles. Commercially produced ZnO nanoparticles (ZnO-Com) were purchased from Meliorum (Rochester, NY). Pure and iron-doped ZnO derivatives were synthesized in-house (University of Bremen, Germany) by flame spray porolysis (FSP) (8). The iron content in nanoparticles varied from 0 to 10% (mol/mol). Nanoparticle stock suspensions (1000 mg/L) were prepared by stirring ZnO nanoparticles vigorously in ultrapure deionized water (DI) followed by sonication. To be more specific, a 250-mL glass conical flask containing 50 mL of suspension was placed in an ultrasonic bath and treated for 30 min at maximum power (FS30H, Fisher Scientific, 100 W, 42 kHz). The actual mixing power delivered to the suspension was not estimated in the sonication bath. The suspensions were sonicated again for at least 1 min before use. All the nanoparticles suspensions were kept in 4 °C and in dark after use. New suspension was prepared every 72 h. To assess the role of soluble ionic Zn2+, the toxicity of ZnCl2 (Fisher Scientific, ACS certified) was examined. All salts, acid, base, and other chemicals were ACS reagent grade (Fisher Scientific, Pittsburgh, PA). 2.3. Water Chemistry. A model fresh water electrolyte was prepared using water-quality parameters previously reported for a fresh river water in Wheeling, WV (20). This water matrix has total ionic strength of 5.6 mM, pH 8.5, total dissolved solid of 219 mg/L and conductivity of 318 µs/cm (hereafter referred as “freshwater” or “FW”). The solution was maintained at a constant ionic strength, (5.6 mM) by replacing specific cation or anion with the ionic strength equivalent of Na+ or Cl- (hereafter referred as “NaCl only”), respectively, in order to assess the influence of inorganic ion composition. See the Supporting Information (Table S1) for detailed ionic composition and properties of the two water matrices. Five different model natural organic macromolecules, namely, humic acid (HA), fulvic acid (FA), alginic acid (AA), tannic acid (TA), and bovine serum albumin (BSA), were used to simulate fresh water natural organic matter (NOM). Stock solutions of model NOM were prepared by dissolving the macromolecules in deionized water followed by filtration through 0.2 µm filter. 756

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 45, NO. 2, 2011

For all the tests in freshwater with organic matter, the pH was adjusted to 8.5. For all the tests in NaCl only solution, the pH was adjusted to 6.4. The pH of the solution was adjusted by HCl or NaOH to desired value. The theoretical maximum free Zn2+ concentration in each water matrix was calculated using USGS software PHREEQC (21). The solubility limit of Zn2+ according to PHREEQC simulation in the freshwater was 0.5 and 5880 mg/L in NaCl solutions. The soluble Zn2+ was governed by pH instead of anions presented. The least soluble species was ZnO; hence, reprecipitation of dissolved species was not expected to occur. 2.4. Nanoparticle Physical-Chemical Characterization. For X-ray diffraction measurements, ZnO nanoparticles were examined in a PANalytical X’Pert MPD PRO diffracting system, equipped with Ni-filtered Cu KR (λ ) 0.154 nm) radiation. The Brunauer-Emmett-Teller (BET) measurements were carried out at 77 K using a Quantachrome NOVA 4000e Autosorb to determine the specific surface areas of the samples. The low-resolution TEM of the sample and the corresponding selected area electron diffractions (SAED) were examined by transmission electron microscopy (TEM) on a Phillips TEM CM 12 instrument with an accelerating voltage of 120 kV. High-resolution microscopic imaging of the specimens were investigated with a FEI Titan 80/300 microscope (8). The average hydrodynamic radii and ζ-potential of nanoparticles were determined by DLS and electrophoresis measurements (BI-90 ZetaPALS, Brookhaven Instruments, NY) in deionized water and freshwater. All the size and ζ-potential experiments were conducted with a concentration of 25 mg/L ZnO nanoparticles suspension. To determine Zn2+ concentration released from ZnO nanoparticles, the dissolved Zn2+ concentrations in various conditions were measured. In brief, known concentrations (0.1 and 1 mg/L) of six types of ZnO nanoparticles were mixed in testing media for 20 h. Then, the suspension was filtered through a 0.1 µm filter, which produced the same separation of dissolved and nanoparticulate Zn as a 3 kDa membrane, centrifugation at 15 000g (Supporting Information, Figure S1), and electrolysis (19). The filtrates were collected and digested in HNO3 according to US-EPA method 3005. The Zn2+ concentration referred to hereafter represents the free Zn2+ in the filtrate determined by ICP-OES (TJA RADIAL IRIS 1000). 2.5. High-Throughput Bacterial Viability Assay. Grampositive bacteria, B. subtilis, and Gram-negative bacteria, P. putida and E. coli, were used as model bacteria. P. putida was cultured in trypticase soy broth (TSB). B. subtilis and E. coli were cultured in Luria-Bertani broth (LB). Bacterial cell culture was grown at 25 °C in an incubator with shaking at 150 rpm and harvested at midexponential growth phase. Cells were washed twice with phosphate-buffered saline (PBS) and resuspended in specific solution to achieve an initial concentration of 108 cells/mL before exposure to ZnO nanoparticles. To rapidly assess the influences of bacteria cell type, aquatic chemistry and iron doping on ZnO nanoparticle toxicity, a high-throughput bactericidal assay was developed. In brief, the viable cell percentage was determined by a Live/ Dead Baclight bacterial viability kit (Molecular Probes, Eugene, OR). Exactly 25 µL of ZnO nanoparticles suspension at 19 different concentrations (from 2 µg/L to 500 mg/L) and 25 µL of bacterial suspensions were automatically dispensed into specific wells of a 384-well clear-bottom polysterene microplate (Greiner Bio-One, Monroe, NC). After 24-h incubation at 35 °C in the dark, SYTO 9 and propidium iodide (Invitrogen Baclight Live/Dead Kit) were added to differentiate live and dead cells accordingly. Simultaneous application of both dyes results in green fluorescence of viable cells with an intact membrane and red fluorescence of dead cells with compromised membranes.

TABLE 1. Properties of ZnO and Iron Doped ZnO Nanoparticles

primary particle Size (XRD), nm primary particle size (BET), nm aggregate sizea (DI), nm aggregate sizea (FW), nm mobility (DI), 10-8 m2/V · s mobility (FW), 10-8 m2/V · s a

ZnO-Com

ZnO-0% Fe

ZnO-2% Fe

ZnO-6% Fe

ZnO-8% Fe

ZnO-10% Fe

30.7 ( 1 18.3 ( 0.2 305 ( 19 955 ( 35 2.41 ( 0.36 0.47 ( 0.11

20.2 ( 1.1 15.8 ( 0.3 160 ( 13 1563 ( 82 1.45 ( 0.29 0.75 ( 0.10

13.9 ( 1.4 10.8 ( 0. 212 ( 28 1674 ( 141 2.10 ( 0.29 0.54 ( 0.09

8.1 ( 0.3 7.0 ( 0.3 254 ( 12 1510 ( 149 1.60 ( 0.13 0.11 ( 0.19

8.2 ( 0.3 6.2 ( 0.4 195 ( 14 1560 ( 108 1.83 ( 0.14 0.15 ( 0.16

8.3 ( 0.4 5.5 ( 0.3 1119 ( 10 1689 ( 127 0.97 ( 0.11 -0.14 ( 0.2

30 min after sonication.

The green-to-red fluorescence ratio detected by fluorescence microreader (Flexstation, Molecular Devices, Sunnyvale, CA) provided the live bacterial percentage in each well (excitation at 485 nm and emission at 630 and 530 nm) according to a calibration curve obtained by blending different ratios of live bacteria and dead bacteria. Bacteria were inactivated by immersion in 70/30% 2-propanol/water mixture. Control replicates (media and bacteria, no nanoparticles) and blank replicates (media and nanoparticles, no bacteria) were tested in the same plate as well. The percentage of live bacteria cells was calculated from the live bacterial percentage for each nanoparticle concentration normalized to the live bacterial percentage for nanoparticles-free controls. Data were fitted using a nonlinear regression by sigmoidal dose response (Prism 4, GraphPad Software), which was used to calculate the half maximal inhibitory concentration (IC50) and 95% confidence intervals. Each combination of nanoparticle, concentration, water chemistry, bacteria was repeated at least twice using independent bacteria culture, and nanoparticle suspensions in separated plates, containing a minimum of four replicate wells in each plate each.

3. Results 3.1. Nanoparticle Physical-Chemical Properties. Characterization of six types of ZnO nanoparticles as produced and in different aquatic media is summarized in Table 1. The BET equivalent primary particle size and the primary particle size from XRD measurements agreed reasonably well, showing a decreasing particle size with increasing iron loading. Further structural evidence was illustrated by EDX analysis, showing a uniform Fe concentration in the homogeneous noncentrosymmetric ZnO matrix with alternating layers of tetrahedrally coordinated O2- and Zn2+ stacked along the c-axis (8). Pure ZnO nanoparticles prepared by flame spray pyrolysis exhibited a primary particle size of 15-20 nm, as revealed by TEM. The presence of natural organic matter is known to influence aggregation of colloids and nanoparticles in aqueous media (22). Pure ZnO (ZnO-0% Fe) mixed with 10 mg/L humic acid (Figure 1b) were coated with a gel-like matrix. The refined SAED ring structure (Figure 1c,d) indicated that the in-house synthesized ZnO are very crystalline (23). The SAED combined with HR-TEM images (Figure 1e,f) confirmed that iron loading up to 10.2% did not affect the lattice spacing, and more importantly, no phase segregation of Fe from Fe-doped ZnO nanoparticles was observed. The lattice distances of pure ZnO were almost identical to 10.2% Fe-doped ZnO, corroborating a negligible effect on the lattice spacing, even at high atomic loadings of Fe in ZnO. With prolonged sonication, ZnO and Fe-doped ZnO-based nanoparticles formed stable aggregates of ∼200-300 nm in DI, except ZnO-10% Fe, which aggregated continuously over the 30-min experiment from ∼300 to ∼1100 nm (Table 1). When the same nanoparticles were exposed to synthetic fresh water, the particle suspension lost its stability immediately and continued to form aggregates due to the presence of

divalent ions, i.e., Ca2+ and Mg2+. All in-house synthesized ZnO nanoparticles continued to aggregate with size increase from around 200 nm initially to about 1500 nm within 30 min. Commercial ZnO nanoparticles aggregated slowly with final aggregate size around 1000 nm after 30 min. All ZnO and iron-doped ZnO derivatives exhibited positive electrophoretic mobilities in the range of 0.97-2.41 (10-8 m/V · s) when dispersed in DI. When nanoparticles were exposed to fresh water, the mobility was reduced in all cases, which partially explained the reduced stability toward aggregation in the more complex electrolyte. 3.2. High-Throughput Bacterial Viability Assay. Although all viable cell numbers decrease with increasing ZnO nanoparticles dose, three bacteria species behaved differently upon exposure to the same levels of nanoparticles suspension (Supporting Information, Figure S2). In general, the two Gram-negative bacteria (P. putida and E. coli) were more resistant to the antibacterial activity of ZnO nanoparticles than the Gram-positive bacteria (B. subtilis). Increasing ZnO-0% Fe concentrations gradually reduced the viable P. putida cell percentage from 85% to 58% when exposed to 500 mg/L nanoparticles. In contrast, exposure to 1 mg/L of ZnO-0% Fe nanoparticles resulted in complete inactivation of B. subtilis. The other Gram-negative bacteria, E. coli, showed an intermediate susceptibility to inactivation by ZnO. Similar trends were observed for all ZnO nanoparticles and water chemistries tested. High-throughput viability tests were conducted for three bacteria cell types, six nanoparticles, and ZnCl2 in fresh water. All the IC50 values were expressed in total Zn2+ added per liter of media when assay was initiated. Smaller IC50 values correspond to higher toxicity and vice versa. All IC50 values in fresh water for B. subtilis were below 1 mg/L level. In contrast, the IC50 values for E. coli were almost 2 orders of magnitude higher. Moreover, no IC50 values could be determined for zinc-tolerant Gram-negative bacteria P. putida with an initial ZnO concentration up to 500 mg/L. The IC50 values of ZnCl2 for all three bacteria were similar to the IC50 values of ZnO nanoparticles expressed as total Zn2+ added (Figure 2), suggesting that released Zn2+ plays a dominating role in ZnO antibacterial activity. 3.3. Bacteria Viability in Different Aquatic Media. The influence of iron contents and water chemistry on ZnO antimicrobial activity was evaluated in the synthetic fresh water and in fresh water NaCl-only water matrices both in the absence and presence of 10 mg/L humic acid. The B. subtilis IC50 values for six different ZnO nanoparticles ranged from 0.26 to 0.52 mg/L (as total Zn2+ added) in fresh water (Figure 3), which agreed with previous research (7). Although, it was reported that the toxicity of ZnO nanoparticles to mammalian cells decreased with increasing iron loading (8), no correlation between the IC50 values and the iron contents was observed here. The IC50 in NaCl-only fresh water was slightly higher than that in the simulated fresh water. A small amount of humic acid slightly reduced the toxicity of ZnO nanoparticles in fresh water. However, in NaCl solution, there was a dramatic impact of humic acid on the toxicity of ZnO VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

757

FIGURE 1. Undoped ZnO NPs (a) in the absence and (b) in the presence of 10 mg/L of humic acid; SAED patterns of (c) undoped ZnO (d) 10% Fe-doped ZnO prepared from FSP and high-resolution TEM of (e) undoped ZnO and (f) 10% Fe-doped ZnO NPs.

FIGURE 2. Twenty-four hour IC50 values (total Zn2+ concentration added) of various ZnO nanoparticles and ZnCl2 using three bacteria species, in synthetic fresh water (*No IC50 could be determined with up to 500 mg/L initial ZnO concentration). 758

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 45, NO. 2, 2011

nanoparticles. The IC50 values increased to ∼3 mg/L, implying a decrease of antibacterial activity in this water chemistry. 3.4. Influence of Organic Macromolecules. The influence of NOMs on the ZnO antimicrobial activity for B. subtilis and P. putida were further evaluated using HA, FA, AA, TA, and BSA at10 mg/L. No IC50 values could be determined for P. putida with ZnO nanoparticle concentration up to 500 mg/L. For B. subtilis, all organics tested reduced the antibacterial activity of ZnO nanoparticles to some degree (Figure 4). However, the IC50 value in the presence of 10 mg/L of tannic acid was 1.2 mg/L, which was a 4-fold increase comparing to that in organic-free fresh water. Results from the parallel ionic Zn2+ antimicrobial assay revealed that the IC50 value of ZnCl2 in the presence of tannic was increased about 5-fold as well, from 0.2 to 1.1 mg/L expressed as Zn2+ (double-dot-dash line in Figure 4). The dissolved metal Zn2+ concentrations released by ZnO (ZnO-0% Fe and ZnO-10% Fe) in simulated fresh water were measured in the presence of 5 NOMs at 10 mg/L (Figure 4). Except for tannic acid, the presence of organic macromolecules did not significantly affect the dissolved Zn2+

FIGURE 3. Twenty-four hour IC50 values (B. subtilis) of ZnO nanoparticles (total Zn2+ added) and dissolved Zn2+ concentration (free Zn2+ passed through a 0.1 µm filter with initial ZnO concentration of 1 mg/L) in various water chemistries. Note: FW ) fresh water, NaCl ) NaCl-only fresh water with equivalent ionic strength, HA ) humic acid; pH of FW, FW + HA solution ) 8.47 and pH of NaCl, NaCl + HA solution ) 6.35; solubility limit of Zn2+ in pH 6.35 solution is 5880 mg/L and in pH 8.47 solution is 0.5 mg/L.

FIGURE 4. Influence of natural organic matter on 24-h B. subtilis IC50 values (total Zn2+ added) and dissolved Zn2+ (free Zn2+ passed 0.1 µm filter with initial ZnO concentration of 0.1 mg/L). Note: FW ) fresh water, pH of all solution was controlled at 8.47. (- · · ) IC50 of ZnCl2 in FW+TA and (- - -) IC50 of ZnCl2 in FW. concentration significantly. In contrast, a small amount of tannic acid dramatically reduced soluble Zn2+ ions. The dissolved Zn2+ concentration was reduced from 75 to 18 µg/L by adding 10 mg/L tannic acid to the fresh water matrix. A Zn2+ uptake experiment using tannic acid and humic acid as complexing agents was conducted with initial Zn2+ concentration of 200 µg/L (from ZnCl2). After 24 h of mixing with 10 mg/L tannic acid and humic acid, the free Zn2+ concentration in the filtrate was analyzed by ICP. The presence of 10 mg/L humic acid slightly reduced the Zn2+ concentration from 200 to 174 µg/L (only 12.8% Zn2+ was bound to HA) (Supporting Information, Figure S4). In contrast, 10 mg/L of tannic acid reduced the concentration from 200 to 32 µg/L (84% of Zn2+ was bound by TA).

4. Discussion According to the HTS bacterial viability assay, three bacteria cell types exhibited fundamentally different susceptibilities to ZnO nanoparticle and ZnCl2 toxicity. The difference in toxicity thresholds between Gram-positive and -negative bacteria might be attributed to cell membrane structure, which influenced the antibacterial efficacy of Zn2+ by controlling access to sites of action. One distinct feature of Gram-negative bacteria is the presence of an LPS layer

extending from the outer cell membrane. This LPS layer contributes greatly to the structural integrity of the bacteria and protects the cell from a variety of toxic molecules, such as some antibiotics (e.g., penicillin), digestive enzymes (e.g., lysozyme), detergents, heavy metals, and dyes (24). The LPS chain molecules form the outer membrane in Gram-negative bacteria, together with phospholipids, peptidoglycan, and exopolyscaccharide layers, containing large amounts of metal-complexing sites, contributing to the metal tolerance and sorption capacity of bacteria (18). P. putida is known to have high resistance to and retention of divalent heavy metals ions, such as Zn2+ and Cu2+. It exhibits high adsorption capacity (mass of zinc per unit mass of humic acid) toward heavy metal ions, such as Zn2+ and Cu2+, especially viable cells (17, 18). Indeed, this bacterial species is often found in metal-contaminated soil and is utilized for heavy metal remediation (25). Besides the LPS layer, the high resistance was attributed to the cell physiology and metabolism of P. putida cells (7). Mechanistic studies reveal that the P-type ATPase-based efflux system enables P. putida cells to actively transport the Zn2+ across the membrane, which efficiently protects intracellular targets from metal poisoning (18, 26). In this testing system, water chemistry played a more important role in ZnO toxicity than iron doping level. The IC50 values for B. subtilis varied almost 6-fold across the different water matrices tested. While the dissolved Zn2+ concentration was about 300 µg/L in all tests solutions with 1 mg/L initial ZnO concentration (Figure 3), the mechanism underlying the change in apparent toxicity is not fully understood. Perhaps a more important consideration is that use of simple 1:1 electrolytes (like NaCl here) may produce anomalous results in toxicity tests. A potentially critical question to consider is, what is an environmentally relevant aquatic media? It seems obvious from the results presented herein that simple 1:1 electrolytes are not. Therefore, regulators should exercise much caution when drawing conclusions from data derived from assays using environmentally irrelevant assay buffers. At ZnO nanoparticle concentrations of 0.1 mg/L, almost all ZnO dissolved, resulting in Zn2+ concentrations of 80 µg/ L, while at 1 mg/L ZnO loading, the dissolved Zn2+ was about 300 µg/L (Supporting Information, Figure S3). The dissolution results here demonstrated that ZnO nanoparticles were capable of releasing high dosage of ionic Zn2+. At the IC50 concentration of ZnO nanoparticles, the concentration of dissolved Zn2+ was expected to be 80-300 µg/L, which was in the range of IC50 values for Zn2+ due to ZnCl2 addition (IC50 value of ZnCl2 was 203 µg/L). The results suggest that released Zn2+ ions could explain the observed bactericidal effects. Thus, we observed no measurable nano effect associated with ZnO bacterial toxicity. At relatively low concentrations, tannic acid increased the IC50 value by almost 4-fold due to Zn2+ uptake by tannic acid. Tannic acid is a polyphenolic compound containing glucose linkages through ether bonds to an average of 9-10 molecules of gallic acid; it is a fairly common constituent of NOM in aquatic environments and is often used as a chelating agent for metals such as Fe(III), Zn(II), and Cu(II) (27-29). It has higher metal-binding capacity than the other organic macromolecules tested because of the highly concentrated o-diphenol groups. Furthermore, the multicatecholate structure of tannic acid also allows the formation of metal-tannin precipitates (27, 30, 31). With low concentration of ZnO, i.e., 0.1 mg/L of ZnO, almost all ZnO dissolved. TA bound the dissolved Zn2+, which effectively reduced the bioavailable Zn2+ and consequently the toxic response induced by dissolved Zn2+. With increase of ZnO concentration, the dissolved Zn2+ saturated the binding sites of TA. The toxic response of Zn2+ appeared VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

759

again under such conditions. Thus, a shift of dose-response curve to higher Zn2+ concentration in TA solution was observed and, subsequently, a higher IC50, as reported in Figure 4. The results demonstrated that both model organism selection and aquatic chemistry are critical for assessing environmental impacts of nanoparticles. Gram-negative bacteria (P. putida and E. coli) are more resistant to the antibacterial activity of ZnO nanoparticles than Grampositive bacteria (B. subtilis). Among the five samples of NOM tested, tannic acid reduced ZnO toxicity most, because it formed stable complexes with Zn2+ ions, thereby reducing the free Zn2+ concentration in the media. The results presented here highlighted the importance of both bacteria cell type and water chemistry in nanotoxicology studies. Although a previous study suggests that iron doping reduces Zn2+ dissolution and toxicity of ZnO nanoparticles for mammalian cells (8), in this study iron doping did not significantly impact ZnO nanoparticle dissolution or bacterial toxicity. A few differences should be mentioned between these two studies. In the previous study, 300 mg/L ZnO concentrations were used in dissolution tests, whereas in this study 1 mg/L ZnO concentrations were used. Also in the present study, ZnO nanoparticles completely dissolved within the 24 h bacterial viability assay incubation period; hence, dissolution kinetics was not considered (32). According to geochemical modeling results, the solubility limit of Zn2+ was governed by solution pH in our testing system; around pH 7, one pH unit difference resulted in nearly 100-fold change in Zn2+ equilibrium concentration (21). The two independent studies offer a cogent example of the complexity of nanotoxicology. The absolute dependence of nanoparticle stability and bioavailability on bacterial cell type and water chemistry makes clear the need for high-throughput combinatorial methods to help assess potential environmental impacts of nanotechnology.

Acknowledgments This research was supported in part by the National Science Foundation and the Environmental Protection Agency through the UC Center for Environmental Implications of Nanotechnology(CooperativeAgreementNumberEF0830117) as well as UC Discovery Industry-University Cooperative Research Program (#GCP07-10239A) with industrial matching funds provided by NanoH2O Inc. The corresponding author has a financial interest in one of the project cosponsors, NanoH2O Inc., through stock ownership and a consulting agreement.

Supporting Information Available Ionic composition of two water matrices, comparison of nanoparticle separation methods, representative doseresponse curves for all three bacterial, dissolved Zn2+ concentration in synthetic fresh water, and the Zn2+ uptake by model NOM. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Usenko, C. Y.; Harper, S. L.; Tanguay, R. L. In vivo evaluation of carbon fullerene toxicity using embryonic zebrafish. Carbon 2007, 45 (9), 1891–1898. (2) Oberdorster, G.; Oberdorster, E.; Oberdorster, J. Nanotoxicology: An emerging discipline evolving from studies of ultrafine particles. Environ. Health Perspect. 2005, 113 (7), 823–839. (3) Lovern, S. B.; Strickler, J. R.; Klaper, R. Behavioral and physiological changes in Daphnia magna when exposed to nanoparticle suspensions (titanium dioxide, nano-C-60, and C(60)HxC(70)Hx). Environ. Sci. Technol. 2007, 41 (12), 4465– 4470. (4) Dahl, J. A.; Maddux, B. L. S.; Hutchison, J. E. Toward greener nanosynthesis. Chem. Rev. 2007, 107 (6), 2228–2269. 760

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 45, NO. 2, 2011

(5) Auffan, M.; Rose, J.; Bottero, J. Y.; Lowry, G. V.; Jolivet, J. P.; Wiesner, M. R. Towards a definition of inorganic nanoparticles from an environmental, health and safety perspective. Nature Nanotechnol. 2009, 4 (10), 634–641. (6) Porter, F. Zinc Handbook: Properties, Processing, and Use in Design; CRC Press: Boca Raton, FL, 1991. (7) Adams, L. K.; Lyon, D. Y.; Alvarez, P. J. J. Comparative ecotoxicity of nanoscale TiO2, SiO2, and ZnO water suspensions. Water Res. 2006, 40 (19), 3527–3532. (8) George, S.; Pokhrel, S.; Xia, T.; Gilbert, B.; Ji, Z. X.; Schowalter, M.; Rosenauer, A.; Damoiseaux, R.; Bradley, K. A.; Ma¨dler, L.; Nel, A. E. Use of a rapid cytotoxicity screening approach to engineer a safer zinc oxide nanoparticle through iron doping. ACS Nano 2010, 4 (1), 15–29. (9) Heinlaan, M.; Ivask, A.; Blinova, I.; Dubourguier, H. C.; Kahru, A. Toxicity of nanosized and bulk ZnO, CuO and TiO2 to bacteria Vibrio fischeri and crustaceans Daphnia magna and Thamnocephalus platyurus. Chemosphere 2008, 71 (7), 1308– 1316. (10) Jeng, H. A.; Swanson, J. Toxicity of metal oxide nanoparticles in mammalian cells. J. Environ. Sci. Health Part A 2006, 41 (12), 2699–2711. (11) Wiench, K.; Wohlleben, W.; Hisgen, V.; Radke, K.; Salinas, E.; Zok, S.; Landsiedel, R. Acute and chronic effects of nano- and non-nano-scale TiO2 and ZnO particles on mobility and reproduction of the freshwater invertebrate Daphnia magna. Chemosphere 2009, 76 (10), 1356–1365. (12) Zhu, X. S.; Zhu, L.; Chen, Y. S.; Tian, S. Y. Acute toxicities of six manufactured nanomaterial suspensions to Daphnia magna. J. Nanopart. Res. 2009, 11 (1), 67–75. (13) Crane, M.; Handy, R. D.; Garrod, J.; Owen, R. Ecotoxicity test methods and environmental hazard assessment for engineered nanoparticles. Ecotoxicology 2008, 17 (5), 421–437. (14) Warheit, D. B. How meaningful are the results of nanotoxicity studies in the absence of adequate material characterization. Toxicol. Sci. 2008, 101 (2), 183–185. (15) Nel, A. E.; Ma¨dler, L.; Velegol, D.; Xia, T.; Hoek, E. M. V.; Somasundaran, P.; Klaessig, F.; Castranova, V.; Thompson, M. Understanding biophysicochemical interactions at the nanobio interface. Nat. Mater. 2009, 8 (7), 543–557. (16) Service, R. F. NanotechnologysCan high-speed tests sort out which nanomaterials are safe. Science 2008, 321 (5892), 1036– 1037. (17) Chen, X.; Wu, W.; Shi, J. Y.; Xu, X. H.; Wang, H.; Chen, Y. X. Adsorption of copper and zinc on Pseudomonas putida CZ1: Particle concentration effect and adsorption reversibility. Colloids Surf., B 2007, 54 (1), 46–52. (18) Guine, V.; Spadini, L.; Sarret, G.; Muris, M.; Delolme, C.; Gaudet, J. P.; Martins, J. M. F. Zinc sorption to three Gram-negative bacteria: Combined titration, modeling, and EXAFS study. Environ. Sci. Technol. 2006, 40 (6), 1806–1813. (19) Franklin, N. M.; Rogers, N. J.; Apte, S. C.; Batley, G. E.; Gadd, G. E.; Casey, P. S. Comparative toxicity of nanoparticulate ZnO, bulk ZnO, and ZnCl2 to a freshwater microalga (Pseudokirchneriella subcapitata): The importance of particle solubility. Environ. Sci. Technol. 2007, 41 (24), 8484–8490. (20) Tchobanoglous, G.; Schroeder, D. Water Quality: Characteristics, Modeling, Modification. Addison-Wesley Pub. Co.: Reading, MA, 1987. (21) USGS. http://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqc/. (22) Chen, K. L.; Mylon, S. E.; Elimelech, M. Aggregation kinetics of alginate-coated hematite nanoparticles in monovalent and divalent electrolytes. Environ. Sci. Technol. 2006, 40 (5), 1516– 1523. (23) Karmakar, D.; Mandal, S. K.; Kadam, R. M.; Paulose, P. L.; Rajarajan, A. K.; Nath, T. K.; Das, A. K.; Dasgupta, I.; Das, G. P. Ferromagnetism in Fe-doped ZnO nanocrystals: Experiment and theory. Phys. Rev. B: Condens. Matter 2007, 75, 14. (24) Kucerka, N.; Papp-Szabo, E.; Nieh, M. P.; Harroun, T. A.; Schooling, S. R.; Pencer, J.; Nicholson, E. A.; Beveridge, T. J.; Katsaras, J. Effect of cations on the structure of bilayers formed by lipopolysaccharides isolated from Pseudomonas aeruginosa PAO1. J. Phys. Chem. B 2008, 112 (27), 8057–8062. (25) Chen, X. C.; Wang, Y. P.; Lin, Q.; Shi, J. Y.; Wu, W. X.; Chen, Y. X. Biosorption of copper(II) and zinc(II) from aqueous solution by Pseudomonas putida CZ1. Colloids Surf., B 2005, 46 (2), 101– 107. (26) Choudhury, R.; Srivastava, S. Mechanism of zinc resistance in Pseudomonas putida strain S4. World J. Microbiol. Biotechnol. 2001, 17 (2), 149–153.

(27) Cruz, B. H.; Diaz-Cruz, J. M.; Arino, C.; Esteban, M. Heavy metal binding by tannic acid: A voltammetric study. Electroanalysis 2000, 12 (14), 1130–1137. (28) Ucer, A.; Uyanik, A.; Aygun, S. F. Adsorption of Cu(II), Cd(II), Zn(II), Mn(II) and Fe(III) ions by tannic acid immobilised activated carbon. Sep. Purif. Technol. 2006, 47 (3), 113–118. (29) Wang, J. N.; Li, A. M.; Xu, L.; Zhou, Y. Adsorption of tannic and gallic acids on a new polymeric adsorbent and the effect of Cu(II) on their removal. J. Hazard. Mater. 2009, 169 (1-3), 794– 800.

(30) McDonald, M.; Mila, I.; Scalbert, A. Precipitation of metal ions by plant polyphenols: Optimal conditions and origin of precipitation. J. Agric. Food. Chem. 1996, 44 (2), 599–606. (31) South, P. K.; Miller, D. D. Iron binding by tannic acid: Effects of selected ligands. Food Chem. 1998, 63 (2), 167–172. (32) Elgersma, F.; Witkamp, G. J.; Vanrosmalen, G. M. Kinetics and mechanism of reductive dissolution of zinc ferrite in H2O and D2O. Hydrometallurgy 1993, 33 (1-2), 165–176.

ES102266G

VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

761