Toxicity of engineered nanomaterials and their transformation products following wastewater treatment on A549 human lung epithelial cells

Graphical abstract


Introduction
The concern that engineered nanomaterials (ENMs) may have adverse effects on human health is increasing as application of nanotechnology in consumer products expands [1,2]. Humans can be exposed to ENMs in the workplace and during use of nano-products, and also through contact with water, soil, or air to which ENMs may have been released [3]. Comprehensive risk assessment of ENMs requires characterization of the toxicity of ENMs under a wide range of exposure conditions, including environmental routes. Cytotoxicity and genotoxicity of common metal and carbon nanomaterials, such as nanoAg, nanoTiO 2 , and carbon nanotubes have been widely studied in human lung, dermal, and visceral cells [4][5][6][7]. However, in assessing the risks of ENMs released into the natural environment, available data on the toxicity of environmentally-relevant forms of ENMs is lacking [8,9]. Extrapolation of toxicity data based on testing of pristine ENMs may not be appropriate because ENMs are highly reactive by nature and can be chemically, physically and biologically transformed in the environment, potentially altering their toxicity [10,11].
Wastewater treatment plants (WWTPs) are a critical route of ENM receipt and release into the natural environment [12][13][14]. The complex wastewater matrix is likely to favor transformation of ENMs. A small portion of ENMs will remain in the wastewater effluent, while the majority will associate with the sludge [15,16] and eventually be disposed of by land-application, landfill, or incineration [12]. During the reuse of treated wastewater and landapplication of waste sludge (biosolids), there is potential for humans to be exposed to transformed ENMs, especially through inhalation of aerosols generated [17,18]. However, to the knowledge of the authors, impacts of human exposure to transformed ENMs following wastewater treatment have not previously been reported.

Cell culture and treatment
A549 human lung alveolar epithelial cells were obtained from ATCC (#CCL-185, Manassas, VA). Cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM, Thermo Scientific HyClone, Logan, UT) containing 10% heat inactivated fetal bovine serum (FBS, Atlanta Biologicals, Flowery Branch, GA) and 1% penicillin-streptomycin (Thermo Scientific HyClone, Logan, UT), and were maintained in a humidified incubator at 37 • C and 5% CO 2 . Cells were seeded at a density of 1 × 10 4 in 100 L culture medium in each well of 96-well plates. At 80% confluency, they were treated for 24 h either with pristine nanomaterials, ionic/bulk analogs, SBR aqueous effluents, or SBR biosolids as indicated. The exposure concentrations of the pristine materials (Table S1) varied from 1 to 67 g/mL for the cytotoxicity assay to estimate IC 50 values. Genotoxicity assays were carried out at key concentration values based on the results of the cytotoxicity assays (Table S1). The exposure concentrations of nanomaterials and ionic/bulk materials in SBR effluents and biosolids are shown in Table 1. All samples were diluted with the culture medium to target concentrations.

Cytotoxicity assay
Cell viability was measured using the WST-1 assay (Roche, Indianapolis, IN) based on quantification of mitochondrial activity as an indicator of cytotoxicity. In viable cells, the tetrazolium salt WST-1 is converted to soluble formazan dye by mitochondrial succinate-tetrazolium reductase, which can be quantified by absorbance. After the A549 cells were treated with the samples for 24 h, WST-1 reagent was added to each well at 1/10 volume of the medium. The absorbance was quantified after incubating at 37 • C for 3 h using a Tecan Safire 2 Microplate Reader (Tecan US In., Research Triangle Pa, NC) at 440 nm with a reference wavelength of 660 nm. All treated A549 cells (a) were tested in triplicate in three independent experiments with three controls: (b) untreated A549 cells in culture medium; (c) samples in culture medium without A549 cells; and (d) culture medium only. In a single experiment, cell viability was calculated as percentage of the average absorbance derived from triplicate runs of treated cells relative to untreated control cells, with absorbances of corresponding controls subtracted out to address possible matrix interferences: fraction cell viabil-

Genotoxicity assay
DNA damage in A549 cells was detected using immunofluorescent labeling of ␥H2AX foci as described elsewhere [19]. At sites of DNA double strand breaks, H2AX, a minor nucleosomal histone protein, is rapidly phosphorylated and forms ␥H2AX [20]. The experimental and imaging procedures are provided in the Supplementary Information. Three independent experiments were conducted with at least 200 cells imaged in a single test. Untreated cells and cells treated with 100 M H 2 O 2 for 10 min were included as negative and positive controls, respectively. Images were analyzed using ImageJ 1.47 (http://rsbweb.nih.gov/ij/) with a macro designed to subtract background and count the number of foci within the defined nucleus masks. Table 1 Concentration of nanomaterials and ionic/bulk materials in SBR effluents and biosolids, and exposure concentration in cytotoxicity and genotoxicity assays.

SBR effluents (g/L)
Exposure concentration (g/L) a Biosolids (g/g total solids) Exposure concentration (g/mL) b a SBR effluents were diluted approximately 1:3 in the culture medium. b SBR biosolids were exposed to A549 cells at 200 g total solids/mL. c Concentrations of nanoTiO2 and bulkTiO2 in SBR effluents were near the detection limit of Ti by ICP-MS and not significantly different from concentrations in the undosed SBR.

Statistical analysis
The data were presented as mean ± standard deviation of three independent experiments. Student's t test or pairwise t test was conducted in R-2.8.1 (http://cran.r-project.org/bin/windows/base/old/2.8.1/) to determine statistical differences between samples at a significance level of 0.05 (p < 0.05).

Toxicity of pristine nanomaterials
Based on the WST-1 assay, the viability of cells exposed to nanoAg, Ag + , NZVI and Fe 2+ for 24 h decreased significantly (p < 0.05) in a dose-dependent manner (Fig. S1). The IC 50 values for these materials were estimated to be 53 ± 2, 21 ± 0.1, 38 ± 2 and 55 ± 2 g/mL, respectively. In particular, the IC 50 of Ag + was significantly lower than that of nanoAg (p = 5 × 10 −4 ), a result that is in agreement with other studies using A549 cell targets [21,4]. By contrast, NZVI was more toxic than its ionic analog, Fe 2+ (p = 6 × 10 −4 ). In another study, the cytotoxicty of NZVI (synthesized through reduction of FeCl 3 by NaBH 4 and coated with Pd) to human bronchial epithelial cells 16HBE14o was not significantly different from Fe 2+ [22]. Differences between the studies could relate to differences in NZVI coatings (manufacturer in present study reports biodegradable organic and inorganic stabilizers) or the different cells used in the assays (A549 versus 16HBE14o). The viability of cells exposed to nanoTiO 2 , nanoCeO 2 and their bulk analogs only decreased at a concentration of 67 g/mL, by 10-20% (Fig. S1, p < 0.05). But in a few other studies, no significant cytotoxicity of nanoTiO 2 and nanoCeO 2 to A549 cells was observed, even at concentrations up to 100 g/mL [23,5,24].
Only cells that were exposed to 33 and 53 g/mL nanoAg and 21 g/mL Ag + showed significantly higher numbers of ␥H2AX foci per cell (Fig. S2A) and greater percentages of cells containing ␥H2AX foci (Fig. S2B) (p < 0.05) relative to untreated control cells, suggesting genotoxicity resulting from DNA double strand breaks. The number of ␥H2AX foci per cell exposed to 9 g/mL nanoAg was not significantly different from that of the control cells (Fig.  S2A, p = 0.12), but the percentage of cells containing ␥H2AX foci was significantly higher than for the control cells (Fig. S2B, p = 0.03). Although IC 50 concentrations of NZVI and Fe 2+ (38 and 55 g/mL, respectively), and nanoTiO 2 , nanoCeO 2 and the bulk analogs at 67 g/mL induced significant cytotoxicity to A549 cells, the number of ␥H2AX foci were not significantly different in cells treated with these materials (Fig. S2, p > 0.05). To the authors' knowledge, genotoxicity of NZVI to human cells has not been investigated previously. Genotoxicities have been observed in A549 cells at concentrations of 2.5-15 g/mL of nanoAg by 32 P post-labeling of DNA adducts [4], 10-50 g/mL of nanoTiO 2 by cytokinesis block micronucleus assay [25], and 0.5-100 g/mL of nanoCeO 2 by alkaline comet assay [23]. However, in this study, the genotoxicity of nanoAg could be demonstrated, but not nanoTiO 2 or nanoCeO 2 at 67 g/mL. The distinct outcome of the nanoTiO 2 in the present study could be attributed to the different genotoxicity assays employed. With respect to nanoCeO 2 , the prior study [23] utilized lab-synthesized nanoCeO 2 with a size range of 16-22 nm, whereas the present study utilized commercial nanoCeO 2 with an average particle size of 33 ± 12 nm.

Toxicity of SBR effluents and biosolids
Based on a previous study (Ma et al. [35]), >99% of nanomaterials and ionic/bulk analogs dosed into the SBRs partitioned into the sludge relative to the influent concentrations. Concentrations of nanomaterials in SBR aqueous effluents and biosolids, as well as exposed to A549 cells, are shown in Table 1. No significant decrease of cell viability or induction of ␥H2AX foci were observed in A549 cells exposed to SBR effluents ( Fig. 1A and 2, p > 0.05). The exposure concentration of nanoAg in this study (9.48 g/L) was much higher than predicted concentrations in WWTP effluents (<0.5 g/L) [12][13][14], while the exposure concentrations of NZVI (2.37 g/L), nanoTiO 2 (0.32 g/L), and nanoCeO 2 (0.08 g/L) were within or lower than the lower bound concentrations predicted (0.7-20 g/L for NZVI [14], 1-70 g/L for nanoTiO 2 [12][13][14], and 0.5 × 10 −4 to 2 g/L for nanoCeO 2 [13,14]. Results of this study indicated limited toxicity of nanoAg in wastewater effluents at higher than environmentally-relevant concentrations to A549 cells, while the effects of higher concentrations of NZVI, nanoTiO 2 and nanoCeO 2 remain to be determined. The concentrations of nanomaterials in biosolids in this study (>2000 g/g dry mass, Table 1) were significantly Fig. 1. Characteristic cytotoxicity of A549 cells exposed to (A) wastewater effluents and (B) biosolids from undosed SBR, and SBRs dosed with nanoAg, Ag + , NZVI, Fe 2+ , nanoTiO2, bulkTiO2, nanoCeO2, and bulkCeO2 for 24 h by WST-1 assay. Exposure concentrations are shown in Table 1. Error bars represent standard deviations of three independent experiments. "*" indicates significant decrease of viability compared with untreated control cells (p < 0.05). higher than concentrations predicted in biosolids from WWTPs (<1000 g/g dry mass) [12][13][14]. Moreover, 200 g total solids/mL represented a high exposure dose relative to other studies of effects of aerosolized biosolids to human lung cells [26]. The viability of cells exposed to biosolids at 200 g total solids/mL decreased by 7-10% relative to untreated control cells (p < 0.05), except for biosolids containing NZVI (p = 0.08) (Fig. 1B). But, there was no significant difference between cells exposed to biosolids from undosed versus dosed SBRs (p > 0.05), indicating that the decrease in cell viability was not likely due to the nanomaterials or ionic/bulk materials, but probably instead to the high concentration of total solids. Cytotoxicity of biosolids was also examined, at 50 and 100 g total solids/mL, and no significant effects were observed relative to control cells (p > 0.05, Fig. S3).
Genotoxicity of biosolids was examined at 200 g total solids/mL. However, no significant differences were observed in terms of the number of ␥H2AX foci per cell or in the percentage of cells containing ␥H2AX foci in cells treated with biosolids relative to untreated control cells (Fig. 3, p > 0.05), suggesting little or no DNA damage to A549 cells at concentrations exceeding most likely aerosol exposure levels.
Uptake of pristine nanoAg, nanoTiO 2 , and nanoCeO 2 by human cells [21,5,24,27], and NZVI by mammalian nerve cells [28] has been observed in previous studies, and the most commonly identified mechanism of toxicity was the generation of reactive oxygen species, which induced oxidative stress [27,4,22,25]. Release of Ag + was considered another potential cause of nanoAg toxicity [29]. Based on TEM-EDS mapping carried out in a previous study of the SBR biosolids (Ma et al. [35]), while a large portion of nanoAg remained dispersed, it mainly formed Ag-S complexes. Sulfidation has been reported to reduce toxicity of nanoAg to microbes, aquatic and terrestrial eukaryotic organisms due to low solubility of Ag-S complexes [30,31]. Similarly, transformation of nanoAg in this study may limit its reactivity and result in little toxicity of SBR biosolids to A549 cells. The majority of NZVI, nanoTiO 2 and nanoCeO 2 were aggregated, but not chemically modified. Epithelial cells are impervious to aggregated nanomaterials by diffusion or macropinocytosis [32]. Therefore, the absence of cytotoxicity and genotoxicity of SBR biosolids in this study could be attributed to the inability of aggregated NZVI, nanoTiO 2 , and nanoCeO 2 to enter cells. However, the size of the nanomaterial aggregate can affect its physiological distribution and kinetics, cellular distribution (for example within the draining lymph node for an aerosolized particle), cellular uptake, and intracellular processing pathways [33]. Also, it is difficult to ascertain the fate of aerosolized nanomaterials from biosolids in the respiratory tract as particles can diffuse and convert during interstitial transport depending on their size [34]. Thus, future work may consider the toxicity under ex vivo and in vivo conditions.