ZnS quantum dot exposure in estuarine fish: Bioavailability, oxidative stress responses, reproduction, and maternal transfer

Aquatic Toxicology 148 (2014) 27–39

Contents lists available at ScienceDirect

Aquatic Toxicology journal homepage: www.elsevier.com/locate/aquatox

Dietary CdSe/ZnS quantum dot exposure in estuarine fish: Bioavailability, oxidative stress responses, reproduction, and maternal transfer T. Michelle Blickley a,b,∗ , Cole W. Matson b,c,1 , Wyatt N. Vreeland d , Daniel Rittschof a,b , Richard T. Di Giulio b,c , Patricia D. McClellan-Green e,f a

Marine Science and Conservation, Duke University Marine Laboratory, Beaufort, NC, United States Integrated Toxicology & Environmental Health Program, Duke University, Durham, NC, United States c Center for the Environmental Implications of NanoTechnology (CEINT), Duke University, Durham, NC, United States d Biochemical Science Div., National Institute of Standards & Technology, Gaithersburg, MD, United States e Dept. of Environmental & Molecular Toxicology, North Carolina State University, Raleigh, NC, United States f Center for Marine Sciences & Technology, North Carolina State University, Morehead City, NC, United States b

a r t i c l e

i n f o

Article history: Received 4 July 2013 Received in revised form 18 December 2013 Accepted 22 December 2013 Keywords: Engineered nanoparticles Quantum dots Fundulus heteroclitus Bioavailability Oxidative stress Maternal transfer

a b s t r a c t Continued development, use, and disposal of quantum dots (QDs) ensure their entrance into aquatic environments where they could pose a risk to biological organisms as whole nanoparticles or as degraded metal constituents. Reproductive Fundulus heteroclitus were fed a control diet with lecithin, diets containing 1 or 10 ␮g of lecithin-encapsulated CdSe/ZnS QD/day, or a diet containing 5.9 ␮g CdCl2 /day for 85 days. Cadmium concentrations in liver, intestine, and eggs were quantified with inductively coupled plasma mass spectrometry. In fish fed 10 ␮g QD/day, QDs or their degradation products traversed the intestinal epithelia and accumulated in the liver. Less than 0.01% of the QD’s cadmium was retained in the liver or intestinal tissues. This compares to 0.9% and 0.5% of the cadmium in the intestine and liver, respectively of fish fed a CdCl2 diet. Cadmium was also detected in the eggs from parents fed 10 ␮g QD/day. No significant changes in hepatic total glutathione, lipid peroxidation, or expression of genes involved in metal metabolism or oxidative stress were observed. While QDs in the diet are minimally bioavailable, unusual levels of vitellogenin transcription in male fish as well as declining fecundity require further investigation to determine if endocrine disruption is of environmental concern. © 2013 Elsevier B.V. All rights reserved.

1. Introduction The release of engineered nanoparticles (ENPs) from commercial products raises concerns about their environmental fate and toxicity (Benn and Westerhoff, 2008; Kaegi et al., 2008). In aquatic systems, ENPs complex with organic debris and adhere to organisms living in the water column (Navarro et al., 2009; Oberdorster et al., 2006). At high salinities (ionic strength), ENPs can form large aggregates that precipitate and interact with benthic organisms and bottom feeders (Blickley and McClellan-Green, 2008; Cumberland and Lead, 2009; French et al., 2009; Jiang et al.,

∗ Corresponding author at: Dow AgroSciences, LLC. 9330 Zionsville Rd., Indianapolis, IN 46268, United States. Tel.: +1 317 337 5257; fax: +1 317 337 4880. E-mail addresses: [email protected] (T.M. Blickley), Cole [email protected] (C.W. Matson), [email protected] (W.N. Vreeland), [email protected] (D. Rittschof), [email protected] (R.T. Di Giulio), [email protected] (P.D. McClellan-Green). 1 Present address: Environmental Science and Center for Reservoir and Aquatic Systems Research (CRASR), Baylor University, Waco, TX, United States. 0166-445X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aquatox.2013.12.021

2009). Little information is available on the bioavailability and toxicity of ENPs or their aggregates in environmentally relevant situations. Quantum dots (QDs) are ENPs that emit tunable, discrete fluorescence based upon particle diameter. They are found in light emitting diodes, inks, solar cells, and telecommunication components. As applications for QDs expand, studies are needed to determine if QDs or their degradation products pose risks to aquatic organisms. Several published reports indicate that QDs are cytotoxic, genotoxic, and generate reactive oxygen species (ROS) (Green and Howman, 2005; Ipe et al., 2005; Kirchner et al., 2005; Liang et al., 2007). Using UV-exposed biotintylated CdSe/ZnS QDs, Green and Howman (2005) hypothesized that DNA damage occurred because the ZnS shell was oxidized to generate SO2 䊉− , which then generated superoxide and hydroxyl radicals. Ipe et al. (2005) reported similar results; irradiated CdSe QDs generated hydroxyl radicals, while irradiated CdS QDs generated superoxide and hydroxyl radicals. Given that QDs and their degradation products generate ROS, oxidative stress could function as a mechanism of QD toxicity.

28

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

In addition to generating ROS, liberated cadmium from unshelled QDs can cause developmental and behavioral abnormalities (Zhang et al., 2012a,b). Mercaptopropionic acid (MPA)-capped CdSe QDs released about 3.5% of the core’s total Cd into the media (60 mg/l Instant Ocean), while thioglycolic acid (TGA)-capped CdTe QDs leached about 8% of their total Cd. Zebrafish (Danio rerio) displayed deformities and altered locomotion congruent with Cd exposure (Cheng et al., 2000; Chow and Cheng, 2003). Shelling the QD’s core delays degradation (Derfus et al., 2004). Although core–shell QDs have not been fully evaluated for sublethal effects, some ENPs such as fullerene (C60 ) and titanium dioxide (TiO2 ) reportedly cause reproductive and developmental impairment (Tsuchiya et al., 1996; Oberdorster et al., 2006; Tao et al., 2009; Takeda et al., 2009). For example, water-stirred fullerene decreased fecundity in water fleas (Daphnia magna) by delaying molting and brood release (Oberdorster et al., 2006). Fullerene solubilized in tetrahydrofuran1 similarly accumulated in neonates and caused reproductive impairment in greater than 90% of the daughter daphnids by delaying reproductive maturity (Tao et al., 2009). In previous studies such as those conducted by Tsuchiya et al. (1996), mouse embryos exposed in utero to fullerene developed lethal head, tail, and yolk sac deformities. Takeda et al. (2009) also demonstrated developmental effects in mice following ENP exposure. They observed nanoscale TiO2 in the brain and testes of male offspring, which subsequently affected genital and cranial nervous system development. These data support the hypothesis of an increased risk to progeny through maternal transfer of ENPs. Fundulus heteroclitus are an established fish model for toxicological studies and their physiology is well documented (Burnett et al., 2007). These common estuarine fish are voracious indiscriminate feeders consuming organisms and organic particulates in the water column and sediments. As a result, they can be expected to eat ENPs that are transported into estuarine environments. The goals of this study were to determine if dietary QDs are potentially harmful to F. heteroclitus, specifically if they alter reproduction, development, or physiological functions. We examined the effects of dietary exposure to lecithin-encapsulated CdSe/ZnS QDs on oxidative stress and fecundity in adult fish. Hepatic total glutathione and lipid peroxidation levels were measured as oxidative stress biomarkers along with changes in the expression of metallothionein, glutathione-stransferase, glutathione peroxidase, Cu/Zn-superoxide dismutase, and Mn-superoxide dismutase. Fecundity was analyzed by monitoring the reproductive output of mating pairs throughout the study period and the expression of the egg yolk precursor protein vitellogenin in the liver tissue at the conclusion of the study. Cadmium concentrations were quantified in the eggs to determine whether ENPs were maternally transferred. Lastly, developmental effects were quantified in a subset of embryos raised to hatching. 2. Methods 2.1. Chemicals CdSe/ZnS Evidots (Cat.# ED-C11-TOL-0620, Lot # LN018ICSB) were purchased from Evident Technologies (Troy, NY). CdCl2 (#202908) and gelatin (G-9382) were purchased from SigmaAldrich (St. Louis, MO). Lecithin (#102147) was purchased from MP Biomedicals, Inc. (Solon, OH). OptimaTM trace-metal free HNO3 and 30–32% H2 O2 were purchased from Fisher (Pittsburgh, PA). The cadmium standard (CLCD-2-2Y) was purchased from Spex CertiPrep (Metuchen, NJ).

Table 1 QD and Cd contenta in the diets. Treatment Standard diet (SD) SD + lecithin Low QD High QD CdCl2

Nominal dose 0 ␮g Lec. 5 ␮g Lec. 5 ␮g Lec. 5 ␮g Lec. 5 ␮g Lec.

0.0 ␮g QD 0.0 ␮g QD 0.5 ␮g QD 5.0 ␮g QD 1.7 ␮g Cd

Measured Cd

CV (%)

0.00 ␮g Cd 0.00 ␮g Cd 0.07 ␮g Cd 0.74 ␮g Cd 1.42 ␮g Cd

3.2 10.7 6.5

CV = Coefficient of variation. a Amounts are for an individual food disc. Each fish received 2 food discs per day.

2.2. Diet preparation Five food diets were prepared: standard diet (SD), SD + lecithin, low QD, high QD, and CdCl2 (Table 1). Each diet contained 112.5 mg/ml ground ZeiglerTM tropical flake food (Aquatic Ecosystems, Apopka, FL), 2% gelatin, 250 ␮l/ml shrimp puree (1 homogenized medium size shrimp/10 ml of deionized water), and heated sterile DI water. The appropriate amount of the lecithin-encapsulated QDs (refer to Supplemental Information 1 for preparation methodology and characterization of the nanoparticles), lecithin, and/or CdCl2 were added to each food mixture and stirred for 5–10 min to obtain homogeneous slurry. The nominal lecithin concentration in the slurry was 20 ␮g/ml; all treatments except the SD had equivalent lecithin concentrations. The food mixtures were pipetted in 250 ␮l aliquots onto waxed paper and allowed to solidify at room temperature. The food discs were then stored at −20 ◦ C until administered. The food was evaluated for the presence of fluorescent nanoparticles to ensure that despite being frozen and thawed, whole functional quantum dots were being fed to the F. heteroclitus. Food discs from the high QD treatment were heated in a waterbath to 48 ◦ C for 20 min, aliquotted onto glass slides, and coverslipped. The slides were viewed with a multiphoton confocal microscope (Leica TCS SP5 with DM6000, Germany). The QDs were excited with a 405 nm UV laser and the emissions were recorded from 500 to 700 nm in 6 nm increments. A xy␭ scan (speed: 100 Hz) was used to generate the spectral images for the regions of interest. While fluorescence is useful in determining QD concentrations in transparent suspensions (Yu et al., 2003), the opaque nature of the diets required using the QD’s cadmium as a tracer. Cadmium levels (Table 1) were measured with inductively coupled plasma mass spectrometry (ICP-MS). Food discs were homogenized in acid-washed polypropylene tubes for 30 seconds with a batteryoperated hand-held grinder in 200 ␮l of 7N (7 mol/l) OptimaTM trace-metal free nitric acid. Then, 800 ␮l of 7 N HNO3 was added to each tube and the samples were allowed to digest overnight. In the morning, the samples were sonicated for 5 min (Branson 2510, Danbury, CT) and 90 ␮l OptimaTM 30−32% H2 O2 was added to each mixture. The samples were vortexed intermittently over the next 6 h and then transferred to trace-metal free 15 ml vials. Each sample was combined with 10 ml of internal standard solution and then centrifuged at ∼1300 × g for 10 min. Supernatant, 1 ml, was transferred to a fresh tube, diluted with 10 ml of internal standard solution, and analyzed on a VG Plasmaquad 3 ICP-MS (Thermo Scientific, Waltham, MA). The internal standard solution consisted of 10 ppb In, Tm, and Bi in ∼2% HNO3 . The detection limit was 0.0005 ␮g of total Cd per sample or 0.002 ng/ml (Taylor, 2001). Recovery was 86.8% of cadmium standard spiked into SD + lecithin food discs. 2.3. Research design

1

nC60 underwent distillation, filtration, and solvent-exchange processes to remove greater than 99.5% of the tetrahydrofuran (THF). Residual solvent was below the GC-MS detection limit of 1 ppb (Tao et al., 2009).

Sexually mature F. heteroclitus were obtained from a tidal creek in Morehead City, NC (GPS coordinates: N 34.737198, W 76.71091).

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

Fish were acclimated to the laboratory for 2 weeks and then 30 mating pairs were housed individually in segmented tanks containing 15 l of aerated, filtered natural seawater at 23 ◦ C and 20 ± 2 ppt salinity with a 16-light:8-dark cycle. Water changes (50%) were performed daily. Each pair received the standard diet for 27 days while their spawning output was observed (F. heteroclitus spawn on a 2 week lunar cycle). Following the observation period, the treatment diets were administered daily for 85 days. The daily dose was achieved by feeding each fish 2 food discs per day. The fish were given additional tropical flake food 1–2 times daily (after the food discs were consumed) and live brine shrimp (Artemia spp.) twice a week. Eggs/embryos were collected in mesh-covered finger bowls (egg traps) to prevent consumption by the adult fish. The eggs were collected daily, washed 3× with clean seawater, and counted. Subsets of eggs raised in clean 20 ppt seawater were monitored for morphological variations in the notochord, eyes, brain, heart, bladder, trunk, and fins (Armstrong and Child, 1965). Embryos with deformities were imaged using a Meiji Techno R2 dissecting microscope (Santa Clara, CA) and with a T45 camera (Diagnostic Images, Inc., Sterling Heights, MI) using Image-Pro Plus software (MediaCybernetics, Inc., Bethesda, MD). Additional eggs, stored in groups of 10 eggs, were preserved at −80 ◦ C for cadmium analysis. F. heteroclitus were euthanized on day 86 with tricaine methanesulfonate (MS-222) and the tissues excised. Intestines were purged with sterile PBS and preserved at −80 ◦ C until analyzed for cadmium content. A 2–3 mm portion of the liver was fixed in 10 volumes of ice-cold Millonig’s buffer + 2% glutaraldehyde. The remaining liver tissue was divided into 3 portions and stored at −80 ◦ C. Liver portions were analyzed for: (a) cadmium content via ICP-MS, (b) oxidative stress biomarkers (total glutathione and lipid peroxidation), and (c) changes in the mRNA expression of metallothionein, glutathione-s-transferase, glutathione peroxidase, superoxide dismutases, and vitellogenin. This study was reviewed and approved by the North Carolina State University Institutional Animal Care and Use Committee prior to initiation (06-058-B).

2.4. Tissue Cd measurements Cadmium concentrations (N = 4–6 individuals/treatment/sex) were measured to infer QD uptake and accumulation. Liver and intestinal tissues were weighed and placed in acid-washed polypropylene tubes. Tissues were homogenized in 200–300 ␮l of 7N HNO3 for 20 s. The samples were brought up to 1 ml of acid and digested overnight at room temperature. Eggs (N = 10 eggs per sample, 3–5 samples per diet) were homogenized in 200 ␮l of 7N HNO3 . The homogenate was transferred to Teflon vials and 800 ␮l of 15.8N HNO3 was added to each sample. Egg samples were digested overnight at 90 ◦ C. The next morning, all samples (liver, intestine, and eggs) were sonicated for 5 min (Branson 2510, Danbury, CT) and then 90 ␮l of 30–32% H2 O2 was added. The samples were vortexed intermittently over the next 6 h and then transferred to tracemetal free 15 ml vials. Each sample was combined with 10 ml of internal standard solution and centrifuged at ∼1300 × g for 10 min to remove undissolved particulates. One milliliter of supernatant was transferred to a fresh tube and diluted with 10 ml of internal standard solution. The samples were analyzed on a VG Plasmaquad 3 ICP-MS and compared to a cadmium standard curve. Twentytwo samples, eighteen of which were below or near the cadmium detection limit, were re-analyzed using 3 ml of supernatant and 10 ml of internal standard solution; the differences in measured Cd concentrations between the dilutions were negligible. Recovery of cadmium standard from NIST bovine liver standard reference material (1577b) was 83.7%. Recovery of cadmium standard spiked into SD + lecithin eggs was 79.0%.

29

2.5. Total glutathione (GSH) assay Total glutathione (N = 4–6 individuals/treatment/sex) was determined using the DTNB-GSSG recycling assay (Andersen, 1985; Ringwood et al., 2003) modified for 96-well plates. Livers were homogenized in 10 volumes of 5% sulfosalicylic acid (SSA) and then centrifuged at 6000 × g for 15 min at 4 ◦ C. The supernatant was diluted 1:10 by volume with SSA. A 10 ␮l aliquot of sample was combined with 195 ␮l of reaction cocktail consisting of 0.238 mg/ml NADPH buffer, 10 mM 5, 5 -dithiobis (2-nitrobenzoic acid), and sterile DI water. Glutathione reductase (50 U/ml) was added to each well and the plate read at 405 nm every 10 s for 2–2.5 min on a SpectraMax 190 UV/VIS spectrophotometer (Sunnyvale, CA, USA). Samples were compared to a GSH standard curve and rates determined via simple linear regression. Samples and standards were run in triplicate. 2.6. Lipid peroxidation (LPo) assay Liver samples (N = 4–6 individuals/treatment/sex) were prepared and analyzed according to the protocols of Ringwood et al. (2003) adapted from Ohkawa et al. (1979). Tissues were homogenized in 4 volumes of 150 mM Tris, pH 7.5, 1 mM EDTA buffer (Orbea et al., 2002) and then centrifuged at 6000 × g for 12 min at 4 ◦ C. A 50 ␮l aliquot of supernatant was combined with 700 ␮l 0.375% thiobarbituric acid and 7 ␮l 2% butylated hydroxytoluene. The samples were heated in a water-bath at 100 ◦ C for 15 min and centrifuged at 13,000 × g for 5 min at room temperature. Then, 200 ␮l aliquots of the supernatant were measured spectrophotometrically at 532 nm and compared against a malondialdehyde tetraethylacetal (MDA) standard curve. Samples and standards were run in triplicate. 2.7. Quantitative real-time PCR (qPCR) All glassware was baked at 100 ◦ C for 1 week or soaked in 3% H2 O2 for 15 min and rinsed in DEPC-treated water. The hand-held grinder, pipettes, and bench-top were cleaned with 70% EtOH and RNaseZap. RNase, DNase-free disposable pestles were rinsed with DEPC-treated water between samples. RNA was isolated from the livers of adult F. heteroclitus using RNA-Bee reagent (Tel-Test, Inc., Friendswood, TX) according to the manufacturer’s protocol with an additional chloroform phase separation and EtOH rinse of the pellet. RNA extraction efficiency was measured spectrophotometrically at 260/280 nm on a SpectraMax 190 microplate reader and the final concentration of the RNA determined (ng RNA/␮l). All samples had a 260/280 ratio of 1.9 or higher. Reverse transcription was carried out using an Omniscript cDNA synthesis kit (Qiagen Inc., Valencia, CA) according to the kit’s directions. Briefly, 500 ng of RNA, 10 ␮M oligo dT primers (Promega Corp., Madison, WI), and RNaseOut inhibitor (Invitrogen Corp., Carlsbad, CA) were combined and incubated at 37 ◦ C for 1 h in an Eppendorf Mastercycler (Hamburg, Germany) to produce the cDNA. The cDNA was then diluted to a 2 ng/␮l working concentration. QPCR was carried out using previously published ␤-actin primers (Wills et al., 2010). Target gene primers (Table 2) were designed using PrimerQuest software (Integrated DNA Technologies, Coralville, IA; http://www.idtdna.com/Scitools/ Applications/Primerquest). Primer efficiencies were tested to confirm that ␤-actin and the target genes amplified with comparable efficiency, greater than 90%. QPCR was performed using SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) following the manufacturer’s protocol. A 25 ␮l reaction consisting of 200 nM of each primer, 12.5 ␮l 2× SYBR Green PCR Master Mix, 9.5 ␮l dH2 O, and 4 ng cDNA template was amplified on an Applied Biosystems 7300 Real-Time PCR System with the following thermal profile: 95 ◦ C for 10 min; 40 cycles of 95 ◦ C for 15 s, 60 ◦ C

30

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

Table 2 Real-time PCR primer sequences for F. heteroclitus. GenBank ID

Primer name

Primer sequence (5 –3 )

␤-Actin

AY735154

Metallothionein (mt)

AB426465

Glutathione-S-transferase, mu class (gstmu)

AY725220

Glutathione peroxidase (gpx)

CN971701

Cu/Zn superoxide dismutase (sod1)

CV824930

Mn superoxide dismutase (sod2)

CN980747

Vitellogenin (vtg)

U07055

Fh-B-Actin-F Fh-B-Actin-R Fh-MT-F1 Fh-MT-R1 Fh-GSTmu-F1 Fh-GSTmu-R1 Fh-GPX-F1 Fh-GPX-R1 Fh-SOD1-F1 Fh-SOD1-R1 Fh-SOD2-F1 Fh-SOD2-R2 Fh-vg-F1 Fh-vg-R1

ACCACACATTTCTCATACACTCGGG CGCCTCCTTCATCGTTCCAGTTT AAAGGGAAGACCTGCGACAA ACAAAGAAAGGCTCGGCGTA TATGTGCGGAGAGACTGAGG TCACAAAGCCGTTTCTGAAG TCACCGACGAGCACTACAGG TGAACAGCGGGAAGGAGACG AGCCGTGTGTGTGCTGAAAG TGCCAAAGGGATTGTAGTGTGG ACACTTCCTGACCTGACATACGAC ATGATGCTTGCTGTGGTGGAG GCTCTTCCTGTTGATGTGCCTGAA GCTTGGATGTGTGTCTCCTGATTGG

Gene a

a

␤-Actin primers originally described by Wills et al. (2010).

for 60 s. A dissociation curve was generated at the completion of the run. All samples were amplified in duplicate. Data were analyzed using ABI PRISM 7000 Software, Version 1.1 (Applied Biosystems Inc., Foster City, CA). Gene expression was calculated using relative quantification by the 2− CT method of Livak and Schmittgen (2001). Target gene expression, following normalization to ␤-actin, was compared to references (SD + lecithin) to estimate average fold induction for each experimental group and each target gene. Each biological replicate represents a single adult F. heteroclitus, and each experimental group is represented by at least four individuals/treatment/sex. Data are presented as mean fold induction ± standard error.

Table 3 Average Cd concentrations in the intestine and liver after 85 days of exposure. Treatment Std diet (SD) SD + lecithin Low QD High QD CdCl2

Intestine (ng Cd/g tissue) 37 ± 6 39 ± 9 56 ± 11 (<0.01%) 103 ± 25 (<0.01%)b 18,734 ± 2012 (0.9%)b

Liver (ng Cd/g tissue) 44 ± 10 20 ± 4a 31 ± 5 (<0.01%) 42 ± 6 (<0.01%)b 5,627 ± 1046 (0.5%)b

Concentrations are given as the mean cadmium level (ng) per gram of tissue ± the standard error. a Denotes a statistically significant difference between the standard diet and SD + lecithin treatments. b Statistically different from the SD + lecithin treatment, ˛ = 0.05, N = 10–12.

2.8. Reproduction data analysis F. heteroclitus spawning is highly variable. One breeding pair may consistently spawn 10 eggs per lunar cycle, while another pair spawns 200 eggs per cycle, thus cumulative egg production was not used as the high spawning pairs obscure trends of the lesser spawning pairs. Fecundity rates (eggs/day) for each spawning period were recorded for each individual breeding pair. The rates were converted to percentages based on the pair’s spawning rate during the 27 day observation period (spawn O, which was normalized to 100%) and averaged for each treatment diet (N = 3–6 pairs). Pairs that did not spawn during the observation period were not used to determine fecundity. If a pair did not complete a spawning interval, their rate was not included in the average for that time-point through the end of the study. 2.9. Statistics The cadmium concentration distribution for liver and intestinal tissues (N = 4–6 individuals/treatment/sex) was right-skewed (range < LOD to 29,394 ng/g tissue; skewness = 2.76), thus they were log transformed and analyzed for significance with a 3-factor ANOVA (treatment, tissue, and sex) and a Tukey HSD post hoc, ˛ = 0.05 (Fig. 2), using JMP, version 8.0.2 (SAS Institute Inc.® , Cary, NC). No statistical difference existed between the sexes due to the treatment, so male and female data were combined to determine the average tissue cadmium concentrations (Table 3). However, given the p-value for the treatment versus tissue ANOVA, the liver and intestinal cadmium concentrations were analyzed with separate Kruskal–Wallis tests and Dunn’s post hoc comparisons (N = 10–12 individuals/treatment) using GraphPad Prism, version 5.0 (GraphPad Software, San Diego, CA). GSH and LPO concentrations were analyzed using a Kruskal–Wallis test with a Dunn’s post hoc comparison to compare across all treatment groups and between sexes. All qPCR data

were analyzed with Kruskal–Wallis tests with a Dunn’s post hoc test to determine significance (N = 4–6 individuals/treatment/sex). Where no statistical differences were seen between the sexes, the data were merged. Fecundity rates were analyzed for significance using a Repeated Measures ANOVA performed by JMP. Growth, hepatosomatic indices, and gonadosomatic indices for the adults (see Supplemental Information 2) were analyzed for significance across all the treatment diets using a Kruskal–Wallis test with a Dunn’s comparison post hoc, ˛ = 0.05. With the exception of fecundity, statistical analyses were performed using GraphPad Prism.

3. Results No unusual behaviors were observed during the study. One female receiving the SD + lecithin diet died during the treatment period and two females from the low QD treatment died during the observational period; torn fins and scale loss were observed in these females indicating male aggression. One high QD female died during the exposure period. She had no physical signs of male aggression.

3.1. QD stability Micrographs of the food (Fig. 1A–C) taken after completion of the study show that QD aggregates retained their ability to fluoresce after being stored at −20 ◦ C for a prolonged period. Spectral analysis (Fig. 1D) indicates a 15 nm left shift of the intensity peaks, from 630 nm (Fig. S1) to 615 nm. This shift could be due to interference of the food matrix or a slight degradation of the particle. The aggregates (arrows, Fig. 1D) are approximately 6.2, 5.3, and 6.7 ␮m in diameter (left to right).

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

31

Fig. 1. Confocal micrographs (A, transmitted; B, fluorescent; C, combined) of a high QD food disc and photoluminescence spectra (bottom) of the three regions of interest seen in the confocal micrographs (arrows) indicate that the discs contain intact nanocrystals.

3.2. QD uptake and bioavailability Examination of liver slices for the presence of whole QDs by confocal and transmission electron microscopy was inconclusive (data not shown). No QD fluorescence was observed in paraffin embedded liver sections. Epon embedded sections from fish fed the high QD diet had dense bodies in lipid vacuoles. Similar dense bodies were observed in SD + lecithin and CdCl2 sections suggesting these may be artifacts from staining or biological inclusions (data not shown). QD uptake was subsequently evaluated by measuring the cadmium concentrations in the food discs, intestines, and liver. Based on the food ICP-MS analysis, the total Cd administered over the course of the 85-day exposure period was 12.5 ␮g Cd/fish in the low QD diet, 125.2 ␮g Cd/fish in the high QD diet, and 240.8 ␮g Cd/fish in the CdCl2 diet. Tissue analysis revealed that the intestinal Cd concentrations in male and female fish fed the CdCl2 diet were significantly elevated compared to the SD + lecithin controls (Fig. 2). CdCl2 -treated females had an average intestinal Cd concentration of 23,017 ± 2111 ng Cd/g tissue versus 46 ± 14 ng Cd/g tissue in females receiving the SD + lecithin diet. Males receiving the CdCl2 diet had an average intestinal concentration of 14,451 ± 2451 ng Cd/g tissue compared with 31 ± 13 ng Cd/g tissue in the SD + lecithin males. Similarly, liver cadmium concentrations of both sexes were significantly increased in the CdCl2 diet versus SD + lecithin diet. In females, the average cadmium concentration in the liver was 7680 ± 1620 ng Cd/g tissue in CdCl2 diet versus 21 ± 3 ng Cd/g tissue in the SD + lecithin treatment. Males receiving the CdCl2 diet had an average liver Cd concentration of 3574 ± 711 ng Cd/g tissue compared to 19 ± 8 ng Cd/g tissue in males fed the SD + lecithin treatment. Due to the large multi-factorial data set and small sample size, cadmium concentrations in the tissues of fish fed the QD diets were not statistically different from those fed the SD + lecithin diet. When uptake was examined with regard to sex-specific differences

(Fig. 2), it was determined that females acquired cadmium more readily than males (P = 0.0013), but there was no treatment effect (P = 0.8096). Because no statistical differences were apparent in cadmium concentrations between males and females within each treatment in either intestine or liver tissues, male and female data were combined and the average cadmium concentrations are listed in Table 3. Cadmium concentrations were affected by tissue type (P < 0.0001), but not in conjunction with treatment (P = 0.1007). Statistical comparisons of liver and intestinal cadmium concentrations shown in Table 3 are non-parametric 1-factor ANOVAs reflecting the need to consider each tissue type separately. A statistically significant increase in cadmium was observed in the intestinal and liver tissues of fish fed the high QD and CdCl2 diets. Although cadmium was detected in the tissues of wild-caught F. heteroclitus fed the SD and SD + lecithin diets (Table 3), the Cd was attributed to accumulation from their natural environment as the cadmium content of the SD and SD + lecithin food cubes was 0.00 ␮g Cd (Table 1). Lecithin supplementation of controls affected hepatic, but not intestinal cadmium concentrations. There was a statistical difference between the hepatic Cd concentrations in animals fed the standard diet versus the SD + lecithin diet; however, the total cadmium content in the livers were similar given the increase in the liver size of the fish fed lecithin (refer to the hepatosomatic index presented in Supplemental Information 2). 3.3. Oxidative stress biomarkers 3.3.1. Biochemical biomarkers Oxidative stress biomarkers were measured in adult liver tissue to determine if there was an effect due to QD or CdCl2 ingestion. Total glutathione and lipid peroxidation levels (Fig. 3) in QD or CdCl2 exposed fish were not statistically different from the SD + lecithin treatment [Note: GSH and LPO were similar between SD and SD + lecithin animals, so only the latter is presented in the graphs]. GSH levels were statistically similar in males and

32

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

Fig. 2. Cd concentrations of intestine and liver tissues in female (open) and male (filled) F. heteroclitus after treatment for 85 days. Bars are the log transformation of the mean ± standard error. A 3-factor ANOVA (˛ = 0.05) was used for statistical analysis, thus direct comparisons can be made between treatments, sexes, and tissues. Bars that do not have a letter in common (A, B, C, or D) are significantly different. The table denotes the p-values for each statistical factor individually as well as the various crosses in the 3-factor ANOVA.

Fig. 3. The average hepatic (A) total glutathione and (B) lipid peroxidation levels in female (open boxes) and male (filled boxes) F. heteroclitus after 85 days of continuous dietary exposure to CdSe/ZnS nanocrystals. Bars are mean ± standard error.

females. Though statistically similar, males showed a trend toward increased MDA levels following exposure, whereas female MDA levels did not show a dose-dependent response. 3.3.2. Gene expression changes We looked for alterations in mRNA expression of genes involved with metal metabolism and oxidative stress pathways (Fig. 4A). ␤-actin — the normalizing gene — was not significantly altered by treatment (ANOVA, P = 0.293). Metallothionein (mt) expression was increased, though insignificant, in response to CdCl2 consumption (3.02 ± 0.7 fold induction versus 1.53 ± 0.4 in the SD + lecithin diet). This increased expression was not observed in QD treated fish (1.14 ± 0.2 fold induction). Expression of glutathione-s-transferase (gstmu), glutathione peroxidase (gpx), Cu/Zn-superoxide dismutase (sod1), and Mn-superoxide dismutase (sod2) were not significantly affected by QD or Cd exposure. However, sod1 was differentially induced between males and females. Trends indicate that females fed the high QD or CdCl2 diets up-regulated sod1, while expression in males did not change (Fig. 4B). Because of high individual variation, males and females in the CdCl2 treatment group had sod1 levels that were not statistically different from their

SD + lecithin controls, but sod1 expression was significantly different between the sexes (0.86 ± 0.2 fold induction in males versus 3.59 ± 0.8 fold induction in females fed the CdCl2 diet). Although not statistically different, the high QD females and males showed the same tendencies in sod1 transcription as fish receiving the CdCl2 diet. 3.4. Reproduction Evaluating reproduction on a spawn-to-spawn basis suggests reproductive effects due to dietary exposure (Fig. 5); however, the dramatic changes in fecundity over time were not statistically significant due to low sample size. All animals were fed equal amounts of food throughout the study and increases in baseline fecundity observed in the laboratory controls (standard diet) were due to the stabilized feeding regimen and progression of the breeding season. In comparison, the SD + lecithin group had an immediate increase in fecundity during the transition spawn, 1018% versus 219% in the SD, and was highly variable thereafter, ranging from 670% to 1967%, not including the final spawn. Lecithin likely hastened reproductive onset by rapidly improving maternal condition. Female growth and

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

B.

4.5

Fold Induction

4.0

*

High QD

3.5

Cd

3.0 2.5 2.0 1.5 1.0

4.0 3.0 2.0 1.0

0.5 0.0

sod1 5.0

SD+Lecithin

Fold Induction

A.

33

mt

gstmu

gpx

sod2

sod1

0.0

SD+Lec High QD

Cd

Fig. 4. Expression changes (A) in metallothionein (mt), glutathione-s-transferase (gstmu), glutathione peroxidase (gpx), Mn-superoxide dismutase (sod2), and Cu/Znsuperoxide dismutase (sod1) hepatic mRNA levels; and (B) sex-specific differences in sod1 expression in female (open boxes) and male (filled boxes) F. heteroclitus after 85 days of dietary exposure. Bars are mean fold induction ± standard error. Asterisks indicate statistical differences between the sexes, ˛ = 0.05.

hepatosomatic indices increased 167% and 173%, respectively in the SD + lecithin exposure group from that observed in the standard diet group. Although lecithin was shown to sharply increase fecundity, pairs fed the high QD diet had a negligible increase in fecundity during the transition (116%) and subsequently declined below the 100% observational rate for the remainder of the exposure. Fish fed the low QD and CdCl2 diets had an increase in fecundity during the transition (931% and 816%, respectively), similar to the SD + lecithin group (1018%), but egg numbers steadily declined post-transition and fell below 100% by the fourth spawn.

Table 4 Development and hatching of parentally exposed embryosa . Treatment

% Smallb

% Abnormalc

% Hatchedd

Standard diet SD + lecithin Low QD High QD CdCl2

2.8 3.9e 1.7f 3.5f 0.0f

1.6 6.7e 0.3f 3.1 f 2.2f

64 68e 84f 81f 98f

a b c d e

3.5. Vitellogenin transcription and gonadosomatic indices Vitellogenin (VTG) is the precursor of egg yolk proteins produced in the liver of oviparous females. Production of VTG normally occurs due to activation of the estrogen receptor by estrogens or environmental estrogen mimics (Miller et al., 2007). We examined the expression of vtg mRNA levels in male and female F. heteroclitus at the conclusion of the study. Hepatic vtg transcription (Fig. 6) was elevated, though statistically insignificant, in males consuming the high QD and CdCl2 diets (13.6 ± 10.5 and 9.3 ± 4.2 fold induction, respectively). These vtg levels were three to five times higher than the levels measured in the SD and SD + lecithin animals (2.0 ± 1.5 and 2.7 ± 1.0 fold induction, respectively). Males fed the QD or CdCl2 diets tended to have smaller testes, as indicated by gonadosomatic indices of 0.82 ± 0.28 for low QD, 0.93 ± 0.27 for high QD, and 1.05 ± 0.25 for CdCl2 diets than those receiving the SD + lecithin diet (1.21 ± 0.34) (see Supplemental Information 2). Female vtg transcription and gonadosomatic indices were not statistically different due to the extreme variability that occurs during reproductive cycles. 3.6. Maternal transfer and embryo development Analysis of eggs/embryos from the second and third treatment spawns had no detectable levels of cadmium regardless of the parental treatment regime. However, cadmium was detected in eggs/embryos from the fourth spawn whose parents received the high QD diet. Concentrations in these eggs ranged from below detectable levels to 0.7 ng Cd per sample with an average of 0.24 ± 0.1 ng Cd/sample, whereas samples from the control diets (SD and SD + lecithin) were below detection levels. Due to insufficient egg production, cadmium content was not analyzed in low QD or CdCl2 embryos after the 3rd treated spawn or in the high QD embryos after the 4th treated spawn.

f

Embryos were collected from spawns T, 2, and 3. No deformities, but are half the size of normal embryos. Includes lethal and sub-lethal deformities. Includes both normally and abnormally developed fry. Significantly different from the SD. Statistically different from the SD + lecithin treatment, ˛ = 0.05.

While the prevalence of abnormalities was low, lecithin in the adult diet resulted in a statistically significant increase in the number of abnormally sized and developed embryos during the first three treatment spawns compared to embryos from the SD treatment group. Embryos from parents fed QDs or CdCl2 exhibited significantly lower percentages of deformities than embryos from the SD + lecithin group (Table 4). Hatching percentages of embryos from the QD and CdCl2 exposures were significantly higher than hatching percentages in the SD + lecithin group. There was insufficient egg production from the final three treated spawns to collect developmental data. 4. Discussion At present, environmental concentrations of ENPs such as QDs are unknown as monitoring instrumentation is not capable of distinguishing between nano-sized and bulk particles. However, as the popularity of engineered nanomaterials grows, they will become ubiquitous environmental contaminants. In this study, we delivered CdSe/ZnS quantum dots to reproductively active adult fish through their diet. Our interests were to determine if dietary delivery of QDs impacted the fecundity of fish and whether the particles or components of the nanomaterials were transferred to the eggs. At the end of the feeding study, we quantified cadmium levels in intestine and liver tissues and determined if dietary exposures had an impact on metal metabolism, oxidative stress pathways, or vitellogenin transcription. Hoshino et al. (2004) reported that the exterior coating agents (mercaptoundecanoic acid and cysteamine), not the ZnS-coated CdSe QDs, caused cytotoxicity in Vero cells. Thus, we encapsulated

34

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

SD + Lecithin 5000

4000

4000

3000

3000

%

%

Standard Diet 5000

2000 1000 0

1000 O

T

2

3 4 Spawn

5

0

6

5000

4000

4000

3000

3000

%

5000

2000

T

O

T

2

3 4 Spawn

5

0

6

O

T

Cadmium

5

6

2

3 4 Spawn

5

6

5

6

High QD 200

4000

150

3000

%

%

3 4 Spawn

2000

5000

2000

100 50

1000 0

2

1000

1000 0

O

High QD

Low QD

%

2000

O

T

2

3 4 Spawn

5

6

0

O

T

2

3 4 Spawn

Fig. 5. Fecundity profiles shown as a percent of the observational (O) spawning rate (eggs/day). Fish were exposed from the transition spawn (T) through the sixth spawn. Note that in all the fecundity profiles the y-axes crosses at 100%, which is the observational spawning rate for that exposure group. The scale of the y-axis of the graph in the lower right corner (high QD) has been adjusted so that the spawning fluctuations that occurred near or below the observational spawning rate are apparent. Bars are mean ± std error.

the hydrophobic quantum dots in lecithin (phosphatidylcholine), a membrane phospholipid found in all organisms, to render them water-miscible for incorporation into the diet formulations. The lecithin was also added to our standard fish diet, SD + lecithin, to determine whether any toxicological effects observed or measured in the study were solely the result of the lecithin micelles. Lecithin is neither a reproductive toxicant, nor a mutagen, and is regarded as virtually non-toxic in oral animal studies; it is generally recognized as safe (GRAS) by the US Food and Drug Administration (Fiume, 2001). Lecithin supplementation in fish diets increases growth, lipid levels within the liver, and feeding efficiency (Kim et al., 2006). In our study, lecithin supplementation altered reproduction, development, and hatching (Fig. 5 and Table 4), thus statistical comparisons examining the effects of QD or CdCl2 exposure were made to the SD + lecithin treatment groups.

4.1. Bioavailability Cadmium analysis confirmed that QDs or their degradation products were taken up via diet and distributed to the liver, but not to the same extent as CdCl2 . While the high QD diet contained roughly half as much cadmium as the CdCl2 diet2 , fish fed the high QD diet had intestinal and liver tissue cadmium concentrations approximately 180 and 130 times lower than fish receiving

2 Cd content of the food discs was lower than expected and did not match the Cd levels in the CdCl2 discs as intended due to a technical oversight. The supplier altered the synthesis of Maple-Red CdSe/ZnS QDs between the preliminary (Lot # 062806 30AMB) and final studies (Lot #LN018ICS-B) thus affecting the overall composition (Siarous, 2008). This highlights the importance of testing each batch of ENPs individually to confirm composition.

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

diet is comparable to other sub-chronic dietary cadmium studies in fish (≤1% Cd accumulated) (Franklin et al., 2005; Harrison and Klaverkamp, 1989). Low retention occurs because a large amount of dietary Cd is bound in the digestive tract tissues (stomach, cecae, and intestines) (Chowdhury et al., 2004) or eliminated via feces (Harrison and Klaverkamp, 1989). Our results suggest that while cadmium accumulation from the diet was relatively low, the physiological responses to CdCl2 versus core–shell CdSe/ZnS QDs may have lead to bioavailability disparities.

vtg females

Fold Induction

200 150 100

4.2. Oxidative stress biomarkers

50 0

SD

SD+Lec.

Low QD High QD

Cd

vtg males 30

Fold Induction

35

25 20 15 10 5

0 SD

SD+Lec. Low QD High QD

Cd

Fig. 6. vtg mRNA levels in female and male F. heteroclitus after 85 days of dietary QD exposure. Bars are mean ± std error.

the Cd diet, respectively. Evaluation of liver preparations by TEM and confocal microscopy were inconclusive, so it remains to be determined if whole QDs were deposited to the liver or if the particles degraded and the cadmium was re-distributed to the liver. At present, there are only a few publications that have examined uptake of ZnS shelled QDs in aquatic organisms. Carboxylated CdSe/ZnS QDs were found in the digestive tracts of water fleas (Ceriodaphnia dubia) (Ingle et al., 2008) and rotifers (Brachionus calyciflorus) (Holbrook et al., 2008); however, tissue distribution was not examined, so it is not known if the QDs were assimilated. Polymer coated CdSe/ZnS QDs were trophically transferred from brine shrimp (Artemia franciscana) to zebrafish, but only about 1% of the QD steady-state dose was retained after a one week clearance period, thus biomagnification did not occur (Lewinski et al., 2011). The fluorescent signal from PEG-functionalized CdSe/ZnS QDs was not observed in aqueously exposed zebrafish embryos/larvae, but cadmium body burdens were elevated after 120 h of exposure confirming Cd uptake (King-Heiden et al., 2009). Karabanovas et al. (2008) used the fluorescent signal of CdSe/ZnS QDs to determine that whole QDs traversed the esophagus and duodenum when administered to Wistar rats. Core–shell QDs were also reportedly degraded by gastric fluids and enzymes in vitro (Karabanovas et al., 2008; Wiecinski et al., 2009). Taken together, this suggests that whole as well as degraded QDs pass into the body from the upper digestive tract. There are notable differences in the cadmium tissue concentrations and the total cadmium retained in the tissues of fish receiving the QD and CdCl2 diets. QD-treated fish retained less than 0.01% of the total dose administered in the intestine and liver (as it pertains to Cd content of the food within the tissues), while those fed the CdCl2 diet retained 0.9% of the total dose in the intestinal tissue and 0.5% in the liver. The amount of cadmium retained from the CdCl2

In vitro studies have reported the generation of ROS and oxidative stress as a consequence of exposure to shelled and unshelled QDs (Cho et al., 2007; Choi et al., 2007; Green and Howman, 2005; Ipe et al., 2005; Li et al., 2009; Lovric et al., 2005). Yet, little is known about the ability of QDs to generate ROS and cause oxidative stress in whole animal models. In our study, total glutathione and lipid peroxidation levels were not altered by QD exposure, which reflects the low uptake of QDs and/or their degradation products. This is consistent with a 35-day dietary exposure of F. heteroclitus to CdSe/ZnS QDs (Blickley, 2010) and a 21-day aqueous exposure of sticklebacks (Gasterosteus aculeatus) to CdS QDs (Sanders et al., 2008). In contrast, Gagne et al. (2008a) reported that LPo levels were significantly elevated in the gills of freshwater mussels acutely exposed to CdTe QDs. The conflicting results are likely due to differences in the type of QD, concentrations, exposure route and duration, and tissue type. The stability of QDs is a function of structure and environmental conditions. Cho et al. (2007) reported that unshelled CdTe QDs leached free Cd2+ , while ZnS-shelled QDs showed no detectable levels of free cadmium at 24 h. However, shelling QDs does not provide absolute protection. A ZnS shell delays, but does not prevent oxidative degradation of QDs (Derfus et al., 2004). The liberation of cadmium from degraded QDs could affect antioxidant and lipid peroxidation levels as free cadmium indirectly generates ROS (Waisberg et al., 2003). As such, unshelled QDs are expected to more effectively induce an oxidative stress response than shelled QDs once a threshold level of free cadmium is attained. The duration of the exposure influences oxidative stress biomarkers. In acute QD exposures, detoxification biomolecules such as metallothionein and glutathione are depleted, while lipid peroxidation levels are elevated (Gagne et al., 2008a,b; Li et al., 2009; Peyrot et al., 2009). In the only sub-chronic QD study published, glutathione disulfide levels in sticklebacks were not significantly altered by aqueous exposure to CdS QDs suggesting limited bioavailability and/or the organisms adapted to the insult (Sanders et al., 2008). With respect to cadmium exposure, changes in glutathione levels are dose and time-dependent. While Pathak and Khandelwal (2006) reported dose-dependent glutathione depletion in acute in vitro cadmium exposures, other acute and chronic Cd studies have shown transient changes in glutathione levels (Basha and Rani, 2003; Lange et al., 2002; Nzengue et al., 2008; Wolf and Baynes, 2007). This adaptive response occurs because metal contaminants initially stimulate the synthesis of additional antioxidants to quench ROS and then metallothionein to bind/sequester excess free metals. Intermediate time points were not examined in our study, thus early fluctuations in glutathione levels could not be observed. Certain forms of selenium have also been linked to the generation of superoxide anion and hydrogen peroxide (Spallholz, 1997; Spallholz et al., 2004; Yan and Spallholz, 1993). Although selenide is one of the major components of CdSe/ZnS QDs, we did not measure selenium in this study because it is an essential micronutrient in fish and accurate measurement is unlikely. While selenide is not a

36

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

form of Se that is linked to ROS generation, possible transformation into one of the toxic forms of selenium is acknowledged. 4.3. Gene expression changes As mentioned previously, oxidative stress results in the upregulation of a suite of ROS-related genes including thioredoxin, metallothionein, glutathione-s-transferase, glutathione peroxidase, and superoxide dismutases (Sheader et al., 2006; Waisberg et al., 2003). We determined that mt mRNA expression was not altered by exposure to the high QD diet, but that fish receiving the CdCl2 diet had twice the amount of mt mRNA as the SD + lecithin diet. This slight increase in mt transcription is supported by other investigations (Lange et al., 2002; Olsson et al., 1989, 1995; Sheader et al., 2006; Wangsongsak et al., 2007). Cadmium’s affinity for metallothionein and its ability to displace zinc in the configuration of metallothionein is well known, as is the increase in metallothionein production that occurs when labile Zn is subsequently increased. Roesijadi et al. (2009) reported that dietary cadmium exposure of F. heteroclitus for 7 days resulted in significantly increased intestinal mt mRNA, but not hepatic mt mRNA levels. It is probable that like Roesijadi et al. (2009), our F. heteroclitus livers had not accumulated sufficient cadmium to induce a significant change in metallothionein transcription. Alternately, recent studies by Asselman et al. (2012, 2013) show that metallothionein mRNA transcription was both timeand homolog-dependent in Daphnia pulex when exposed to various environmental stressors. Exposure to 0.5 ␮g Cd/l for 16 days significantly induced mt1 and mt3 gene transcription on day 4 and mt4 transcription on day 8; however, mRNA expression was not statistically different from controls at termination on day 16. In our study, mt induction was only evaluated in one sequence after 85 consecutive days of dietary exposure. As additional mt homologs are identified in F. heteroclitus, it would be prudent to evaluate induction in response to acute QD exposure. Although none of the oxidative stress-related genes were significantly altered by QD or Cd exposure, there were sex-specific differences in sod1 mRNA levels. Females fed the high QD and Cd diet had elevated sod1 levels, though not statistically different, whereas males had sod1 mRNA levels that were consistent with that of the SD + lecithin males. This increase in sod1 could explain the unusual trends observed in female LPo levels. Though we did not measure actual SOD1 protein levels, if they were elevated as suggested by the upregulation of mRNA expression, this would combat lipid peroxidation by quenching radicals and altering lipid radical formation. 4.4. Reproduction Reproduction is an energetically demanding physiological process. In the natural environment, specifically in North Carolina, F. heteroclitus spawn multiple batches of eggs from April through September on a semi-lunar cycle. F. heteroclitus fecundity is variable from individual to individual, and is dependent on a variety of factors including the age of the fish, nutritional status, temperature, salinity, and photoperiod (Kamler, 2005). Our results suggest that lecithin may have hastened reproductive onset by improving maternal condition. Female growth and hepatosomatic indices (Figure S4) were approximately 1.7 times higher in the SD + lecithin exposure group versus the standard diet group indicating that lecithin served as a nutrient by improving both body and relative liver size. Significantly improved growth was reported in shrimp (Penaeus vannamei) fed diets supplemented with 1.5% soybean lecithin (Coutteau et al., 2000) and pikeperch (Sander lucioperca) fed increasing amounts of soybean lecithin (1.4–9.5% of dry matter) (Hamza et al., 2008). A similar relationship appears to exist

with the F. heteroclitus employed in this study, where the lecithin only constituted 0.01% of the dietary dry matter. Roberts et al. (2007) suggested that Daphnia magna exposed to lysophophatidylcholine-coated single-walled nanotubes (LPCSWNTs) utilized the digested lipid as a food source at concentrations up to 0.5 mg LPC-SWNTs/l. At concentrations greater than 0.5 mg/l, the nutritional benefit of the LPC was overcome by the toxicity of the SWNTs as evidenced by increased mortality. Benefits observed by the addition of lecithin in our study were also negated by chronic dietary exposure to CdSe/ZnS QDs or cadmium. The level of egg production after the transition period decreased in the low QD, high QD, and CdCl2 exposures indicating a toxic and possible endocrine mediated response. 4.5. Vitellogenin transcription Based on the increase in vtg transcription observed in male F. heteroclitus in this study, it appears that QD and Cd ingestion potentially feminized the male fish. This differs from a previously published report where aqueous exposure to methyl polyethylene glycol coated-CdS QDs (0 to 500 ␮g/l nCdS) for 21 days had no effect on vitellogenin expression in male sticklebacks (Sanders et al., 2008). Exposure of the sticklebacks to the MPEG-coated CdS QDs did produce other endocrine disrupting effects namely a reduction in the number of male fish building nests. Sanders et al. (2008) hypothesized that Cd2+ concentrations used in their studies were below the physiological threshold required to induce a change in vitellogenin levels, but high enough to elicit behavioral changes. In other studies, Haux et al. (1988) reported that aqueous exposure to 100 ␮g/l Cd for 4 weeks decreased plasma vitellogenin levels in female rainbow trout (Oncorhynchus mykiss). Tilton et al. (2003) reported no changes in plasma vitellogenin levels in male or female Japanese medaka (Oryzias latipes) exposed to Cd for 7 weeks at concentrations ranging from 1 to 10 ␮g/l. However, they did observe significant changes in gonadal steroid release in both sexes with 5 ␮g/l Cd increasing plasma E2 while 10 ␮g/l resulted in a decrease in plasma E2. These studies were similar to Thomas (1989), who reported increases in circulating estradiol levels in Atlantic croaker exposed to 1 mg/l Cd. Together these studies imply a species differential in cadmium sensitivity and a concentration dependent response occurring along the hypothalamus-pituitarygonadal (HPG) axis following cadmium exposure (Tilton et al., 2003). It is probable this response occurs due to binding and activation of the estrogen receptor as reported by Stoica et al. (2000a). They demonstrated that cadmium interacts with the hormonebinding domain of ER-␣ resulting in a 4-fold increase in reporter gene activity. While our study focused on the cadmium in QDs, it is possible that selenide contributed to the changes in vtg transcription. In a companion to their previous study, Stoica et al. (2000b) demonstrated that selenium compounds, including selenite, also interact with the ER-␣ hormone-binding domain eliciting a three to fivefold increase in reporter gene activity. Although Se concentrations were not measured in the adult F. heteroclitus, it is possible that if QDs were degrading in vivo, selenium could contribute to endocrine disruption. Further studies including quantification of circulating VTG levels are necessary to confirm an effect. 4.6. Embryonic development and maternal transfer Early life stages are usually more susceptible to toxicity, and alterations in embryonic development are often a sentinel for contaminant exposure. King-Heiden et al. (2009) showed that polyethylene glycol (PEG5000 ) coated CdSe/ZnS QDs (expressed as ␮M Cd equivalents) were an order of magnitude more toxic than CdCl2 in aqueous exposures of zebrafish. The zebrafish

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

larvae developed numerous morphological defects typical of aqueous cadmium exposure, including edema, bent spine, tail malformation, and yolk sac malformation (Cheng et al., 2000). This suggests that exposure to unaggregated QDs produces deleterious consequences during development. In our study, the QD aggregates first passed through the adult F. heteroclitus digestive system, which may have impeded deposition to the developing embryos and/or enzymatically degraded the QDs and their coating, thus releasing their teratogenic constituents. However, no significant malformations indicative of cadmium (Cheng et al., 2000) or selenium toxicity (Pyron and Beitinger, 1989) were noted in our studies. While QD exposure did not cause teratogenic effects, cadmium was detected in embryos that were spawned during the fourth spawn (6–8 weeks of parental exposure), but not before. Pharmacokinetic studies using rodents have shown that QDs accumulate within the liver (Fischer et al., 2006; Yang et al., 2007). The fish liver is where the egg yolk protein precursor vitellogenin is synthesized prior to deposition into the maturing ova. Numerous studies have shown that vitellogenin can carry environmental contaminants such as metals and organics into the developing ova (Ghosh and Thomas, 1995; Monteverdi and Di Giulio, 2000a). Ovo-deposition of contaminants, including metals, functions as a non-specific means for decreasing the contaminant load of many female oviparous animals (Schultz and Hermanutz, 1990; Hall and Oris, 1991; Monteverdi and Di Giulio, 2000b; Hammerschmidt and Sandeinrich, 2005; Sellin and Kolok, 2006; Nyholm et al., 2008). Recent studies have shown that certain types of ENPs are bioavailable, and that they are transferred from parent to the offspring (Tao et al., 2009; Lin et al., 2009; Snell and Hicks, 2010; Meyer et al., 2010). Snell and Hicks (2010) observed that 37 nm unreactive fluorescent nanoparticles passed through the intestinal wall of rotifers (Branchionus manjavacas) and were deposited into the eggs. Similarly, Meyer et al. (2010) reported that nematodes (Caenorhabditis elegans) ingested citrate-coated nano-silver with food and the internalized particles were detected in the internally developing embryos. 5. Conclusion In this dietary study, CdSe/ZnS quantum dots that were consumed by fish were bioavailable, but did not elicit a statistically significant change in the oxidative stress response. Declining trends in fecundity suggest that while estuarine fish may be able to tolerate acute dietary exposure to QDs, chronic QD exposure could have consequences at the population level. QD uptake, though limited, and maternal transfer of QDs or their degradation products to developing progeny may pose a threat to future generations of aquatic organisms. Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. Acknowledgements We thank Dr. R. David Holbrook of the National Institute of Standards & Technology (NIST) in Gaithersburg, MD for performing the ESEM/EDX and confocal microscopy. This work was conducted with assistance from the Oak Foundation, Duke University Marine Laboratory, and NC State University Center for Marine Sciences and Technology (CMAST). This research was also partially supported by the National Science Foundation and the Environmental Protection Agency under NSF Cooperative Agreement EF-0830093, Center for the Environmental Implications of NanoTechnology (CEINT)

37

(CM, RD). 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 National Science Foundation or the Environmental Protection Agency. This work has not been subjected to EPA review and no official endorsement should be inferred. T.M. Blickley was supported by the Society of Environmental Toxicology and Chemistry (SETAC) Doctoral Fellowship, sponsored by the Procter & Gamble Company. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.aquatox. 2013.12.021. References Andersen, M.E., 1985. Determination of glutatione and glutathione disulfide in biological samples. Method Enzymol. 113, 548–555. Armstrong, P.B., Child, J.S., 1965. Stages in the normal development of Fundulus Heteroclitus. Biol. Bull. 128, 143–167. Asselman, J., Galholt, S.P., Smith, Z., Smagghe, G., Janssen, C.R., Colbourne, J.K., Shaw, J.R., De Schamphelaere, K.A., 2012. Functional characterization of four metallothionein genes in Daphnia pulex exposed to environmental stressors. Aquat. Toxicol. 110–111, 54–65. Asselman, J., Shaw, J.R., Galholt, S.P., Colbourne, J.K., De Schamphelaere, K.A., 2013. Transcription patterns of genes encoding four metallothionein homologs in Daphnia pulex exposed to copper and cadmium are time-and homologdependent. Aquat. Toxicol. 142–143, 422–430. Basha, P.S., Rani, A.U., 2003. Cadmium-induced antioxidant defense mechanism in freshwater teleost Oreochromis mossambicus (Tilapia). Ecotoxicol. Environ. Saf. 56, 218–221. Benn, T.M., Westerhoff, P., 2008. Nanoparticle silver released into water from commercially available sock fabrics. Environ. Sci. Technol. 42, 4133–4139. Blickley, T.M., McClellan-Green, P., 2008. Toxicity of aqueous fullerene in adult and larval Fundulus heteroclitus. Environ. Toxicol. Chem. 27, 1964–1971. Blickley, T.M., 2010. The Toxicological Effects of Engineered Nanoparticles, Qunatum Dots, in Estuarine Fish. Doctoral dissertation. Duke University, Durham, NC. Burnett, K.G., Bain, L.J., Baldwin, W.S., Callard, G.V., Cohen, S., Di Giulio, R.T., Evans, D.H., Gomez-Chiarri, M., Hahn, M.E., Hoover, C.A., Karchner, S.I., Katoh, F., MacLatchy, D.L., Marshall, W.S., Meyer, J.N., Nacci, D.E., Oleksiak, M.F., Rees, B.B., Sinder, T.D., Stegeman, J.J., Towle, D.W., Van Veld, P.A., Vogelbein, W.K., Whitehead, A., Winn, R.N., Crawford, D.L., 2007. Fundulus as the primier teleost model in environmental biology: opportunities for new insights using genomics. Comp. Biochem. Physiol. D 2, 257–286. Cheng, S.H., Wai, A.W.K., So, C.H., Wu, R.S.S., 2000. Cellular and molecular basis of cadmium-induced deformities in zebrafish embryos. Environ. Toxicol. Chem. 19, 3024–3031. Cho, S.J., Maysinger, S., Jain, M., Roder, B., Hackbarth, S., Winnik, F.M., 2007. Longterm exposure to CdTe quantum dots causes functional impairments in live cells. Langmuir 23, 1974–1980. Choi, A.O., Cho, S.J., Desbarats, J., Lovric, J., Maysinger, D., 2007. Quantum dot-induced cell death involves Fas upregulation and lipid peroxidation in human neuroblastoma cells. J. Nanobiotechnol. 5, Available: http://www.jnanobiotechnology.com/contents/5/1/1. Chow, E.S.K., Cheng, S.H., 2003. Cadmium affects muscle type development and axon growth in zebrafish embryonic somitogenesis. Toxicol. Sci. 73, 149–159. Chowdhury, M.J., McDonald, D.G., Wood, C.M., 2004. Gastrointestinal uptake and fate of cadmium in rainbow trout acclimated to sublethal dietary cadmium. Aquat. Toxicol. 69, 149–163. Coutteau, P., Kontara, E.K.M., Sorgeloos, P., 2000. Comparison of phosphatidylcholine purified from soybean and marine fish roe in the diet of postlarval Penaeus vannamei. Aquaculture 181, 331–345. Cumberland, S.A., Lead, J.R., 2009. Particle size distribution of silver nanoparticles at environmentally relevant conditions. J. Chromotogr. A 1216, 9099–9105. Derfus, A.M., Chan, W.C.W., Bhatia, S.N., 2004. Probing the cytotoxicity of semiconductor dots. Nano Lett. 4, 11–18. Fischer, H.C., Liu, L., Pang, K.S., Chan, W.C.W., 2006. Pharmacokinetics of nanoscale quantum dots: in vivo distribution, sequestration, and clearance in the rat. Adv. Funct. Mater. 16, 1299–1305. Franklin, N.M., Glover, C.N., Nicol, J.A., Wood, C.M., 2005. Calcium/cadmium interaction at uptake surfaces in rainbow trout: waterborne versus dietary routes of exposure. Environ. Toxicol. Chem. 24, 2954–2964. French, R.A., Jacobson, A.R., Kim, B., Isley, S.L., Penn, R.L., Baveye, P.C., 2009. Influence of ionic strength, pH, and cation valence on aggregation kinetics of titanium dioxide nanoparticles. Environ. Sci. Technol. 43, 1354–1359. Fiume, Z., 2001. Final report on the safety assessment of lecithin and hydrogenated lecithin. Int J. Toxicol. 20, 21–45. Gagne, F., Auclair, J., Turcotte, P., Fournier, M., Gagnon, C., Suave, S., Blaise, C., 2008a. Ecotoxicity of CdTe quantum dots to freshwater mussels: Impacts on immune system, oxidative stress and genotoxicity. Aquat. Toxicol. 86, 333–340.

38

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39

Gagne, F., Maysinger, D., Andre, C., Blaise, C., 2008b. Cytotoxicity of aged cadmiumtelluride quantum dots to rainbow trout hepatocytes. Nanotoxicology 2, 113–120. Ghosh, P., Thomas, P., 1995. Binding of metals to red drum vitellogenin and incorporation into ooctes. Mar. Environ. Res. 39, 165–168. Green, M., Howman, E., 2005. Semiconductor quantum dots and free radical induced DNA nicking. Chem. Commun., 121–123 [serial online]. Available: http://rsc.org/chemcomm on the INTERNET. Accessed 2011 March 18. Hall, A.T., Oris, J.T., 1991. Anthracene reduces reproductive potential and is maternally transferred during long-term exposure in fathead minnows. Aquat. Toxicol. 19, 249–264. Hammerschmidt, C.R., Sandeinrich, M.B., 2005. Maternal diet during oogenesis is the major source of methylmercury in fish embryos. Environ. Sci. Technol. 39, 3580–3584. Hamza, N., Mhelti, M., Khemis, I.B., Cahu, C., Kestemont, P., 2008. Effect of dietary phospholipid levels on performance, enzyme activities, and fatty acid composition of pikeperch (Sander lucioperca) larvae. Aquaculture 275, 274–282. Harrison, S.E., Klaverkamp, J.F., 1989. Uptake, elimination, and tissue distribution of dietary and aqueous cadmium by rainbow trout (Salmo gairdneri) and lake whitefish (Coregonus clupeaformis). Environ. Toxicol. Chem. 8, 87–97. Haux, C., Bjornsson, B.T., Forlin, L., 1988. Influence of cadmium exposure on plasma calcium, vitellogenin and calcitonin in vitellogenic rainbow trout. Mar. Environ. Res. 24, 199–202. Holbrook, R.D., Murphy, K.E., Morrow, J.B., Cole, K.D., 2008. Trophic transfer of nanoparticles in a simplified invertebrate food web. Nat. Nanotechnol. 3, 352–355. Hoshino, A., Fujioka, K., Oku, T., Suga, M., Saski, Y.F., Ohta, T., Yasuhara, M., Suzuki, K., Yamamoto, K., 2004. Physiocochemical properties and cellular toxicity of nanocrystal quantum dots depend on their surface modification. Nano Lett. 4, 2163–2169. Ingle, T.M., Alexander, R., Bouldin, J., Buchanan, R.A., 2008. Absorption of semiconductor nanocrystals by the aquatic invertebrate Ceriodaphnia dubia. Bull. Environ. Contam. Toxicol. 81, 249–252. Ipe, B.I., Lehnig, M., Niemeyer, C.M., 2005. On the generation of free radical species from quantum dots. Small 1, 706–709. Jiang, J., Oberdorster, G., Biswas, P., 2009. Characterization of size, surface charge, and agglomeration state of nanoparticle dispersions for toxicological studies. J Nanopart. Res. 11, 77–89. Kaegi, R., Ulrich, A., Sinnet, B., Vonbank, R., Wichser, A., Zuleeg, S., Simmler, H., Brunner, S., Vonmont, H., Burkhardt, M., Boller, M., 2008. Synthetic TiO2 nanoparticle emission from exterior facades into the aquatic environment. Environ. Pollut. 156, 233–239. Kamler, E., 2005. Parent-egg-progeny relationships in teleost fishes: an energetics perspective. Rev. Fish. Biol. Fish. 15, 399–421. Karabanovas, V., Zakarevicius, E., Sukackaite, A., Streckyte, G., Rotomskis, R., 2008. Examination of the stability of hydrophobic (CdSe)ZnS quantum dots in the digestive tract of rats. Photochem. Photobiol. Sci. 7, 725–729. Kim, K.D., Kim, K.M., Kim, K.W., Kang, Y.J., Lee, S.M., 2006. Influence of lipid level and supplemental lecithin in diet on growth, feed utilization and body composition of juvenile flounder (Paralichthys olivaceus) in suboptimal water temperatures. Aquaculture 251, 484–490. King-Heiden, T.C., Wiecinski, P.N., Mangham, A.N., Metz, K.M., Nesbit, D., Pedersen, J.A., Hamers, R.J., Heideman, W., Peterson, R.E., 2009. Quantum dot nanotoxicity assessment using zebrafish embryo. Environ. Sci. Technol. 43, 1605–1611. Kirchner, C., Liedl, T., Kudera, S., Pellegrino, T., Javier, A.M., Gaub, H.E., Stolzle, S., Fertig, N., Parak, W.J., 2005. Cytotoxicity of colloidal CdSe and CdSe/ZnS nanoparticles. Nano Lett. 5, 331–338. Lange, A., Ausseil, O., Segner, H., 2002. Alterations of tissue glutatione levels and metallothionein mRNA in rainbow trout during single and combined exposure to cadmium and zinc. Comp. Biochem. Physiol. C 131, 231–243. Lewinski, N.A., Zhu, H., Ouyang, C.R., Conner, G.P., Wagner, D.S., Colvin, V.L., Drezek, R.A., 2011. Trophic transfer of amphiphilic polymer coated CdSe/ZnS quantum dots to Danio rerio. Nanoscale 3, 3080–3083. Li, K.G., Chen, J.T., Bai, S.S., Wen, X., Song, S.Y., Yu, Q., Li, J., Wang, Y.Q., 2009. Intracellular oxidative stress and cadmium ions release induce cytotoxicity of unmodified cadmium sulfide quantum dots. Toxicol. In Vitro 23, 1007–1013. Liang, J., He, Z., Zhang, S., Huang, S., Ai, Z., Yang, H., Han, H., 2007. Study of DNA damage induced by CdSe quantum dots using nucleic acid molecular light switches as probe. Talanta 71, 1675–1678. Lin, S., Reppert, J., Hu, Q., Hudson, J.S., Reid, M.L., Ratnikova, T.A., Rao, A.M., Luo, H., Ke, P.C., 2009. Uptake, translocation, and transmission of carbon nanomaterials in rice plant. Small 5, 1128–1132. Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2-CT method. Methods 25, 402–408. Lovric, J., Cho, S.J., Winnik, F.M., Maysinger, D., 2005. Unmodified cadmium telluride quantum dots induce reactive oxygen species formation leading to multiple organelle damage and cell death. Chem. Biol. 12, 1227–1234. Meyer, J.N., Lord, C.A., Yang, X.Y., Turner, E.A., Badireddy, A.R., Marinakos, S.M., Chilkoti, A., Wiesner, M.R., Auffan, M., 2010. Intracellular uptake and associated toxicity of silver nanoparticles in Caenorhabditis elegans. Aquat. Toxicol. 100, 140–150. Miller, D.H., Jensen, K.M., Villeneuve, D.L., Kahl, M.D., Makynen, E.A., Durhan, E.J., Ankley, G.T., 2007. Linkage of biochemical responses to population-level effects:

a case study with vitellogenin in the fathead minnow (Pimephales promelas). Environ. Toxicol. Chem. 26, 521–527. Monteverdi, G.H., Di Giulio, R.T., 2000a. In vitro and in vivo associateion of 2,3,7,8tetrachlorinodibenzo-p-dioxin and benzo[a]pyrene with the yolk-precursor protein vitellogenin. Environ. Toxicol. Chem. 19, 2502–2511. Monteverdi, G.H., Di Giulio, R.T., 2000b. Oocytic accumulation and tissue deposition of 2,3,7,8-tetrachlorodibenzene-p-dioxin and benzo[a]pyrene in gravid Fundulus heteroclitus. Environ. Toxicol. Chem. 19, 2512–2518. Navarro, D.A., Watson, D.F., Aga, D.S., Banerjee, S., 2009. Natural organic mattermediated phase transfer of quantum dots in the aquatic environment. Environ. Sci. Technol. 43, 677–682. Nyholm, J.R., Norman, A., Norrgren, L., Haglund, P., Andersson, P.L., 2008. Maternal transfer of brominated flame retardants in zebrafish (Danio rerio). Chemosphere 73, 203–208. Nzengue, Y., Steiman, R., Garrel, C., Lefebvre, E., Guiraud, P., 2008. Oxidative stress and DNA damage induced by cadmium in the human keratinocyte HaCaT cell line: role of glutathione in the resistance to cadmium. Toxicology 243, 193–206. Oberdorster, E., Zhu, S., Blickley, T.M., McClellan-Green, P., Haasch, M.L., 2006. Ecotoxicology of carbon-based engineered nanoparticles: effects of fullerene on aquatic organisms. Carbon 44, 1112–1120. Ohkawa, H., Ohishi, N., Yagi, K., 1979. Assay for lipid peroxides in animal tissues by thiobarbituric acid reaction. Anal. Biochem. 95, 351–358. Olsson, P.-E., Kling, P., Petterson, C., Silversand, C., 1995. Interaction of cadmium and oestradiol-17B on metallothionein and vitellogenin synthesis in rainbow trout (Oncorhynchus mykiss). Biochem. J. 307, 197–203. Olsson, P.-E., Larsson, A., Maage, A., Haux, C., Bonham, D., Zafarullah, M., Gedamu, L., 1989. Induction of metallothionein synthesis in rainbow trout, Salmo gairdneri, during long-term exposure to waterborne cadmium. Fish. Physiol. Biochem. 6, 221–229. Orbea, A., Ortiz-Zarragoitia, M., Sole, M., Porte, C., Cajaraville, M.P., 2002. Antioxidant enzymes and peroxisome proliferation in relation to contaminant body burdens of PAHs and PCBs in bivalve mollusks, crabs, and fish from the Urdaibai and Plentzia estuaries (Bay of Biscay). Aquat. Toxicol. 58, 75–98. Pathak, N., Khandelwal, S., 2006. Influence of cadmium on murine thymocytes: Potentiation of apoptosis and oxidative stress. Toxicol. Lett. 165, 121–132. Peyrot, C., Gagnon, C., Gagne, F., Willkinson, K.J., Turcotte, P., Sauve, S., 2009. Effects of cadmium telluride quantum dots on cadmium bioaccumulation and metallothionein production to the freshwater mussel, Elliptio complanata. Comp. Biochem. Physiol. C 150, 246–251. Pyron, M., Beitinger, T.L., 1989. Effect of selenium on reproductive behavior and fry of fathead minnows. Bull. Environ. Contam. Toxicol. 42, 609–613. Ringwood, A.H., Hoguet, J., Keppler, C.J., Gielazyn, M.L., Ward, B.P., Rourk, A.R., 2003. Cellular Biomarkers (Lysosomal Destabilization, Glutathione & Lipid Peroxidation) in Three Common Estuarine Species: A Methods Handbook. Available: http://ciceet.unh.edu/powerpoint/ringwood-methods.pdf via the INTERNET. Accessed 2011 Mar 18. Roberts, A.P., Mount, A.S., Seda, B., Souther, J., Qiao, R., Lin, S., Ke, P.C., Rao, A.M., Klaine, S.J., 2007. In vivo biomodification of lipid-coated carbon nanotubes by Daphnia magna. Environ. Sci. Technol. 41, 3025–3029. Roesijadi, G., Rezvankhah, S., Perez-Matus, A., Mitelberg, A., Torruellas, K., Van Veld, P.A., 2009. Dietary cadmium and benzo(a)pyrene increased intestinal metallothionein expression in the fish Fundulus heteroclitus. Mar. Environ. Res. 67, 25–30. Sanders, M.B., Sebire, M., Sturve, J., Christian, P., Katsiadaki, I., Lyons, B.P., Sheahan, D., Weeks, J.M., Feist, S.W., 2008. Exposure of sticklebacks (Gasterosteus aculeatus) to cadmium sulfide nanoparticles: Biological effects and the importance of experimental design. Mar. Environ. Res. 66, 161–163. Schultz, R., Hermanutz, R., 1990. Transfer of toxic concentrations of selenium from parent to progeny in fathead minnow (Pimephales promelas). Bull. Environ. Contam. Toxicol. 45, 568–573. Sellin, M.K., Kolok, A.S., 2006. Cadmium exposures during early development: do they lead to reproductive impairment in fathead minnows? Environ. Toxicol. Chem. 25, 2957–2963. Sheader, D.L., Williams, T.D., Lyons, B.P., Chipman, J.K., 2006. Oxidative stress response of European flounder (Platichthys flesus) to cadmium determined by a custom cDNA microarray. Mar. Environ. Res. 62, 33–44. Siarous, N., Jan. 17, 2008. Personal communication. Snell, T.W., Hicks, D.G., 2010. Assessing toxicity of nanoparticles using Brachionus manjavacas (Rotifera). Environ. Toxicol. 26, 146–152. Spallholz, J.E., Palace, V.P., Reid, T.W., 2004. Methioninase and selenomethionine but not Se-methylselenocysteine generate methylselenol and superoxide in an in vitro chemiluminescent assay: implications for the nutritional carcinostatic activity of selenoamino acids. Biochem. Pharm. 67, 547–554. Spallholz, J.E., 1997. Free radical generation by selenium compounds and their prooxidant toxicity. Biomed. Environ. Sci. 10, 260–270. Stoica, A., Katzenellenbogen, B.S., Martin, M.B., 2000a. Activation of estrogen receptor-a by the heavy metal cadmium. Mol. Endocrinol. 14, 545–553. Stoica, A., Pentecost, E., Martin, M.B., 2000b. Effects of selnite on estrogen receptora expression and activity in MCF-7 breast cancer cells. J. Cell. Biochem. 79, 282–292. Takeda, K., Suzuki, K.-I., Ishihara, A., Kubo-Irie, M., Fujimoto, R., Tabata, M., Oshio, S., Nihei, R., Ihara, R., Sugamata, M., 2009. Nanoparticles transferred from pregnant mice to their offspring can damage the genital and cranial nerve systems. J. Health Sci. 55, 95–102.

T.M. Blickley et al. / Aquatic Toxicology 148 (2014) 27–39 Tao, X., Fortner, J.D., Zhang, B., He, Y., Chen, Y., Hughes, J.B., 2009. Effects of aqueous stable fullerene nanocrystals (nC60) on Daphnia magna: evaluation of sub-lethal reproductive responses and accumulation. Chemosphere 77, 1482–1487. Taylor, H.E., 2001. Inductively Coupled Plasma Mass Spectrometry: Practices and Techniques. Academic Press, New York. Thomas, P., 1989. Effects of Aroclor 1254 and cadmium on reproductive endocrine function and ovarian growth in Atlantic croaker. Mar. Environ. Res. 28, 499–503. Tilton, S., Foran, C.M., Benson, W.B., 2003. Effect of cadmium on the reproductive axis of Japanese medaka (Oryzias latipes). Comp. Biochem. Physiol. C 136, 265–276. Tsuchiya, T., Oguri, I., Nakajima, Y., Miyata, N., 1996. Novel harmful effects of [60] fullerene on mouse embryos in vitro and in vivo. FEBS Lett. 393, 139–145. Waisberg, M., Joseph, P., Hale, B., Beyersmann, D., 2003. Molecular and cellular mechanisms of cadmium carcinogenesis. Toxicology 192, 95–117. Wangsongsak, A., Utarnpongsa, S., Druatrachue, M., Ponglikitmonkol, M., Pokethitiyook, P., Sumranwanich, T., 2007. Alterations of organ histopathology and metallothionein mRNA expression in silver barb, Puntius gonionotus during subchronic cadmium exposure. J. Environ. Sci. 19, 1341–1348. Wiecinski, P.N., Metz, K.M., Mangham, A.N., Jacobson, K.H., Hamers, R.J., Pedersen, J.A., 2009. Gastrointestinal biodurability of engineered nanoparticles: development on an in vitro assay. Nanotoxicology 3, 202–214.

39

Wills, L.P., Matson, C.W., Landon, C.D., Di Giulio, R.T., 2010. Characterization of recalcitrant CYP1 phenotype found in Atlantic killifish (Fundulus heteroclitus) inhabiting a Superfund site on the Elizabeth River, VA. Aquat. Toxicol. 99, 33–41. Wolf, M.B., Baynes, J.W., 2007. Cadmium and mercury cause an oxidative stressinduced endothelial dysfunction. BioMetals 20, 73–81. Yan, L., Spallholz, J.E., 1993. Generation of reactive oxygen species from the reaction of selenium compounds with thiols and mammary tumor cells. Biochem. Pharm. 45, 429–437. Yang, R.S.H., Chang, L.W., Wu, J.-P., Tsai, M.-H., Wang, H.-J., Kuo, Y.-C., Yeh, T.-K., Yang, C.S., Lin, P., 2007. Persistent tissue kinetics and redistribution of nanoparticles, quantum dot 705, in mice: ICP-MS quantitative assessment. Environ. Health Perspec. 115, 1339–1343. Yu, W.W., Qu, L.H., Guo, W.Z., Peng, X.G., 2003. Experimental determination of the extinction coefficient of CdTe, CdSe, and CdS nanocrystals. Chem. Mater. 15, 2854–2860. Zhang, W., Lin, K., Miao, Y., Dong, Q., Huang, C., Wang, H., Guo, M., Cui, X., 2012a. Toxicity assessment of zebrafish following exposure to CdTe QDs. J. Hazard Mater. 213–214, 413–420. Zhang, W., Lin, K., Sun, X., Dong, Q., Huang, C., Wang, H., Guo, M., Cui, X., 2012b. Toxicological effect of MPA-CdSe QDs exposure on zebrafish embryo and larvae. Chemosphere 89, 52–59.