Cytotoxicity evaluation of silica nanoparticlesusing fish cell lines

Nguyen T. K. Vo & Mary R. Bufalino & Kurtis D. Hartlen &

Vladimir Kitaev & Lucy E. J. Lee

Received: 19 August 2013 /Accepted: 2 December 2013 / Editor: T. Okamoto# The Society for In Vitro Biology 2013

Abstract Nanoparticles (NPs) have extensive industrial, bio-technological, and biomedical/pharmaceutical applications,leading to concerns over health risks to humans and biota.Among various types of nanoparticles, silica nanoparticles(SiO2 NPs) have become popular as nanostructuring, drugdelivery, and optical imaging agents. SiO2 NPs are highlystable and could bioaccumulate in the environment.Although toxicity studies of SiO2 NPs to human and mam-malian cells have been reported, their effects on aquatic biota,especially fish, have not been significantly studied. Twelve

adherent fish cell lines derived from six species (rainbowtrout, fathead minnow, zebrafish, goldfish, haddock, andAmerican eel) were used to comparatively evaluate viabilityof cells by measuring metabolic impairment using AlamarBlue. Toxicity of SiO2 NPs appeared to be size-, time-,temperature-, and dose-dependent as well as tissue-specific.However, dosages greater than 100 μg/mL were needed toachieve 24 h EC50 values (effective concentrations needed toreduce cell viability by 50%). Smaller SiO2 NPs (16 nm) wererelatively more toxic than larger sized ones (24 and 44 nm)and external lining epithelial tissue (skin, gills)-derived cellswere more sensitive than cells derived from internal tissues(liver, brain, intestine, gonads) or embryos. Higher EC50

values were achieved when toxicity assessment was per-formed at higher incubation temperatures. These findings arein overall agreement with similar human and mouse cellstudies reported to date. Thus, fish cell lines could be valuablefor screening emerging contaminants in aquatic environmentsincluding NPs through rapid high-throughput cytotoxicitybioassays.

Keywords Alamar Blue . Cytotoxicity . Fish cell lines .

Nanotoxicology . Silica nanoparticles

Introduction

Engineered nanomaterials (ENMs—particles that have at leastone dimension <100 nm) and nanoparticles (NPs—particlesthat are <100 nm in all three dimensions) have become ofpopular research and industrial interest due to their novelphysicochemical properties beyond their bulk material coun-terparts (Wilson 2002; McIntyre 2012). Their unique proper-ties have allowed ENMs and NPs to be used in a wide range of

N. T. K. Vo :M. R. Bufalino : L. E. J. LeeDepartment of Biology, Wilfrid Laurier University,75 University Avenue West, Waterloo, ON N2L 3C5, Canada

N. T. K. Vo :K. D. Hartlen :V. KitaevDepartment of Chemistry, Wilfrid Laurier University,75 University Avenue West, Waterloo, ON N2L 3C5, Canada

L. E. J. Lee (*)Faculty of Science, University of the Fraser Valley,Abbotsford, BC, Canada V2S 7M8e-mail: lucy.lee@ufv.ca

Present Address:N. T. K. VoDepartment of Biology, University of Waterloo,Waterloo, ON, Canada

Present Address:M. R. BufalinoDepartment of Medical Biophysics, University of Toronto,Toronto, ON, Canada

Present Address:K. D. HartlenImperial Oil Ltd, Sarnia, ON, Canada

In Vitro Cell.Dev.Biol.—AnimalDOI 10.1007/s11626-013-9720-3

industrial, biotechnological, biomedical, and pharmaceuticalapplications. However, health risk concerns for plausible tox-icity effects of these novel new materials on humans and biotahave not been fully investigated (Oberdorster et al. 2005).Among various types of NPs, silica nanoparticles (SiO2

NPs) have been extensively studied and widely used asnanostructuring, in common household components, drugdelivery, and optical imaging agents (Singh et al. 2009).SiO2 is the most abundant compound in the Earth’s crustand it is estimated that over one million tons of SiO2 NPsare used annually in consumer products (Jarvie et al. 2009a)and much of that is believed to get washed down the drainsinto the aquatic milieu (Jarvie et al. 2009b). These amorphousNPs are highly stable and could potentially bioaccumulate inthe environment with possible hazardous effects (Napierskaet al. 2010).

Toxicity evaluation of SiO2 NPs to human and mammaliancells had been previously reported both in vivo and in vitro(see comprehensive review by Napierska et al. 2010), butresults have been variable. In vivo studies found that SiO2

NPs were in general nontoxic (He et al. 2008; Kumar et al.2010) but could accumulate in liver, lung, and spleen tissuesand potentially damage liver functions (Xie et al. 2010).Generally, toxicity of SiO2 NPs was shown to be dose-,time-, and size-dependent in studies with human and mam-malian cells (Napierska et al. 2009, 2010; Yu et al. 2009) withhigh doses, longer exposures, and smaller particles being mosttoxic. Exposure to SiO2 NPs increased levels of reactiveoxygen species while decreasing glutathione levels, resultingin beyond-threshold oxidative stress (Eom and Choi 2009;Yang et al. 2009). Sun et al. (2011) also found that mitochon-drial damage leading to oxidative stress and apoptosis was thedirect cause of cytotoxicity by SiO2 NPs in human hepatomaHepG2 cells. Spherical SiO2 NPs of less than 200 nm wereshown to be able to enter the nucleus and bind to DNAphosphate backbone, causing DNA damage, and inhibitingDNA replication (Eom and Choi, 2009; Singh et al. 2009).Most studies on toxicity effects of NPs have been per-formed using mammalian models, and investigation ofNP effects on fish health is comparatively small (Handyet al. 2008; Casado et al. 2013). Ecotoxicological testsusing fish models could be helpful when linking toxico-logical testing of mammalian models (Bols et al. 1994)because most mammals are either primary or secondaryconsumers of aquatic species.

In the present study, we report the evaluation of the cyto-toxicity of SiO2 NPs using established fish cell lines. Weemployed a synthesis method previously developed byHartlen et al. (2008) to produce highly monodispersearginine-stabilized SiO2 NPs with standard deviation (SD) ofdiameter below 5%. Warheit (2010), as well as Napierskaet al. (2010) noted that additional to particle size, dose, andexposure times, the surface reactivity of NPs may play a key

role in defining NP toxicity; thus, the use of arginine-coatedSiO2 NPs for our study. For mammalian cells, Thomassenet al. (2010) recently evaluated lysine-coated SiO2 NPs toavoid the surface reactivity effects that commonly coatingagents like surfactants add to the toxic responses, and reportedthat increasing dosage and decreasing particle size were pri-marily linked to the toxic effects seen in the evaluated humanmacrophages and endothelial cells. Nanoparticle uptake isknown to be tissue-dependent (Lai et al. 2008), andnanoparticle-induced cytotoxity responses could also be sen-sitive to species and cell types. Thus, for comparison pur-poses, we evaluated 12 fish cell lines derived from differenttissues of six different species: rainbow trout (Oncorhynchusmykiss ), fathead minnow (Pimephales promelas ), goldfish(Carassius auratus ), zebrafish (Danio rerio ), haddock(Melanogrammus aeglefinius ), and American eel (Anguillarostrata ).

The toxicity of SiO2 NPs has not been widely evaluatedusing aquatic vertebrate models. To the best of our knowledge,aside from a few in vivo studies using zebrafish (Ramesh et al.2013), their embryos (Fent et al. 2010; Nelson et al. 2010),and medaka embryos and larvae (Lee et al. 2011), there areonly two SiO2 NP studies with fish cells in vitro. Theseinclude fathead minnow fibroblasts (Christen and Fent 2012)and a rainbow trout cell line (Casado et al. 2013). For othertypes of NPs, fish cell lines are becoming popular tools toassess effects, most notably titanium dioxide NPs (Reeveset al. 2008; Vevers and Jha 2008), tungsten carbide NPs(Kuhnel et al. 2009), silver nanoparticles (George et al.2012), and nanospheres (Wise et al. 2010). Moreover, anoutstanding advantage of using fish cell cultures overwhole fish animal testing is that experimental data iscollected and analyzed in a faster, simpler, cheaper, andmore high-throughput fashion. For this study, cell viabil-ity was evaluated by fluorometric means using AlamarBlue assay, which measured cellular metabolic activities(Dayeh et al. 2005).

Materials and Methods

Fish cell cultures. Twelve established fish cell lines derivedfrom various tissues of six different species were evaluated(Table 1). These cell lines have been characterized and pub-lished as denoted in the table except for RTBrain, EelBrain,and Glofish which are cell lines that are at least 6 yr old orolder and manuscripts are in preparation. All cell lines wereauthenticated to their respective species of origin by DNAbarcoding at the Biodiversity Institute of Ontario, Universityof Guelph as per Cooper et al. (2007). Cell lines were main-tained in Leibovitz L-15 media (Sigma-Aldrich, Oakville,ON, Canada) supplemented with 5–20% (v /v ) fetal bovineserum (FBS, Sigma), 1% (v /v ) penicillin/streptomycin

VO ETAL.

solution with 2 mM glutamine (10,000 units/mL penicillin,10 mg/mL streptomycin, Sigma), and 0.1% (v /v ) Tylosin(Sigma). Cells were subcultured every 1–2 wk using TryplE,a recombinant form of trypsin trademarked by Invitrogen(Carlsbad, CA), as described in Servili et al. (2009).

Synthesis of SiO2 NPs. Three SiO2 NP samples, designatedas Si120 (16 nm), Si168 (24 nm), and Si173 (44 nm), weresynthesized for this study. These SiO2 NPs were produced bya method developed by Hartlen et al. (2008). They weregenerally spherical and did not possess any characteristiccolor in the visible light/UV spectrum, although slight turbid-ity could be noted in stock solutions of larger diameter SiO2

NPs (Table 2). These NPs were quite distinct in their stabili-zation with a natural amino acid arginine and were producedin aqueous solution (Hartlen et al. 2008) at 2–3 mM arginine(stock SiO2 NP solutions). The highest arginine concentrationin the tested NPs (diluted solutions) was 20 μM, which isnontoxic to the cells, as the normal growth medium L-15normally contains 2.87 mM arginine in its formulation(Sigma). The particle sizes of the SiO2 NPs were determinedby either a scanning or transmission electron microscope(Hitachi S-5200) and representative electron micrographs of

these SiO2 NPs have been published (Hartlen et al. 2008).General characteristics of the selected SiO2 NPs are presentedin Table 2.

Standardizing the use of Alamar Blue assays in cytotoxicityassessment of SiO2 NPs. The turbidity of SiO2 NPs raisedconcerns that they could possibly interfere with absorptionand emission wavelengths during fluorescence events with theAlamar Blue (AB) assay. Both resazurin and resorufin, whichare respectively non-fluorescent Alamar Blue (nAB) dye andits fluorescent form (fAB), were tested. In order to obtainresorufin, an older flask of FHMT-W1 at passage 27 was useddue to its high metabolic activity previously determined byAB assay (data not shown). Media was aspirated and rinsedtwice with 5 mL of L-15/ex solution. The working solutioncontaining 9 mL of L-15/ex and 525.5 μL of the stock ABsolution was added into the culture flask and allowed toincubate for 45 min in the dark. WS was collected, sterilefiltered, covered with aluminum foil, and placed in the darkprior to use. Silica NPs (at 1:100 dilution to the stock concen-trations given in Table 2) were added into the wells of one 96-well microplate containing either nAB or fAB so that the finalconcentrations were 100–200 μg/mL in 100 μL/well. Once

Table 1 Description of fish cell lines used for cytotoxicity evaluation of SiO2 NPs. All cell lines were derived from normal adult tissues except for HEWand ZEB2J and had been established for a minimum of 4 yr or longer

Cell line Species Tissue origin Morphology Growth conditions Reference

% FBS Tculture (°C)

RTgill-W1 Rainbow trout Oncorhynchus mykiss Gill Epithelial 10 18 Bols et al. (1994)

RTgut-GC Intestine Epithelial 10 18 Kawano et al. (2010)

RTL-W1 Liver Epithelial 5 18 Lee et al. (1993)

RTBrain Brain Neuronal-like 20 21 Unpublished

FHML2-6 Fathead minnow Pimephales promelas Liver Epithelial 5 24 Lee et al. (2009)

FHMT-W1 Testis Epithelial 10 26 Vo et al. (2010)

GFB3C Goldfish Carassius auratus Cerebrum Glial-like 10 27 Lee et al. (2008)

GFSk-S1 Skin Epithelial 10 21 Lee et al. (1997)

GloFish Zebrafish Danio rerio Abdomen Epithelial 10 26 unpublished

ZEB2J Blastula Epithelial 10 26 Xing et al. (2008)

EelBrain American Eel Anguilla rostrata Brain Glial-like 10 21 Unpublished

HEW Haddock Melanogrammus aeglefinius Embryo Fibroblastic 10 21 Bryson et al. (2006)

Table 2 Silica nanoparticles used for toxicity testing. Turbidity level SiO2 NP samples were visually evaluated by number system on the scale of 0–10,where turbidity level 0 was equivalently compared to transparency of standard MiliΩ water and turbidity level 10 to commercial 2% fat milk

SiO2 NP sample Size (nm) Stock concentration(mg/mL)

Appearance of solution Turbidity level Solvent Other possible contaminants

Si120 16±4.9% 15.02 Clear 0 Water Ethanol, cyclohexane(in trace amount)Si168 24±4.5% 18.92 Clear 0

Si173 44±4.2% 15.62 Slightly cloudy 2

CYTOTOXICITY EVALUATION OF SILICA NPS IN FISH CELL CULTURES

ready, the microplate was processed for Alamar Blue assayand relative fluorescence unit (RFU) values obtained. AllRFU values of nAB and fAB were normalized to RFU read-ings of controls (without the addition of SiO2 NPs).

Assessment of cell viability with Alamar Blue assay. The SiO2

NPs were originally synthesized in water and used as originaldispersion, and were diluted 1:2 with a 2× L-15/ex media priorto further dilutions with 1× L-15/ex solution. The SiO2 NPsamples were sterile filtered using 10-mL sterile syringesfitted with 0.2-μm polyethersulfone (PES) or celluloseacetate-membrane sterile syringe filters (VWR Canlab,Mississauga, Canada). Dilutions of SiO2 NP solutions weremade in 50-mL sterile glass brown vials. The L-15/ex solutionwas prepared as described in Dayeh et al. (2005) and wassterile filtered via a 0.2-μm PES-membrane vacuum filtrationunits (VWR).

AB is the trade name for resazurin, which has been widelyused in cell viability assays (O’Brien et al. 2000).Metabolically active cells reduce the non-fluorescentresazurin to the fluorescent resorufin via intracellular ester-ases. AB was purchased from Invitrogen and the viabilityassay was performed as described by Dayeh et al. (2005).AB has been previously used to assess viability of fish celllines in toxicity assays, including the evaluation of tungstenNPs (Kuhnel et al. 2009). Cell lines were seeded at 50,000–60,000 cells/well in 96-well microplates so that confluentmonolayers could be achieved overnight. After 24 h, the wellswere washed with L-15/ex solution, a simplified L-15 formu-lation used for exposures, that contains no vitamins or aminoacids. SiO2 NPs at different concentrations (1.8–1,800 μg/mL) suspended in L-15/ex were added into six to eight repli-cate wells. After 24 h, the wells were rinsed with L-15/ex and100 μL of a working AB solution (525.5μL of AB solution in10 mL of L-15/ex) was applied to the plate wells. The assayplates were then incubated for an hour at room temperatureand RFUs were measured with a fluorescence microplatereader (Spectramax Gemini XS; Molecular Devices,Sunnyvale, CA). Excitation and emission wavelengths wereset at 530 and 590 nm, respectively. A dose–response curvewas obtained with the averaged RFUs and a 24 h EC50 wasthen computationally calculated using Prism (GraphPad, SanDiego, CA). One-way analysis of variance and Dunnett’s testcomparisons of control and treatments were used for signifi-cance testing if p values were less than 0.05 using InStat(GraphPad).

Temperature-dependent cytotoxicity of SiO2 NPs. The eval-uated fish cell lines were maintained at various incubationtemperatures (18–27°C) (Table 1). Slight temperature changescould increase or decrease the metabolic activities of the cells,thus facilitating or slowing down NP uptake into the cells andinfluencing toxicity outcome. SiO2 NPs are naturally stable at

the range of tested temperatures (18–27°C). Proof of thisstability, versatility, and biocompatibility was provided by arecent report in which proteins were encapsulated in SiO2 NPs(Cao et al. 2010) and functionality was demonstrated withencapsulated green fluorescent protein after desiccation at40°C and resuspension in water at room temperature.Because GFSk-S1 cells appeared to be the most sensitive toSiO2 NPs and could be grown at 27°C, they were used toinvestigate temperature effects. GFSk-S1 cells were dosed at50,000 cells/well. Two 96-well plates were prepared andincubated at 21°C for 48 h. Cells were exposed to Si120 atvarious concentrations because it was found that SiO2 NPs ofthis size was the most toxic among the tested sizes. One of twoplates was then incubated at room temperature and the otherone at 27°C for 24 h. Two plates were then evaluated for ABassay.

Effect of silica nanoparticles on the morphologies of fishcells . Cells were plated at three different cell densities(250,000, 125,000, and 50,000 cells/well) in 24-well micro-plates (Falcon). Cells were allowed to attach for 24 h in FBS-supplemented L-15. Media was then removed and wells wererinsed with 1 mL of L-15/ex solution. Cells were then appro-priately exposed to two different types of SiO2 NPs (Si120and Si173 samples) at three different concentrations (39, 195,and 390 μg/mL). Phase contrast micrographs were taken at 1,3, 12, and 24 h post-exposure using a Nikon T300 invertedmicroscope with a Nikon 990 Coolpix camera (Nikon, Tokyo,Japan).

Results

No fluorescence interference of SiO2 NPs in Alamar Blueassays. Concentrated stock solutions of large size SiO2 NPsdisplayed slight turbidity (Table 2); thus, the NP solutionswere evaluated for possible fluorescence interference.Dispersions of SiO2 NPs in L-15/ex alone served as negativecontrols (without cells). These had no fluorescence emissionon their own at the tested concentration and size range (Fig. 1).Exposure of the NP solutions to non-fluorescent and fluores-cent AB solutions showed no significant deviation in fluores-cence units (p >0.05) from the controls (Fig. 1). Thus, the NPsalone did not interfere with the viability assay fluorescenceemissions, making it applicable for assessing SiO2 NP toxicityin the tested fish cell lines. This assay had previously beenused for evaluating the toxicity of tungsten carbide NPs onRTgill-W1 cells (Kuhnel et al. 2009) and proved also usefulfor evaluating SiO2 NP toxicity with the 12 fish cell linestested.

Cytotoxicity of SiO2 NPs to fish cell lines . Effective concen-trations needed to reduce cell viability by 50% (24 h EC50) in

VO ETAL.

the tested fish cell lines with the three sized SiO2 NPs were allgreater than 100 μg/mL (Table 3), indicating very low cyto-toxicity under the guidelines of U.S. EnvironmentalProtection Agency (EPA)’s ecotoxicity classification level ofconcerns for aquatic organisms (US EPA 2012). Dose–re-sponse curves needed to generate full mortality required dos-ages in the milligram per milliliter range (Fig. 2), which areunlikely to be encountered in the aquatic environment, thusoverall, SiO2 NPs pose very low risk to aquatic biota.However, it is possible that fish experience mild surfaceirritation as cell lines derived from external lining epithelialtissues (skin and gills) were more sensitive to SiO2 NPs thancell lines derived from embryos and some internal tissues suchas brain, gonads, and intestine (Table 3). The ZEB2J cells (cellline derived from blastula stage zebrafish) (Xing et al.2008) were the most tolerant to SiO2 NPs, and GFSk-S1 (Goldfish skin cells) (Lee et al. 1997) appeared mostsensitive, while EelBrain cells gave inconsistent results due tocell clumping.

Time- and size-dependent cytotoxicity of SiO2 NPs. In theabsence of SiO2 NPs, control cells formed confluent mono-layers with their characteristic cell shapes. However, follow-ing exposure, under the phase contrast microscope, morpho-logical alterations such as cell shivering, cell detachment fromthe tissue culture surfaces, nuclear condensation, disruption ofcell monolayers, and accumulation of cell fragmented debriscould be noted within the first 12 h. For FHMT-W1 cells(Fig. 3), cell death could clearly be discerned after 12 hexposure to Si120 (16 nm) or Si173 (44 nm) NPs at thecytotoxic doses compared to after 1 h exposure (Fig. 3).Cells appear to experience stress-induced morphologicalchanges with increasing NP dose and cell debris became moreevident with time. For the Glofish cell line, derived from atransgenic green fluorescent protein (GFP) expressing adult

zebrafish (Fig. 4), exposure to high doses of Si168 (24 nm) ledto visible decrease of GFP expression concomitant to a de-crease in cell viability which was most apparent with dosagesabove their EC50 (Fig. 4, panels C, D). Similar morphologicalchanges were reported by Napierska et al. (2009) using thehuman endothelium cell line, EAHY926. SiO2 NPs of smallersize (16 nm) appeared to be slightly more toxic than those oflarger size NPs (44 nm), as judged by lower EC50 values.Thus, toxicity of SiO2 NPs was dependent on exposure timeand dosage, which was also concluded by these authors. In

Figure 1. Evaluation of SiO2 NPs for possible interference with AlamarBlue assay. SiO2 NPs at 1:100 dilution from stock solution (see Table 2)were dosed in normal non-fluorescent Alamar Blue working solution(Normal AB) or fluorescent Alamar Blue working solution (Fluor. AB)in L-15/ex in the absence of cells. Each data point represents the meanrelative fluorescence (RFUs) of six-well replicates (n=6) as percent of

controls with no nanoparticles (No NP) and is presented with standarddeviations. No significant differences in fluorescence were noted for anyof the NP solutions compared to respective controls (p >0.05). Stocksilica nanoparticles in L-15/ex without any Alamar Blue were alsoevaluated for fluorescence along with L-15/ex as controls, and theseshowed no fluorescence on their own (data not shown).

Table 3 Cytotoxicity of SiO2 NPs of varying sizes (16, 24, and 44 nm) tofish cell lines. EC50 is the 24 h effective concentration that reducesviability of cells compared to the respective unexposed controls asmeasured by Alamar Blue assay

Cell line EC50 (μg/mL) based on Alamar Blue assays±SD [n]

Si120 (16 nm) Si168 (24 nm) Si173 (44 nm)

RTgill-W1 120.0 ±14.0 [3] 221.6±62.2 [4] 175.9±54.6 [4]

RTgut-GC 207.6±59.1 [3] 314.1±70.7 [3] 666.9±176.5 [3]

RTL-W1 160.6±63.0 [3] 208.3±41.2 [3] 209.8±47.9 [2]

RTBrain 269.1±29.2 [3] 206.6±72.8 [2] 591.4 [1]

FHML2-6 205.1±9.35 [3] 217.1±73.9 [3] 291.1±29.6 [2]

FHMT-W1 141.3±12.9 [3] 258.7±34.6 [4] 248.0 [1]

GFB3C 219.1±59.6 [5] 252.1±37.6 [3] 438.6±116 [3]

GFSk-S1 118.8±19.1 [4] 148.2±37.2 [4] 270.4±77.2 [3]

GloFish 288.5±74.1 [3] 402.8±7.11 [3] 562.3±193 [5]

ZEB2J 696.5±14.4 [2]* 1050 [1] 1844±107 [2]

HEW 313.2±98.2 [4] 348.7±79.2 [3] 448.8 [1]

EelBrain 896.9±402 [3]* 449.9±285 [2] 1856±1245 [2]

SD standard deviation, n number of independent replicate experimentsperformed for each cell line

*p <0.05 (significantly different from all other cell lines (except to eachother) within the tested size of NP)

CYTOTOXICITY EVALUATION OF SILICA NPS IN FISH CELL CULTURES

addition, the cells were altered more severely and the cellmonolayers were damaged more quickly when exposed toSi120 (16 nm) than to Si173 (44 nm) NPs after 1 h. Thisimplies SiO2 NP toxicity was also dependent on the NP sizesbased on morphological observation of the cells. Similarconclusions could be drawn from the calculated EC50 valuesin Table 3.

Tissue-dependent cytotoxicity of SiO2 NPs. Dependence oftissue types on the toxicity of SiO2 NPs was also investigated.Comparisons among four rainbow trout-derived cell lines(Table 3) provided an indication that the gill epithelial cellline RTgill-W1 was most sensitive (EC50 120 μg/mL forSi120), while for two goldfish cell lines, the skin-derived cellsGFSK-S1 were almost twofold more sensitive than the brain-

Figure 2. Representative dose–response curves to Si168 NP samples (24 nm) for six different fish cell lines after 24 h exposures.Data points are shownas means of six-well replicates (n=6) with standard deviations. Calculated EC50 values based on Alamar Blue assays are included for each graph.

VO ETAL.

derived GFB3C cell line (Table 3). Morphological evaluationof exposed cell monolayers to Si120 NPs (16 nm) providedclear dose–response indication of monolayer disruption

within 1 h of exposure (Fig. 5) between RTgill-W1 andRTgut-GC, a trout gut epithelium-derived cell line. In partic-ular, after 1 h of exposure, at 195–390μg/mL, cell monolayers

Figure 3. Time- and size-dependent toxicity of SiO2 NPs as observed bymorphological changes in FHMT-W1 cells. FHMT-W1 cells were ex-posed to Si120 (16 nm) and Si173 (44 nm) in 24-well plates and observedat time intervals. Phase contrast micrographs were taken after 1 and 12 h.

FHMT-W1 cells were exposed to three different concentrations of SiO2

NPs: 39, 195, and 390 μg/mL and the morphologies were compared tothe control cells. Scale bar=100 μm.

Figure 4. Morphological changes of Glofish cells exposed to 218,437, and 946 μg/mL of Si168 NPs (24 nm). Glofish cells expressgreen fluorescent protein (GFP) when viable. A decrease in cellviability correlated to a decrease in GFP fluorescence signals.

Pictures were taken directly from emptied wells of the 96-wellplate after Alamar Blue viability assay. Control Glofish cells werein silica nanoparticle-free L-15/ex. Scale bar =100 μm.

CYTOTOXICITY EVALUATION OF SILICA NPS IN FISH CELL CULTURES

were already disrupted and the cells were dying in RTgill-W1but not in RTgut-GC (Fig. 5). In terms of EC50 values forSi120NPs, gill-derived RTgill-W1 cell line was about twofoldmore sensitive than the gut-derived RTgut-GC cell line(Table 3). The sensitivity of the gill cells was consistent forthe other sized NPs as well, with lower EC50s for RTgill-W1than RTgut-GC (Table 3).

Species-dependent cytotoxicity of SiO2 NPs. Comparison ofcytotoxicity responses for the same tissue-derived cells (e.g.,liver or brain) from two or more different species provided anindication to species differences. For liver cell lines derivedfrom rainbow trout (RTL-W1) and fathead minnow (FHML2-6), SiO2 NPs, for all three tested sizes, were slightly morecytotoxic to the trout-derived liver cells than those derivedfrom fathead minnow (Table 3), with 24 h EC50 values con-sistently smaller for RTL-W1 cells than for FHML2-6 cells.However, for brain-derived cell lines from trout and goldfish,no consistent species trend differences could be noted whileeel brain-derived cells appeared least sensitive. The incubationtemperature for the trout cell lines was 18 to 21°C (roomtemperature), similar to the incubation temperature for theeel brain cell line (21°C). But for fathead minnow and gold-fish, warmer incubation temperatures (24 to 27°C) were usedthat may affect metabolic responses; thus, temperature effectswere investigated.

Effect of incubation temperature on the toxicity of SiO2

NPs. Due to the notoriously wide temperature tolerance ofgoldfish (Ford and Beitinger 2005), GFSk-S1 cell line, whichappeared most sensitive to SiO2 NPs, was selected to investi-gate the effect of incubation temperature on the toxicity ofSiO2 NPs. Si120 (16 nm) was more toxic to GFSk-S1 cells at

higher temperature (27°C) than at room temperature (21°C)(Fig. 6) with an almost twofold difference in EC50 values. Thiscould be because goldfish cells are metabolically more activeat higher temperatures and may be more susceptible to toxi-cant action, but further investigation is needed.

Discussion

Nanotechnology has been an increasingly popular new sci-ence discipline and industrial field (Wilson 2002; Handy et al.2008; McIntyre 2012). The development of ENMs and NPsare leading novel applications that functionally exceed thoseof conventional materials (Wilson 2002; McIntyre 2012).Among these ENMs are SiO2 NPs that have been extensivelyresearched and used in biopharmaceutical/medical and indus-trial applications (Knopp et al. 2009; Napierska et al. 2010).There have been many reports evaluating the toxicity of SiO2

NPs using human and mammalian models both in vivo(Kumar et al. 2010; Xie et al. 2010) and in vitro (Napierskaet al. 2009; Park et al. 2009; Yang et al. 2009; Yu et al. 2009),and results have been variable (Napierska et al. 2010).Massive amounts of SiO2 NPs are produced annually, andtherefore discharged into water bodies (Jarvie et al. 2009b),possibly causing undesired effects on fish and aquatic inver-tebrates. However, very few reports have used aquatic verte-brate models: zebrafish (Ramesh et al. 2013), zebrafish em-bryos (Fent et al. 2010; Nelson et al. 2010), medaka embryosand larvae (Lee et al. 2011), fathead minnow fibroblasts(Christen and Fent 2012), and a rainbow trout cell line(Casado et al. 2013). Fent et al. (2010) reported that no “overtembryotoxicity” could be detected at least with the 60 and

Figure 5. Tissue-dependent toxicity of SiO2 NPs. RTgill-W1 andRTgut-GC cells derived from rainbow trout gill and gut respectively wereselected for tissue comparisons. Cells were plated in a 24-well plate for24 h in L-15 10% FBS media and kept at room temperature. Wells were

rinsed with L-15/ex and cells were dosed with three different concentra-tions of Si120 NPs (16 nm): 39, 195, and 390 μg/mL compared to thecontrol cells. Phase contrast photomicrographs were taken after 1 h ofexposure. Scale bar =100 μm.

VO ETAL.

200 nm SiO2 NPs that they tested, while in the more recentstudy by Christen and Fent (2012), SiO2 NPs were found toperturb normal functioning of endoplasmic reticulum in fat-head minnow cells as well as in human hepatoma cells andthat further investigations are warranted. Additionally, Casadoet al. (2013), using the trout cell line RTG-2, reported cyto-toxicity only at higher dosages but they stated that theircytotoxicity results were unreliable as they noted agglomera-tion and sedimentation. Many variables could affect toxicityof nanoparticles (Warheit 2010) and thus methodological con-siderations must be kept in mind when performing in vitroevaluations (Stone et al. 2009). Therefore, we evaluated stablemonodisperse arginine-coated SiO2 NPs of three sizes forcytotoxicity in 12 fish cell lines under controlled conditionsusing a quick fluorometric assay.

SiO2 NPs are one of the most stable among NPs, even indifferent organic solvents as well as aqueous solutions withhigh ionic strength (Lison et al. 2008; Knopp et al. 2009;Napierska et al. 2010). The exposure medium L-15/ex is aphysiological salt solution (Dayeh et al. 2005), and the eval-uated arginine-coated SiO2 NPs remained stable. Imaging ofthe SiO2 NPs in L-15/ex using SEM or TEM showed noalterations to the original morphological shapes and proper-ties, and did not aggregate (data not shown). This is in agree-ment to similar studies performed by Yu et al. (2009) in F-12exposure media, and by Thomassen et al. (2010) with lysine-coated SiO2 NPs in DMEM culture medium. The stability ofSiO2 NPs is in contrast to other NPs, like tungsten carbide NPsthat aggregated in L-15/ex (Kuhnel et al. 2009), althoughneither uptake into fish cells nor the expression of toxicity inRTgill-W1 cells appeared to be affected by the aggregation.

We first ensured that the NPs did not interfere with thefluorometric assay as recommended by Stone et al. (2009) toprevent misinterpretation of data. The tested SiO2 NPs on theirown did not emit any fluorescence and did not interefere with

the fluorescence emission of Alamar Blue (Fig. 1). In terms ofpharmacological doses, the tested SiO2 NPs had very lowtoxicity on the evaluated cell lines, since all the experimental24 h EC50 values obtained were greater than 100 μg/mL(Table 3). These values agreed well with the most reportedtoxicity data in human and mammalian models (Chang et al.2007; Park et al. 2009; Yang et al. 2009; Yu et al. 2009),although a couple of studies deduced 24 h EC50 values of alittle less than 100 μg/mL (e.g., Wang et al. 2009), possiblydue to variations in experimental setup parameters (nanopar-ticle sources, nanoparticle surface modification, seeding celldensity, toxicity endpoints). Overall, compared with othernanomaterials, SiO2 NPs have shown less cytotoxicity andthe hazards of nanosilica particles for at least the terrestrialorganisms reviewed so far appear variable (Napierska et al.2010). Furthermore, surface coatings of the NPs must be takeninto consideration as these may change the physicochemicalcharacteristics of the particles. For mammalian cells,Thomassen et al. (2010) recently evaluated lysine-coatedSiO2 NPs to avoid the surface reactivity effects that commonlycoating agents like surfactants add to the toxic responses.Withthis under control, Thomaseen et al. (2010) reported thatincreasing SiO2 NP dosage and decreasing particle size wereprimarily linked to the toxic effects seen in the evaluatedhuman macrophages and endothelial cells. These authors didnote differential susceptibility of macrophages over endothe-lial cells; thus, tissue sensitivity needs to be evaluated. In thisstudy, most of the dose–response curves had a graduallyincreasing slope as can be observed with the representativedose–response curves for the 24-nm SiO2 NPs to 6 of the 12cell lines tested (Fig. 2). This implies that the cytotoxicity ofSiO2 NPs spans over a wide range of concentrations, furthersuggesting that toxicity of SiO2 NPs would more likely causechronic effects in vivo when NPs bioaccumulate to higherconcentrations.

Figure 6. Effect of incubationtemperature on the cytotoxicity ofSi120 NPs (15–20 nm) in GFSk-S1 cells. Percent cell viability wasdetermined by Alamar Blueassay. Each data point is themean of six replicates (n =6) andare shown with standarddeviations. EC50 values forexposure at 21 and 27°C weredetermined to be 274.4 and143.6 μg/mL, respectively.

CYTOTOXICITY EVALUATION OF SILICA NPS IN FISH CELL CULTURES

In addition, by morphological observations, cytotoxicity ofSiO2 NPs was shown to be dependent on exposure time, NPsize, temperature, tissue types, and species. Longer exposuretime caused higher toxicity (Fig. 3), which is usually the casewith most chemicals as irreparable damages accumulate ki-netically. Larger SiO2 NPs (44 nm) were less toxic thanNPs ofsmaller sizes (16 nm) (Table 3 and Fig. 3). This confirmed thewell-documented time and size-dependent cytotoxicity natureof SiO2 NPs in the literature (Napierska et al. 2009, 2010; Yuet al. 2009). Moreover, lining epithelial cell lines were foundto be more sensitive to SiO2 NPs than cell lines derived fromembryos or some internalized tissues such as brain, gonads,and intestine (Table 3). In particular, GFSk-S1 and RTGill-W1cells were the most sensitive to SiO2 NPs whereas ZEB2Jcells were the least sensitive, along with the neural cell linesGFB3C, RTBrain, and EelBrain cells. EC50 values for theEelBrain cells yielded inconsistent results due to frequent cellclumping, thus the high SDs.

It is plausible that lining tissues such as gills and skin tissuewhich are constantly in contact with the aquatic environmentare more sensitive to abrasive chemicals than internal tissues orthat rapidly renewing cells are more affected because the NPscould interfere with their metabolic activities. However, this isin contrast to the report by Chang et al. (2007) who concludedthat metabolically active cells were less sensitive to SiO2 NPsbased on a comparison of three metabolically slow normalhuman-derived cell lines and three metabolically active humantumor-derived cells. Although they did conclude that cancercells are more resistant, a comparison of normal human cells ofvarying metabolic activities would have been desirable. Furtherresearch is needed to elucidate these points. Nevertheless, ourfindings are in overall agreement with studies performed withmammalian cells (Chang et al. 2007; Li et al. 2011; Sun et al.2011) in that greater than 100 μg/mL dosages were needed toobserve adverse effects. At low concentrations below EC50

values, SiO2 NPs induced minimal morphological changes(Figs. 3 and 4). At high concentrations, SiO2 NPs causedcellular fragmentation and even cell death. The latter is inagreement with Ramesh et al. (2013) findings with zebrafishwhere nuclear fragmentation was observed.

Temperature is an important regulatory factor that canaffect cellular processes and mediate the toxicity of nanopar-ticles. For poikilothermic species like fishes whose physio-logical temperatures are heavily dependent on their surround-ing environmental climate, temperature has been always partof many experimental designs. Arginine-coated SiO2 NPssynthesized and used in this study were functionally stablefrom 4°C to over 40°C (Hartlen et al. 2008), which is therange of temperature teleost fishes and fish cell lines aregenerally known to be capable of tolerating, surviving, andgrowing (Ford and Beitinger 2005; Elliott and Elliott 2010; Voet al. 2013). Thus, toxicity of SiO2 NPs would be largely dueto the interaction of the NPs with cells. In this study, by using

skin cells GFSk-S1 from goldfish, it was found that toxicity ofSiO2 NPs was higher at 27°C than at 21°C (Fig. 6). Despitethis narrow window of temperature difference, toxicity ofSiO2 NPs was already observed in GFSk-S1 cells. This reso-nated well with a recent work by Mahmoudi et al. (2013)showing that cellular uptake of nanoparticles and their toxicitywas rapidly affected by only small changes in incubationtemperatures. Thus, our data showed that increased toxicityof SiO2 NPs was directly proportional to the increase in theexposure temperatures. This result may have helpful ecolog-ical implications of nanoparticle toxicity during the currentlyalarmed global warming.

An important point that must be emphasized is that thesein vitro toxicity findings may not reflect the equivalent in vivosystem. Cell lines are isolated cellular models and may havelost some tissue and species-specific characteristics or devel-oped new traits, one of which is indefinite growth in vitro. Forinstance, the GFSk-S1 cell line was the most sensitive to SiO2

NPs. However, fish skin tissues naturally produce mucus,which can act as a potential barrier, trapping and preventingSiO2 NPs from interfacing with skin cells. Thus, in vivomodels could even be less sensitive than in vitro cell lines.The recent work by Fent et al. (2010) stating no embryotoxicityof fluorescent SiO2 NPs at up to 200 mg/L to zebrafish em-bryos but interfering with the stress response in the ER ofcultured fathead minnow and human hepatoma cells(Christen and Fent 2012) may be reflective of this. Finally,the work of George et al. (2012) with silver NPs statingexcellent correlation between in vivo (fish embryos) andin vitro (fish cell line) toxicological evaluations must also betaken into account. Consequently, careful screening and doserange testing with cell lines followed by mechanistic evalua-tion at the cellular level could reduce/refine the need of wholeorganismal testing.

Overall, it was determined that cytotoxicity of SiO2 NPswas dependent on NP sizes, exposure temperature, exposuretime, dosage, cell types, and species of study. Calculated EC50

values greater than 100 μg/mL were needed for cytotoxicity,indicating that SiO2 NPs had very low toxicity levels for fishin vitro. Fish cell cultures, much like mammalian cell culturesfor human toxicity screening, would serve as convenientaquatic in vitro models in assessing new potential emergingenvironmental contaminants such as ENMs and NPs.

Acknowledgments The authors wish to thank the Natural Sciences andEngineering Research Council of Canada for operating grants to LEJLand VK. This work was also partially supported by an undergraduateresearch award to NTKVprovided by the former Science and TechnologyEndowment Program at Wilfrid Laurier University. We gratefully ac-knowledge Heather Braid and Robert Hanner (Biodiversity Institute,University of Guelph) for the barcoding authentication of the cell linesto species of origin, and Telly Athanasopoulos for his technical assistancein synthesizing the silica nanoparticles used in this study. We also appre-ciated the Centre for Nanostructure Imaging (University of Toronto) forElectron Microscopy facilities.

VO ETAL.

References

Bols N. C.; Barlian A.; Chirino-Trejo M.; Caldwell S. J.; Goegan P.; LeeL. E. J. Development of a cell line from primary cultures of rainbowtrout,Oncorhynchus mykiss (Walbaum), gills. J. Fish Dis. 17: 602–611; 1994.

Bryson S. P.; Joyce E. M.; Martell D. J.; Lee L. E. J.; Holt S. E.; Kales S.C.; Fujiki K.; Dixon B.; Bols N. C. A cell line (HEW) from embryosof haddock (Melanogrammus aeglefinius) and its capacity to toler-ate environmental extremes.Marine Biotechnol. 8: 641–653; 2006.

Cao L.; Ye Z.; Cai Z.; Dong E.; Yang Z.; Liu G.; DengX.;Wang Y.; YangS. T.; Wang H.; Wu M.; Liu Y. A facile method to encapsulateproteins in silica nanoparticles: encapsulated green fluorescent pro-tein as a robust fluorescence probe. Angew. Chem. Int. Ed. Engl. 49:3022–3025; 2010.

Casado M. P.; Macken A.; Byrne H. J. Ecotoxicological assessment ofsilica and polystyrene nanoparticles assessed by a multitrophic testbattery. Environ. Int. 51: 97–105; 2013.

Chang J. S.; Chang K. L. B.; Hwang D. F.; Kong Z. L. In vitro cytotox-icity of silica nanoparticles at high concentrations strongly dependson the metabolic activity type of the cell line. Environ. Sci. Technol.41: 2064–2068; 2007.

Christen V.; Fent K. Silica nanoparticles and silver-doped silica nanopar-ticles induce endoplasmic reticulum stress response and alter cyto-chrome P4501A activity. Chemosphere 87: 423–434; 2012.

Cooper J. K.; Sykes G.; King S.; Cottrill K.; Ivanova N. V.; Hanner R.;Ikonomi P. Species identification in cell culture: a two-prongedmolecular approach. In Vitro Cell. Dev. Bio.Anim. 43: 344–351;2007.

Dayeh V. R.; Schirmer K.; Lee L. E. J.; Bols N. C. Rainbow trout gill cellline microplate cytotoxicity test. In: Blaise C.; Férard J. F. (eds)Small-scale freshwater environment toxicity test methods, vol. 15.Kluwer Academic Publishers, Dordrecht, pp 473–500; 2005.

Elliott J. M.; Elliott J. A. Temperature requirements of Atlantic salmonSalmo salar, brown trout Salmo trutta and Arctic charr Salvelinusalpinus : predicting the effects of climate change. J. Fish Biol. 77:1793–1817; 2010.

Eom H. J.; Choi J. Oxidative stress of silica nanoparticles in humanbronchial epithelial cell, Beas-2B. Toxicol. in Vitro 23: 1326–1332; 2009.

Fent K.; Weisbrod C. J.; Wirth-Heller A.; Pieles U. Assessment ofuptake and toxicity of fluorescent silica nanoparticles inzebrafish (Danio rerio) early life stages. Aquat. Toxicol. 100: 218–228; 2010.

Ford T.; Beitinger T. L. Temperature tolerance in the goldfish, Carassiusauratus. J. Therm. Biol. 30: 147–152; 2005.

George S.; Lin S.; Ji X.; Thomas C. R.; Li L. J.; MecklenburgM.; Meng H.; Wang X.; Zhang H.; Xia T.; Hohman J. N.;Lin S.; Zink J. I.; Weiss P. S.; Nel A. E. Surface defects onplate-shaped silver nanoparticles contribute to its hazard po-tential in a fish gill cell line and zebrafish embryos. ACSNano 6: 3745–3759; 2012.

Handy R. D.; Henry T. B.; Scown T. M.; Johnston B. D.; Tyler C. R.Manufactured nanoparticles: their uptake and effects on fish—amechanistic analysis. Ecotoxicology 17: 396–409; 2008.

Hartlen K. D.; Athanasopoulos A. P. T.; Kitaev V. Facile preparation ofhighly monodisperse small silica spheres (15 to >200 nm) suitablefor colloidal templating and formation of ordered arrays. Langmuir24: 1714–1720; 2008.

He X.; Nie H.; Wang K.; Tan W.; Wu X.; Zhang P. In vivo study ofbiodistribution and urinary excretion of surface-modified silicananoparticles. Anal. Chem. 80: 9597–960; 2008.

Jarvie H. P.; Al-Obaidi H.; King S. M.; Bowes M. J.; Lawrence M. J.;Drake A. F.; Green M. A.; Dobson P. J. Managing nanoparticlewaste in sewage. ISIS Science and Technology Facilities Council.

http://www.isis.stfc.ac.uk/science/earth-science-and-environment/managing-nanoparticle-waste-in-sewage9094.html; 2009a.

Jarvie H. P.; Al-Obaidi H.; King S. M.; Bowes M. J.; Lawrence M. J.;DrakeA. F.; GreenM. A.; Dobson P. J. Fate of silica nanoparticles insimulated primary wastewater treatment. Environ. Sci. Technol. 43:8622–8628; 2009b.

Kawano A.; Haiduk C.; Schirmer K.; Hanner R.; Lee L. E. J.; Dixon B.;Bols N. C. Development of a rainbow trout intestinal epithelial cellline and its response to lipopolysaccharide. Aquac. Nut. 17: e241–e252; 2010.

Knopp D.; Tang D.; Niessner R. Review: bioanalytical applications ofbiomolecule-functionalized nanometer sized doped silica nanopar-ticles. Anal. Chim. Acta. 647: 14–30; 2009.

Kuhnel D.; Busch W.; Meibner T.; Springer A.; Potthoff A.; Richter V.;Gelinsky M.; Scholz S.; Schirmer K. Agglomeration of tungstencarbide nanoparticles in exposure medium does not prevent uptakeand toxicity toward a rainbow trout gill cell line. Aquat. Toxicol. 93:91–99; 2009.

Kumar R.; Roy I.; Ohulchanskky T. Y.; Vathy L. A.; Bergey E. J.; SajjadM.; Prasad P. N. In vivo biodistribution and clearance studies usingmultimodal organically modified silica nanoparticles. ACS Nano 4:699–708; 2010.

Lai Y.; Chiang P.; Blom J. D.; Li N.; Shevlin K.; Brayman T. G.; Hu Y.;Selbo J. G.; Hu L. G. Comparison of in vitro nanoparticles uptake invarious cell lines and in vivo pulmonary cellular transport inintratracheally dosed rat model. Nanoscale Res. Lett. 3: 321–329;2008.

Lee L. E. J.; Bufalino M. R.; Wilkie M. P. Comparison of growthcharacteristics, neurochemical parameters, and response to toxicantsfor neural-tissue-derived cell lines from goldfish and from crayfish.In Vitro Cell. Dev. Biol. Anim. 44(S1): A2011; 2008.

Lee L. E. J.; Caldwell S. J.; Gibbons J. Development of a cell line fromskin of goldfish, Carassius auratus , and effects of ascorbic acid oncollagen deposition. Histochem. J. 29: 31–43; 1997.

Lee L. E. J.; Clemons J. H.; Bechtel D. G.; Caldwell S. J.; Han K. B.;Pasitschniak-Arts M.; Mosser D. D.; Bols N. C. Development andcharacterization of a rainbow trout liver cell line expressing cyto-chrome P450-dependent monooxygenase activity. Cell Biol.Toxicol. 9: 279–294; 1993.

Lee L. E. J.; Vo N. T. K.; Werner J.; Weil R.; Denslow N. D.; Law R. D.Development of a liver cell line from fathead minnow, Pimephalespromelas, and their molecular and biochemical characterization. InVitro Cell. Dev. Biol. Anim 45(S1): A2025; 2009.

LeeW. M.; Ha S.W.; Yang C. Y.; Lee J. K.; An Y. J. Effect of fluorescentsilica nanoparticles in embryo and larva of Oryzias latipes : soniceffect in nanoparticle dispersion. Chemosphere 82: 451–459; 2011.

Li Y.; Sun L.; Jin M.; Du Z.; Liu X.; Guo C.; Li Y.; Huang P.;Sun Z. Size-dependent cytotoxicity of amorphous silica nano-particles in human hepatoma HepG2 cells. Toxicol. In Vitro 25:1343–1352; 2011.

Lison D.; Thomassen L. C. J.; Rabolli V.; Gonzalez L.; Napierska D.; SeoJ.W.; Kirsh-Volders M.; Hoet P.; Kirschhock C. E. A.; Martens J. A.Nominal and effective dosimetry of silica nanoparticles in cytotox-icity assays. Toxicol. Sci. 104: 155–162; 2008.

MahmoudiM.; ShokrgozarM. A.; Behzadi S. Slight temperature changesaffect protein affinity and cellular uptake/toxicity of nanoparticles.Nanoscale 5: 3240–3244; 2013.

McIntyre R. A. Common nanomaterials and their use in real worldapplications. Sci. Prog. 95: 1–22; 2012.

Napierska D.; Thomassen L. C. J.; Lison D.; Martens J. A.; Hoet P. H.The nanosilica hazard: another variable entity. Part. Fibre Toxicol.7: 39–71; 2010.

Napierska D.; Thomassen L. C. J.; Rabolli V.; Lison D.; Gonzalez L.;Kirsch-Volders M.; Martens J. A.; Hoet P. H. Size-dependent cyto-toxicity of monodisperse silica nanoparticles in human endothelialcells. Small 5: 846–853; 2009.

CYTOTOXICITY EVALUATION OF SILICA NPS IN FISH CELL CULTURES

Nelson S. M.; Mahmoud T.; Beaux II M.; Shapiro P.; McIlroy D. N.;Stenkamp D. L. Toxic and teratogenic silica nanowires in develop-ing vertebrate embryos. Nanomedicine 6: 93–102; 2010.

Oberdorster G.; Oberdorster E.; Oberdorster J. Nanotoxicology: anemerging discipline evolving from studies of ultrafine particles.Environ. Health Perspect. 113: 823–839; 2005.

O’Brien J.; Wilson I.; Orton T.; Pognan F. Investigation of the AlamarBlue (resazurin) fluorescent dye for the assessment of mammaliancell cytotoxicity. Eur. J. Biochem. 267: 5421–5426; 2000.

Park M. V.; AnnemaW.; Salvati A.; Lesniak A.; Elsaesser A.; Barnes C.;McKerr G.; Howard C. V.; Lynch I.; Dawson K. A.; Piersma A. H.;de Jong W. H. In vitro developmental toxicity test detects inhibitionof stem cell differentiation by silica nanoparticles. Toxicol. Appl.Pharmacol. 240: 108–116; 2009.

Ramesh R.; Kavitha P.; Kanipandian N.; Arun S.; Thirumurugan R.;Subramanian P. Alteration of antioxidant enzymes and impairmentof DNA in the SiO2 nanoparticles exposed zebra fish (Danio rerio).Environ. Monit. Assess. 185: 5873–5881; 2013.

Reeves J. F.; Davies S. J.; Dodd N. J. F.; Jha A. N. Hydroxyl radicals(OH) are associated with titanium dioxide (TiO2) nanoparticle in-duced cytotoxicity and oxidative DNA damage in fish cells.Mutat.Res. 640: 113–122; 2008.

Servili A.; Bufalino M. R.; Nishikawa R.; de Sanchez Melo I.; Muñoz-Cueto J. A.; Lee L. E. J. Establishment of long term cultures ofneural stem cells from adult sea bass, Dicentrarchus labrax. Comp.Biochem. Physiol. A 152: 245–254; 2009.

Singh N.; Manshian B.; Jenkins G. J. S.; Griffiths S. M.; Williams P. M.;Maffeis T. G. G.; Wright C. J.; Doak S. H. Nanogenotoxicology: theDNA damaging potential of engineered nanomaterials. Biomaterials30: 3891–3914; 2009.

Stone V.; Johnston H.; Schins R. P. Development of in vitro systems fornanotoxicology: methodological considerations. Crit. Rev. Toxicol.39: 613–626; 2009.

Sun L.; Li Y.; Liu X.; Jin M.; Zhang L.; Du Z.; Guo C.; Huang P.; Sun Z.Cytotoxicity and mitochondrial damage caused by silica nanoparti-cles. Toxicol. In Vitro 25: 1619–1629; 2011.

Thomassen L. C. J.; Aerts A.; Rabolli V.; Lison D.; Gonzales L.; Kirsch-Volders M.; Napierska D.; Hoet P. H.; Kirschhock C. E. A.; MartensJ. A. Synthesis and characterization of stable monodisperse silicananoparticle sols for in vitro cytotoxicity testing. Langmuir 26:328–335; 2010.

US EPA. Technical overview of ecological risk assessment analysisphase: ecological effects characterization. http://www.epa.gov/oppefed1/ecorisk_ders/toera_analysis_eco.htm#Ecotox; 2012.

Vevers W. F.; Jha A. N. Genotoxic and cytotoxic potential of titaniumdioxide (TiO2) nanoparticles on fish cells in vitro. Ecotoxicology 17:410–420; 2008.

Vo N. T. K.; Bender A. W.; Lee L. E. J.; Lumsden J. S.; Lorenzen N.;Dixon B.; Bols N. C. Development of a walleye cell line and use tostudy the effects of temperature on infection by viral hemorrhagicsepticaemia virus (VHSV) group IVb. J. Fish Dis. doi:10.1111/jfd.12208. (in press); 2013

Vo N. T. K.; Sansom B.; Kozlowski E.; Bloch S. R.; Lee L. E. J.Development of a permanent cell line derived from fathead minnowtestis and their applications in environmental toxicology. MDIBLBull. 49: 83–86; 2010.

Wang F.; Gao F.; Lan M.; Yuan H.; Huang Y.; Liu J. Oxidative stresscontributes to silica nanoparticles-induced toxicity in human embry-onic kidney cells. Toxicol. in Vitro 23: 808–815; 2009.

Warheit D. B. Debunking some misconceptions about nanotoxicology.Nano Lett. 10: 4777–4782; 2010.

Wilson M. Nanotechnology: basic science and emerging technologies.Chapman and Hall, CRC Press, Boca Raton; 2002: 271 pp

Wise J. P.; Goodale B. C.; Wise S. S.; Craig G. A.; Ponga A. F.; Walter R.B.; Thompson W. D.; Ng A. K.; Aboueissa A. M.; Mitani H.;Spalding M. J.; Mason M. D. Silver nanospheres are cytotoxic andgenotoxic to fish cells. Aquat. Toxicol. 97: 34–41; 2010.

Xie G.; Sun J.; ZhongG.; Shi L.; ZhangD. Biodistribution and toxicity ofintravenously administered silica nanoparticles in mice. Arch.Toxicol. 84: 183–190; 2010.

Xing J. G.; Lee L. E. J.; Fan L.; Collodi P.; Holt S. E.; Bols N. C. Initiationof a zebrafish blastula cell line on rainbow trout stromal cells andsubsequent development under feeder-free conditions into a cellline, ZEB2J. Zebrafish 5: 49–63; 2008.

Yang H.; Liu C.; Yang D.; Zhang H.; Xi Z. Comparative study ofcytotoxicity, oxidative stress and genotoxicity induced by four typ-ical nanomaterials: the role of particle size, shape and composition.J. Appl. Toxicol. 29: 69–78; 2009.

Yu F. O.; Grabinski C. M.; Schrand A. M.; Murdock R. C.; WangW.; GuB.; Schlager J. J.; Hussain S. B. Toxicity of amorphous silicananoparticles in mouse keratinocytes. J. Nanoparticle. Res. 11:15–24; 2009.

VO ETAL.