at SciVerse ScienceDirect

Marine Environmental Research 84 (2013) 51e59

Contents lists available

Marine Environmental Research

journal homepage: www.elsevier .com/locate/marenvrev

Genotoxicity of copper oxide and silver nanoparticles in the musselMytilus galloprovincialis

Tânia Gomes, Olinda Araújo, Rita Pereira, Ana C. Almeida, Alexandra Cravo, Maria João Bebianno*

CIMA, Faculty of Science and Technology, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal

a r t i c l e i n f o

Article history:Received 24 August 2012Received in revised form22 November 2012Accepted 30 November 2012

Keywords:CuO NPsAg NPsDNA damageComet assayOxidative stressMytilus galloprovincialis

* Corresponding author. Tel.: þ351 289800100; faxE-mail address: mbebian@ualg.pt (M.J. Bebianno).

0141-1136/$ e see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.marenvres.2012.11.009

a b s t r a c t

Though there is some information on cytotoxicity of copper nanoparticles and silver nanoparticles onhuman cell lines, there is no information on their genotoxic and cytotoxic behaviour in bivalve molluscs.The aim of this study was to investigate the genotoxic impact of copper oxide and silver nanoparticlesusing musselsMytilus galloprovincialis. Mussels were exposed to 10 mg L�1 of CuO nanoparticles and Cu2þ

and Ag nanoparticles and Agþ for 15 days to assess genotoxic effects in hemocytes using the comet assay.The results obtained indicated that copper and silver forms (nanoparticles and ionic) induced DNAdamage in hemolymph cells and a time-response effect was evident when compared to unexposedmussels. Ionic forms presented higher genotoxicity than nanoparticles, suggesting different mechanismsof action that may be mediated through oxidative stress. DNA strand breaks proved to be a usefulbiomarker of exposure to genotoxic effects of CuO and Ag nanoparticles in marine molluscs.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Nanoparticles exhibit novel electrical, catalytic, chemical andmagnetic properties that can be applied in several commercial andindustrial applications that have led to its increased production anduse. Inevitably these particles will end up in the environment,accumulate and interact with the biota and their potential geno-toxic effects could have short and long-term consequences inaquatic organisms (Handy et al., 2008; Moore, 2006; Singh et al.,2009).

Copper nanoparticles (CuO NPs) are commonly used in manyindustrial and consumer applications mostly because of its anti-microbial properties, as well as for its elevated thermal and elec-trical conductive efficiency. These nanoparticles are used in air andliquid filtration, in coatings on integrated circuits and batteries,catalysts, microelectronics and cosmetics, among others (Ahamedet al., 2010; Fahmy and Cormier, 2009; Griffitt et al., 2007;Karlsson et al., 2008). The use of silver nanoparticles (Ag NPs)incorporated in consumer products has become common in the lastyears because of the antimicrobial and antibacterial effect of thesilver ion (e.g. Luoma, 2008). Industry makes use of this newtechnology in food contact applications, in building materials,wound dressings, socks, and other textiles, air filters, medical

: þ351 289800069.

ll rights reserved.

equipment and textiles, toothpaste, baby products, vacuumcleaners and washing machines (www.nanoproject.org). Accord-ingly, these NPs may already exist in the environment, but atpresent there is no information regarding its levels in the aquaticcompartment.

Overproduction of reactive oxygen species (ROS) is commonlyconsidered themajor source of genotoxicity that has been observedafter exposure to many types of NPs that have been found to causeDNA-strand breaks, point mutations, oxidative DNA adducts, andchromosomal fragmentation (Bhatt and Tripathi, 2011; Handy et al.,2008; Karlsson, 2010; Singh et al., 2009). CuO NPs possess redoxcycling properties with the capacity to generate intra- and extra-cellular generation of reactive oxygen species (ROS) due toa combination between the particle effect and the dissociation ofcopper ions from the NPs (Fahmy and Cormier, 2009; Gomes et al.,2011; Griffitt et al., 2009). Ag NPs also generates ROS resulting fromthe release of silver ions, NPs properties (e.g. size and/or shape) ora combination of both (Asharani et al., 2008; Fabrega et al., 2011;Griffitt et al., 2009). Oxidative stress is a significant mechanism oftoxicity of CuO and Ag NPs, but their underlying mechanism ofaction are still uncertain (Ahamed et al., 2010; Asharani et al., 2009;Fahmy and Cormier, 2009; Gomes et al., 2011). Organisms havedeveloped enzymatic and non-enzymatic antioxidant defencemechanisms to prevent and intercept ROS, as well as repair systemsfor oxidized components. When the rate of ROS production exceedsthe antioxidant defences and repair mechanisms, oxidative stressoccurs leading to oxidation of key cellular components as proteins,

T. Gomes et al. / Marine Environmental Research 84 (2013) 51e5952

DNA, lipids and carbohydrates (Halliwell and Gutteridge, 1984;Valavanidis et al., 2006). In mussel gills and digestive glands,oxidative stress after long-term exposure to CuO NPs and Ag NPswas evidenced by induction or reduction of the antioxidant defencesystem, lipid peroxidation, as well as metallothionein induction(Gomes et al., 2011, 2012; Gomes et al., own data). Given thecapacity of both CuO and Ag NPs to generate oxidative stress andoxidative damage, it becomes imperative to address the genotoxicpotential of these particles, principally in bivalve species wherescarce information exists. Though there is some information oncytotoxicity of Cu and Ag NPs on human cell lines, the informationon genotoxic and cytotoxic behaviour of these nanoparticles inaquatic organisms is scarce (e.g. Ahamed et al., 2010; Asharani et al.,2009; Fahmy and Cormier, 2009; Park and Choi, 2010). Only Gagnéet al. (2008) and Kádar et al. (2011) addressed DNA damage in themussels Elliptio complanata and Mytilus galloprovincialis afterexposure to cadmium telluride quantum dots and nano iron,respectively. DNA strand breaks (single and double) represent oneof the major oxidative damage to DNA via oxidative stress that isgenerally assessed by the Comet assay. This methodology has beenextensively used to evaluate the genotoxic effects of contaminants(as metals) in bivalves’ hemocytes (e.g. Almeida et al., 2011; Al-Subiai et al., 2011; Steinert, 1995; Valavanidis et al., 2006) thatare a potential target for nanoparticles genotoxicity (Canesi et al.,2010, 2012; Gagné et al., 2008; Kádar et al., 2011; Moore et al.,2009).

Themain goal of the current workwas to assess of the genotoxicpotential of CuO NPs and Ag NPs in the mussel M. galloprovincialisexposed to 10 mg L�1 either in nano or in the ionic form. The cometassay was applied to detect DNA damage (single, double strandbreaks and alkali labile sites) in hemolymph cells (hemocytes) ofmussels.

2. Materials and methods

2.1. Nanoparticles preparation and characterization

CuO and Ag nanoparticles were obtained from SigmaeAldrich(Germany) with the particle size specified as <50 nm and<100 nm, respectively. NPs stock solutions (CuO NPs and Ag NPs)were prepared in bi-distilledwater (10 mg L�1), sonicated for 30min(Ultrasonic bath VWR International, 230 V, 200 W, 45 kHzfrequency) and kept in constant shaking. Ionic stock solution (Cu2þ

and Agþ; 10 mg L�1) were prepared identically but not sonicated.NPs were characterized in terms of particle size, aggregationbehaviour in natural seawater using transmission electronmicroscopy (TEM) (JEOL JEM-1230) equipped with a digital camera(model 785 ES1000W) and by dynamic light scattering (DLS) usingan ALV apparatus with Ar ion lased (514.5 nm). More details on thecharacterization methods are described in Gomes et al. (2011).

2.2. Experimental design

MusselsM. galloprovincialis (Cu exposure: 61.7� 8.4 mm and Agexposure: 63.2� 5.8mm)were collected in the Ria Formosa Lagoon(South of Portugal) and acclimated for 7 days in natural seawater atconstant temperature and aeration. Fifty mussels were placed in25 L aquaria, filled with 20 L of natural seawater (around2.5 mussels/L) in a triplicate design (3 tanks per treatment) andexposed to 10 mg L�1 of CuO NPs and Ag NPs and correspondingionic form, along with a control group for a period of 15 days. TheCu and Ag concentrations chosen are environmentally relevant andreported for several aquatic systems (Bryan and Langston, 1992;Luoma, 2008). Water, copper and silver solutions were renewedevery 12 h to avoid nanoparticles aggregation. Before each renewal,

CuO and Ag NPs solutions were sonicated for 30 min (Ultrasonicbath VWR International, 230 V, 200 W, 45 kHz frequency) to breakdown the size of aggregates. Physico-chemical parameters weremeasured daily for Cu exposure: temperature (17.8 � 1.1 �C),salinity (36.3 � 0.2), percentage of oxygen saturation (97.8 � 4.9%)and pH (7.8 � 0.07); and Ag exposure: temperature (17.6 � 0.3 �C),salinity (36.3 � 0.1), percentage of oxygen saturation (96.9 � 3.3%)and pH (7.8 � 0.05). Multiple sampling was performed (3, 7 and 15days) where mussels were collected and biotic parametersmeasured. Mussels were not fed and no mortality was detectedduring the exposure period.

2.3. Metal analysis

Copper and silver were analysed inwater samples collected 12 hbefore water renewal and re-dosing from the NPs and ionic expo-sure groups, as previously described in Gomes et al. (2011). Briefly,total metal concentrations from both exposures were determinedafter acid digestion with 2% nitric acid (HNO3), while dissolved Cuand Ag from CuO and Ag NPs exposures were determined afterwater filtration (0.02 mm filter, Anotop 25, Whatman) and aciddigestion (Griffitt et al., 2009). Copper and silver concentrations intotal edible tissues of five mussels were determined on driedsamples (80 �C) after wet digestion with nitric acid by atomicabsorption spectrophotometry (AAS Analyst 800, Perkin-Elmer).

2.4. Condition index

To assess the physiological status of control and exposedmussels to CuO and Ag NPs and Cu2þ and Agþ during the course ofthe experiment, the soft tissues and shells of ten individuals wereweighted and the condition index (CI) estimated as the percentage(%) of the ratio between drained weight of the soft tissues (g) andtotal weight (g).

2.5. DNA damage using the comet assay

Hemolymph of ten mussels collected after 0, 3, 7 and 15 days ofexposure to CuO and Ag NPs and corresponding ionic forms wasextracted from the posterior adductor muscle with a sterile hypo-dermic syringe. Cell viability was assessed by staining a 100 ml sub-sample from each experimental condition with 100 ml of trypanblue and measuring the percentage of live cells, by randomlycounting 100 cells. DNA damage was estimated using the cometassay in a slightlymodified version of that by Singh et al. (1988) anddescribed in Almeida et al. (2011). Briefly, microscopic slides werecoated with 0.65% normal melting point agarose (NMA) in Tris-acetate EDTA. After collection, hemolymph cells for each musselwere centrifuged at 3000 rpm for 3 min (4 �C) and the pellets withisolated cells suspended in 0.65% low melting point agarose (LMA,in Kenny’s salt solution) and casted on the microscope slides.Afterwards, the slides with the embedded cells were immersed ina lysis buffer (2.5 M NaCl, 100 mM EDTA, 10 mM Tris, 1% Triton X-100, 10% Dimethylsulfoxide, 1% Sarcosil, pH 10, 4 �C) for 1 h, for thediffusion of cellular components and DNA immobilization inagarose. Following the lysis step, slides were gently placed in anelectrophoresis chamber containing electrophoresis buffer(300 mM NaOH, 1 mM EDTA, adjusted at pH 13, 4 �C). The slideswere gently submerged and left in this solution for 15min to permitDNA unwinding. The electrophoresis was carried out for 5 min at25 V and 300 mA. Once the electrophoresis concluded, the slideswere removed and immersed in neutralizing solution (0.4 mM Tris,pH 7.5), rinsed with bi-distilled water and left to dry overnight.Afterwards, slides were stained with 4,6-diamidino-2-phenylindole (DAPI, 1 mg mL�1) and the presence of comets was

T. Gomes et al. / Marine Environmental Research 84 (2013) 51e59 53

analysed using an optical fluorescence microscope (Axiovert S100)coupled to a camera (Sony). The Komet 5.5 image analysis system(Kinetic Imaging Ltd) was used to score 50 randomly chosen cellsfor each slide (25 in each gel from each individual mussel) at a totalmagnification of �400. Different parameters of the comet,including the olive tail moment (product of comet tail length andproportion DNA in comet tail), comet tail length (in micrometres,measured from the edge of the comet head) and amount of DNA inthe comet tail (proportion based on tail intensity) were used.Results are expressed as mean � SEM. Cells were also categorizedfor grade of damage (using tail % DNA) based on the criteria referredin Almeida et al. (2011): grade of damage: zero or minimal 10% tail,low damage 10e25%, mid damage 25e50%, high damage 50e75%,and extreme damage >75%. During the entire procedure, greatcare was taken to avoid exposing cells and slides to light and heat.

2.6. Statistical analysis

One-way analysis of variance (ANOVA) was applied to the cometparameters and metal concentrations, as variances were homoge-neous and distribution normal, followed by the Tukey pair-wisemultiple comparison test, in order to evaluate differencesbetween control and exposed mussels. Linear regression analysiswas also applied to verify existing relationships between variablesand determine the rates of metal accumulation and cometparameters versus time. Statistical significance was defined atp < 0.05 and analysis performed with SigmaPlot10.

3. Results

3.1. Nanoparticles characterization

The distribution, size and shape of CuO NPs and Ag NPs wereobtained by TEM analysis and DLS analysis (Table 1, Fig. 1). CuO NPsare spherical in shape and presented a mean size of 31 � 10 nm(Fig. 1A, Gomes et al., 2011). Ag NPs are mainly spherical in shapeand polydisperse (Fig. 1B). The size of Ag NPs reported by themanufacturer (<100 nm) is broadly in agreement with the sizeobtained by TEM. Themean particle size was also determined usingDLS, confirming the formation of CuO NPs (284 � 21 nm) and AgNPs (144.2 � 39.2 nm) large aggregates in seawater that increasewith time of exposure (Fig. 1C and D). Additionally, high poly-dispersity indexes were observed for CuO NPs (0.35 � 0.03) and AgNPs (0.44 � 0.03), suggesting that under the exposure conditions,both nanoparticles tendency to aggregate produces suspensionswith the presence of both single particles and large aggregates,with size ranging from 30 to 338 nm for CuO NPs and 97 to 690 nmfor Ag NPs. The tendency of both NPs to form aggregates while insuspension have previously been established by other authorsusing these particles, some of which from the same provenience(e.g. Ahamed et al., 2010; Choi et al., 2010; Griffitt et al., 2007; Parkand Choi, 2010).

Table 1Characterization of CuO and Ag nanoparticles using different techniques. Values aremean � std.

Particle characterization Method CuO NPs Ag NPs

Particle size (nm) TEM <50a <100a

Primary particle size distribution (nm) TEM 31 � 10b e

Polydispersity index DLS 0.35 � 0.03c 0.44 � 0.03c

Mean particle diameter (nm) DLS 284.0 � 21.2c 144.2 � 39.2c

Specific surface area (m2/g) e 29.0a 5.0a

a Information from the manufacturer SigmaeAldrich.b CuO NPs dispersed in ultrapure water. Average diameter of 250 particles.c CuO and Ag NPs dispersed in natural seawater during a 12 h cycle.

3.2. Condition index

No significant changes were obtained in the condition index ofunexposed and exposed mussels throughout the exposure period,ranging from 18.4 � 1.5% to 20.6 � 3.8% (p > 0.05) in Cu exposure(Fig. 2A) and from 16.0 � 2.6% to 22.2 � 4.2% (p > 0.05) in the Agexposure (Fig. 2B).

3.3. Metal concentrations

Copper analysis of water showed that after 12 h of exposure(between water renewal) more than 50% of the Cu added in thenano and ionic form was removed from the water column, asalready referred in Gomes et al. (2011). The dissolution of Cu2þ fromCuO NPs <1% indicated that most of the Cu present is in thenanoparticulate form. As for silver, more than 80% of the nominalconcentration of 10 mg Ag L�1 added in the nano or ionic form wasremoved from the water column after the 12-h exposure (84% forAg NPs and 89% for Agþ). Approximately 24% of the initial added AgNPs dose was in the dissolved form, indicating that majority of theAg present in solution is in the nanoparticulate form. The loss ofthis amount of Cu and Ag may be due either to the presence of themussels, to metal dissolution or nanoparticles aggregation andsedimentation.

The exposure to both forms of Cu and Ag resulted in a significantCu and Ag accumulation in mussel whole soft tissues throughoutthe exposure period when compared to unexposedmussels (Fig. 2Cand D). In CuO NPs exposure, Cu concentrations in mussel tissuesincreased linearly with time of exposure (Table 2, Fig. 2C). As forCu2þ exposure, Cu was significantly accumulated in mussel tissuesin the first 3 days, reaching a steady state during the first week ofexposure (p > 0.05), followed by a decrease by the end of theexperiment (Fig. 2C). No differences were detected between Cuforms after 3 and 7 days of exposure (p> 0.05), however by the endof the experiment, Cu concentrations in mussels exposed to CuONPs continued to increasewhile those exposed to Cu2þ significantlydecreased (p < 0.05).

Mussels exposed to both Ag forms presented a similar trend ofaccumulation throughout the exposure period (Fig. 2D). In the first3 days of exposure, Agþ exposed mussels’ accumulated higher Aglevels than those exposed to Ag NPs (p < 0.05), while in theremaining exposure period no significant differences were found(p > 0.05). After the first week, Ag continued to accumulate inwhole soft tissues, reaching similar levels, followed by a decreaseby the end of the experiment for Ag NPs and Agþ exposed mussels.

3.4. DNA damage

Cell viability was assessed using the trypan blue exclusion test;cell viability was >87% in all cases. In Fig. 3 some examples ofcomets in hemocytes of mussels from control and exposed to Cuand Ag are reported. The olive tail moment (OTM), the % of tail DNAand tail length were the comet parameters chosen to reflect theDNA damage caused by CuONPs and Cu2þ exposure (Fig. 4A, C, E) aswell as Ag NPs and Agþ in hemolymph cells (Fig. 4B, D, F).

In hemocytes from unexposed mussels (either in Cu and Agexperiments) no significant changes were detected in any of theparameters during the course of the experiment (p > 0.05, Fig. 4).The OTM, % of tail DNA and tail length were always higher in Cu-exposed mussels than in unexposed ones, except for the third dayof exposure, where no significant changes occurred (p > 0.05).Additionally, significant differences existed between levels of DNAdamage of both nano and ionic copper and silver (at 7 and 15 days),with higher damage inmussels exposed to Cu2þ and Agþ (p< 0.05).Exposure to CuO NPs showed an increase in the OTM following 7

Fig. 1. Transmission electron microscopic image of CuO NPs (A) and Ag NPs (B) at 32 ppm in Milli-Q water. Particle size distribution (nm) during a 12 h cycle by dynamic lightscattering for CuO NPs (C) and Ag NPs (D).

T. Gomes et al. / Marine Environmental Research 84 (2013) 51e5954

days of exposure, that remained unchanged until the end of theexperiment (p > 0.05, Fig. 4A). The same pattern was found for the% of tail DNA and tail length (Fig. 4C and E), with no significantchanges during the remaining period of exposure (p > 0.05).Regression analysis showed a linear relationship between time ofexposure and OTM (Fig. 4A) and comet tail length (Fig. 4E)following exposure to Cu2þ (Table 2, p < 0.05). As for % of tail DNA,

Fig. 2. Condition index of mussels from controls and exposed to CuO and Cu2þ (A) and Ag Nwhole soft tissues of musselsM. galloprovincialis (mg g�1 d.w.) from controls and exposed to Cstatistical differences between treatments in each exposure day and for each treatment du

an increase was also detected in exposed mussels from the begin-ning until the 15th day of exposure (p < 0.05), although not linear(Fig. 4C).

For Ag, the OTM, % of tail DNA and tail length were alwayshigher in exposed mussels than in unexposed ones, except forthe 3th day of exposure where no significant changes occurredin OTM and % of Tail DNA in Ag NPs exposed ones (p > 0.05).

Ps and Agþ (B) for 15 days. Copper concentrations (C) and silver concentrations (D) inu and Ag (NPs and ionic) for 15 days (average� Std). Capital and lower letters representring the exposure duration, respectively (p < 0.05).

Table 2Results of regression analysis between metal concentrations in tissue and cometparameters for Cu and Ag exposures along the exposure period (p < 0.05).

Exposure Parameter Induction rate r2

CuO NPs Concentration 0.3 mg g�1 d�1 0.99Cu2þ OTM 0.9 au d�1 0.99

Tail length 6.0 mm d�1 0.98Ag NPs OTM 0.2 au d�1 0.99

Tail length 1.2 mm d�1 0.99Agþ OTM 0.7 au d�1 0.99

Tail length 4.0 mm d�1 0.99

T. Gomes et al. / Marine Environmental Research 84 (2013) 51e59 55

Exposure to Ag NPs showed a linear increase in OTM (Fig. 4B)and tail length (Fig. 4F) over time (Table 2). The % of tail DNAalso increased with time of exposure, although not linearly(Fig. 4D). Similarly to Ag NPs, following exposure to Agþ, OTM(Fig. 4B) and comet tail length (Fig. 4F) increased linearly alongthe course of the experiment (Table 2, p < 0.05), whereas for %of tail DNA, an increase was also detected in exposed mussels(p < 0.05, Fig. 4D).

Fig. 3. Examples of comet assay images, recorded with an optical fluorescence microscopeControl M. galloprovincialis hemocytes showing a comet head (nucleoid core) with no DNAAgþ exposure, respectively, showing a comet head (nucleoid core) with broken DNA fragm

Although DNA damage in mussel hemocytes increased withtime of exposure, no significant relationship was found betweenDNA damage, copper and silver concentrations (p > 0.05).

The percentage of DNA in the tail was used to categorize thegrade of damage in unexposed and exposed mussels (Table 3). Themajority of control cells showed minimal and low grade of damage(>95% of the cells), characterized by zero or minimal DNA migra-tion toward the anode. A small proportion of the control hemocytesexhibited mid damage, probably related to some strand breaksoccurring in the field (e.g. environmental and/or pollutant stress) ororiginated during the steps of the comet assay (e.g. cell isolationand/or processing). The results obtained for control mussels are inthe same range as those found in mussels collected in the southcoast of Portugal (Almeida et al., 2011). As for mussels exposed toCuO NPs and Ag NPs, >90% of hemocytes also presented minimaland low DNA damage, while for those exposed to Cu2þ and Agþ

fewer cells showed minimal and low damage with increasing timeof exposure. For both CuO and Ag NPs and respective ionic forms,a significant percentage of cells with mid and high damageincreased with time, namely after 15 days of exposure.

(Axiovert S100) coupled to a camera (Sony) using a total magnification of �400. AeB)migrating into the tail region; CeD) CuO NPs and Cu2þ exposure and EeF) Ag NPs andents or damaged DNA migrating away from the nucleus into the tail region.

Fig. 4. DNA damage in hemocytes of musselsM. galloprovincialis from control and exposed to CuO NPs and Cu2þ and Ag NPs and Agþ for 15 days (average� SEM, n ¼ 500) expressedas Olive Tail Moment (A, B), Tail DNA (C, D) and Tail Length (E, F), respectively. Capital and lower letters represent statistical differences between treatments in each day of exposureand for each treatment during the exposure duration, respectively (p < 0.05).

Table 3Percentage of cells distributed by grade of DNA damage in hemocytes of musselsM. galloprovincialis from controls (CT) and exposed to CuO NPs, Cu2þ, Ag NPs andAgþ for 15 days (n ¼ 500).

Time of exposure DNA damage criteria

Days Minimal Low Mid High Extreme

0 CT 68.8 29.8 1.4 0 03 CT 69.8 28.6 1.6 0 0

CuO NPs 77.6 19.6 2.8 0 0Cu2þ 73.4 24.0 2.6 0 0

7 CT 74.4 24.0 1.6 0 0CuO NPs 63.0 29.4 7.6 0 0Cu2þ 57.4 31.2 11.0 0.4 0

15 CT 81.8 18.1 0.1 0 0CuO NPs 70.4 23.4 5.6 0.6 0Cu2þ 38.6 28.6 29.3 3.5 0

0 CT 68.0 30.2 1.8 0 03 CT 68.0 29.0 3.0 0 0

Ag NPs 66.2 30.6 3.2 0 0Agþ 57.6 40.4 2.0 0 0

7 CT 64.8 32.8 2.4 0 0Ag NPs 61.4 29.8 8.8 0 0Agþ 56.3 30.6 12.6 0.5 0

15 CT 73.6 24.6 1.8 0 0Ag NPs 56.0 34.6 8.8 0.6 0Agþ 53.4 28.2 16.8 1.6 0

T. Gomes et al. / Marine Environmental Research 84 (2013) 51e5956

4. Discussion

Nanomaterials have unpredictable genotoxic properties withnumerous mechanisms controlling their capacity to promote DNAdamage, making essential to investigate their genotoxic potential.Direct or indirectly, damage can occur not only generated by directparticles influence through their high reactivity and surface areaand/or physical and chemical properties, by transition metalscomprised or released from the particles (e.g. Cu, Fe, Cd) that havethe ability to produce ROS and generate oxidative stress. The maingenotoxic effect of nanoparticles comes from the induction ofcellular responses or stimulation of target cells that can producecompoundswith oxidant or genotoxic capacity, leading to oxidativeand inflammatory processes. Additionally, cellular internalizationof nanoparticles may promote direct interaction with DNA insidethe nucleus or even during mitosis where they can induce variousDNA damages (Bhatt and Tripathi, 2011; Handy et al., 2008;Karlsson, 2010; Singh et al., 2009).

The results obtained from the comet assay showed that CuO NPsare genotoxic to mussels’ hemocytes after 7 days of exposure. Atime-lag response effect of CuO NPs on DNA damage was evidentand resulted in a significant increase in the Olive tail moment, % oftail DNA and tail length when compared to control mussels, exceptat the first 3 days of exposure. Cu2þ presented a similar but highergenotoxic trend as its nano counterpart, shown not only by the

T. Gomes et al. / Marine Environmental Research 84 (2013) 51e59 57

significant linear increase in DNA strand breaks with time(demonstrated by the comet parameters) but also by the significantincrease in the percentage of cells with mid and high damage,namely after 15 days of exposure. These results are the opposite ofthose reported by other authors, that showed higher cytotoxic andgenotoxic capacity of CuO NPs compared to bulk and micrometreCuO and Cu2þ (e.g. Griffitt et al., 2007; Karlsson et al., 2008). Thegenotoxicity of Cu2þ in mussels hemocytes may be related to itsinvolvement in the formation of ROS and subsequent oxidativedamage as reported for different aquatic species, namely in theform of DNA strand breaks, oxidation of bases and apoptotic DNAfragmentation (Al-Subiai et al., 2011; Bolognesi et al., 1999; Steinert,1995). As for CuO NPs, some information exist on their genotoxicpotential to human cells, where DNA damage and oxidative DNAlesions by these particles seem to be mediated through oxidativeDNA damage rather than direct interaction with the DNA (Ahamedet al., 2010; Karlsson et al., 2008; Midander et al., 2009).

As for Ag, the present results demonstrate that both Ag formsexerted genotoxic effects in mussels’ hemocytes, which is inaccordancewith previous studies on the genotoxicity of Ag NPs andAgþ (Ahamed et al., 2010; Asharani et al., 2009; Choi et al., 2010;Hackenberg et al., 2011; Park and Choi, 2010). DNA strand breaks(OTM, % Tail DNA and Tail Length) increased with time of exposurein Ag NPs-exposed mussels, even though OTM and % Tail DNAweresimilar to controls at day 3. On the other hand, the degree ofincrease in DNA damage was higher in Agþ-exposed mussels thanin those exposed to Ag NPs (and control mussels), reflected not onlyby the significant increase in comet parameters but also by theincrease in the percentage of cells with mid and high damage(particularly at the end of the experiment). There have beencontradictory discussions regarding the comparative toxicity of AgNPs and Agþ, where Agþ seems to have a higher genotoxic potentialthan Ag NPs in various cell types (e.g. Asharani et al., 2008; Bar-Ilanet al., 2009; Park and Choi, 2010). Based on the greater tendency ofAgþ ion to strongly interact with sulfhydryl groups of vital enzymesand phosphorus-containing bases, it is likely that Agþ can interactwith DNA directly by the formation of ROS, causing damage bycovalent binding to DNA (Hossain and Huq, 2002) or by inhibitingDNA synthesis (Hidalgo and Dominguez,1998), thus preventing celldivision and DNA replication (Bar-Ilan et al., 2009; Singh et al.,2009). As for Ag NPs, the exact outcome of these NPs in thenucleus is not clear yet, but it is expected to interfere with DNAsynthesis, DNA damage, chromosomalmorphology and segregationand DNA and cell division. Genotoxic potential of Ag NPs todifferent cell types either by direct (direct interaction of NPs withDNA) or indirect mechanisms (oxidative stress and ROS), resultingin oxidative lesions as DNA strand breaks, induction of specificmolecular markers (e.g. p53 protein) and DNA repair proteins (e.g.Rad51 protein), chromosomal aberrations and formation ofmicronuclei were reported (Ahamed et al., 2010; Asharani et al.,2009; Choi et al., 2010; Hackenberg et al., 2011; Park and Choi,2010). The different extent in DNA damaging effects in musselhemocytes by both CuO NPs and Ag NPs and corresponding ionicforms suggest different toxic mechanisms involved in the occur-rence of DNA damage or different degrees of reactivity of Cu and Agthat are dependent on the metal form (nano vs. ionic). In fact,different surface chemistry of NPs can result in different effects ongenotoxicity (Ahamed et al., 2010). If extracellular ions dissolutionexplained the effects seen, the genotoxic response of both nano-metals should be identical to the ones of Cu2þ and Agþ (same massadded). This is in accordance with the results that reported that thecytotoxicity of these NPs is not caused by the soluble fraction alone(e.g. Asharani et al., 2008; Griffitt et al., 2007, 2009; Midander et al.,2009). In fact, most of the CuO and Ag NPs in solution are in thenanoparticulate form, with less than 1% of the initial added dose in

the dissolved form for CuO NPs (Gomes et al., 2011) and about 24%for Ag NPs. These findings suggest that CuO NPs and Ag NPsexposed hemocytes are exposed to fewer metal ions than hemo-cytes exposed with corresponding concentrations of Cu2þ and Agþ,even considering the potential intracellular release of these ionsfrom the NPs and the inefficient absorption of Cu2þ and Agþ ionsinto the cells due to the efficient barrier capacity of cellularmembranes formost ions. Furthermore, reduced amounts of metalsions are entering the nucleus and interacting with DNA. Thus, thegenotoxic potential of both NPs are not solely explained by therelease of metal ions from the particles but also by the effects of theparticle properties (e.g. particle size, surface charge processes). Itwas already suggested that CuO and Ag NPs are taken up into cellsand generate radicals that alter or interfere with the metabolicactivity of various organelles and inevitably lead to DNA damage(Asharani et al., 2009; Bhatt and Tripathi, 2011; Gomes et al., 2011;Karlsson et al., 2008). Other mechanisms as the extracellularrelease of Cu and Ag ions from the particles or even particle surfaceprocesses are also possible mechanisms by which CuO and Ag NPscan indirectly interact with DNA. Additionally, the hypothesis thatCuO NPs and Ag NPs pass through the cell nucleus cannot beexcluded, since Cu2þ is known to accumulate in the nucleus andbind to DNA to form adducts (Sagripanti et al., 1991) and Ag NPs canaccumulate in the nuclei of various cell types and lead to genomicdamage, instability and chromosomal aberrations (Asharani et al.,2008, 2009; Hackenberg et al., 2011). On the other hand, Cu andAg NPs can also induce mitochondrial dysfunction and induction ofoxidative stress, which in turn set off DNA damage (e.g. Asharaniet al., 2009; Manna et al., 2012). Nevertheless, no clear data existson the exact mechanism by which CuO and Ag NPs promote DNAstrand breaks. Based on what is already known on these nano-metals toxicity in mussels (Gomes et al., 2011; Gomes et al., owndata), it is proposed that the DNA damage seen in mussel hemo-cytes is mediated by oxidative stress. Hemocytes, as components ofthe open circulatory system of molluscs, are highly susceptible toNPs uptake and toxicity (e.g. nano carbon black, C60 fullerene) andchanges in the immune parameters and ROS production were alsoreported (Canesi et al., 2012 and literature cited therein).

At present little information is available on the bioavailability ofnanoparticles to aquatic organisms, route of exposure and uptakemechanisms, as well as their ingestion rates, internal exposureconcentrations and cell and tissue distribution. The challenge lies inthe development of methods that allow accurate detection andquantification of the uptake of nanoparticles, mode of action, anddistribution inside tissues, cells and sub-cellular components of theorganisms. Additionally, numerous constraints arise when usingnanoparticles, related not only to their size and quantity but also toinherent particle properties (Bhatt and Tripathi, 2011; Canesi et al.,2012; Ward and Kach, 2009). The current literature suggestsnanoparticles uptake in bivalve molluscs (see Canesi et al., 2012),but little is known on the cellular uptake of CuO and Ag NPs andtheir accumulation and effects in bivalve tissues. The presentresults on bioaccumulation in mussel tissues showed that both Cuand Ag are bioavailable to mussels regardless of the form used (NPsor ionic). Exposure to CuO NPs resulted in an increasing rate of Cuaccumulation in mussel tissues with time, while for Cu2þ by the15th day of exposure Cu was already eliminated. These results arein line with those observed in mussels gills exposed to the samenanoparticles, where the elimination rate of CuO NPs was slowerthat its accumulation, in contrast to Cu2þ exposure (Gomes et al.,2011). As for Ag, exposure to Ag NPs and Agþ resulted in a signifi-cant accumulation in the first week of exposure while by the end ofthe experiment both Ag forms were eliminated from musseltissues. Bivalve species have developed various regulation/detoxi-fication mechanisms to cope with excess of metals (as Cu and Ag),

T. Gomes et al. / Marine Environmental Research 84 (2013) 51e5958

as the binding to metallothioneins (Amiard et al., 2006; Bolognesiet al., 1999; Langston et al., 1998), the accumulation of Ag ininsoluble sulfide complexes (Berthet et al., 1992; Shi et al., 2003), orthe preferential distribution of metals to other parts of the bodysuch as the valves, byssus threads andmucus excretion (Gundacker,1999; McDonald et al., 1993; Nicholson and Lam, 2005; Zuykovet al., 2011), which could explain the pattern for Cu2þ, Ag NPsand Agþ accumulation. Mussels were more efficient in eliminatingAg NPs than CuO NPs, probably due to a higher propensity of theseparticles to dissolve with time (24%) compared to CuO NPs (<1%),evidencing the importance of dissolved metals in DNA damagingeffects. The increasing DNA damage of CuO NPs, Cu2þ, Ag NPs andAgþ did not follow metal accumulation in mussels’ whole softtissues. This may be related to different accumulation patterns thatreflect distinct physiological and metabolic functions of tissues thatwhen measured in the whole tissues are not able to discriminatetissue responses. As filter-feeding organisms, bivalves have thecapacity to distribute nanoparticles throughout their organs andrespective cells, where capture and ingestion are the major routesof internal exposure and potential effects. Once exposed, cells willtake up NPs through different mechanisms, as phagocytosis,pinocytosis, endocytosis and direct penetration via disturbance ofmembrane components (Bhatt and Tripathi, 2011; Canesi et al.,2012; Handy et al., 2008; Moore, 2006; Ward and Kach, 2009). Inshort-term exposure where most of the nanoparticles are presentin the form of aggregates, as in the present experimental condi-tions, NPs aggregates are taken up mainly by the digestive systemwhile gills are more sensitive to dissolved metals (Canesi et al.,2010, 2012; Gomes et al., 2011, 2012; Gomes et al., own data;Ward and Kach, 2009). Nevertheless, evidence exists of a sortingprocess in the gills, followed by the transport of larger particles andaggregates to the digestive gland. During this transport, aggregatesare broken down in smaller particles, accumulate and transferredto the hemolymph and circulating hemocytes (Browne et al., 2008;Canesi et al., 2010; Moore et al., 2009; Ward and Kach, 2009). Eventhough bivalve hemocytes participate in the transport of metalsand nanoparticles, particle translocation is dependent on theparticle size or other particle properties (Moore et al., 2009; Zuykovet al., 2011). In the present case, the existence of large aggregates(as seen by the DLS studies) explain the lack of CuO and Ag NPsavailability and uptake by the hemocytes (that probably areingested/incorporated by the gills and digestive system), resultingin a lower genotoxic potential of CuO and Ag NPs when comparedto its ionic forms. In Mytilus edulis exposed to polystyrene micro-spheres, the particle size influenced their capacity to translocatefrom the gut cavity to the hemolymph (Browne et al., 2008). Thelack of relationship between Cu and Ag concentrations in musselwhole tissues and DNA damage in hemocytes reinforces the ideathat NPs availability between tissues is different (Gomes et al., 2011,2012; Gomes et al., own data). Time is also an important factor inDNA damaging effects and accumulation, and further work isneeded to determine if and how quick particles translocate fromwater/organs to the hemolymph and the mechanisms by whichthese particles are taken up by hemocytes and accumulate in thehemolymph. Further studies using doseeresponse relationshipsafter exposure to CuO NPs and Ag NPs of different sizes will help todefine target cell types and endpoints. Despite these evidences(Gomes et al., 2011, 2012 and the present study) no clear insightsexist on CuO and Ag NPs mediated genotoxicity and accumulationin mussels’ hemocytes.

5. Conclusions

Overall, the present results suggested that CuO NPs and Ag NPsare genotoxic towards aquatic organisms (as Cu2þ and Agþ), namely

filter-feeding bivalves, contributing to the knowledge on aquatictoxicology of two of the most widely used NPs. Both particlespossessed DNA damaging potential in hemocytes that may bemediated through oxidative stress, even though less toxic than itsionic counterparts. The measurement of genotoxic effects ofemerging nanomaterials using the comet assay is a useful tool formonitoring toxicitydue tonanoparticles in themarineenvironment.

Acknowledgements

This work was supported by a Foundation of Science andTechnology PhD Grant (SFRH/BD/41605/2007) and the NANO-ECOTOX (PTDC/AAC-AMB/121650/2010) and GENERA (FP7-PEOPLE-2009-IRSES) projects.

References

Ahamed, M., Siddiqui, M., Akhtar, M.J., Ahmad, I., Pant, A.B., Alhadlaq, H.A., 2010.Genotoxic potential of copper oxide nanoparticles in human lung epithelialcells. Biochem. Biophys. Res. Commun. 396, 578e583.

Almeida, C., Pereira, C., Gomes, T., Bebianno, M.J., Cravo, A., 2011. DNA damage asa biomarker of genotoxic contamination in Mytilus galloprovincialis from thesouth coast of Portugal. J. Environ. Monit. 13, 2559e2567.

Al-Subiai, S.N., Moody, A.J., Mustafa, S.A., Jha, A.N., 2011. A multiple biomarkerapproach to investigate the effects of copper on the marine bivalve mollusc,Mytilus edulis. Ecotoxicol. Environ. Saf. 74, 1913e1920.

Amiard, J.-C., Amiard-Triquet, C., Barka, S., Pellerin, J., Rainbow, P.S., 2006. Metal-lothioneins in aquatic invertebrates: their role in metal detoxification and theiruse as biomarkers. Aquat. Toxicol. 76, 160e202.

Asharani, P.V., Wu, Y.L., Gong, Z., Valiyaveettil, S., 2008. Toxicity of silver nano-particles in zebrafish models. Nanotechnology 19, 255102.

Asharani, P.V., Mun, G.L.K., Hande, M.P., Valiyaveettil, S., 2009. Cytotoxicity andgenotoxicity of silver nanoparticles in human cells. ACS Nano 3 (2), 279e290.

Bar-Ilan, O., Albrecht, R.M., Fako, V.E., Furgeson, D.Y., 2009. Toxicity assessments ofmultisized gold and silver nanoparticles in zebrafish embryos. Small 5 (16),1897e1910.

Berthet, B., Amiard, J.C., Amiard-Triquet, C., Martoja, M., Jeantet, A.Y., 1992. Bio-accumulation, toxicity and physico-chemical speciation of silver in bivalvemolluscs: ecotoxicological and health consequences. Sci. Total. Environ. 125,97e122.

Bhatt, I., Tripathi, B.N., 2011. Interaction of engineered nanoparticles with variouscomponents of the environment and possible strategies for their risk assess-ment. Chemosphere 82 (3), 308e317.

Bolognesi, C., Landini, E., Roggieri, P., Fabbri, R., Viarengo, A., 1999. Genotoxicitybiomarkers in the assessment of heavy metal effects in mussels: experimentalstudies. Environ. Mol. Mutagen. 33, 287e292.

Browne, M.A., Dissanayake, A., Galloway, T.S., Lowe, D.M., Thompson, R.C., 2008.Ingested microscopic plastic translocates to the circulatory system of themussel, Mytilus edulis (L.). Environ. Sci. Technol. 42, 5026e5031.

Bryan, G.W., Langston, W.J., 1992. Bioavailability, accumulation and effects of heavymetals in sediments with special reference to United Kingdom estuaries:a review. Environ. Pollut. 76, 89e131.

Canesi, L., Fabbri, R., Vallotto, D., Marcomini, A., Pojana, G., 2010. Biomarkers inMytilus galloprovincialis exposed to suspensions of selected nanoparticles(Nano carbon black, C60 fullerene, Nano-TiO2, Nano-SiO2). Aquat. Toxicol. 100,168e177.

Canesi, L., Ciacci, C., Fabbri, R., Marcomini, A., Pojana, G., Gallo, G., 2012. Bivalvemolluscs as a unique target group for nanoparticle toxicity. Mar. Environ. Res.76, 16e21.

Choi, J.E., Kim, S., Ahn, J.H., Youn, P., Kang, J.S., Park, K., Yi, J., Ryu, D.-Y., 2010.Induction of oxidative stress and apoptosis by silver nanoparticles in the liver ofadult zebrafish. Aquat. Toxicol. 100, 151e159.

Fabrega, J., Luoma, S.N., Tyler, C.R., Galloway, T.M., Lead, J.R., 2011. Silver nano-particles: behaviour and effects in the aquatic environment. Environ. Int. 37 (2),517e531.

Fahmy, B., Cormier, S.A., 2009. Copper oxide nanoparticles induce oxidative stressand cytotoxicity in airway epithelial cells. Toxicol. in Vitro 23, 1365e1371.

Gagné, F., Auclair, J., Turcotte, P., Fournier, M., Gagnona, C., Sauvé, S., Blaise, C., 2008.Ecotoxicity of CdTe quantum dots to freshwater mussels: impacts on immunesystem, oxidative stress and genotoxicity. Aquat. Toxicol. 86, 333e340.

Gomes, T., Pinheiro, J.P., Cancio, I., Pereira, C.G., Cardoso, C., Bebianno, M.J., 2011.Effects of copper nanoparticles exposure in the mussel Mytilus galloprovincialis.Environ. Sci. Technol. 45 (21), 9356e9362.

Gomes, T., Pereira, C.G., Cardoso, C., Pinheiro, J.P., Cancio, I., Bebianno, M.J., 2012.Accumulation and toxicity of copper oxide nanoparticles in the digestive glandof Mytilus galloprovincialis. Aquat. Toxicol. 118e119, 72e79.

Griffitt, R.J., Weil, R., Hyndman, K.A., Denslow, N.D., Powers, K., Taylor, D.,Barber, S.D., 2007. Exposure to copper nanoparticles causes gill injury and acutelethality in zebrafish (Danio rerio). Environ. Sci. Technol. 41, 8178e8186.

T. Gomes et al. / Marine Environmental Research 84 (2013) 51e59 59

Griffitt, R.J., Hyndman, K., Denslow, N.D., Barber, D.S., 2009. Comparison of molec-ular and histological changes in zebrafish gills exposed to metallic nano-particles. Toxicol. Sci. 107 (2), 404e415.

Gundacker, C., 1999. Tissue-specific heavy metal (Cd, Pb, Cu, Zn) deposition ina natural population of the zebra mussel Dreissena polymorpha Pallas. Che-mosphere 38, 3339e3356.

Hackenberg, S., Scherzed, A., Kessler, M., Hummel, S., Technau, A., Froelich, K.,Ginzkey, C., Koehler, C., Hagen, R., Kleinsasser, N., 2011. Silver nanoparticles:evaluation of DNA damage, toxicity and functional impairment in humanmesenchymal stem cells. Toxicol. Lett. 201, 27e33.

Halliwell, B., Gutteridge, J.M.C., 1984. Oxygen toxicity, oxygen radicals, transitionmetals and disease. Biochem. J. 219, 1e14.

Handy, R.D., Kammer, F., Lead, J.R., Hassellöv, M., Owen, R., Crane, M., 2008. Theecotoxicology and chemistry of manufactured nanoparticles. Ecotoxicology 17,287e314.

Hidalgo, E., Dominguez, C., 1998. Study of cytotoxicity mechanisms of silver nitratein human dermal fibroblasts. Toxicol. Lett. 98, 169e179.

Hossain, Z., Huq, F., 2002. Studies on the interaction between Agþ and DNA. J. Inorg.Biochem. 91, 398e404.

Kádar, E., Tarran, G.A., Jha, A.N., Al-Subiai, S.N., 2011. Stabilization of engineeredzero-valent nanoiron with Na-acrylic copolymer enhances spermiotoxicity.Environ. Sci. Technol. 45, 3245e3251.

Karlsson, H.L., Cronholm, P., Gustafsson, J., Möller, L., 2008. Copper oxide nano-particles are highly toxic: a comparison between metal oxide nanoparticles andcarbon nanotubes. Chem. Res. Toxicol. 21, 1726e1732.

Karlsson, H.L., 2010. The comet assay in nanotoxicology research. Anal. Bioanal.Chem. 398, 651e666.

Langston, W.J., Bebianno, M.J., Burt, G.R., 1998. Metal handling strategies inmolluscs. In: Langston, W.J., Bebianno, M.J. (Eds.), Metal Metabolism in AquaticEnvironments. Chapman & Hall, London, pp. 219e284 (Chapter 8).

Luoma, S.N., 2008. Silver Nanotechnologies and the Environment: Old Problems orNew Challenges. Project on Emerging Nanotechnologies. Publication 15.Woodrow Wilson International Center for Scholars and PEW Charitable Trusts,Washington, DC.

Manna, P., Ghosh, M., Ghosh, J., Das, J., Sil, P.C., 2012. Contribution of nano-copperparticles to in vivo liver dysfunction and cellular damage: role of IkBa/NF-kB,MAPKS and mitochondrial signal. Nanotoxicology 6 (1), 1e21.

McDonald, P., Baxter, M.S., Fowler, S.W., 1993. Distribution of radionuclides inmussels, winkles and prawns. Part 2. Study of organisms under laboratoryconditions using alpha-autoradiography. J. Environ. Radioact. 18, 203e228.

Midander, K., Cronholm, P., Karlsson, H.L., Elihn, K., Möller, L., Leygraf, C.,Wallinder, I.O., 2009. Surface characteristics, copper release, and toxicity ofnano- and micrometer-sized copper and copper(II) oxide particles: a cross-disciplinary study. Small 5 (3), 389e399.

Moore, M.N., 2006. Do nanoparticles present ecotoxicological risks for the health ofthe aquatic environment? Environ. Int. 32, 967e976.

Moore, M.N., Readman, J.A.J., Readman, J.W., Lowe, D.M., Frickers, P.E., Beesley, A.,2009. Lysosomal cytotoxicity of carbon nanoparticles in cells of the molluscanimmune system: an in vitro study. Nanotoxicology 3 (1), 40e45.

Nicholson, S., Lam, P.K.S., 2005. Pollution monitoring in Southeast Asia usingbiomarkers in the mytilid mussel Perna viridis (Mytilidae: Bivalvia). Environ. Int.31, 121e132.

Park, S.-Y., Choi, J., 2010. Geno- and ecotoxicity evaluation of silver nanoparticles infreshwater crustacean Daphnia magna. Environ. Eng. Res. 15 (1), 23e27.

Sagripanti, J.L., Goering, P.L., Lamanna, A., 1991. Interaction of copper with DNA andantagonism by other metals. Toxicol. Appl. Pharmacol. 110, 477e485.

Shi, D., Blackmore, G., Wang, W.-X., 2003. Effects of aqueous and dietary pre-exposure and resulting tissue concentration on silver biokinetics in the greenmussel Perna viridis. Environ. Sci. Technol. 37, 936e943.

Singh, N.P., McCoy, M.T., Tice, R.R., Schneider, E.L., 1988. A simple technique forquantitation of low levels of DNA damage in individual cells. Exp. Cell. Res. 175(1), 184e191.

Singh, N., Manshian, B., Jenkins, G.J.S., Griffiths, S.M., Williams, P.M., Maffeis, T.G.G.,Wrigh, C.J., Doak, S.H., 2009. NanoGenotoxicology: the DNA damaging potentialof engineered nanomaterials. Biomaterials 30, 3891e3914.

Steinert, S.A., 1995. Contribution of apoptosis to observed DNA damage in musselcells. Mar. Environ. Res. 42 (1e4), 253e259.

Valavanidis, A., Vlahogianni, T., Dassenakis, M., Scoullos, M., 2006. Molecularbiomarkers of oxidative stress in aquatic organisms in relation to toxic envi-ronmental pollutants. Ecotoxicol. Environ. Saf. 64, 178e189.

Ward, J.E., Kach, D.J., 2009. Marine aggregates facilitate ingestion of nanoparticlesby suspension-feeding bivalves. Mar. Environ. Res. 68, 137e142.

Zuykov, M., Pelletier, E., Demers, S., 2011. Colloidal complexed silver and silver nano-particles in extrapallial fluid ofMytilus edulis. Mar. Environ. Res. 71 (1), 17e21.