Multibiomarker assessment of cerium dioxide nanoparticle (nCeO2) sublethal effects on two freshwater invertebrates, Dreissena polymorpha and Gammarus roeseli

Aquatic Toxicology 158 (2014) 63–74

Contents lists available at ScienceDirect

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

Multibiomarker assessment of cerium dioxide nanoparticle (nCeO2 ) sublethal effects on two freshwater invertebrates, Dreissena polymorpha and Gammarus roeseli M. Garaud a,b , J. Trapp a , S. Devin a , C. Cossu-Leguille a , S. Pain-Devin a , V. Felten a , L. Giamberini a,b,∗ a Université de Lorraine, CNRS UMR 7360, Laboratoire Interdisciplinaire des Environnements Continentaux (LIEC), Campus Bridoux, Rue du Général Delestraint, 57070 Metz, France b International Consortium for the Environmental Implications of Nanotechnology (iCEINT), Aix en Provence, France

a r t i c l e

i n f o

Article history: Received 20 June 2014 Received in revised form 4 November 2014 Accepted 5 November 2014 Keywords: Nanoparticle Cerium Biomarker Ecotoxicology Dreissena polymorpha Gammarus roeseli

a b s t r a c t Cerium nanoparticles (nCeO2 ) are widely used in everyday products, as fuel and paint additives. Meanwhile, very few studies on nCeO2 sublethal effects on aquatic organisms are available. We tried to fill this knowledge gap by investigating short-term effects of nCeO2 at environmentally realistic concentrations on two freshwater invertebrates; the amphipod Gammarus roeseli and the bivalve Dreissena polymorpha, using an integrated multibiomarker approach to detect early adverse effects of nCeO2 on organism biology. Differences in the behaviour of the organisms and of nanoparticles in the water column led to differential nCeO2 bioaccumulations, G. roeseli accumulating more cerium than D. polymorpha. Exposure to nCeO2 led to decreases in the size of the lysosomal system, catalase activity and lipoperoxidation in mussel digestive glands that could result from nCeO2 antioxidant properties, but also negatively impacted haemolymph ion concentrations. At the same time, no strong adverse effects of nCeO2 could be observed on G. roeseli. Further experiments will be necessary to confirm the absence of severe nCeO2 adverse effects in long-term environmentally realistic conditions. © 2014 Elsevier B.V. All rights reserved.

1. Introduction As nanotechnologies are quickly developing, emission of manufactured nanomaterials into the environment and subsequent biota exposure are bound to drastically increase, as well as concerns about their potential adverse impact (Moore, 2006; Sun et al., 2014). The French MESONNET ANR (2010–2014) program aims to evaluate the fate and impact of manufactured nanoparticles in terrestrial and aquatic ecosystems at environmentally relevant concentrations, and with silver and titanium, cerium nanoparticles (nCeO2 ) are among the priority nanomaterials to evaluate (OECD, 2010). Macrometric cerium (Ce) oxides have been used for a long time for polishing optical components. Promising nCeO2 uses have been expanding quickly in recent decades, mostly as fuel additives to decrease CO, NOx and ultrafine particle emissions (Jung et al., 2005;

∗ Corresponding author at: Université de Lorraine, CNRS UMR 7360, Laboratoire Interdisciplinaire des Environnements Continentaux (LIEC), Campus Bridoux, Rue du Général Delestraint, 57070 Metz, France. Tel.: +33 3 87378415. E-mail address: [email protected] (L. Giamberini). http://dx.doi.org/10.1016/j.aquatox.2014.11.004 0166-445X/© 2014 Elsevier B.V. All rights reserved.

Selvan et al., 2009) but also in wood stains or in cosmetics for its UV-shielding properties (Zholobak et al., 2011; Auffan et al., 2014). Taking only into account its use in fuels, the highest PEC calculation gave a 0.1 ␮g/L to 1 ␮g/L concentration interval in surface waters (O’Brien and Cummins, 2010). The large set of toxicological studies on nCeO2 demonstrated the ability of nCeO2 to act as a reactive oxygen species (ROS) regulator depending on the intracellular pH (Alili et al., 2010; Amin et al., 2011; Wason et al., 2013). Thus, nCeO2 constitute a promising cancer treatment (Alili et al., 2010; Colon et al., 2010, 2009), and could also find applications for their neuroprotective (Das et al., 2007), wound healing (Chaudhury et al., 2012; Chigurupati et al., 2013) or angiogenesis promoting properties (Das et al., 2012). Antioxidant properties of nCeO2 could result from oxygen vacancies in the crystal lattice surface caused by the presence of Ce in the trivalent state that could provide reaction sites for ROS trapping (Ciofani et al., 2014; Korsvik et al., 2007; Xue et al., 2011). Molecular mechanism investigations of nCeO2 antioxidant properties have shown their catalase (CAT) and superoxide dismutase (SOD) mimetic activities (Das et al., 2007; Korsvik et al., 2007; Pirmohamed et al., 2010) while mRNA expression data have confirmed their ROS trapping

64

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

properties (Ciofani et al., 2014). Most of these studies have been conducted in vitro, but those antioxidant properties have also been observed in vivo on murine models (Amin et al., 2011). In contrast, only a few aquatic ecotoxicological studies investigating environmental effects of nCeO2 have been published before now. In addition to that, most of them have focused on the acute toxicity of nCeO2 and have reported low toxicity in bacteria, algae, invertebrates and fishes, with EC50 of at least thousand times the PEC (Manier et al., 2011; Pelletier et al., 2010; Van Hoecke et al., 2011, 2009). Nevertheless, a multibiomarker approach would be useful to screen nCeO2 potential effects on aquatic organisms. In fact, this approach is increasingly used as a sensitive tool to assess precocious impacts of stressors on organisms, with the aim of predicting long-term effects of those stressors at high levels of biological organization (Guerlet et al., 2010; Moore et al., 2004; Sornom et al., 2010; Vasseur and Cossu-Leguille, 2003). In addition to that, the study of biomarkers related to different organism functions (antioxidant system, osmoregulation, etc.), and at different levels of the biological scale (molecular, cellular, individual, etc.) allows a better understanding of the mechanisms underlying the effects of stressors (Moore et al., 2004). The usefulness of such complementary multibiomarker batteries as early warning tools has been confirmed for freshwater invertebrate health assessment studies (Guerlet et al., 2006, 2007). In fact, studies on zebra mussels have shown that cellular endpoints responded early to environmental contamination and preceded effects at the individual level (Guerlet et al., 2010). Moreover, the complex dataset obtained can be further valued by an integrated biomarker analysis in order to synthesize and highlight stressor effects (Guerlet et al., 2010, 2007). One of the few biomarker studies dealing with nCeO2 effects on organisms report no effects on antioxidant defences and osmo-regulatory biomarkers under 40 mg/L in fish (Xia et al., 2013). Therefore, there is clearly a lack of data on the potential sublethal effects of nCeO2 on aquatic organisms at environmentally realistic concentrations. To fill this knowledge gap, we investigated sublethal effects of a waterborne exposure to bare nCeO2 (pristine nanoparticles) at concentrations close to the PEC on two freshwater invertebrate species, the amphipod Gammarus roeseli and the bivalve Dreissena polymorpha. Species ecological traits make them very relevant for nanoparticle toxicity assessment. In fact, D. polymorpha is a filter-feeding organism with highly developed phagocytosis and endocytosis digestive capacities making it a unique target for nanoparticle toxicity (Canesi et al., 2012; Moore, 2006), and the amphipod G. roeseli is a shredder recognized in important ecological processes such as leaf-litter breakdown. In addition to that, we used a very coherent multibiomarker approach at different levels of the biological organization scale to detect precocious effects of nCeO2 on the tested species. Selected biomarkers investigated effects of nCeO2 at molecular, cellular and individual levels, measuring antioxidant and antitoxic defences, cellular damage, iono-osmoregulation, energy reserves and functional parameters on both species. Biomarker responses were synthesized using principal component analysis (PCA).

2. Materials and methods 2.1. Test material The nCeO2 used were bare nanoparticles commercially available supplied as a 130 g/L (750 mM) stable suspension. nCeO2 were synthesized by aqueous precipitation of Ce4+ (NO3 − )4 salt at acidic pH (Auffan, 2007). TEM observations showed well-crystallized pseudo-spherical clusters of cerianite (95–98% purity) with a diameter of 3 ± 1 nm (Artells et al., 2013). Their average hydrodynamic

diameters in the stock suspension were centred on 8 nm and the point of zero charge (PZC) was estimated between 7 and 7.5 (Diot, 2012). nCeO2 suspensions in pure water (100 mg/L or 581 ␮M) were found to be colloidally stable and exhibited negative zeta potential of −40 ± 5 mV at pH 4 (Artells et al., 2013). Therefore, serial dilutions (200 and 20 mg/L or 1162 and 116.2 ␮M) of the stock suspension (130 g/L) were prepared in ultrapure water before each injection to contaminate experimental tanks.

2.2. Collection of organisms D. polymorpha were hand-collected on the walls of a lock on the “Canal de l’Est” (Vadonville, Meuse, France, 48◦ 45 N and 05◦ 36 E) in April 2011 by cutting their byssus, and then transferred quickly to the laboratory. They were selected to have a body length between 20 and 25 mm, their shells were gently cleaned up and they were acclimatized to the experimental conditions for 3 days in continuously aerated spring water (Cristaline® , [Ca2+ ] 106 mg/L; [Mg2+ ] 4.2 mg/L; [Na+ ] 3.5 mg/L; [K+ ] 1.5 mg/L; [HCO3 − ] 272 mg/L; [SO4 2− ] 50 mg/L; [Cl− ] 3.8 mg/L; [F− ] 0.9 mg/L) without feeding. G. roeseli (males only, 7–12 mm of body length) were collected in the Nied river (Remilly, northeastern France, 49◦ 00 N and 6◦ 23 E) using traps and nets, transferred to the laboratory and acclimatized in the same manner.

2.3. Experimental design The experimental design was summarized in a flow chart (Fig. 1). Briefly, after the acclimation period, mussels were randomly divided into 6 experimental groups of 100 individuals each and put into six plastic tanks (31.5 cm × 16.5 cm × 18 cm in height) onto two ceramic tiles (15 cm × 15 cm) per aquarium, used as mussel byssus attachment support. Six experimental groups of 250 gammarids were formed likewise and put into six separate similar tanks. Tanks were filled with 8 L of spring water (Cristaline® ) and kept in the thermostatic chamber at 15 ◦ C and at a photoperiod of 14:10 light: dark ratio. The water was continuously aerated and the organisms were not fed during the exposure period. The experimental groups of each species were separately exposed for 24 h and 96 h to 0 (control), 10 and 100 ␮g/L/d of nCeO2 (0.581 and 0.058 ␮mol/L/d), concentrations corresponding respectively to 10 and 100 times the PEC values (O’Brien and Cummins, 2010). Experimental tanks of organisms exposed for 96 h were contaminated daily by the addition of freshly prepared nCeO2 suspensions. Organisms of both species were sampled at the end of the 24 h and 96 h exposure periods.

2.4. nCeO2 internal concentration analysis Three mussels per condition were dissected and their byssus cut off. The whole soft body was rinsed, dried on absorbent paper, placed in cryotubes and dry-freezed at −80 ◦ C in liquid nitrogen before processing. Three pools of five entire gammarids per condition were constituted and processed the same way. Samples were lyophilized, weighed and digested in 69% HNO3 for 24 h at 60 ◦ C. Digests were then evaporated and 95% sulphuric acid, supplemented with 3 M ammonium sulphate, was introduced and boiled at 250 ◦ C for 1 h to dissolve Ce nanoparticles. Samples were finally transferred to a graduated flask filled to 10 mL with 2% HNO3 . Ce concentrations were analyzed by ICP-MS (PerkinElmer Elan DRCe). The results were expressed in ␮g Ce g−1 dw (dw = dry weight).

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

65

Fig. 1. Flow chart of the experimental design (DG = digestive gland).

2.5. Iono-osmoregulation parameters 2.5.1. Analysis of haemolymph Cl− , Na+ and Ca2+ concentrations Samples of haemolymph were taken from mussels (n = 8 per condition) by pericardiac punction using a micro-syringe and transferred into a gauged 5-␮L capillary. For the gammarids, after drying on filter paper, samples of haemolymph were taken from the telson of each individual (n = 10) using a gauged 5-␮L capillary. The capillaries were centrifuged for 5 min at 6660 × g. After centrifugation, the haemolymph volume was determined, diluted in 2 mL of MQ water and kept at 4 ◦ C before analysis (Felten and Guerold, 2004). Chloride concentrations were determined by ionic chromatography (Dionex 4500i equipped with a Ion Pac AS4A column) whereas sodium and calcium concentrations were measured by atomic absorption spectrometry (Perkin Elmer Analyst 100) (Felten et al., 2008). The results were expressed in mmol/L. 2.5.2. Gill Na+ /K+ -ATPase activity The two gills of each mussel (n = 16) were sampled and placed into tubes containing 250 ␮L of ice-cold buffer (100 mM NaCl, 100 mM Imidazole and 0.1% sodium deoxycholate; pH 7.2). For gammarids (n = 4 pools of 3 organisms per condition), the five first coxal gill pairs were removed, pooled into tubes containing 240 ␮L of ice-cold buffer (NaCl 100 mM; HEPES 100 mM; sodium deoxycholate 0.1%, v/v; pH = 7.2) and stored at −80 ◦ C until analysis. The method used to measure Na+ /K+ -ATPase activity was adapted from Felten et al. (2008) and lies on the specific inhibition of Na+ /K+ -ATPase activity by ouabain. Samples were defrosted on ice and homogenized in a glass potter. 150 ␮L of ice-cold buffer were added and the homogenate was recovered, sonicated at 4 ◦ C for 1 min to break cellular membranes and then centrifuged at 2000 × g for 20 min at 4 ◦ C. The supernatant, containing the plasma membrane fragments, was collected and protein concentration was determined by a colorimetric method (Bio-Rad, France) using BSA as reference. Protein concentrations were adjusted to 3 mg/mL for mussels and to 0.5 mg/mL for gammarids. For the two species, 6 replicates (30 ␮L) per sample were placed into microtubes and incubated 20 min at 4 ◦ C with 220 ␮L of non-inhibiting A medium for 3 of them (100 mM NaCl, 15 mM KCl, 10 mM MgCl2 , 100 mM Hepes, pH 7.2) and with 220 ␮L of the Na+ /K+ -ATPase activity inhibiting B medium for the 3 other (medium A without KCl and with optimal ouabain concentration of 10 nM). 250 ␮L of 5 mM ATP solution was subsequently added to all of the replicates and the medium was incubated at 37 ◦ C for 20 min. The reaction was stopped and the inorganic phosphate (Pi) was revealed by the addition of 500 ␮L

Bonting reagent (8.1 mM ammonium molybdate; H2 SO4 3.07% (v/v); 321.6 mM FeSO4 ). After 20 min at 4 ◦ C, the concentration of Pi was determined spectrophotometrically at 700 nm against a Pi standard calibration curve. Na+ /K+ -ATPase specific activity was calculated as the difference in the amount of Pi produced in absence and in presence of ouabain, and expressed in mmol Pi g−1 protein h−1 . 2.6. Antioxidant defences and cellular damages 2.6.1. Homogenate preparation Six pools of five digestive glands (DG) of mussels and six pools of twelve gammarids per condition were formed, frozen in liquid nitrogen and stored at −80 ◦ C until use for antioxidant defence, cellular damage, and energetic reserve measurements. Prior to analysis pools were treated as described in Sroda and CossuLeguille (2011). Pools were defrosted, weighed and crushed at 4 ◦ C in phosphate buffer (50 mM, pH 7.6 supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF) and 1 mM l-serine borate mixture as protease inhibitors) at a 2/1 (v/w) ratio for gammarids and a 4/1 (v/w) ratio for mussel DG. One part of the homogenate was kept for cellular damage and energetic reserve measurements while the other part was centrifuged at 1000 × g for 20 min at 4 ◦ C. The supernatant was recovered and centrifuged again at 20,000 × g for 50 min at 4 ◦ C. The final supernatant, corresponding to the cytosolic fraction, was recovered for antioxidant enzyme activity assessment. 2.6.2. Antioxidant enzyme activities: total glutathione peroxidase (GPxtot) and catalase activities GPxtot and catalase activities were measured following the method described in Sroda and Cossu-Leguille (2011). For GPxtot activity, cytosolic fractions of gammarids were diluted to 1/8 in 100 mM phosphate buffer (pH = 7.6) while that of mussels were diluted to 1/2. NADPH consumption was then followed at 340 nm wavelength for 2 min in a mixture containing reduced glutathione (2 mM), NADPH (0.1 mM), glutathione reductase (1 U/mL), 1/10 (v/v) of diluted cytosolic fraction, and cumene hydroperoxyde (1.1 mM) as substrate. Enzymatic activities were expressed in ␮mol NADPH g−1 protein min−1 (ε NADPH = 6220 M−1 cm−1 ). For catalase activity, cytosolic fractions of both organisms were diluted to 1/8 in 100 mM phosphate buffer (pH = 7.6). The decomposition of hydrogen peroxide (30 mM) catalyzed by catalase was followed at 240 nm during 30 s (1/10, v/v of diluted cytosolic fraction introduced into the reactive mixture).

66

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

Enzyme activity was expressed in mmol H2 O2 g−1 protein min−1 (ε H2 O2 = 40 M−1 cm−1 ). Protein concentration in the cytosolic fraction was determined by a colorimetric method (Bio-Rad, France) using BSA as reference.

Protein concentration was determined in the homogenate by a colorimetric method (Bio-Rad, France) using BSA as reference and expressed in mg protein g−1 fw. 2.9. Lysosomal system morphology

2.6.3. Cellular damage assessment Cellular damages were assessed by evaluating lipidic membrane peroxidation by MDA (malondialdehyde) HPLC measurement using the method described in Behrens and Madère (1991). Mussel and gammarid homogenates were deproteinized in 3/4 (v/v) ethanol for 90 min at 4 ◦ C and then centrifuged at 18,000 × g for 30 min at 4 ◦ C. 20 ␮L of the resulting supernatant was injected into an HPLC system with UV detection at 267 nm. The HPLC separation was performed at 25 ◦ C on a reverse-phase LiChrospher 100RP18e column (length, 25 cm; inner diameter, 4 mm) and elution was carried out with sodium phosphate buffer (pH 6.5) containing 25% ethanol and 0.5 mM tetradecyltrimethylammonium bromide as ion pairing. A MDA standard calibration curve was constructed and MDA concentration was expressed in nmol g−1 protein.

2.7. Antitoxic defences: glutathione-S-transferase (GST) activity The digestive glands of eight mussels taken individually and 6 pools of ten gammarids were frozen in liquid nitrogen and stored at −80 ◦ C. Prior to analysis, samples were defrosted, weighed and crushed at 4 ◦ C in Tris buffer (10 mM; pH 7.5; supplemented with 0.1 mM PMSF and 1 mM dithiotritol) at a 1/10 (w/v) ratio. Homogenates were then centrifuged at 10,000 × g for 10 min at 4 ◦ C and supernatants were collected and either processed directly or stored at −80 ◦ C before analysis. GST activity was then measured by monitoring at 340 nm during 4 min the GST-catalyzed conjugation of 1-chloro2,4-dinitrobenzene (CDNB; 1 mM) with GSH (1 mM) in phosphate buffer (0.1 M; pH 6.5) with 1/3 (v/v) of diluted sample (1/15 in Tris buffer) introduced in the reactive mixture. GST activity was expressed in ␮mol CDNB g−1 protein min−1 .

2.8. Energetic reserves: glycogen, lipid and protein contents Energetic reserves were determined on tissue homogenates using methods described in Sroda and Cossu-Leguille (2011). 40 ␮L of mussel and gammarid homogenates were treated with 20 ␮L of 2% sodium sulphate and 540 ␮L of chloroform/methanol 1/2 (v/v) for 60 min at 4 ◦ C to extract lipids. Samples were then centrifuged at 3000 × g for 5 min at 4 ◦ C. The supernatant was recovered for lipid content measurement while the pellet was kept for glycogen content measurement. Supernatants were placed in Pyrex clean culture tubes and heated at 95 ◦ C in a dry bath to evaporate the solvent. 200 ␮L of sulphuric acid were subsequently added and the samples were left for another 10 min in the dry bath. Finally, tubes were cooled on ice before the addition of 4.8 mL of phospho-vanillin reagent (960 mg vanillin; 160 mL 100% ethanol, 640 mL 85% phosphoric acid) and 10 min incubation. The lipid content was read at 535 nm against a cholesterol standard (dissolved in chloroform) calibration curve and the results were expressed in mg lipids g−1 fw (fw = fresh weight). Pellets were resuspended in 400 ␮L MQ water and sonicated for 9 min. 100 ␮L of those suspensions were placed into Pyrex clean culture tubes with 4.9 mL anthrone reagent (1.13 g anthrone; 573 mL 95% H2 SO4 ; 170 mL MQ water) and heated on a dry bath for 17 min at 95 ◦ C. Tubes were cooled down on ice, then the glycogen content was read at 625 nm against a glucose standard calibration curve and the results were expressed in mg glycogen g−1 fw.

The digestive glands of eight mussels and the hepatopancreatic caeca of ten gammarids per condition were excised and prepared for histochemistry as described in Giambérini and Cajaraville (2005). The lysosomal system was located in the digestive cells by the histochemical revelation of ␤-glucuronidase demonstrated in unfixed cryostat sections according to Cajaraville et al. (1991) and adapted to freshwater organisms by Giambérini and Cajaraville (2005). Cellular biomarkers were quantified on digestive tissue sections (8 ␮m thick) by image analysis (Analysis pro 3.2, Olympus) using a Sony DP 50 colour video camera connected to an Olympus BX 41 microscope with a 100× objective. Five fields of view were randomly analyzed on one section per individual in each experimental group (total sampling area per organism: 63,172 ␮m2 ). Only areas belonging to digestive tissues were analyzed. Four stereological parameters were calculated: the surface density (Sv = SL /VC ), the volume density (Vv = VL /VC ), the surface to volume ratio (S/V = SL /VL , interpreted as the inverse of lysosome size) and the numerical density (Nv = NL /VC ) where the upper letters N correspond to number, S to surface and V to volume, and the index letters C and L correspond to digestive cell cytoplasm and to lysosomes respectively. 2.10. Locomotor and ventilatory activities of gammarids The locomotor activity was monitored on 10 gammarids per condition by counting the number of moving animals in a 200 mL glass crystallizing dish containing 80 mL of exposure media following the method of Felten et al. (2008). A piece of net was added to provide a resting surface (mesh size 200 ␮m, length × width: 6 cm × 5 cm). After a 5 min acclimation, the number of moving individuals was measured 35 times for 1 s and the results were expressed as the percentage of moving individuals. The ventilatory activity was recorded individually on the same 10 gammarids by counting the number of pleopod beating during 1 min according to Hervant et al. (1997). Each gammarid was transferred in a glass tube (length: 25 cm, Ø: 0.75 cm) and after 30 s of acclimatization, pleopod beats were visually counted using a manual cell counter only when the animal was at rest. Results were expressed in number of pleopod beats min−1 . Those two measurements were performed at the same period of the day in similar light conditions and without noise. 2.11. Statistical analysis The statistical analyses were conducted using R (R development core team). Homoscedasticity and normality were verified respectively by Levene and Shapiro–Wilk tests. When the conditions were filled, two-way ANOVA were performed to evaluate combined effects of nCeO2 concentration and exposure duration with a threshold of p ≤ 0.05 considered as significant. Post hoc Tukey HSD tests were done to verify differences between pairs of values. Non-parametric data were analyzed using Kruskal–Wallis test followed by Kruskal–Wallis post hoc test. A principal component analysis (PCA) was performed on the set of biomarkers to evidence differences in biomarker responses according to the experimental conditions. Between Correspondence Analysis were then used, combined to a Monte-Carlo test, to assess significant differences between groups defined by the combination of time × nanoparticle concentration.

11

30 28 46 46 46 46 27 52

18 30

30

0.0120*

0.0431* 0.6530 0.6306 0.6701 0.6490 0.7185 0.3230 0.1566

0.8828 0.0040**

0.0769

34 34 34 34 30 40 40 40 80 30 30 30

***

*

**

1.54 8.37 0.49 5.46 6.31 3.88 2.27 3.78 1.78 0.63 0.30 2.55

Significant effect of the studied parameter, p < 0.05. Significant effect of the studied parameter, p < 0.01. Significant effect of the studied parameter, p < 0.001.

0.1493 0.1693 0.0048** 0.0119* 0.0206* 0.0960 0.0443* 0.4629 0.9850 0.1959 0.4251 0.3360

30 0.7651 7.87

Pr(F) F Pr(F)

0.0000*** 4.70 0.0531 6.79 Kruskal–Wallis 2 = 9.4925. df = 5. p-value = 0.0910 0.29 0.7520 0.14 0.7066 3.50 0.43 0.6525 0.03 0.8572 0.43 * 2.32 0.1101 4.48 0.3965 0.47 * 1.91 0.1595 4.35 0.0426 0.40 5.09 0.0101* 3.01 0.0893 0.44 0.86 0.4318 3.78 0.0582 0.33 0.35 0.7075 0.38 0.5417 1.18 0.06 0.9376 2.21 0.1428 1.92 Kruskal–Wallis 2 = 15.1229. df = 5. p-value = 0.0099** 2 Kruskal–Wallis  = 14.135. df = 5. p-value = 0.0148* 0.02 0.9843 0.10 0.7556 0.13 9.38 0.0007*** 5.01 0.0328* 6.68 Kruskal–Wallis 2 = 18.882. df = 5. p-value = 0.0020** 3.07 0.0612 7.31 0.0112* 2.80 Kruskal–Wallis 2 = 10.8659. df = 5. p-value = 0.0541 Kruskal–Wallis 2 = 13.3141. df = 5. p-value = 0.0206*

F Pr(F) F Pr(F)

0.0000*** 9.29 0.0057** 2.40 Kruskal–Wallis 2 = 6.6415. df = 5. p-value = 0.2487 0.0018** 0.01 0.9154 0.27 Kruskal–Wallis 2 = 17.6409. df = 5. p-value = 0.0034** ** 0.2287 9.58 0.0039 2.01 0.0011** 5.79 0.0217* 1.87 0.6192 9.65 0.0038** 6.28 0.0088** 2.20 0.1472 5.07 0.0052** 0.51 0.4806 4.43 0.2876* 2.49 0.8281 2.49 0.1164 1.58 0.2159 3.37 * 0.0315 0.11 0.7371 0.79 * 0.1753 6.73 0.0113 0.02 0.5405 2.21 0.1475 1.72 0.7442 11.80 0.0018** 0.88 0.0947 0.27 0.6102 1.13 Not applicable Not applicable

Pr(F)

0.1129

23

F

Pr(F) F F

14.43

[Ce] GPx CAT GST Vv Sv S/V Nv [MDA] [Ca2+ ] [Na+ ] [Cl− ] NaK. [Prot.] [Glyc.] [Lip.] Loco. Vent.

df Interaction (df = 2)

67

417.0

df Interaction (df = 2) Time (df = 1) Gammarus roeseli

3.2.1. Antioxidant and antitoxic defences Biomarker results are listed in Table 2. We did not observe significant impacts of nCeO2 exposure on GPxtot activity in mussel digestive glands (Table 1). All exposure durations combined, a significant decrease in CAT activities for mussels exposed to 10 and 100 ␮g/L/d was evidenced compared to the control (p = 0.011 and 0.003 respectively). A significantly lowered activity (−35%) was noticed in mussels exposed to 100 ␮g/L/d after 4 days (p = 0.048). We also observed non-significant 25% decreases in CAT activities for all other exposed experimental groups at days 1 and 4 (p < 0.29) compared to the control. Finally, GST activities were similar between control and exposed mussels at days 1 and 4, but increased significantly between day 1 and day 4 in mussels exposed to the highest nCeO2 concentration. In gammarids, GST and GPxtot activities were not significantly impacted by nCeO2 exposure (Tables 1 and 2). Interactive effects between exposure time and concentration were observed on CAT activity (Table 1) but post hoc testing did not show any significant differences between experimental conditions although a 17% decrease was observed in gammarids exposed to 100 ␮g/L/d compared to the control at day 1 (p = 0.234).

Concentration (df = 2)

3.2. Biomarker responses

Time (df = 1)

For D. polymorpha, both exposure time and nanoparticle concentration exerted a significant effect on cerium bioaccumulation but no interactive effects were observed (Table 1). All exposure durations taken together, mussels exposed to 100 ␮g/L/d of nCeO2 accumulated significantly more Ce than control (p < 0.001) and 10 ␮g/L/d exposed mussels (p = 0.003). After one day of exposure, Ce internal concentrations were 1.3-fold and 3.0-fold higher in mussels exposed to 10 and 100 ␮g/L/d of nCeO2 respectively compared to the control (Fig. 2A), which presented a total body burden of 2.7 ␮g/gdw. However, no significant differences could be evidenced between experimental groups at day 1 (p = 0.999 and 0.273 for 10 and 100 ␮g/L/d respectively compared to the control). A timedependent significant increase in Ce concentrations was observed between day 1 and day 4 (p = 0.006). Compared to the control, which presented a stable concentration (2.75 ␮g/gdw), Ce concentrations after 4 days of exposure were 2.9-fold and 5.5-fold higher in mussels exposed to 10 and 100 ␮g/L/d of nCeO2 respectively. Ce concentration in mussels exposed to the highest nCeO2 concentration rose significantly between days 1 and 4 (p = 0.041) to reach 15 ␮g/gdw, a value significantly higher than control one (p < 0.001). For G. roeseli, nanoparticle concentration and the interaction between exposure time and concentration exerted a significant effect on cerium bioaccumulation (Table 1). As for mussels, the Ce internal concentrations in control gammarids were stable, and reached 1.35 and 1.14 ␮g/gdw at day 1 and 4 respectively (Fig. 2B). After one day, gammarids exposed to 100 ␮g/L/d of nCeO2 showed a significant 44-fold increase in Ce internal concentration compared to the control (p < 0.001). Between days 1 and 4, Ce internal concentration in gammarids exposed to 10 ␮g/L/d doubled (p < 0.001) to reach a concentration significantly higher than that of the control (p < 0.001). Ce internal concentration in gammarids exposed to the highest nCeO2 concentration stayed stable at 59 ␮g/gdw (p = 0.999). While Ce accumulation in gammarids is similar to that in mussels for the lowest nCeO2 concentration, gammarids exposed to 100 ␮g/L/d nCeO2 accumulated 4-times more Ce than mussels.

Dreissena polymorpha

3.1. nCeO2 internal concentrations

Concentration (df = 2)

3. Results

Table 1 Dreissena polymorpha and Gammarus roeseli results of the statistical analysis (2 ways-ANOVA or Kruskal–Wallis for non-parametric data) to assess respective effects of nanoparticle concentration and exposure duration and their interactive effects on measured biomarkers and cerium internal concentration after 1 and 4 days of exposure to nCeO2 .

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

68

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

Table 2 Dreissena polymorpha and Gammarus roeseli biomarker responses after 1 and 4 days of exposure to nCeO2 (Ct: control; C1: 10 ␮g nCeO2 /L/d; C2: 100 ␮g nCeO2 /L/d) – data are expressed as mean + SD – for each species, values that do not share a common letter are significantly different (p < 0.05). Biomarkers

Antioxidant and antitoxic defences GPx (␮molNADPH/gprot/min) Ct C1 C2 CAT (mmolH2 O2 /gprot/min) Ct C1 C2 GST (␮molCDNB/gprot/min) Ct C1 C2 Lysosomal system morphology Vv (10−4 ␮m3 /␮m3 ) Ct C1 C2 Sv (10−3 ␮m2 /␮m3 ) Ct C1 C2 S/V (␮m2 /␮m3 ) Ct C1 C2 Nv (10−2 /␮m3 ) Ct C1 C2 Cellular damages [MDA] (nmol/gprot) Ct C1 C2

Dreissena polymorpha Exposure duration 1 day

4 days

21 ± 5 18 ± 5 17 ± 3

16 ± 6 13 ± 6 13 ± 7

170 ± 33 203 ± 63 197 ± 27

207 ± 33 291 ± 63 291 ± 27

75 ± 18ab 56 ± 12ab 56 ± 7ab

79 ± 16a 59 ± 22ab 51 ± 15b

198 ± 26 191 ± 15 164 ± 31

179 ± 26 185 ± 16 198 ± 34

98 ± 36ab 99 ± 43ab 85 ± 14a

123 ± 43ab 132 ± 24ab 154 ± 33b

58 ± 10 61 ± 17 57 ± 8

53 ± 10 60 ± 10 62 ± 18

9.1 ± 3.5a 9.0 ± 3.6a 8.5 ± 2.2ab

8.0 ± 1.8ab 4.1 ± 1.7b 6.9 ± 3.1ab

2.0 ± 0.8ab 2.4 ± 0.5a 1.9 ± 0.5ab

1.9 ± 0.8ab 2.0 ± 1.0ab 1.5 ± 0.8b

19.6 ± 2.5a 16.2 ± 4.9ab 18.8 ± 1.9a

19.9 ± 3.8a 11.7 ± 3.2b 14.2 ± 4.5ab

3.1 ± 1.2 3.6 ± 0.8 3.0 ± 0.7

2.9 ± 1.2 3.1 ± 1.4 2.4 ± 1.1

23.2 ± 5.8ab 18.8 ± 3.1a 23.1 ± 5.3ab

25.1 ± 3.4ab 30.6 ± 6.8b 22.2 ± 4.8a

15.7 ± 0.6 15.0 ± 0.6 15.8 ± 0.5

15.8 ± 0.9 15.5 ± 1.1 16.4 ± 1.0

10.8 ± 4.3ab 5.5 ± 2.1b 10.4 ± 3.6ab

12.8 ± 4.6a 10.7 ± 3.8ab 6.7 ± 1.5b

0.7 ± 0.2 0.7 ± 0.1 0.7 ± 0.1

0.6 ± 0.2 0.6 ± 0.2 0.5 ± 0.2

50 ± 19a 27 ± 9b 23 ± 3b

33 ± 4ab 33 ± 9ab 30 ± 5ab

18 ± 8 19 ± 5 14 ± 4

15 ± 5 15 ± 6 16 ± 5

9±2 9±2 6±2

20 ± 13 16 ± 6 19 ± 7

20 ± 9 25 ± 8 19 ± 9

Iono-osmoregulation Haemolymph [Ca2+ ] (mmol/L) Ct 10 ± 3 C1 7±3 C2 8±2 + Haemolymph [Na ] (mmol/L) 19 ± 5ab Ct 18 ± 5ab C1 20 ± 3ab C2 Haemolymph [Cl− ] (mmol/L) 19 ± 4 Ct 17 ± 4 C1 19 ± 2 C2 + + Na /K ATPase activity (mmolPi/gprot/h) Ct 0.7 ± 0.3 0.7 ± 0.4 C1 0.6 ± 0.4 C2 Energetic reserves [Protein] (mg/gFW) Ct C1 C2 [Glycogen] (mg/gFW) Ct C1 C2 [Lipid] (mg/gFW) Ct C1 C2

Gammarus roeseli Exposure duration 1 day

4 days

24 ± 3a 20 ± 5ab 17 ± 4b

127 ± 7a 131 ± 15a 122 ± 11ab

118 ± 14ab 108 ± 18b 124 ± 6ab

20 ± 3 17 ± 4 16 ± 3

113 ± 9a 107 ± 9ab 101 ± 16ab

98 ± 18ab 91 ± 21b 106 ± 10ab

0.9 ± 0.3 0.9 ± 0.3 0.8 ± 0.4

11.1 ± 1.7 11.6 ± 2.1 11.5 ± 1.9

12.1 ± 2.4 11.6 ± 3.2 11.3 ± 3.4

50 ± 7 55 ± 5 54 ± 2

51 ± 6 48 ± 4 52 ± 3

46 ± 6ab 38 ± 4b 55 ± 8c

51 ± 5ac 49 ± 3ac 51 ± 5ac

7.6 ± 2.1 8.8 ± 0.9 8.6 ± 2.0

6.8 ± 1.0 6.6 ± 2.0 6.1 ± 1.0

1.7 ± 0.2ab 1.2 ± 0.2b 2.1 ± 0.5ab

2.2 ± 0.9a 3.3 ± 1.3a 2.1 ± 1.1ab

23.6 ± 5.5 20.1 ± 2.3 23.2 ± 2.6

25.5 ± 3.1 22.1 ± 3.6 21.2 ± 4.5

11.2 ± 1.5ab 9.6 ± 1.8b 12.2 ± 1.5ab

11.5 ± 0.7ab 12.6 ± 1.3a 12.9 ± 2.0a

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

69

Table 2 (Continued) Biomarkers

Dreissena polymorpha Exposure duration 1 day

Gammarus roeseli Exposure duration 4 days

Behavioural parameters (gammarids only) Locomotion (% moving) Ct C1 Not applicable C2 Ventilation (beats/min) Ct Not applicable C1 C2

3.2.2. Lysosomal system morphology In mussel digestive glands, lysosomal volumic density Vv and surface density Sv were significantly lower at day 4 than at day 1 all concentrations taken together (see supplementary data, Figure 1 for pictures illustrating lysosomal system morphology). We also observed significantly lower Sv values for mussels exposed to 10 ␮g/L/d compared to the control (p < 0.001) all exposure duration combined (Table 2). At day 4, Sv in mussels exposed to 10 ␮g/L/d was significantly lower (−41%) than control levels (p = 0.007), while a decreasing trend (−29%) was observed at the highest concentration (p = 0.082). Interactive effects were observed for the lysosomal size (S/V ratio) and number Nv (Table 1). However, we observed no differences in S/V for exposed mussels compared to the control, but only an increase at 10 ␮g/L/d between day 1 and 4 (p < 0.001). Meanwhile, a significant decrease in Nv (-48%) was observed in mussels exposed to 100 ␮g/L/d at day 4 (p = 0.031). Supplementary figure related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.aquatox.2014.11.004. In gammarids digestive caeca, lysosomal volumic density Vv and surface density Sv decreased significantly between days 1 and 4 all concentrations take together (Table 1). No significant effects were noticed on lysosome number Nv . Finally, S/V was significantly elevated in gammarids exposed to 100 ␮g/L/d of nCeO2 compared to the ones exposed to 10 ␮g/L/d (p = 0.008). 3.2.3. Cellular damages In mussel digestive glands, a significant interactive effect was observed (Table 1). A significant decrease in MDA levels was shown for mussels exposed to both nCeO2 concentrations at day 1 compared to the control (Table 2), with level 46% and 54% lower at 10 and 100 ␮g/L/d respectively (p = 0.033 and 0.001). At day 4, no differences in MDA levels between experimental groups could be observed anymore. In gammarids, no significant differences were evidenced at any exposure times.

1 day

4 days

47 ± 12 43 ± 11 44 ± 11

45 ± 14 42 ± 14 38 ± 11

220 ± 11 211 ± 9 223 ± 6

226 ± 11 221 ± 24 219 ± 12

3.2.4. Iono-osmoregulation parameters Nanoparticle concentration exerted a significant effect on mussel haemolymphatic Ca2+ and Cl− concentrations (Table 1), with Ca2+ levels in mussels exposed to 100 ␮g/L/d lower than control ones all exposure durations taken together (p = 0.024, Table 2). An interactive effect, and a significant decrease (−28%) in Na+ haemolymphatic concentrations were observed at day 4 for mussels exposed to the highest concentration of nCeO2 (p = 0.032). Gill Na+ /K+ ATPase activity experienced a time dependent increase all concentrations combined between days 1 and 4 (Table 1). In gammarids, no differences between control and exposed individuals were noticed during the experiment for all of the ionoosmoregulation biomarkers measured (Tables 1 and 2).

3.2.5. Energetic reserves In mussel digestive glands, no differences in protein and lipid contents were observed between control and exposed organisms at both exposure times. We noticed a general significant decrease in glycogen concentration between the two sampling dates (Tables 1 and 2). No differences in lipid and glycogen contents were observed between control and exposed gammarids at both exposure times, and lipid concentration rose significantly over the exposure period (Table 1). An interactive effect was shown on protein content (Table 1) with elevated protein content for gammarids exposed to 100 ␮g/L/d at day 1 compared to the control (p = 0.035) and to the lowest nCeO2 concentration (p < 0.001).

3.2.6. Gammarid behavioural parameters No significant differences between exposed and control gammarids were observed at both exposure times in locomotor or in ventilation activities (Tables 1 and 2).

Fig. 2. Cerium bioaccumulation in whole Dreissena polymorpha (A) and Gammarus roeseli (B) after 1 and 4 days of exposure to nCeO2 – data are represented as mean + SD – values that do not share a common letter are significantly different (p < 0.05).

70

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

Fig. 3. Principal component analysis of the biomarker responses in mussels (A – all conditions represented; B – effects of nanoparticles investigated alone) and in gammarids (C) exposed for 1 day (D1) and 4 days (D4) to 10 ␮g/L/d (10) and 100 ␮g/L/d (100) of nCeO2 (Ctl = control).

3.2.7. Principal component analysis (PCA) PCA were performed on the whole set of biomarkers measured in the two species. Percentage of 64% and 69% of the total inertia were explained on the first four axes for mussels and gammarids respectively. For mussels, when the six experimental conditions were represented on the PCA, the four groups of exposed mussels appeared separated from the two control groups (Fig. 3A). The inter-group separation was even more pronounced when the effect of nanoparticles was investigated alone (Fig. 3B), and Monte-Carlo test showed significant differences between groups. In this case, groups of mussels exposed to 10 and 100 ␮g/L/d of nCeO2 overlapped but clearly differed from control group. For gammarids, while Monte-Carlo test evidenced differences between groups, PCA did not allow discrimination of the experimental groups in relation with nanoparticle exposure (Fig. 3C). The investigation of column inertia on the first four axis of the PCA for both species showed that lysosomal parameters were the most structuring variables, followed by Na+ and Cl− haemolymph concentration, catalase activity and, to a lesser extent, MDA levels (more than 60% of explained inertia for each variable). 4. Discussion 4.1. nCeO2 bioaccumulation The measurements of Ce internal concentrations, carried out on whole organisms, showed bioaccumulation of metal in both invertebrate species following their short-term exposure to nCeO2 . At the end of the exposure period, the gammarids exposed to the highest concentration of nCeO2 tested have accumulated cerium 4-fold more than mussels exposed to the same concentration. At that time and for that concentration, we calculated bioconcentration factors (BCF) of 38 and 176 L/kg for mussels and gammarids respectively, and using organism numbers, Ce internal concentration, and total nCeO2 introduced mass, we also calculated that less than 2.4% of

the total Ce could be found in mussel bodies while that percentage was higher than 27% for gammarids. While we can assume that a part of nCeO2 were simply transiting through the intestine or adsorbed on gills and mantle tissues in zebra mussels, several studies suggest that nCeO2 could have been internalized (Browne et al., 2008; Koehler et al., 2008). As a filterfeeding organism, D. polymorpha filter large quantities of water and trap particles from it using cilia covering their gills. Particles are then directed towards the mouth, to the stomach where they can be directed towards digestive gland tubules, an organ that has been shown to be the preferential site of nanoparticle accumulation in the organism (Browne et al., 2008; Koehler et al., 2008; Canesi et al., 2010). However, it is not excluded that a part of the nanoparticles penetrates into the organism through gill and mantle epithelium (Koehler et al., 2008; Canesi et al., 2012). The differences observed in the bioaccumulation between mussels and gammarids are probably related to differences in species behaviour. D. polymorpha are sessile invertebrates, filtering the water column to feed. Therefore, mussels can only accumulate nCeO2 present in the water column. At the same time, nCeO2 has been shown to aggregate quickly in water to reach size >300 nm in a few minutes (Artells et al., 2013; Tella et al., 2014) and the aggregates formed become non-bioavailable to mussels once they have sedimented. Moreover, through their filtration activity, mussels could immobilize a large part of the introduced nCeO2 as pseudofaeces. In fact, several studies have shown the ability of mussels to sort trapped particles before ingestion in the stomach (see review by Evan Ward and Shumway, 2004), and the inorganic nanoparticles of nCeO2 could have been quickly rejected as pseudofaeces (Baker et al., 1998; Winkel and Davids, 1982). Mass balances done after experiments on marine mussels have shown that mussels were able to filter out almost all of the nCeO2 from the water column, and to immobilize more than 90% of them in pseudofaeces (Conway et al., 2014; Montes et al., 2012). The very low Ce accumulation (1–3%) observed for those marine mussels (Montes et al., 2012) is coherent

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

with our estimated accumulation of 2.4% of the total introduced Ce in mussels. In strong contrast with D. polymorpha, G. roeseli is a non-sessile invertebrate part of the shredder functional group, meaning that it feeds principally on rough organic matter debris on the bottom of freshwater streams. By moving on the bottom of the tanks, gammarids could remobilize settled nCeO2 , which has been shown to homo-aggregate quickly in spring water (Artells et al., 2013; Tella et al., 2014). Beside this re-suspension process, gammarid motion also enhances collision frequency with nCeO2 and their subsequent mechanical trapping on gammarid cuticle and gills, as showed on Daphnia sp. exposed to nCeO2 in the same water (Artells et al., 2013). A much higher mass proportion of nCeO2 is then bioavailable for gammarids compared to mussels, and that could explain the higher bioaccumulation in gammarids. Meanwhile, experiments on D. magna showed nCeO2 accumulation in the digestive tract but no uptake indigestive cells (Artells et al., 2013), suggesting that Ce bioaccumulation in gammarids could also be due to metal adsorption on gammarid body surfaces or to nCeO2 transiting through the digestive tract. In summary, bioavailability and exposure pathways to nCeO2 in gammarids probably differed widely from those observed for mussels, for which the bioavailability of nCeO2 depends on its stability in the water column and could be significantly lowered through fast nCeO2 aggregation and sedimentation processes (Artells et al., 2013; Tella et al., 2014), but also as a result of quick nCeO2 immobilization in pseudofaeces (Conway et al., 2014; Montes et al., 2012). 4.2. Biomarker responses in D. polymorpha Catalase activities in D. polymorpha were significantly lowered by nCeO2 exposure at both concentrations compared to the control. Exposure to pollutants like metallic nanoparticles can trigger excessive ROS production, and cells possess enzymatic antioxidant defences like catalase to neutralize ROS, which could otherwise react with cellular components and exert deleterious oxidative stress on organisms (Lushchak, 2001; Oktyabrsky and Smirnova, 2007; Xia et al., 2013). Catalase inhibition is generally observed in cases of strong pollutant toxicity, which triggers oxidative stress by overwhelming cellular defences, and is associated with cellular damages, showed in particular by lipoperoxidation (Cossu et al., 1997; Osman et al., 2007; Xia et al., 2013). However, in our experiment, no cellular damages were associated with the decreasing catalase activity. On the contrary, lipoperoxidation levels in exposed mussels at day 1 were significantly lower than control one, suggesting the oxidative stress level and the associated cellular damages in exposed mussels were lower at that exposure time. In fact, the oxidative stress experienced by mussels at day 1 could be alleviated in those exposed to nCeO2 , as revealed by biomarker measurements, thanks to the antioxidant activity of nCeO2 shown by many studies (Alili et al., 2010; Colon et al., 2010, 2009; Das et al., 2012; Ting et al., 2013; Zholobak et al., 2011). In fact, strong SOD (Das et al., 2007; Korsvik et al., 2007) and catalase mimetic activities (Das et al., 2007; Pirmohamed et al., 2010) of nCeO2 could drastically reduce intracellular ROS level of exposed D. polymorpha thus attenuating endogenous antioxidant defences, and could most notably explain the observed lowered catalase activity, possibly through gene down-regulation resulting from decreased H2 O2 contents (Ciofani et al., 2014). nCeO2 ROS scavenging could in turn explain the observed decrease in cellular damages, a decrease which has also been shown in several works in irradiated cells pre-treated with nCeO2 (Colon et al., 2010, 2009; Zholobak et al., 2011). The results observed for lysosomal system morphology corroborates the lack of nCeO2 adverse effect on mussels, as we saw

71

no increase in the size of the lysosomal system either at day 1 or at day 4. In fact, exposure to xenobiotics results generally in the increase of the size of the lysosomal system because of sequestration/detoxication processes of xenobiotics and damaged cellular components (Giambérini and Cajaraville, 2005; Giambérini and Pihan, 1997; Guerlet et al., 2010). The observed decrease in lysosomal system size in mussel exposed to nCeO2 at day 4 could give another clue of their antioxidant activity, which could reduce cellular component degradation and the associated recycling/sequestrating autophagic processing of those damaged components by lysosomes (Giambérini and Pihan, 1997; Guerlet et al., 2010). In fact, antioxidant and cellular damage biomarkers showed that control mussels experienced higher oxidative stress level at day 1, and that could translate into a more developed lysosomal system, as observed at day 4, as a result of delayed damaged cellular component accumulation (most notably lipoperoxidation residues) into lysosomes, probably as lipofuscines. The delay observed between the occurrence of oxidative damages and the accumulation of lipofuscines into lysosomes has already been observed for D. polymorpha (Guerlet et al., 2010). Meanwhile, some of the iono-osmoregulation parameters, namely the Ca2+ and Na+ haemolymph concentrations, seemed to be negatively impacted by exposure to the highest nCeO2 concentration. Osmo-regulation disruption can be consecutive to gill or excretory organ structural damages, cell membrane or membrane ionic pump perturbations and could indicate severe toxic effects on tested organisms (Arce Funck et al., 2013). Very few studies are available, but data on nTiO2 and nCuO effects in fish showed that nanoparticles could exert indirect negative impacts on iono-osmoregulation parameters by increasing mucus secretion on the gills, or even causing injuries, and thus reducing ion uptake (Federici et al., 2007; Griffitt et al., 2007). However, in contrast with the finding of those studies, we did not observe any inhibition of the Na+ /K+ -ATPase ionic pump functioning that could account for haemolymph Na+ concentration decrease. For Ca2+ haemolymph concentrations the observed decrease could result from the intracellular formation of Ca concretions to immobilize and eliminate nCeO2 , a metal detoxication mechanism very common in bivalves (Marigómez et al., 2002; Viarengo and Nott, 1993). Finally, energy consumption was not impacted significantly in exposed or control mussels, despite the iono-osmoregulatory or antioxidant effects observed. Energy availability is fluctuant in natural systems and organisms constitute energy reserves as protein, lipids and glycogen to face those variations and to ensure sufficient allocation of energy to homeostasis maintenance, growth and reproduction (Palais et al., 2011). By inducing energy demanding cellular defences or by disrupting homeostasis, exposure to pollutants can modify energy allocation and mobilize reserves (Smolders et al., 2004; Voets et al., 2006). In the case of severe perturbation of the energy budget, severe impacts on growth and reproduction can occur and jeopardize organism individual and population survival in the long-term (De Wolf et al., 2007; Smolders et al., 2004; Voets et al., 2006). Disturbance of energetic biomarkers generally reveals severe toxic effects on the tested organisms that could lead to individual and then to populational effects. In this low-concentration and short-term exposure context, severe disturbances are unlikely to be observed and further long-term studies should be conducted to investigate the impacts on energetic biomarkers in chronic exposure conditions. The PCA carried out on the whole set of biomarker confirmed the effects of nCeO2 on mussels, as exposed groups were clearly separated from control. Interestingly, while exposed to different concentrations, exposed groups showed similar responses, suggesting that nCeO2 could exert non concentration-dependent effects after a threshold concentration. In addition to its ability to discriminate groups, PCA also allowed us to evidence the individual

72

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

biomarkers and the large biological functions impacted by nCeO2 exposure in mussels, namely the lysosomal compartment, followed by antioxidant defences and haemolymph ion concentrations. This particular set of impacted biomarkers will be of particular interest for future studies on nCeO2 ecotoxicology. As a conclusion, effects of nCeO2 exposure on D. polymorpha were evidenced on several individual biomarkers and confirmed by an integrated analysis (PCA). Two of the three most impacted biological functions were the lysosomal and antioxidant systems, for which biomarker responses suggested a potential antioxidant effect of nCeO2 . These potential antioxidant properties need to be investigated more thoroughly with further experiments, in particular by co-exposure with pollutants inducing strong pro-oxidant effects. Such co-exposures have been conducted on the fish Carassius auratus (Xia et al., 2013) and seemed to confirm our findings. At the same time, nCeO2 could have negative impacts by indirectly disrupting haemolymph iono-osmoregulation balance. Nevertheless, none of those impacts seemed to impact mussel energetic parameters in the short-term.

4.3. Biomarker responses in G. roeseli and inter-species comparison In sharp contrast with biomarker responses in D. polymorpha, G. roeseli were largely unaffected by nCeO2 exposure at days 1 and 4, even for antioxidant defences, cellular damages, lysosomal system morphology or iono-osmoregulation parameters, which responded to nCeO2 exposure in mussels. Thus, comparing the two species, no adverse effects of nCeO2 could be observed on G. roeseli. Nevertheless, we highlighted in Section 4.1 that gammarids accumulated more Ce than mussels. That could imply three non-exclusive hypotheses: (i) nCeO2 internal distribution in gammarids and mussels and the subsequent biological effects vary, (ii) the larger size fraction of nCeO2 to which gammarids were exposed did not exhibit the antioxidant properties of the finer fraction to which the mussels were exposed, (iii) gammarids are less sensitive to nCeO2 than mussels. Indeed, the hypothesis (i) is supported by the fact that because of their filter-feeding behaviour and high phagocytosis and endocytosis capabilities, mussels were probably more likely to internalize nCeO2 than gammarids (Canesi et al., 2012), for which accumulated nCeO2 could be adsorbed on cuticle and gills, or in the intestine rather than inside the organism, as shown in other swimming crustacean species (Artells et al., 2013). nCeO2 in lower concentration but localized in strategic tissue would then exert antioxidant activity inside mussel cells, while the nCeO2 in higher concentration in gammarids would not be able to interact with cellular components and exert the same antioxidant activity. Alternatively, the hypothesis (ii) could explain the differences observed in the biological effects between gammarids and mussels. In fact, nCeO2 has been shown to aggregate quickly in spring water, and the swimming motion of gammarids could have further increased this aggregation process (Artells et al., 2013). In addition to that, because of their activity at the bottom of the tanks, gammarids could have been preferentially exposed to aggregated nCeO2 that sedimented, while filter-feeding mussels could have been preferentially exposed to the finer nCeO2 particles staying in suspension in the water column. Meanwhile, Celardo et al. (2011), which studied the antioxidant effects of bare nCeO2 on leukocytes cell lines, showed that the supernatant fraction of nCeO2 suspension, representing only 10 wt%, concentrated all the antioxidant activity of the suspension, while the settled nanoparticles did not show any antioxidant effects (Celardo et al., 2011). The same effect could explain why no antioxidant activity is observed for gammarids in our experiment.

Finally, the PCA done on the biomarker responses in gammarids, while showing no inter-group differences, highlighted that the most impacted biological compartments were the same as in mussels. That could support the hypothesis (iii), implying that nCeO2 would impact the same biological compartment in gammarids but to a lesser extent. More interestingly, it showed that the biological functions impacted by nCeO2 are similar for two invertebrate species belonging to two distinct functional groups. It would be of great interest to see if we could generalize those findings for other species. 5. Conclusions No strong adverse effects of waterborne nCeO2 exposure at low concentration (10 and 100 ␮g/L/d) were observed on the two freshwater invertebrates tested, despite metal bioaccumulation that may be related to the species functional characteristics and nanoparticle behaviour in the medium. The integrated multibiomarker approach allowed us to clearly evidence effects of nCeO2 on mussel biology, and to identify the most impacted biological function, that were common to both invertebrates: lysosomal compartment, antioxidant defences and haemolymph ion concentrations. Biomarkers related to those functions will be of particular interest for further studies on nCeO2 effects, and more generally for nanoparticles studies. The effects observed on lysosomal and antioxidant systems suggested that nCeO2 may exhibit antioxidant protecting activity in mussels exposed in vivo. At the same time, potential negative impacts on mussel haemolymph ion concentrations could result from nCeO2 exposure. None of those impacts were observed on gammarids. Further studies in more complex environmental conditions and with dietary exposure should be conducted to ensure nCeO2 present no risks to aquatic ecosystems in a long term perspective. Co-exposures with pro-oxidant substances should also be conducted to confirm the potential antioxidant properties of nCeO2 in freshwater invertebrates exposed in vivo. This work also confirms the relevance of working on several model species with various bio-ecological traits, especially filter-feeding species, to better assess nanoparticle ecotoxicity. Acknowledgements Financial supports were provided by the Agence Nationale de la Recherche (ANR-10-NANO-0006/MESONNET project) for M. Garaud PhD salary and running costs and CPER Lorraine-ZAM (Contrat Projet Etat Région Lorraine, Zone Atelier Moselle). This work is a contribution to the Labex Ressources 21, ANR-10-LABX-2101 (Strategic metal resources of the 21st century). The authors gratefully acknowledge CNRS and CEA for funding the iCEINT International Consortium for the Environmental Implications of NanoTechnology. Sharon Kruger is gratefully acknowledged for her English corrections. References Alili, L., Sack, M., Karakoti, A.S., Teuber, S., Puschmann, K., Hirst, S.M., Reilly, C.M., Zanger, K., Stahl, W., Das, S., Seal, S., Brenneisen, P., 2010. Combined cytotoxic and anti-invasive properties of redox-active nanoparticles in tumor–stroma interactions. Biomaterials 32, 2918–2929. Amin, K.A., Hassan, M.S., Awad, e.-S.T., Hashem, K.S., 2011. The protective effects of cerium oxide nanoparticles against hepatic oxidative damage induced by monocrotaline. Int. J. Nanomed. 6, 143–149. Arce Funck, J., Danger, M., Gismondi, E., Cossu-Leguille, C., Guérold, F., Felten, V., 2013. Behavioural and physiological responses of Gammarus fossarum (Crustacea Amphipoda) exposed to silver. Aquat. Toxicol. 142–143, 73–84. Artells, E., Issartel, J., Auffan, M., Borschneck, D., Thill, A., Tella, M., Brousset, L., Rose, J., Bottero, J.-Y., Thiéry, A., 2013. Exposure to cerium dioxide nanoparticles differently affect swimming performance and survival in two daphnid species. PLoS ONE 8, e71260.

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74 Auffan, M., Masion, A., Labille, J., Diot, M.-A., Liu, W., Olivi, L., Proux, O., Ziarelli, F., Chaurand, P., Geantet, C., Bottero, J.-Y., Rose, J., 2014. Long-term aging of a CeO2 based nanocomposite used for wood protection. Environ. Pollut. 188, 1–7. Auffan, M., (PhD géosciences de l‘environnement) 2007. Nanoparticules d‘oxydes métalliques: relations entre la réactivité de surface et des réponses biologiques. Université Paul Cézanne Aix-Marseille III, 204p. Baker, S.M., Levinton, J.S., Kurdziel, J.P., Shumway, S.E., 1998. Selective feeding and biodeposition by zebra mussels and their relation to changes in phytoplankton composition and seston loads. J. Shellfish Res. 17, 1207–1213. Behrens, W., Madère, R., 1991. Malonaldehyde determination in tissues and biological fluids by ion-pairing high-performance liquid chromatography. Lipids 26, 232–236. Browne, M.A., Dissanayake, A., Galloway, T.S., Lowe, D.M., Thompson, R.C., 2008. Ingested microscopic plastic translocates to the circulatory system of the mussel, Mytilus edulis (L.). Environ. Sci. Technol. 42, 5026–5031. Cajaraville, M., Marigomez, J.A., Angulo, E., 1991. Automated measurement of lysosomal structure alterations in oocytes of mussels exposed to petroleum hydrocarbons. Arch. Environ. Contam. Toxicol. 21, 395–400. Canesi, L., Ciacci, C., Fabbri, R., Marcomini, A., Pojana, G., Gallo, G., 2012. Bivalve molluscs as a unique target group for nanoparticle toxicity. Mar. Environ. Res. 76, 16–21. Canesi, L., Fabbri, R., Gallo, G., Vallotto, D., Marcomini, A., Pojana, G., 2010. Biomarkers in Mytilus galloprovincialis exposed to suspensions of selected nanoparticles (Nano carbon black, C60 fullerene, Nano-TiO2 , Nano-SiO2 ). Aquat. Toxicol. 100, 168–177. Celardo, I., De Nicola, M., Mandoli, C., Pedersen, J.Z., Traversa, E., Ghibelli, L., 2011. Ce3+ ions determine redox-dependent anti-apoptotic effect of cerium oxide nanoparticles. ACS Nano 5 (6), 4537–4549. Chaudhury, K., Babu, K.N., Singh, A.K., Das, S., Kumar, A., Seal, S., 2012. Mitigation of endometriosis using regenerative cerium oxide nanoparticles. Nanomed. Nanotechnol. Biol. Med. 9, 439–448. Chigurupati, S., Mughal, M.R., Okun, E., Das, S., Kumar, A., McCaffery, M., Seal, S., Mattson, M.P., 2013. Effects of cerium oxide nanoparticles on the growth of keratinocytes, fibroblasts and vascular endothelial cells in cutaneous wound healing. Biomaterials 34, 2194–2201. Ciofani, G., Genchi, G.G., Mazzolai, B., Mattoli, V., 2014. Transcriptional profile of genes involved in oxidative stress and antioxidant defense in PC12 cells following treatment with cerium oxide nanoparticles. Biochim. Biophys. Acta (BBA) – Gen. Subj. 1840, 495–506. Colon, J., Herrera, L., Smith, J., Patil, S., Komanski, C., Kupelian, P., Seal, S., Jenkins, D.W., Baker, C.H., 2009. Protection from radiation-induced pneumonitis using cerium oxide nanoparticles. Nanomed. Nanotechnol. Biol. Med. 5, 225–231. Colon, J., Hsieh, N., Ferguson, A., Kupelian, P., Seal, S., Jenkins, D.W., Baker, C.H., 2010. Cerium oxide nanoparticles protect gastrointestinal epithelium from radiationinduced damage by reduction of reactive oxygen species and upregulation of superoxide dismutase 2. Nanomed. Nanotechnol. Biol. Med. 6, 698–705. Conway, J.R., Hanna, S.K., Lenihan, H.S., Keller, A.A., 2014. Effects and implications of trophic transfer and accumulation of CeO2 nanoparticles in a marine mussel. Environ. Sci. Technol. 48, 1517–1524. Cossu, C., Doyotte, A., Jacquin, M.C., Babut, M., Exinger, A., Vasseur, P., 1997. Glutathione reductase, selenium-dependent glutathione peroxidase, glutathione levels, and lipid peroxidation in freshwater bivalves, Unio tumidus, as biomarkers of aquatic contamination in field studies. Ecotoxicol. Environ. Saf. 38, 122–131. Das, M., Patil, S., Bhargava, N., Kang, J.-F., Riedel, L.M., Seal, S., Hickman, J.J., 2007. Auto-catalytic ceria nanoparticles offer neuroprotection to adult rat spinal cord neurons. Biomaterials 28, 1918–1925. Das, S., Singh, S., Dowding, J.M., Oommen, S., Kumar, A., Sayle, T.X.T., Saraf, S., Patra, C.R., Vlahakis, N.E., Sayle, D.C., Self, W.T., Seal, S., 2012. The induction of angiogenesis by cerium oxide nanoparticles through the modulation of oxygen in intracellular environments. Biomaterials 33, 7746–7755. De Wolf, H., Lukambuzi, L., Van den Broeck, H., Smolders, R., 2007. Spatio-temporal variation in energy content of the periwinkle Littorina littorea, along a pollution gradient in the western Scheldt estuary, The Netherlands. J. Exp. Mar. Biol. Ecol. 344, 78–87. Diot, M.-A., (PhD) 2012. Etude du vieillissement et de l’altération de nanocomposites de la vie courante. Université Paul Cézanne Aix Marseille III, Aix en Provence. Evan Ward, J., Shumway, S.E., 2004. Separating the grain from the chaff: particle selection in suspension- and deposit-feeding bivalves. J. Exp. Mar. Biol. Ecol. 300, 83–130. Federici, G., Shaw, B.J., Handy, R.D., 2007. Toxicity of titanium dioxide nanoparticles to rainbow trout (Oncorhynchus mykiss): gill injury, oxidative stress, and other physiological effects. Aquat. Toxicol. 84, 415–430. Felten, V., Charmantier, G., Mons, R., Geffard, A., Rousselle, P., Coquery, M., Garric, J., Geffard, O., 2008. Physiological and behavioural responses of Gammarus pulex (Crustacea: Amphipoda) exposed to cadmium. Aquat. Toxicol. 86, 413–425. Felten, V., Guerold, F., 2004. Haemolymph [Na+ ] and [Cl− ] loss in Gammarus fossarum exposed in situ to a wide range of acidic streams. Dis. Aquat. Organ. 61, 113–121. Giambérini, L., Cajaraville, M.P., 2005. Lysosomal responses in the digestive gland of the freshwater mussel, Dreissena polymorpha, experimentally exposed to cadmium. Environ. Res. 98, 210–214. Giambérini, L., Pihan, J.C., 1997. Lysosomal changes in the hemocytes of the freshwater mussel Dreissena polymorpha experimentally exposed to lead and zinc. Dis. Aquat. Organ. 28, 221–227. Griffitt, R.J., Weil, R., Hyndman, K.A., Denslow, N.D., Powers, K., Taylor, D., Barber, D.S., 2007. Exposure to copper nanoparticles causes gill injury and acute lethality in zebrafish (Danio rerio). Environ. Sci. Technol. 41, 8178–8186.

73

Guerlet, E., Ledy, K., Giambérini, L., 2006. Field application of a set of cellular biomarkers in the digestive gland of the freshwater snail Radix peregra (Gastropoda, Pulmonata). Aquat. Toxicol. 77, 19–32. Guerlet, E., Ledy, K., Meyer, A., Giambérini, L., 2007. Towards a validation of a cellular biomarker suite in native and transplanted zebra mussels: a 2-year integrative field study of seasonal and pollution-induced variations. Aquat. Toxicol. 81, 377–388. Guerlet, E., Vasseur, P., Giambérini, L., 2010. Spatial and temporal variations of biological responses to environmental pollution in the freshwater zebra mussel. Ecotoxicol. Environ. Saf. 73, 1170–1181. Hervant, F., Mathieu, J., Barré, H., Simon, K., Pinon, C., 1997. Comparative study on the behavioral, ventilatory, and respiratory responses of hypogean and epigean crustaceans to long-term starvation and subsequent feeding. Comp. Biochem. Physiol. Part A: Physiol. 118, 1277–1283. Jung, H., Kittelson, D.B., Zachariah, M.R., 2005. The influence of a cerium additive on ultrafine diesel particle emissions and kinetics of oxidation. Combust. Flame 142, 276–288. Koehler, A., Marx, U., Broeg, K., Bahns, S., Bressling, J., 2008. Effects of nanoparticles in Mytilus edulis gills and hepatopancreas: a new threat to marine life? Mar. Environ. Res. 66, 12–14. Korsvik, C., Patil, S., Seal, S., Self, W.T., 2007. Superoxide dismutase mimetic properties exhibited by vacancy engineered ceria nanoparticles. Chem. Commun., 1056–1058. Lushchak, V.I., 2001. Oxidative stress and mechanisms of protection against it in bacteria. Biochemistry (Moscow) 66, 476–489. Manier, N., Garaud, M., Delalain, P., Aguerre-Chariol, O., Pandard, P., 2011. Behaviour of ceria nanoparticles in standardized test media: influence on the results of ecotoxicological tests. J. Phys.: Conf. Ser. 304, 012058. Marigómez, I., Soto, M., Cajaraville, M.P., Angulo, E., Giamberini, L., 2002. Cellular and subcellular distribution of metals in molluscs. Microsc. Res. Tech. 56 (5), 358–392. Montes, M.O., Hanna, S.K., Lenihan, H.S., Keller, A.A., 2012. Uptake, accumulation, and biotransformation of metal oxide nanoparticles by a marine suspension-feeder. J. Hazard. Mater. 225–226, 139–145. Moore, M.N., 2006. Do nanoparticles present ecotoxicological risks for the health of the aquatic environment? Environ. Int. 32, 967–976. Moore, M.N., Depledge, M.H., Readman, J.W., Paul Leonard, D.R., 2004. An integrated biomarker-based strategy for ecotoxicological evaluation of risk in environmental management. Mut. Res./Fundam. Mol. Mech. Mutagen. 552, 247–268. O’Brien, N.J., Cummins, E.J., 2010. A risk assessment framework for assessing metallic nanomaterials of environmental concern: aquatic exposure and behavior. Risk Anal. 31, 706–726. OECD, 2010. List of manufactured nanomaterials and list of endpoints for phase one of the sponsorship programme for the testing of manufactured nanomaterials: revision. In: Series on the Safety of Manufactured Nanomaterials No. 27. Oktyabrsky, O.N., Smirnova, G.V., 2007. Redox regulation of cellular functions. Biochemistry (Moscow) 72, 132–145. Osman, A.M., Heuvel, H.v.d., Noort, P.C.M.v., 2007. Differential responses of biomarkers in tissues of a freshwater mussel, Dreissena polymorpha, to the exposure of sediment extracts with different levels of contamination. J. Appl. Toxicol. 27, 51–59. Palais, F., Mouneyrac, C., Dedourge-Geffard, O., Giambérini, L., Biagianti-Risbourg, S., Geffard, A., 2011. One-year monitoring of reproductive and energy reserve cycles in transplanted zebra mussels (Dreissena polymorpha). Chemosphere 83, 1062–1073. Pelletier, D.A., Suresh, A.K., Holton, G.A., McKeown, C.K., Wang, W., Gu, B., Mortensen, N.P., Allison, D.P., Joy, D.C., Allison, M.R., Brown, S.D., Phelps, T.J., Doktycz, M.J., 2010. Effects of engineered cerium oxide nanoparticles on bacterial growth and viability. Appl. Environ. Microbiol. 76, 7981–7989. Pirmohamed, T., Dowding, J.M., Singh, S., Wasserman, B., Heckert, E., Karakoti, A.S., King, J.E.S., Seal, S., Self, W.T., 2010. Nanoceria exhibit redox state-dependent catalase mimetic activity. Chem. Commun. 46, 2736–2738. Selvan, V.A.M., Anand, R.B., Dayakumar, M.U., 2009. Effects of cerium oxide nanoparticle addition in diesel and diesel–biodiesel–ethanol blends on the performance and emission characteristics of a CI engine. ARPN J. Eng. Appl. Sci., 4. Smolders, R., Bervoets, L., De Coen, W., Blust, R., 2004. Cellular energy allocation in zebra mussels exposed along a pollution gradient: linking cellular effects to higher levels of biological organization. Environ. Pollut. 129, 99–112. Sornom, P., Felten, V., Médoc, V., Sroda, S., Rousselle, P., Beisel, J.-N., 2010. Effect of gender on physiological and behavioural responses of Gammarus roeseli (Crustacea Amphipoda) to salinity and temperature. Environ. Pollut. 158, 1288–1295. Sroda, S., Cossu-Leguille, C., 2011. Seasonal variability of antioxidant biomarkers and energy reserves in the freshwater gammarid Gammarus roeseli. Chemosphere 83, 538–544. Sun, T.Y., Gottschalk, F., Hungerbühler, K., Nowack, B., 2014. Comprehensive probabilistic modelling of environmental emissions of engineered nanomaterials. Environ. Pollut. 185, 69–76. Tella, M., Auffan, M., Brousset, L., Issartel, J., Kieffer, I., Pailles, C., Morel, E., Santaella, C., Angeletti, B., Artells, E., Rose, J., Thiéry, A., Bottero, J.-Y., 2014. Transfer, transformation, and impacts of ceria nanomaterials in aquatic mesocosms simulating a pond ecosystem. Environ. Sci. Technol. 48, 9004–9013. Ting, S.R.S., Whitelock, J.M., Tomic, R., Gunawan, C., Teoh, W.Y., Amal, R., Lord, M.S., 2013. Cellular uptake and activity of heparin functionalised cerium oxide nanoparticles in monocytes. Biomaterials 34, 4377–4386. Van Hoecke, K., De Schamphelaere, K.A.C., Van der Meeren, P., Smagghe, G., Janssen, C.R., 2011. Aggregation and ecotoxicity of CeO2 nanoparticles in synthetic

74

M. Garaud et al. / Aquatic Toxicology 158 (2014) 63–74

and natural waters with variable pH, organic matter concentration and ionic strength. Environ. Pollut. 159, 970–976. Van Hoecke, K., Quik, J.T.K., Mankiewicz-Boczek, J., Schamphelaere, K.A.C.D., Elsaesser, A., Meeren, P.V.d., Barnes, C., McKerr, G., Howard, C.V., Meent, D.V.D., Rydzynski, K., Dawson, K.A., Salvati, A., Lesniak, A., Lynch, I., Silversmit, G., Samber, B.D., Vincze, L., Janssen, C.R., 2009. Fate and effects of CeO2 nanoparticles in aquatic ecotoxicity tests. Environ. Sci. Technol. 43, 4537–4546. Vasseur, P., Cossu-Leguille, C., 2003. Biomarkers and community indices as complementary tools for environmental safety. Environ. Int. 28, 711–717. Viarengo, A., Nott, J.A., 1993. Mechanisms of heavy metal cation homeostasis in marine invertebrates. Comp. Biochem. Physiol. Part C: Comp. Pharmacol. 104, 355–372. Voets, J., Talloen, W., de Tender, T., van Dongen, S., Covaci, A., Blust, R., Bervoets, L., 2006. Microcontaminant accumulation, physiological condition and bilateral asymmetry in zebra mussels (Dreissena polymorpha) from clean and contaminated surface waters. Aquat. Toxicol. 79, 213–225.

Wason, M.S., Colon, J., Das, S., Seal, S., Turkson, J., Zhao, J., Baker, C.H., 2013. Sensitization of pancreatic cancer cells to radiation by cerium oxide nanoparticle-induced ROS production. Nanomed. Nanotechnol. Biol. Med. 9, 558–569. Winkel, E.H.T., Davids, C., 1982. Food selection by Dreissena polymorpha Pallas (Mollusca: Bivalvia). Freshw. Biol. 12, 553–558. Xia, J., Zhao, H.Z., Lu, G.H., 2013. Effects of selected metal oxide nanoparticles on multiple biomarkers in Carassius auratus. Biomed. Environ. Sci. 26, 742–749. Xue, Y., Luan, Q., Yang, D., Yao, X., Zhou, K., 2011. Direct evidence for hydroxyl radical scavenging activity of cerium oxide nanoparticles. J. Phys. Chem. C 115, 4433–4438. Zholobak, N.M., Ivanov, V.K., Shcherbakov, A.B., Shaporev, A.S., Polezhaeva, O.S., Baranchikov, A.Y., Spivak, N.Y., Tretyakov, Y.D., 2011. UV-shielding property, photocatalytic activity and photocytotoxicity of ceria colloid solutions. J. Photochem. Photobiol. B: Biol. 102, 32–38.