The influence of surface coatings on the toxicity of silver nanoparticle in rainbow trout

Journal Pre-proof Toxicity of silver nanoparticle coatings in rainbow trout

J. Auclair, P. Turcotte, C. Gagnon, C. Peyrot, K.J. Wilkinson, F. Gagné PII:

S1532-0456(19)30339-4

DOI:

https://doi.org/10.1016/j.cbpc.2019.108623

Reference:

CBC 108623

To appear in:

Comparative Biochemistry and Physiology, Part C

Received date:

9 August 2019

Revised date:

4 September 2019

Accepted date:

5 September 2019

Please cite this article as: J. Auclair, P. Turcotte, C. Gagnon, et al., Toxicity of silver nanoparticle coatings in rainbow trout, Comparative Biochemistry and Physiology, Part C(2019), https://doi.org/10.1016/j.cbpc.2019.108623

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier.

Journal Pre-proof

Toxicity of silver nanoparticle coatings in rainbow trout. Auclair, J., Turcotte, P., Gagnon, C., Peyrot, C., Wilkinson KJ, Gagné, F. Aquatic Contaminants Research Division, Environment and Climate Change Canada, 105 McGill, Montréal, QC, Canada. H2Y 2E7.

Abstract

na

lP

re

-p

ro

of

Silver nanoparticles (nAg) are often produced with different coatings that could influence bioavailability and toxicity in aquatic organisms. The purpose of this study was to examine the influence of 4 surface coatings of nAg of the same core size towards bioavailability and toxicity in juvenile rainbow trout (Oncorhynchus mykiss). Juveniles were exposed to 50 µg/L of 50 nm diameter nAg for 96 h at 15oC with the following coatings: branched polyethylenimine (bPEI), citrate, polyvinylpyrrolidone (PVP) and silicate (Si). The data revealed that the coatings influenced hepatic Ag loadings in the following trend PVP>citrate>bPEI and Si with estimated bioavailability factors of 28, 18, 6 and 2 L/kg respectively. Hepatic Ag levels were significantly associated with DNA damage and inflammation as determined by arachidonate cyclooxygenase activity. The bPEI and citrate-coated nAg consistently produced the observed effects above in addition to increased mitochondrial electron transport activity and glutathione S-transferase activity. The absence of metallothionein and lipid peroxidation suggests that mechanisms other than the liberation of Ag+ were at play. In conclusion, surface coatings were shown to significantly influence bioavailability and toxic properties of nAg to rainbow trout juveniles.

Introduction

ur

Key words: silver nanoparticles, coatings, inflammation, genotoxicity.

Jo

Silver nanoparticles (nAg) represent one of the most popular nanomaterials owing to their antibacterial properties. It is estimated that over 55 million tons of nAg are produced annually for industrial applications (Bondarenko et al., 2013). The antimicrobial properties reside in the sustained release of ionic Ag+ and are conveniently incorporated in many consumer products such as clothes, medical devices, cosmetics, food-packaging materials, and household appliances (Wijnhoven et al., 2009). The increased use of nAg has raised concerns on the potential impacts of to the aquatic ecosystems. In municipal wastewaters using physico-chemical treatments, Ag was found at concentrations reaching 70 ng/L with colloidal Ag (nAg) representing 10% of the Ag loadings (Cervantes-Avilés et al., 2019). Although the addition of a secondary treatment

Journal Pre-proof process reduced the colloidal Ag loadings in the effluent, about 10-20 % of the colloidal Ag remained in the effluent which will be ultimately released in the environment. In another study, nAg was shown to degrade/dissolve in municipal effluent where 70 % of added nAg remained as nanoparticles in a primary-treated municipal effluent spiked with 80 nm citrate stabilized-nAg (Proulx et al., 2019). Surface modifications of nAg and other nanoparticles are introduced in the attempt to provide additional properties such as optimized persistence, toxicity and interaction

of

with biological targets such as microorganisms. Hence, the coatings could also influence the

ro

bioavailability and toxicity of nAg in addition to the size and form. Indeed, it was found that

-p

surface coatings, surface charge and particle size all contributed to toxicity in Allium cepa roots

nAg

could

differ

with

the

re

(Cvjetko et al., 2017). While Ag+ is generally considered more toxic than nAg, the toxicity of different

coatings

(citrate,

polyvinylpyrrolidone

and

lP

cetyltrimethylammonium bromide or CTAB). The positively-charged CTAB-coated nAg was

na

more toxic than the above coatings in the roots. Fish was also sensitive to nAg where the mean acute fish toxicity of nAg was at 1.4 mg/L which was lower than Ag+ with a mean acute lethality

ur

of 0.06 mg/L (Bondarenko et al., 2013). However these values could be influenced by the

Jo

surface coatings in addition to the exposure water properties such as pH, conductivity and organic matter content. The toxicity of nAg and Ag+ was examined in many studies which revealed that nAg/Ag+ can induce metallothioneins (MT), inflammation, oxidative stress and DNA damage in aquatic organisms (Ale et al., 2019; Gagné et al., 2012). For example, exposure to the African catfish Clarias gariepinus to either Ag+ or nAg led to decreased hepatosomatic index (HSI) and induced both hematocrit and white blood cells counts suggesting inflammation (Mekkawy et al., 2019). In Atlantic salmon exposed to 20 µg/L of nAg coated with citrate, increased plasma glucose, Na/K-ATPase (osmoregulation), heat shock proteins and MT gene

Journal Pre-proof expression were observed in gills (Farmen et al. 2012). The induction of MT suggests the release of Ag+ from nAg which could arise with electrostatically-bound coatings such as citrate. Hence, nAg capped with different coatings could have variable bioavailability and toxicity not only via the release of Ag+ but by other mechanisms as well. A toxicogenomic investigation of rainbow trout exposed to citrate-coated nAg and Ag+ revealed that nAg increased gene expression of inflammation which was not explained by Ag+ (Gagné et al., 2012). In another

of

study with zebrafish larvae, the 10 nm PVP-coated nAg was shown to decrease swimming

ro

activity while the 50 nm PVP-coated nAg increased swimming activity in zebrafish which

-p

suggest that both the size and coatings of nAg could influence the neuro-behavioral system

re

(Powers et al., 2011). In freshwater mussels exposed to nAg and Ag+, the nanoparticle produced changes in protein-ubiquitin levels which are involved in autophagosome-mediated process for

lP

the removal of denatured proteins by tagging with ubiquitin (Gagné et al., 2013a). Interestingly,

na

ubiquitin (an 8Kd protein binding to lysine residues on denatured proteins) was also shown to form a corona on the surface of nAg which suggests that alterations in this pathway was

ur

independent from Ag+ release (Mao et al., 2016). In respect to surface coatings, the presence of

Jo

negative charges at the surface of the nanoparticles reduced toxicity compared to a positivelycharged surface (Cvjetko et al., 2017). In another study, the toxicity of positively-coated cerium oxide nanoparticles (diethylaminoethyl dextran) were much more harmful towards survival and reproduction than negatively-charged (carboxymethyl dextran) and neutral (dextran) coatings (Arnt et al., 2017). The influence of surface coatings and charge of nAg on the biovailability and toxicity in fish requires more attention to better understand the risks of these products to the environment.

Journal Pre-proof The purpose of this study was therefore to examine the toxicity of nAg with 4 different coatings in respect to bioavailability and toxicity in rainbow trout Oncorhynchus mykiss juveniles. Four coatings of nAg were selected based on the restriction of having similar size and form of nAg (50 nm, spherical): citrate, silicate (Si), polyvinylpyrrolidone (PVP) and branched polyethyleneimine (bPEI) coatings. Citrate coating confer a negative charge at the surface although the citrate is weakly associated to nAg by electrostatic interaction with Ag+ at the

of

surface. The bPEI coating confers a positive charge at the surface of nAg and the PVP- and Si-

ro

coatings are considered neutral. The null hypothesis is that the coatings have no bearing on

-p

bioavailability and toxicity responses. Toxicity was examined at the inflammation, oxidative

re

stress, DNA damage and energy allocation levels to determine the health impacts in fish.

lP

Materials and methods

na

Preparation and characterisation of silver nanoparticles Silver nanoparticles (nAg) encapsulated with 4 different coatings were purchased from

ur

Nanocomposix Inc. (USA). The nAg of same size (50 nm) and shape (spheric) were selected

Jo

with the following coatings: branched polyethylenimine (bPEI), citrate, polyvinylpyrrolidone (PVP) and silicate (Si). The nanoparticle suspensions were prediluted to 1 % in MilliQ water and aquarium water to minimize aggregation. The hydrodynamic diameter and Zeta potential of nAg suspensions were analyzed in triplicate using a light scattering instrument (Mobius Instrument, Wyatt Technologies, ithSanta Barbara, CA, USA) operating with a laser at a wavelength of 532 nm. Prior to measurements, the instrument was calibrated using standard suspensions of the National Institute of Standards and Technology’s Traceable Particle Size Standards of monodisperse polystyrene spheres from Bangs Laboratories Inc. (USA).

Journal Pre-proof Exposure experiments and fish handling Fingerling rainbow trout (5.5±0.4 cm fork length; 2.8±0.3 g) were used to determine the acute toxicity of coated nAg (bPEI, citrate, PVP and Si nAg) following a standardized protocol by Environment and Climate Change Canada (Environmental Protection Series, 1990). Rainbow trout juveniles (N=10 per treatment) were exposed to 50 µg/L of each nAg for 96h at 15oC under constant aeration. Previous experiments revealed that uncoated nAg was not toxic to rainbow

of

trout at a concentration of 50 µg/L after 96 h exposure at 15oC. The selected concentration of

ro

nAg was based on the observation that heavy metals loadings of treated municipal effluents are

-p

in the order of 1-10 µg/L (Gagnon et al., 2006) and could reach 50-100 µg/L in some cases.

re

Thus, the concentration represents an upper limit for metal loading in a typical primary effluent

lP

and corresponds to circa 5 and 500 times the reported maximal total silver and nano-silver loadings in a municipal treated effluent respectively (Goullé et al., 2012; Proulx et al., 2016).

na

The exposure vessel consisted of 60 L containers lined with polyethylene bags. The controls

ur

consisted of aquarium water of tap water from the City of Montreal (QC, Canada) dechlorinated with UV treatment and charcoal filtration. The exposure experiments were repeated 3 times (3

Jo

vessels per treatment concentration). A positive control with ZnSO4 (LC50=0.5 mg/L) was also conducted to ensure that the fish batch was equally sensitive to a metal during the repetition of the exposure experiments. The concentration of each nAg coatings in aquarium water after one h at 15oC was verified by measuring total Ag using ICP-mass spectrometry (XSERIES 2 ICP-MS, Thermofisher Scientific USA) after acidification in 1% v/v HNO3 Seastar grade (BC, Canada). After the exposure period, the fish were euthanized in 50 mg/L tricaine methanesulfonate (4 L, buffered to pH 7.2 with NaHCO3) for 2 min and immediately placed on ice for dissection. The fish weight and fork length were determined and the livers were removed, washed in phosphate

Journal Pre-proof buffered saline (140 mM NaCl, 5 mM KH2PO4, pH 7.4, 1 mM NaHCO3 and 1 mM EDTA) and blotted on paper prior to weighting. The livers were then homogenized using a Teflon pestle tissue grinder in ice-cold phosphate buffered saline containing 0.1 mM dithiothreitol and 1 µg/mL apoprotinin protease inhibitor. The homogenate was sieved (200 µm mesh) and a portion of the homogenate was centrifuged at 15 000 x g for 20 min at 2oC. The supernatant (S15) was carefully removed and stored at -85oC for biomarker analyses. Total proteins were determined in

of

the homogenate and S15 fraction using the Coomassie Brillant Blue dye using standard solutions

ro

of serum bovine albumin (Bradford, 1976).

-p

For tissue analysis of Ag and the following essential elements copper (Cu), zinc (Zn) and iron

re

(Fe) by ICP-mass spectroscopy, the liver was mixed with 0.8 mL of concentrated HNO3, 0.1 mL

lP

of concentrated HCl (Seastar Chemical, BC, Canada), 0.2 mL of 30% H2O2 (Seastar Chemical) and adjusted to 1.2 mL with MilliQwater. The tissues were then digested for 2 h using a gradient

na

microwave digestion system. Total Ag, Cu, Zn and Fe were then determined by ICP-mass

Jo

reproductivity of <5%.

ur

spectrometry as described above. The data were expressed as metals in µg/g wet tissue with a

Sublethal effects assessments

The levels of DNA strand breaks were determined in the homogenate using the alkaline precipitation assay (Olive, 1988) using fluorescent detection of DNA strands (Bester et al., 1994; Gagné and Blaise, 1995). Briefly, 25 µL of the homogenate were mixed with 200 µL of 2% SDS containing 40 mM NaOH, 10 mM Tris base and 10 mM EDTA. After heating for 10 min at 60oC, 200 µL of 0.12M KCl were added, placed on ice for 10 min and centrifuged at 8 000 x g for 10 min at 4oC. An aliquot of the supernatant was collected for DNA quantitation using 150

Journal Pre-proof µL of 1 µg/mL Hoechst (bisbenziminde) in 0.4 M NaCl, 4 mM cholate and 0.1 M Tris-acetate pH 8.5 to reduce the interference from remaining traces of SDS and pH. The data were expressed as µg DNA breaks/mg proteins in the homogenate. The levels of lipid peroxidation (LPO) were also determined in the homogenate using the thiobarbituric acid reactants (TBARS) methodology (Wills, 1987). Standard solutions of tetramethoxypropane was used for calibration and TBARS were measured by fluorescence at 540 nm excitation and 600 nm emission. The data were

of

expressed as µg TBARS/mg proteins in the homogenate. The total levels of lipids and

ro

mitochondrial electron transport activity were determined as previously described (Gagné,

-p

2014a). The lipids levels were determined in the homogenate and crude mitochondrial fraction was prepared as follows for electron transport activity assessments. The homogenate was

re

centrifuged at 1 800 x g for 10 min at 4oC and the resulting supernatant was centrifuged at 9 000

lP

x g for 30 min at 4oC. The crude mitochondrial pellet was resuspended in homogenization buffer

na

and stored at -85oC.

The activity in arachidonate cyclooxygenase (COX) activity was determined by the coupled

ur

dehydrofluorescein/peroxidase reaction in the presence of arachidonic acid (Gagné, 2014b).

Jo

Briefly, 50 μL of the S15 fraction were mixed with 150 μL of the assay mixture composed of 50 μM of arachidonate, 2 μM of dichlorofluorescein, and 0.1 μg/mL of horseradish peroxidase in 50 mM of Tris–HCl buffer, pH 8, containing 0.05% Tween 20. The reaction mixture was incubated for 0, 10, 20 and 30 min at 30 °C, and fluorescence was measured at 485 nm for excitation and 520 nm for emission using a Synergy-4 microplate reader (Biotek instruments Inc, Vermont, USA). The data were expressed as increase in relative fluorescence units/(min × mg proteins). GST activity was determined using the spectrophotometric assay using 1-chloro-24dinitrobenzene and reduced glutathione as the substrates as previously described (Boryslawskyj

Journal Pre-proof et al., 1988). The assay was performed in the S15 fraction and the data were expressed as increase in absorbance at 340 nm/(min x mg proteins in the S15 fraction). The relative levels of MT were determined by the Ag saturation assay with graphite furnace atomic absorption detection (Gagné, 2014c). Briefly, 25 µL of the S15 fraction were mixed with one volume of 2.5 µg/mL AgNO3 in 100 mM glycine-NaOH, pH 8.5 for 10 min. After the

of

incubation period, the volume was adjusted to 250 uL with the glycine buffer and 25 µL of 2% hemoglobin (in water) was added for 5 min at room temperature. The mixture was then heated at

ro

100oC for 2 min, cooled on ice and centrifuged at 10 000 x g for 5 min. The surpernatant was

-p

collected from the (heat denatured) protein pellet and the hemoglobin/heating/centrifugation

re

steps were repeated one more time. Blanks consisted of homogenization buffer and standard

lP

solutions of rat metallothionein-2 (Sigma Chemical Company, ON, Canada) were also included for calibration. The levels of Ag remaining in the supernatant were determined by graphite

na

furnace atomic absorption spectrometer equipped with Zeeman effect background correction (Agilent, USA). The samples were diluted 1/20 in MilliQ water and 2 % (NH4)2HPO4 was used

ur

as the matrix modifier. Standards of Ag+ were used for calibration and a ratio of 17 moles of Ag

Data analysis

Jo

bound/mole MT was used (Scheuhammer and Cherian, 1986).

The exposure vessels contained 10 fish per treatment and the experiment was repeated three times. Data normality and homogeneity of variance were confirmed by the Shapiro-Wilk and Bartlett tests respectively. When the data proved not normal or displayed heterogenous variance, the data were log-transformed. The data were then subjected to an analysis of variance and critical difference between the nAg coating and controls were determined using the Least Square Difference (LSD) test. Correlation analysis was performed using the Pearson-moment procedure.

Journal Pre-proof Multivariate analysis of the biomarker data was performed using principal component and discriminant function analyses. Significance was set at p<0.05. All statistical tests were preformed using the SYSTAT software package (version13.2).

of

Results The basic physico-chemical properties of the coated nAg were provided (Table 1). The reported

ro

diameters of the nanoparticles were in the range of 50 nm based on electron microscopy. The

-p

evaluation of mean particle diameter was determined by DLS and revealed that most coated nAg

re

were in the 40-60 nm diameter range with the exception of Si-coated nAg with a measured

lP

diameter of 101 nm in MilliQ water. The Zeta potentials were somewhat lower at -8.9 mvolt with the PVP-coated nAg compared to the other coatings. Aggregation in aquarium water

na

(dechlorinated tap water of the City of Montreal) was not important for most nAg coatings with

ur

the exception of bPEI-coated nAg where the mean diameter rose to 355 nm compared to 41 nm

freshwater.

Jo

in MilliQ water. This suggests that cationic bPEI-coated nAg readily forms aggregates in

Exposure of rainbow trout to 50 µg/L nAg of different coatings led to significantly increased Ag in the liver and the coatings had a significant effect as determined by ANOVA (Figure 1A). The Si-coated nAg was less bioavailable in trout liver compared to the other coatings with a value of 0.1 µg Ag/g liver weight. The levels rose to 0.4, 1 and 1.5 µg Ag/g liver for bPEI-, citrate- and PVP-coated nAg respectively. The hepatic Ag content of PVP-coated nAg was therefore 15 times higher than Si-coated nAg. Based on the assumption that the liver is the principal organ for nAg accumulation, a provisional bioavailability factor could be derived with 6, 18, 28 and 2

Journal Pre-proof L/Kg for bPEI-, citrate-, PVP- and Si-coated nAg. This suggests that citrate- and PVP-coated nAg were more bioavailable than Si- and bPEI-coated nAg in juvenile fish. The levels of endogenous metals (Fe, Cu, Zn) were determined because their physiological role in redox homeostatis (Figure 1B). The analysis revealed that hepatic levels Fe levels were marginally decreased with citrate-coated nAg compared to bPEI coated nAg (0.1
of

Si-coated nAg, Cu levels were also significantly lower than those in fish exposed to PVP-,

ro

citrate-, and bPEI-coated nAg. For hepatic Zn levels, the levels dropped in fish exposed to

-p

citrate- and bPEI-coated nAg compared to controls. There was no significant change in the

re

condition factor (fish weight/fork length) between treatments (Figure 2 A). The HSI was significantly affected (ANOVA p=0.05) by the nAg coatings (Figure 2B). The HSI was

lP

marginally increased in fish exposed to PVP-coated nAg compared to controls (0.1
na

significantly differed between citrate-coated and PVP-coated nAg and between citrate-coated

ur

and Si-coated nAg (LSD test p<0.05).

The levels of DNA strand breaks were determined in the liver of trout exposed to the selected

Jo

coatings of nAg (Figure 3). The data revealed that DNA strand breaks were significantly induced by bPEI- and citrate-coated nAg in respect to controls (ANOVA and LSD tests p<0.05) suggesting that presence of electrostatic charges at the surface could damage DNA in the liver of rainbow trout. Correlation analysis revealed that hepatic DNA strand breaks were significantly correlated with liver Ag levels (r=0.43) and the condition factor (r=-0.37). Oxidative stress was also determined by following changes in arachidonate COX, GST activities and LPO levels (Figures 4A-C). COX activity was significantly induced by bPEI-, citrate- and PVP-coated nAg compared to controls (Figure 4A). No significant changes were obtained with Si-coated nAg

Journal Pre-proof compared to controls. Correlation analysis revealed that COX activity was significantly correlated with hepatic Ag and Zn levels. GST activity was also significantly induced by bPEI-, citrate- and Si-coated nAg in respect to controls (Figure 4B). The apparent increase in GST activity for the PVP-coated nAg was not significant (LSD test p>0.05). GST activity was significantly correlated with COX (r=0.41) and hepatic Zn levels (r=-0.45). No significant changes were obtained with LPO levels suggesting no oxidative damage. Correlation analysis

of

revealed that LPO levels were significantly correlated with GST activity (r=0.37) and condition

ro

factor (r=-0.41). There was no significant changes in MT levels although the MT levels were

-p

correlated with LPO levels (r=0.6; p<0.05, not shown). Changes in energy status were

re

determined by following changes in liver mitochondrial electron transport (MET) activity and lipid contents in fish exposed to the coated nAg (Figures 5A and B). MET activity was

lP

significantly increased in fish exposed to bPEI- and citrate-coted nAg in respect to controls

na

(Figure 5A). No changes in mitochondrial activity was obtained with PVP- and Si-coated nAg in respect to controls. Correlation analysis revealed that MET activity was significantly correlated

ur

with DNA damage (r=0.42), COX activity (r=0.43), GST activity (r=0.44) and LPO (r=0.45).

Jo

The total levels of lipid did not change with any type of surface coatings of nAg (Figure 5B). However, correlation analysis revealed that hepatic lipids were significantly correlated with LPO levels (r=0.52), GST activity (r=0.37), MET (r=0.37) and HSI (r=-0.42). In an attempt to have a more global view of the various responses in fish exposed to the surface coatings of nAg, multivariate analysis was performed using principal component and discriminant analyses. Principal component analysis revealed that 60% of the variance was explained by 3 factors (Figure 6A). The most important biomarkers (factorial weights ≥0.65) were MET activity, DNA strand breaks, GST activity and LPO levels for factor 1, GST and

Journal Pre-proof COX activities and hepatic Ag for factor 2, and hepatic Cu and Zn levels for the third factor. Discriminant function analysis was also performed to seek out differences between the surface coatings (Figure 6B). The analysis revealed that the surface coatings were completely discriminated with each other (mean classification efficiency of 100%) with the aforementioned biomarkers. Hepatic Ag levels, COX activity and LPO were the major endpoints for component 1 while HSI, condition factor and total hepatic Zn were the most important factors for component

of

2.

-p

ro

Discussion

The 4 different coatings of nAg in this study of nAg were bioavailable to different degrees to

re

rainbow trout juveniles although low availability was observed with Si-coated nAg. Since nAg is

lP

mainly accumulated in the liver (Jung et al., 2014), it was possible to estimate the bioavailability

order:

na

factor for the selected coatings of nAg. The estimated bioavaibility factors were in decreasing 28, 18, 6 and 2 L/Kg for PVP-, citrate-, bPEI-and Si-coatings respectively. The

ur

bioavaibility factors obtained in this study were in the same range but lower than those of

Jo

Japanese Medaka for PVP- (40 L/Kg) and citrate-coated nAg (43 L/Kg). The values obtained were determined in fish which were depurated in clean water overnight suggesting that nAg were not readily eliminated in fish. PVP-coated nAg where found to be the most bioavailable form of Ag which levels were significantly correlated with toxicity in embryonic zebrafish (Kim et al., 2013). Exposure to PVP-coated nAg significantly increased the HSI compared to fish exposed to citrate-coated nAg in rainbow trout. It also increased COX activity in exposed fish, which suggests inflammation. The bioavailability factor of citrate-coated nAg (18 L/Kg) with a diameter of 50 nm was similar (20 L/Kg) to a 20 nm diameter citrate-coated nAg in freshwater

Journal Pre-proof mussels (Gagné et al., 2013b) and was also mainly accumulated in the digestive gland (analogous to hepatic tissues in bivalves). Citrate-coated nAg and bPEI-coated nAg proved to be genotoxic in the present study. Moreover, DNA damage was significantly correlated with hepatic total Ag levels in fish. These coatings involve anionic (citrate) and cationic (bPEI) charges at the surface of nAg. The presence of

of

cationic charges is consistent of interactions with negatively charged DNA. Indeed, nucleic acids are negatively charged (phosphate backbone) at neutral pH where it can associate with positively

ro

charged ligands. Citrate-coated nAg consists of negatively charged citrate electrostatically and

-p

weakly bound to Ag+ at the surface of the nanoparticle which could be exchanged by other

re

anionic binding sites such as DNA. The toxicity of bPEI-coated nAg was also significantly

lP

higher in Bacillus sp (negatively charged Gram + bacteria) compared to negatively charged coatings such as citrate nAg and neutral PVP coating. The data thus corroborate the notion that

na

the surface charge is one of the important factors governing toxicity (El Badawy et al., 2011). In

ur

another study with zebrafish (Danio rerio), the toxicity of citrate-coated nAg was greater than PVP-coated nAg showing again the importance of surface charge towards toxicity (Liu et al.,

Jo

2019). However, the study also revealed that PVP-coated nAg was more bioavailable than citrate-coated nAg in zebrafish digestive system which was also observed in the present study. The genotoxicity of citrate-coated nanoAg was previously reported in freshwater mussels (Gagné et al., 2013a). In mussels, the citrate-coated nAg was able to induce MT and LPO in addition to increased DNA strand breaks indicating the involvement, in part at least, of Ag+ release during the manifestation of toxicity. In the present study, the induction of MT and LPO did not significantly change with the citrate-coated nAg and the other coatings suggesting mechanisms other than Ag+ release were at play. It was previously reported that nAg toxicity was coating-

Journal Pre-proof dependent and involved pathways other than the release of Ag+ which subsequently led to oxidative stress (Kwo et al., 2016). The toxicity involved disruption of sodium regulation by Na/K-ATPase and was related to the size of nAg aggregates and the coating properties. Based on DLS analysis, bPEI-coated nAg produced large aggregates (355 nm) compared to PVP- and citrate-coated nAg which suggests that coatings could also modulate the size of aggregates. The involvement of altered protein-ubiquitin binding involved in autophagy in mussels exposed to

of

citrate-coated nAg was also observed in mussels and this response was not induced by Ag+

ro

(Gagné et al., 2013a). Autophagy was proposed as a marker of exposure and effects for

-p

nanomaterials including nAg (Mao et al., 2016). Nanomaterials in cells could interact with

re

ubiquitin and proteins forming a proteinaceous corona which is removed by autophagosomes to be ultimately incorporated in lysozomes. Autophagy plays an important function for the selective

lP

removal of damaged or denatured proteins and injured organelles, and defect in this pathway

na

could lead to aggravated cytotoxic responses. A recent study in our laboratory with mussels exposed to the same coated nAg showed a significant decrease in protein-ubiquitin levels that

ur

were affected by the different nAg coatings which supports the hypothesis that this defective

Jo

protein turnover pathway is caused by nanomaterials (Auclair et al., 2019). DNA damage was also related with COX activity and lipids which suggests the involvement of inflammation and lipid mobilization respectively involving oxidative stress. Effects other than the liberation of Ag+ were also reported for citrate-coated nAg at the sodium uptake sites in juvenile rainbow trout (Schultz et al., 2013). The involvement of inflammation reaction of citrate-coated nAg was reported in rainbow trout at the toxicogenomic level (Gagné et al., 2012) and was not explained by Ag+. Indeed, it was found that citrate-coated nAg changed the expression of genes involved in inflammation while Ag+ influenced more genes in oxidative stress and protein denaturation (heat

Journal Pre-proof shock proteins). Exposure of PC12 cells to citrate-, PVP- and Si-coated nAg was examined to determine the influence of surface coatings (Powers et al., 2011). In undifferentiated cells, citrate- and PVP-coated nAg impaired DNA synthesis and produced oxidative stress leading to decrease differentiation into the acetylcholine phenotype. However, Si-coated nAg did not produce these effects and were biologically more neutral than the other coatings. This is consistent with the reduced bioavailability and toxicity in the present study for Si-coated nAg

of

i.e., no induction of COX activity and genotoxicity. The only observed effects of Si-coated nAg

ro

was increased GST activity compared to controls and increased HSI compared to citrate-coated

-p

nAg.

re

In conclusion, the surface coatings were shown to influence the bioavailability and generally

lP

influenced GST activity, inflammation and genotoxicity. Negatively (citrate) and positively (bPEI) charged coatings of nAg were more inflammatory and genotoxic than the neutral coatings

na

(Si and PVP) nAg. The absence of significant changes in MT induction and LPO levels suggests

ur

that toxicity was not related to released Ag+ and oxidative stress but to other properties. The data suggest that the surface coatings could influence the bioavailability and toxicity of nAg in

Jo

environmental risk assessments.

Acknowledgements

The research was funded by the Chemical Management Plan of Environment and Climate Change Canada. The technical assistance of Joanna Kowalczyk for the biochemical assays is recognized.

References Ale A, Liberatori G, Vannuccini ML, Bergami E, Ancora S, Mariotti G, Bianchi N, Galdopórpora JM, Desimone MF, Cazenave J, Corsi I. 2019. Exposure to a nanosilver-enabled

Journal Pre-proof consumer product results in similar accumulation and toxicity of silver nanoparticles in the marine mussel Mytilus galloprovincialis. Aquat Toxicol. 211: 46-56. Arndt DA, Oostveen EK, Triplett J, Butterfield DA, Tsyusko OV, Collin B, Starnes DL, Cai J, Klein JB, Nass R, Unrine JM. 2017. The role of charge in the toxicity of polymer-coated cerium oxide nanomaterials to Caenorhabditis elegans. Comp Biochem Physiol C Toxicol Pharmacol. 201:1-10. Auclair J., Turcotte P., Gagnon C., Peyrot C., Wilkinson K.J., Gagné F. 2019. The influence of surface coatings of silver nanoparticles on the bioavailability and toxicity of freshwater mussels. J Nanomat, in press.

ro

of

Bester MJ, Potgieter HC, Vermaak WJH 1994. Cholate and pH reduce interference by SDS in the determination of DNA with Hoescht. Anal Biochem 223, 299-305.

-p

Bondarenko O, Juganson K, Ivask A, Kasemets K, Mortimer M, Kahru A,2013. Toxicity of Ag, CuO and ZnO nanoparticles to selected environmentally relevant test organisms and mammalian cells in vitro: a critical review. Arch Toxicol 87: 1181–1200.

lP

re

Boryslawskyj M, Garrood AC, Pearson JT, Woodhead D 1988. Elevation of glutathione-Stransferase activity as a stress response to organochlorine compounds, in the freshwater mussel, Sphaerium corneum. Marine Environ Res 24:101-104.

na

Bradford MM 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254.

ur

Cervantes-Avilés P, Huang Y, Keller AA. 2019. Incidence and persistence of silver nanoparticles throughout the wastewater treatment process. Water Res. 156: 188-198.

Jo

Cvjetko P, Milošić A, Domijan AM, Vinković Vrček I, Tolić S, Peharec Štefanić P, LetofskyPapst I, Tkalec M, Balen B. 2017. Toxicity of silver ions and differently coated silver nanoparticles in Allium cepa roots. Ecotoxicol Environ Saf. 137: 18-28. El Badawy AM, Silva RG , Morris B, Scheckel KG, Suidan MT, Toalymat, TM. 2011. Surface Charge-Dependent Toxicity of Silver Nanoparticles. Environ. Sci. Technol. 2011, 45, 283–287. Environmental Protection Series. 1990. Biological test method: acute lethality test using rainbow trout. Method Development and Applications Section, Environmental Technology Centre, Environment Canada. Report EPS 1/RM/9. Farmen E, Evensen O, Mikkelsen HN, Einset J, Heier IS, Rosseland BO, Salbu B, Tollefsen KE, Oughton DH 2012. Acute and sub-lethal effects in juvenile Atlantic salmon exposed to low μg/L concentrations of Ag nanoparticles. Aquatic Toxicology 108, 78-84.

Journal Pre-proof Gagné, F. Blaise, C. 1995. Genotoxicity of environmental contaminants in sediments to rainbow trout hepatocytes. Environ. Toxicol. Wat. Qual 10, 217-229. Gagné, F., André, C., Skirrow, R., Gélinas, M., Auclair, J., Van Aggelen, G., Turcotte, P., Gagnon, C. 2012. Toxicity of silver nanoparticles to rainbow trout: A toxicogenomic approach. Chemosphere 89, 615-622. Gagné, F., Auclair, J., Turcotte, P., Gagnon, C. 2013a. Sublethal Effects of Silver Nanoparticles and Dissolved Silver in Freshwater Mussels. J. Toxicol Environ Health 76A, 479-490.

of

Gagné F, Auclair J, Fortier M, Bruneau A, Fournier M, Turcotte P, Pilote M, Gagnon C. 2013b. Bioavailability and immunotoxicity of silver nanoparticles to the freshwater mussel Elliptio complanata. J Toxicol Environ Health A. 76:767-777.

ro

Gagné, F., 2014a. Cellular energy allocation (Chapter 8). Biochemical Ecotoxicology— Principles and Methods, First Edition Elsevier Inc., USA.

re

-p

Gagné, F., 2014b. Biomarkers of infection and diseases (Chapter 11). Biochemical Ecotoxicology—Principles and Methods, First Edition Elsevier Inc., USA.

lP

Gagné, F., 2014c. Metal metabolism and detoxification (Chapter 5). Biochemical Ecotoxicology—Principles and Methods, First Edition Elsevier Inc., USA.

na

Gagnon C, Gagné F, Turcotte P, Saulnier I, Blaise C, Salazar MH, Salazar SM. 2006. Exposure of caged mussels to metals in a primary-treated municipal wastewater plume. Chemosphere 62: 998-1010.

Jo

ur

Goullé JP, Saussereau E, Mahieu L, Cellier D, Spiroux J, Guerbet M. 2012. Importance of anthropogenic metals in hospital and urban wastewater: its significance for the environment. Bull Environ Contam Toxicol. 89, 1220-1224. Jung YJ, Kim KT, Kim JY, Yang SY, Lee BG, Kim SD 2014. Bioconcentration and distribution of silver nanoparticles in Japanese medaka (Oryzias latipes). J Hazard Mater 267, 206-213. Kim KT, Truong L, Wehmas L, Tanguay RL. 2013. Silver nanoparticle toxicity in the embryonic zebrafish is governed by particle dispersion and ionic environment. Nanotechnology 24: 115101. Kwok KW, Dong W, Marinakos SM, Liu J, Chilkoti A, Wiesner MR, Chernick M, Hinton DE. 2016. Silver nanoparticle toxicity is related to coating materials and disruption of sodium concentration regulation. Nanotoxicology 10: 1306-1317. Liu H, Wang X, Wu Y, Hou J, Zhang S, Zhou N, Wang X.2019.Toxicity responses of different organs of zebrafish (Danio rerio) to silver nanoparticles with different particle sizes and surface coatings. Environ Pollut. 246: 414-422.

Journal Pre-proof Mao BH, Tsai JC, Chen CW, Yan SJ, Wang YJ 2016. Mechanisms of silver nanoparticleinduced toxicity and important role of autophagy. Nanotoxicology 10:1021-1040. Mekkawy IA, Mahmoud UM, Hana MN, Sayed AEH. 2019. Cytotoxic and hemotoxic effects of silver nanoparticles on the African Catfish, Clarias gariepinus (Burchell, 1822). Ecotoxicol Environ Saf. 171: 638-646. Olive PL 1988. DNA precipitation assay: a rapid and simple method for detecting DNA damage in mammalian cells Environ Mol Mutagen 11, 487-495.

of

Powers CM, Badireddy AR, Ryde IT, Seidler FJ, Slotkin TA. 2011. Silver nanoparticles compromise neurodevelopment in PC12 cells: critical contributions of silver ion, particle size, coating, and composition. Environ Health Perspect. 119: 37-44.

re

-p

ro

Powers CM, Slotkin TA, Seidler FJ, Badireddy AR, Padilla S. 2011. Silver nanoparticles alter zebrafish development and larval behavior: distinct roles for particle size, coating and composition. Neurotoxicol Teratol. 33: 708-714.

na

lP

Proulx K, Hadioui M, Wilkinson KJ. 2016. Separation, detection and characterization of nanomaterials in municipal wastewaters using hydrodynamic chromatography coupled to ICPMS and single particle ICPMS. Anal Bioanal Chem. 408:5147-5155.

ur

Scheuhammer, A. M., Cherian, M. G. 1986. Quantitation of metallothioneins by a silver saturation method. Toxicol. Appl. Pharmacol. 82, 417-425.

Jo

Schultz, A.G., Ong, K.J., MacCormack, T., Ma, G., Veinot, J.G.C. and Goss, G.G. 2013. Silver Nanoparticles Inhibit Sodium Uptake in Juvenile Rainbow Trout (Oncorhynchus mykiss). Environ Sci Technol 46, 10295-10301.

Wills ED (1987) Evaluation of lipid peroxidation in lipids and biological membranes. In: Snell K, Mullock B. (Eds.), Biochemical Toxicology: A Practical Approach. IRL Press, Washington, USA:127. Wijnhoven SWP, Peijnenburg W, Herberts CA et al (2009) Nano-silver – a review of available data and knowledge gaps in human and environmental risk assessment. Nanotoxicology 3, 109– 138.

Journal Pre-proof A 2,0

*

Hepatic Ag (ug/g)

1,5

* 1,0

of

0,5

* 0,0 Si

PVP

Cit

BPEI

-p

Control

ro

*

Coatings (50 ug/L nAg)

re

B

64

lP

70

Fe

Cu

na

Zn

52 46

ur

Hepatic metals (ug/g)

58

*

40

Jo

34 28

*

22

*

16 Control

Si

PVP

Cit

BPEI

Coatings (50 ug/L nAg)

Figure 1. Hepatic Ag levels in trout exposed to 4 different coatings of nAg. The levels of Ag (A) and endogenous Fe, Cu and Zn (B) were determined in the liver of exposed fish using ICP-mass spectrometry. The star symbol * indicates significance at p<0.05 from the control. The data are expressed as ug metal/g liver wet weight.

Journal Pre-proof 0,09

1,1

B

A

0,9

HSI

Fish weight/fork length)

1,0 0,08

0,07

0,8

0,7 0,06 0,6

Coatings (50 ug/L nAg)

BP EI

C it

PV P

Si

C on tro l

BP EI

C it

C on tro l

PV P

0,5 Si

0,05

Coatings (50 ug/L nAg)

of

Figure 2. Morphometric analysis of trout exposed to various coated nAg.

Jo

ur

na

lP

re

-p

ro

The condition factor (A) and hepatic somatic index (B) were determined in fish exposed for 96 h to 4 coatings of nAg. The star symbol * indicates significance at p<0.05 from the control.

Journal Pre-proof

0,10

*

of

*

-p

ro

0,05

0,00 Control

re

DNA breaks (ug DNA/mg proteins)

0,15

Si

PVP

Cit

BPEI

lP

Coatings (50 ug/L nAg) Figure 3. DNA damage of various coated nAg in trout liver.

Jo

ur

na

The levels of DNA strand breaks were determined in the liver of exposed fish. The star symbol * indicates significance at p<0.05 from the control.

Journal Pre-proof A

200

* *

* 150

100 Si

PVP

Cit

BPEI

ro

Control

of

COX activity (RFU/min/mg proteins)

250

Coatings (50 ug/L nAg)

-p

B

re lP

*

4

*

*

Cit

BPEI

na

GST activity (Abs/min/mg proteins)

5

Jo

ur

3

2 Control

Si

PVP

Coatings (50 ug/L nAg)

Figure 4. Changes in oxidative stress in fish exposed to the different nAg coatings. |The activitiy of COX (A) and GST (B) are shown. COX activity was expressed as relative fluorescence units/min/mg proteins and GST activity was expressed as change in absorbance/min/mg proteins. The star symbol * indicates significance at p<0.05 from the control.

Journal Pre-proof A 15

13 12

* *

11 10 9

of

8 7 6 5 Control

Si

PVP

ro

MET activity (Abs/mimn/mg proteins)

14

Cit

BPEI

-p

Coatings (50 ug/L nAg)

re

B

lP

na

25

20

Jo

ur

Hepatic lipids (ug lipids/mg proteins)

30

15

10 Control

Si

PVP

Cit

BPEI

Coatings (50 ug/L nAg)

Figure 5. Change in cellular energy allocation in fish exposed to coatings of nAg Mitochondrial electron transport (A) and hepatic lipids (B) were determined in fish exposed to coated nAg. The star symbol * indicates significance at p<0.05 from the control.

Journal Pre-proof A Factor Loadings Plot

1,0 Cu

Factor(3)

0,5

ZN Ag

CF

DNA

Lip MT

COX

MET

0,0

LPO

Fe

GST

of

-0,5

-0 ,5

0, 0

Fa cto r(2

-0,5

-1,0

re

-1 ,0

)

-p

0, 5

1, 0

ro

HSI

0,0

1) tor(

Fac

lP

B

1,0 0,5

5,00 3,75

Controls

PVP

1,25 0,00

Jo

Component 2

ur

2,50

na

Mean classification:100%

-1,25

Si Citrate

-2,50 -3,75 bPEI -5,00 -30

-18

-6

6

18

30

Component 1 Figure 6. Multivariate analysis of biomarker data. Principal component (A) and discriminant function (B) analyses were performed with the biomarker data. In A), the major biomarkers (factorial weight >0.7) were tissues Ag, HSI, CF, COX, DNA damage and MT. In B), the classification efficiency was 100% and the mean scores are shown. The major biomarkers were hepatic Ag> COX>LPO>Zn and HSI> Zn>COX>COX for component 1 (x axis) and component 2 (y axis) respectively.

Journal Pre-proof

Table 1. Physico-chemical properties of the coated silver nanoparticles. Diameter1 (nm) 51 ± 6

Citrate

polyvinylpyrolidone (PVP) branched polyehtyleneimine (bPEI)

Zeta potential2 (mvolt)

42 ± 3 (MilliQ) 56 ± 1 (Aqarium) 87 ± 4 (MilliQ) 87 ± 1 (Aquarium) 60 ± 2 (MilliQ) 53 ± 3 (Aquarium)

NA -11.4 ± 2 NA -12.2 ± 3 NA -9 ± 2

41 ± 0.5 (MilliQ) 355 ± 16 (Aquarium)

NA -22 ± 2

0.021

49±6

Silica

Measured diameter2 (nm)

Total Ag (mg/mL)

1.07

51 ± 5

0.021

47 ± 5

0.022

of

Nanosilver coatings

Jo

ur

na

lP

re

-p

ro

1. According to manufacturer’s data sheet based on transmittance electron microscopy. 2. Measured by DLS as described in Methods at 50 µg/L concentration in either MilliQ or aquarium waters.

Journal Pre-proof Table 2. Correlation analysis DNA COX GST Lip LPO MT MET CF HSI Fe Cu Ag Zn ----------------------------------------------------------------------------------------------------------------------------------------------------------------------DNA 1 COX 0.43* 1 GST 0.31 0.41* 1 Lip 0.37* 0.20 0.29 1 LPO 0.32 -0.01 0.37* 0.52* 1 MT 0.26 -0.15 0.34 0.23 0.56* 1 MET 0.42* 0.43* 0.44* 0.37* 0.45* 0.0 1 CF -0.37* -0.30 -0.21 -0.09 -0.03 -0.15 -0.06 1 HSI -0.18 0.01 -0.05 -0.13 -0.41* -0.34 -0.31 -0.23 1 Fe -0.24 -0.25 -0.30 -0.42* -0.34 -0.13 -0.28 0.22 0.08 1 Cu 0.34 0.28 -0.34 -0.01 -0.11 -0.0 -0.0 0.08 - 0.19 0.0 1 Ag 0.43* 0.47* -0.02 0.10 -0.02 0.06 0.21 0.02 0.07 -0.09 0.46* 1 Zn -0.06 -0.40* -0.45* 0.11 -0.19 0.07 -0.36 0.20 -0.21 0.19 0.32 0.15 1 ----------------------------------------------------------------------------------------------------------------------------------------------------------------------The star * symbol indicates significance at p<0.05. CF: condition factor (fish weigh/fork length); HSI: hepatic somatic index.

f o

l a

Jo

n r u

r P

e

o r p

Journal Pre-proof Conflicts of interest declaration

The authors declare no conflict of interest during the submission processes either financial or otherwise regarding the manuscript untitled: Toxicity

of silver nanoparticle coatings in rainbow

trout.

Jo

ur

na

lP

re

-p

ro

of

F. Gagné on behalf of the co-authors.

lP

re

-p

ro

of

Journal Pre-proof

Jo

ur

na

Graphical abstract

Journal Pre-proof Highlights The coatings of silver nanoparticles influence availability and toxicity in fish Silver nanoparticles are available in the liver and produce inflammation and DNA damage

Jo

ur

na

lP

re

-p

ro

of

Charged coatings produce stronger effects in fish liver