1 Interference of goethite in the effects of glyphosate and Roundup® on ZFL cell line Natara D. G. da Silva a, Cristiane E. A. Carneiro b, Estefânia V. R. Campos c, Jhones L. de Oliveira c, Wagner E. Risso d, Leonardo F. Fraceto c, Dimas A. M. Zaia b, Cláudia B. R. Martinez a,d,* a Programa de Pós-Graduação em Ciências Biológicas, Universidade Estadual de Londrina – UEL, Londrina, Paraná, Brasil. b Departamento de Química, Universidade Estadual de Londrina – UEL, Londrina, Paraná, Brasil. c Departamento de Engenharia Ambiental, Universidade Estadual Paulista – UNESP, Sorocaba, São Paulo, Brasil. d Departamento de Ciências Fisiológicas, Universidade Estadual de Londrina - UEL, Londrina, Paraná, Brasil. *Corresponding author at: Departamento de Ciências Fisiológicas, Universidade Estadual de Londrina, Londrina, Paraná 86057-970, Brasil. E-mail address: cbueno@uel.br (C.B.R. Martinez). 2 1 Abstract 2 3 Goethite (α-FeOOH) brings important perspectives in environmental 4 remediation, as, due to its physicochemical properties, this iron oxide can 5 adsorb a wide variety of compounds, including glyphosate. This study aimed to 6 evaluate the effects of goethite nanoparticles (NPs), glyphosate (Gly), 7 Roundup® (Rd), and co-exposures (Gly+NPs and Rd+NPs) on zebrafish liver 8 cell line (ZFL). ZFL cells were exposed to NPs (1, 10, and 100 mg L-1), Gly (3.6 9 mg L-1), Rd (10 mg L-1), and co-exposures (Gly+NPs and Rd+NPs), or only to 10 saline for 1, 6, and 12 h. Cell viability was assessed by Trypan blue, MTT, and 11 neutral red assays. The generation of reactive oxygen species and total 12 antioxidant capacity were also determined, while genotoxicity was quantified by 13 the comet assay. Both NPs and Rd in isolation produced cytotoxic effects at 6 h 14 and genotoxic effects at 1 and 6 h. Rd+NPs resulted in synergistic effects, 15 intensifying the toxicity. Cells exposed to Gly did not present toxic effects and 16 Gly+NPs resulted in the suppression of toxic effects observed for NPs. The 17 presence of other components in Roundup® seems to favor its toxicity 18 compared to the active ingredient. In conclusion, according to the in vitro model, 19 the concentrations used were not safe for the ZFL lineage. 20 21 Keywords: iron oxide; nanotoxicology; in vitro assays; Danio rerio; cell viability; 22 DNA damage. 3 1 1. Introduction 2 3 Glyphosate, N-(phosphonomethyl) glycine, is an active ingredient that 4 occupies first place in the ranking of commercialized herbicides worldwide 5 (Benbrook, 2016), and is widely used in the control of invasive plants, both in 6 agriculture and domestic use (Benbrook, 2016; De Gerónimo et al., 2018; 7 Hanke et al., 2010; Tang et al., 2015). For this reason, the presence of 8 glyphosate is increasingly common in aquatic environments (Bruggen et al., 9 2018; Maqueda et al., 2017), leading to contamination of surface and 10 groundwater, mainly through processes such as surface runoff and soil leaching 11 (Borggaard and Gimsing, 2008; Bruggen et al., 2018; Marques et al., 2014). 12 According to the World Health Organization, glyphosate toxicity is low for 13 non-target organisms (WHO, 2005). This is based on the mode of action of 14 glyphosate, which consists of its ability to reduce aromatic amino acids 15 synthesis through specific inhibition of the enzyme 5-enolpyruvylshikimate-3- 16 phosphate synthase, present in a small range of organisms, primarily green 17 plants (Lopes et al., 2014; Lushchak et al., 2009). However, several authors 18 suggest higher toxicity of glyphosate formulated products for aquatic organisms, 19 demonstrating alterations in the activity of antioxidant enzymes (Lushchak et al., 20 2009; Modesto and Martinez, 2010b), cytotoxic effects (Goulart et al., 2015; 21 Koller et al., 2012; Peixoto, 2005; Pereira et al., 2018), genotoxic effects 22 (Cavalcante et al., 2008; Marques et al., 2014; Moreno et al., 2014), altered 23 acetylcholinesterase activity (Modesto and Martinez, 2010b), and reproductive 24 and behavioral changes (Bridi et al., 2017; Vanlaeys et al., 2018). 25 Roundup® is the commercial name of a popular glyphosate-based 26 herbicide extensively used around the world (Caballero-Gallardo et al., 2016). 27 Its composition includes, besides glyphosate (360 g L-1), a mixture of 28 surfactants, containing 15% polyoxyethylene amine (POEA), which appears to 29 be related to the increased toxicity of the product (Lushchak et al., 2009; 30 Moreno et al., 2014; Peixoto, 2005; Tsui and Chu, 2003; Vanlaeys et al., 2018). 31 In this context, Navarro and Martinez (2014) reported several damages related 32 to imbalance in the redox state of the neotropical fish Prochilodus lineatus, such 33 as hemolysis, DNA damage, and lipid peroxidation after exposure to surfactant 34 POEA. In addition, glyphosate-based herbicides with POEA were more toxic to 4 1 the microcrustacean Artemia salina and the zebrafish Danio rerio than 2 formulations without POEA, both with 360 g of glyphosate L-1 (Rodrigues et al., 3 2017). 4 Considering the potential for contamination of water bodies by 5 glyphosate, iron oxide nanoparticles present important perspectives in the 6 remediation of contaminated aquatic environments (Liu et al., 2014). Due to 7 their special physicochemical properties (large surface area and high surface 8 reactivity), iron oxide nanoparticles can adsorb a variety of compounds such as 9 anions, organic compounds/organic acids, and cations (Kharisov et al., 2012; 10 Liu et al., 2014; O’Carroll et al., 2013; Tosco et al., 2014). 11 Among the most commonly used iron oxides, goethite (α-FeOOH) is 12 highlighted as the most thermodynamically stable (Cornell and Schwertmann, 13 2003), as an abundant and cheap natural material (Liu et al., 2014), as well as 14 for presenting great adsorption capacity for glyphosate (Jonsson et al., 2008; 15 Liu et al., 2014; Sheals et al., 2002; Yang et al., 2018). Goethite may form 16 monodentate complexes with glyphosate by binding to the phosphonate group 17 or form bidentate bridges on the iron oxide surface in an inner sphere mode, 18 while the carboxylate and amino group are noncoordinated to the surface (Barja 19 and dos Santos Afonso, 2005). 20 Although several studies have evaluated the structure and 21 characterization of goethite (Cornell and Schwertmann, 2003; Dultz et al., 2018; 22 Liu et al., 2014), its adsorption capacity for glyphosate (Barja and dos Santos 23 Afonso, 2005; Jonsson et al., 2008; Pessagno et al., 2008; Sheals et al., 2002; 24 Waiman et al., 2012), and its potential for application in environmental 25 protection (Liu et al., 2014; Orcelli et al., 2018), there is a lack of studies 26 evaluating biological responses to exposure to goethite and, therefore, its 27 toxicological impact, mainly for aquatic organisms, remains unknown. 28 In this context, studies at the cellular level are extremely important, since 29 they provide information on the mechanisms of action of the agents tested 30 and/or the cellular response and, thus, can be used preventively (Bols et al., 31 2005; Fent, 2007). The development of in vitro models has increased 32 significantly in recent years and these play a key role in toxicological research, 33 offering ethical, scientific, and economic advantages (Jennings, 2015; Singh et 34 al., 2018; Srivastava et al., 2018). Fish cell lines are of growing importance in 5 1 eco/genotoxicology as they represent standardized experimental systems that 2 are carried out in a controlled environment, giving fast, affordable and ethically 3 eligible results (Žegura and Filipič, 2019). 4 In the present study, we used the hepatocyte cell line from Danio rerio 5 (zebrafish), initially obtained by Ghosh et al. (1994), which presents 6 characteristics of hepatic parenchyma cells. As the liver is the main organ of 7 metabolism, detoxification, and homeostasis, the applicability of this lineage 8 becomes relevant in in vitro toxicological studies. In addition, several studies 9 have demonstrated the sensitivity of this lineage to metal toxicity (Chen and 10 Chan, 2018; Sandrini et al., 2009; Tang et al., 2013), nanomaterials (Azevedo 11 Costa et al., 2012; Morozesk et al., 2018; Thit et al., 2017), pharmaceutical 12 products (Gajski et al., 2016; Novak et al., 2017), biodiesel (Cavalcante et al., 13 2014), gasoline (Lachner et al., 2015), and herbicides (Goulart et al., 2015; 14 Lopes et al., 2018). 15 Thus, considering the potential risk of contamination of aquatic 16 organisms by glyphosate and the limited studies evaluating biological 17 responses to goethite, this work aimed to evaluate the possible cytotoxic, 18 biochemical, and genotoxic effects of goethite nanoparticles (NPs), glyphosate, 19 and the formulated product Roundup® on the hepatocyte cell line of D. rerio 20 (ZFL), as well as to evaluate the interference of the nanomaterial in the effects 21 of these herbicides. 22 23 2. Material and Methods 24 25 2.1. Preparation of goethite nanoparticles (NPs) 26 The goethite NPs were produced according to the protocol described by 27 Schwertmann and Cornell (2000). A solution containing goethite NPs was 28 dispersed in ultra-pure water and sonicated. The concentration of NPs in the 29 stock solution was estimated by dry weight and the corresponding iron 30 concentration. This solution was autoclaved and used for the preparation of 31 exposure solutions at final concentrations of 1, 10, and 100 mg L-1 of NPs 32 diluted in Dulbecco´s phosphate buffered saline with calcium, magnesium and 33 glucose (Dulbecco's PBS: 136.9 mM NaCl; 2.68 mM KCl; 0.90 mM CaCl2; 0.49 34 mM MgCl2•6H2O; 7.58 mM Na2HPO4; 1.47 mM KH2PO4; 5.55 mM glucose; pH 6 1 7.4). Considering the lack of data reporting the effects of goethite NPs at 2 cellular level, the concentrations used in the present work were based on 3 cytotoxicity results found in the literature for different iron oxide nanoparticles 4 (Ankamwar et al., 2010; Bhattacharya et al., 2012; Karlsson et al., 2008; Singh 5 et al., 2010). Furthermore, previous tests (MTT and NR) were performed in a 6 range of 1 – 200 mg L-1 (data not shown) to select appropriate concentrations. 7 Based on these results, the concentrations of 1, 10, and 100 mg L-1 were 8 chosen to further investigate the toxic effects of goethite NPs in ZFL cells. 9 10 2.2. Characterization of goethite NPs 11 The goethite NPs exposure solutions (1, 10, and 100 mg L-1) were 12 characterized for both the hydrodynamic diameter and the polydispersity index 13 using the dynamic light scattering technique (DLS) and zeta potential (ZP) 14 through the electrophoresis light scattering technique. For this, a Malvern ZS90 15 particle analyzer (Malvern®) was used at a fixed angle of 90° to 25°C. The 16 characterization tests were conducted in triplicate at 0, 24, and 48 h in the 17 absence of stirring, in order to approximate the exposure performed in the 18 toxicity tests. 19 20 2.3. Quantification of iron (Fe) 21 Total and dissolved Fe concentrations were analyzed in the exposure 22 solutions. For the analysis of total Fe, the solutions were fixed in nitric acid 23 (HNO3 0.5% - Fmaia, Brazil) and for dissolved Fe solutions, they were filtered 24 (0.45 μm) and then fixed. The samples were analyzed using an atomic 25 absorption spectrophotometer (EAA - Analyst 700, Perkin Elmer®, USA) through 26 flame atomization. 27 28 2.4. ZFL cell line 29 The ZFL cell line was grown in 25 cm2 flasks using medium containing 30 50% Leibovitz L-15 (Gibco®), 40% RPMI 1640 (Gibco®), and 10% fetal bovine 31 serum (FBS) (Gibco®). The flasks were kept in a dry oven without addition of 32 CO2 at 28°C. 33 34 2.5. In vitro exposures 7 1 Concentrations of the selected herbicides, Roundup® (360 g glyphosate 2 L-1, Monsanto do Brasil LTDA) and glyphosate (CAS no. 1071-83-6, Milenia 3 Agrociências S/A), were defined taking into account previous studies that our 4 research group have already developed with the neotropical fish P. lineatus 5 exposed to Roundup Transorb®, Roundup®, and glyphosate (Cavalcante et al., 6 2008; Modesto and Martinez, 2010a, b; Moreno et al., 2014). 7 Stock solutions of Roundup® and glyphosate were prepared at a 8 concentration 100 times higher (100 and 36 mg L-1, respectively) than the final 9 exposure and diluted in Dulbecco's PBS before distribution to the plate wells to 10 reached 10 and 3.6 mg L-1, respectively. Exposure concentrations of Roundup® 11 and glyphosate were based on the concentration of the active ingredient 12 present in the formulation (360 g glyphosate L-1). 13 All the experimental solutions were prepared in Dulbecco's PBS to avoid 14 any interactions between the treatments and the compounds present in the 15 culture medium, which could influence the results. In addition, preliminary tests 16 demonstrated that Dulbecco's PBS was able to maintain ZFL cell viability 17 (mitochondrial and lysosomal activity) for at least twelve hours (data not 18 shown), which corresponds to the maximum exposure time used in this study. 19 ZFL cells were seeded at a density of 106 cells mL-1 on a transparent 96- 20 well plate (TPP®) for the cytotoxic assays, on a black 96-well plate with a clear 21 bottom (Perkin Elmer®) for the biochemical assays, and on a transparent 24- 22 well plate (TPP®) for the genotoxic assay. After 24 h of cell attachment at 28°C, 23 the cells were exposed for 1, 6, and 12 h to the following treatments: goethite 24 NPs at 1 mg L-1 (N1), 10 mg L-1 (N2), and 100 mg L-1 (N3); glyphosate at 3.6 mg 25 L-1 (Gly), and co-exposures to Gly plus goethite NPs (GlyN1, GlyN2, GlyN3); 26 Roundup® at 10 mg L-1 (Rd); and Rd plus goethite NPs (RdN1, RdN2, RdN3). 27 The cells of the control groups (CTR) were only exposed to Dulbecco's PBS. 28 Three independent experiments were performed, and for each assay four plates 29 were assembled per experimental time and, depending on the assay, the 30 number of replicates per treatment was variable (cytotoxic and biochemical 31 assays: eight replicates per treatment; genotoxic assay: two replicates per 32 treatment). 33 34 2.6. Cytotoxic assays 8 1 Cytotoxicity was evaluated using different assays. To verify the integrity 2 of the plasma membrane, the Trypan blue exclusion test (TB) was used prior to 3 the comet assay. After exposure, the cells were detached from the wells, 4 homogenized with 0.4% of the TB dye, and a total count of 100 cells was 5 performed using a Neubauer chamber. Cell viability was expressed as the 6 percentage of viable cells and the treatments with cell viability equal to or 7 greater than 80% (Tice et al., 2000) were selected for ROS, ACAP, and comet 8 assay. 9 Mitochondrial activity was assessed through the reduction in 3-4,5- 10 dimethylthiazol-2-yl-2,5-diphenyltetrazolium bromide (MTT) salt according to the 11 protocol of Mosmann (1983), with modifications. After exposure, the MTT salt 12 was added to the wells for 4 h at a final concentration of 0.80 mM. 13 Subsequently, the plate was centrifuged for 5 min at 200 g, dimethyl sulfoxide 14 (DMSO, 99.5%) was added for solubilization of the formazan crystals, and the 15 absorbance was determined on a microplate reader (Victor 3, Perkin Elmer®) at 16 540 nm. 17 Lysosomal integrity was also assessed by the neutral red (NR) retention 18 assay. After exposure, the cells were incubated with NR dye (40 μg mL-1) for 3 19 h. Next, the plate was centrifuged for 5 min at 200 g, fixed with formaldehyde 20 (0.5%) in calcium chloride solution (1%) for 2 min, and exposed to an acid 21 alcohol solution (1% acetic acid in 50 % ethyl alcohol) under constant stirring for 22 15 min. The absorbance was determined on a microplate reader (Victor 3, 23 Perkin Elmer®) at 540 nm and, for MTT and NR assays, the results of the 24 different treatments were given in relation to the CTR, which was considered as 25 100% cell viability. 26 27 2.7. Biochemical assays 28 The generation of reactive oxygen species (ROS) and the total 29 antioxidant capacity against peroxyl radicals (ACAP) were determined 30 according to the protocol of Amado et al. (2009) with modifications. After the 31 exposures, the solutions were withdrawn and reaction medium (30 mM HEPES, 32 200 mM KCl, 1 mM MgCl2, pH 7.2) was added to all wells. For each 33 experimental treatment, eight replicates were performed per plate; four wells 34 were treated with potassium phosphate buffer (0.1 M, pH 7.4) and the other four 9 1 received peroxyl, 2,2'-azobis radicals (2-methylpropinamide) dihydrochloride (10 2 mM ABAP, pH 7.2). The autofluorescence reading was performed on a 3 microplate reader (Victor 3, Perkin Elmer®). This was carried out by adding the 4 2'-7'-dichlorofluorescein diacetate compound (1 mM H2CDF-DA) and re-reading 5 at 35°C for 70 min with readings every 5 min, excitation of 485 nm, and 6 emission of 520 nm (Azevedo Costa et al., 2012). To quantify the ROS and 7 ACAP, the fluorescence data were adjusted to a second-order polynomial 8 function and the integral value was calculated. For ROS quantification, the 9 values of the integrals of the samples treated with potassium phosphate buffer 10 in isolation were analyzed and the results expressed as unit area of 11 fluorescence over time (FU x min). To evaluate the ACAP, the difference in the 12 area values of the samples treated and non-treated with ABAP were calculated. 13 A greater difference between the areas indicated lower antioxidant capacity of 14 the sample. To facilitate visualization of the results, ACAP data were inverted 15 (1/relative area). 16 17 2.8. Genotoxic assay 18 To evaluate DNA damage, the comet alkaline test was performed 19 according to Singh et al. (1988) with modifications. For this assay, a positive 20 control (PC) was also performed using 0.5 mM methyl methanesulfonate 21 (MMS). After exposure, aliquots of 20 μL of each sample were homogenized 22 with low melting point agarose (5%) for the preparation of slides previously 23 prepared with normal melting point agarose (1.5%), covered with coverslips, 24 and placed in a refrigerator for 40 min. After this period, the coverslips were 25 removed and the slides were immersed in lysis solution (2.5 mM NaCl, 100 mM 26 EDTA, 10 mM Tris, 1% Triton X-100, 10% DMSO, pH 10) for 2 h. After the lysis, 27 the slides were transferred to an electrophoresis cube, containing fresh and ice- 28 cold alkaline buffer (1 mM EDTA and 300 mM NaOH, pH > 13) for 35 min. The 29 electrophoresis was conducted at 25 V and 300 mA for 20 min. The slides were 30 then neutralized (0.4 M Tris, pH 7.5) for 15 min and fixed with 100% ethanol for 31 10 min. The slides were stained with GelRed (Biotium®) and 100 nucleoids per 32 slide were analyzed in a blinded test under a fluorescence microscope (Leica®, 33 DM 2500) in a 40X objective. DNA damage was visually classified, according to 34 the migration of the DNA fragments into four classes (Collins et al., 2008): class 10 1 0 (nucleoid without tail with few fragments around), class 1 (tail smaller than the 2 nucleoid diameter), class 2 (tail with a length one to two times the diameter of 3 the nucleoid), and class 3 (tail with a length greater than twice the diameter of 4 the nucleoid). The damage score was obtained by multiplying the number of 5 nucleoids observed in each class of damage analyzed by the value of the class. 6 7 2.9. Statistical analyzes 8 For each parameter evaluated, the results were compared between the 9 different treatments CTR x N1 x N2 x N3, CTR x Gly x GlyN1 x GlyN2 x GlyN3, 10 CTR x Rd x RdN1 x RdN2 x RdN3, and CTR x PC (where applicable) for each 11 experimental time (1, 6, and 12 h) through parametric (ANOVA) and 12 nonparametric analysis of variance (Kruskal-Wallis), according to distribution of 13 data (normality and homogeneity of variance). When necessary, the differences 14 were identified by the Student-Newman-Keuls (SNK) multiple comparisons test. 15 Values of P < 0.05 were considered significant and the results are expressed as 16 mean ± standard error (SE). 17 18 3. Results 19 20 3.1. Characterization of goethite NPs 21 The results indicate a significant increase in the mean diameter of 22 goethite nanoparticles in N2 and N3 treatments at 0 h (P < 0.001) and 48 h (P = 23 0.005) (Figure 1A). For 24 h (P = 0.080), no difference in this parameter was 24 verified. The DLS methodology indicates that there were aggregates, which 25 could influence the size of the nanoparticles. 26 For the polydispersity index results, although significant differences were 27 not observed for concentrations and times evaluated (0 h: P = 0.975; 24 h: P = 28 0.132; 48 h: P = 0.071) (Figure 1B), the values found were high (> 0.5), 29 indicating a highly polydisperse system. In this case, the high polydispersity of 30 the system also indicates the formation of aggregates in the solution. 31 Finally, the zeta potential results indicated some significant changes in 32 the values as a function of time (0 h: P < 0.001; 24 h: P = 0.003; 48 h: P < 33 0.001) (Figure 1C); however, the values were lower than 30 mV, characterizing 34 the solutions as unstable dispersions. The zeta potential infers on the surface 11 1 charge of the particles, an important parameter of stability. Thus, these results 2 demonstrate system instability in solution, leading to aggregation and also 3 increased polydispersity. 4 5 3.2. Quantification of iron 6 According to Table 1, the results showed an increase in total Fe 7 concentrations as the NPs concentrations increased both in isolation and in co- 8 exposure with glyphosate and Roundup® when compared to the respective 9 CTR. Although these values were lower than the nominal values of the 10 concentrations of 1, 10, and 100 mg L-1 of goethite, the intended gradient was 11 produced. Similarly, the concentrations of dissolved Fe found in the different 12 solutions containing NPs were lower when compared with the total Fe 13 concentrations analyzed, and the highest concentrations of dissolved Fe were 14 for solutions containing the highest concentration of NPs (N3, GlyN3, and 15 RdN3) when compared to the respective CTR. In addition, the concentration of 16 dissolved Fe found for N3 (8.36) was approximately 12 to 16 times higher, and 17 for GlyN3 (7.10) and RdN3 (4.97), approximately 7 to 10 times higher when 18 compared to the dissolved Fe values of the different treatments. 19 20 3.3. Cytotoxicity 21 The integrity of the plasma membrane (i.e. viability), assessed by TB, 22 was greater than 90% for all times and experimental treatments tested. No 23 statistical differences were found for NPs (1 h: P = 0.220; 6 h: P = 0.622; 12 h: 24 P = 0.153), Gly and co-exposures (1 h: P = 0.623; 12 h: P = 0.046), and Rd and 25 co-exposures (1 h: P = 0.184; 6 h: P = 0.039; 12 h: P = 0.056) in this parameter. 26 Despite a significant decrease at 6 h (P = 0.048) for treatment GlyN3, the 27 viability was still greater than 90% (Fig. 2A, B, and C). 28 When the MTT assay was employed, the results indicated that the N1 29 and N2 treatments were cytotoxic at 6 h (P = 0.002), resulting in a significant 30 decrease in mitochondrial activity of the cells in relation to the CTR, with no 31 change in this parameter at 1 and 12 h (P = 0.655 and P = 0.466, respectively) 32 (Fig. 2D). For glyphosate alone (Gly) or in combination (GlyN1, GlyN2, and 33 GlyN3), no significant alterations in viability were found (1 h: P = 0.096; 6 h: P = 34 0.549; 12 h: P = 0.081). The concentrations of NPs which demonstrated 12 1 cytotoxic effects when isolated (N1 and N2) showed that in association with 2 glyphosate the effect was suppressed (Fig. 2E). For Roundup®, all treatments 3 (Rd, RdN1, RdN2, and RdN3) at 6 h (P < 0.001) produced cytotoxic effects for 4 ZFL cells, with a significant reduction in the viability of these organelles. In 5 addition, the association of NPs with the herbicide did not reverse the 6 cytotoxicity caused by the Roundup®. In contrast, the concentrations of NPs 7 used seem to have negatively influenced this cytotoxicity, since N1 and N2 8 treatments also resulted in a decrease in viability, with a pronounced decrease 9 in the co-exposure RdN2. No alteration in mitochondrial activity was observed 10 for times of 1 and 12 h (P = 0.279 and P = 0.652, respectively) (Fig. 2F). 11 The results of lysosomal integrity through the NR retention test indicated 12 that N1, N2, and N3 promoted a significant increase in viability of the cells at 1 h 13 (P = 0.003) and this pattern of increase in viability was maintained in the N3 14 treatment at 6 h (P < 0.001). In contrast, N1 and N2 were cytotoxic at 6 h, and 15 this cytotoxicity was more pronounced for N2. No alteration in this parameter 16 was observed at 12 h (P = 0.258) (Fig. 2G). When glyphosate-containing 17 treatments were evaluated, the same pattern for the MTT assay was repeated: 18 glyphosate in isolation or in combination did not induce alterations in the viability 19 of the ZFL line for any period tested (1 h: P = 0.504; 6 h: P = 0.054; 12 h: P = 20 0.129), and although the isolated NPs were cytotoxic, when associated with 21 glyphosate this cytotoxic effect disappeared (Fig. 2H). In the treatments 22 containing Roundup®, it was possible to verify that only the co-exposure RdN2 23 resulted in a significant reduction in the viability of these organelles at 6 h (P < 24 0.001). In this case, it is possible to suggest that N2, which was cytotoxic to ZFL 25 cells at the same experimental time, appears to negatively influence the co- 26 exposure cytotoxicity. For 1 and 12 h (P = 0.629 and P = 0.103, respectively), 27 no alterations in this parameter were found (Fig. 2I). 28 29 3.4. ROS and ACAP 30 Regarding the generation of reactive oxygen species (ROS), no 31 significant difference was found for NPs (1 h: P = 0.134; 6 h: P = 0.174; 12 h: P 32 = 0.637) or Gly and co-exposures (1 h: P = 0.045; 6 h: P = 0.806; 12 h: P = 33 0.023) in all experimental times evaluated (Figs 3A and B). The results 34 indicated that only the co-exposure RdN3 resulted in a significant increase in 13 1 ROS at 6 h (P = 0.034) and no significant alterations were observed in this 2 parameter at 1 h and 12 h (P = 0.479 and P = 0.344, respectively) (Fig. 3C). 3 Concerning total antioxidant capacity against peroxyl radicals (ACAP), no 4 significant difference between the treatments was found for NPs in the three 5 different times of exposure (1 h: P = 0.028; 6 h: P = 0.241; 12 h: P = 0.334) 6 (Fig. 4A). When glyphosate-containing treatments were evaluated, only 7 treatment GlyN1 showed a significant increase in ACAP at 1 h (P = 0.028) and 8 no significant alteration was observed for 6 and 12 h (P = 0.991 and P = 0.094, 9 respectively) (Fig. 4B). Similarly, in the treatments containing Roundup®, only 10 RdN1 and RdN3 produced a significant increase in this parameter at 1 h (P = 11 0.019) and no alterations were found for 6 and 12 h (P = 0.357 and P = 0.554, 12 respectively) (Fig. 4C). 13 14 3.5. Genotoxicity 15 The DNA damage in ZFL cells exposed to the concentration of 0.5 mM 16 MMS (PC) were significantly higher than their respective CTR at the three times 17 tested (1 h: P < 0.001; 6 h: P = 0.029; 12 h: P = 0.029) (Fig. 5A), ensuring the 18 efficiency of the procedure and validation of the methodology used. 19 The results for goethite NPs showed a significant increase in the DNA 20 damage score, compared to the respective CTR, for the ZFL cells exposed to 21 N1 and N2 after 1 h (P = 0.007), and the N3 cells after 6 h (P = 0.003). After 12 22 h (P = 0.157), no alteration in DNA score was observed (Fig. 5B). When the 23 ZFL line was exposed to glyphosate in isolation (Gly) or in combination (GlyN1, 24 GlyN2, and GlyN3), no significant alteration in the DNA damage score was 25 identified (1 h: P = 0.056; 6 h: P = 0.463; 12 h: P = 0.327) (Fig. 5C). All the 26 treatments containing Roundup® (Rd, RdN1, RdN2, and RdN3) resulted in a 27 significant increase in the DNA damage score in relation to the CTR after 1 h (P 28 = 0.008), with a more pronounced genotoxic effect for the co-exposures. After 6 29 h (P < 0.001), only Rd and RdN3 promoted a significant increase in DNA 30 damage of the ZFL. On the other hand, after 12 h (P = 0.001) all co-exposures 31 of Rd with NPs resulted in a significant increase in damage score, with RdN1 32 and RdN3 causing the greatest damage (Fig. 5D). Different comet classes 33 observed in ZFL cells were shown in Fig. 5E. 34 14 1 4. Discussion 2 3 Considering the potential application of goethite NPs in environmental 4 remediation, the present study evaluated the effects of this iron oxide and its 5 interference in the effects of glyphosate and Roundup® on the hepatocyte cell 6 line of D. rerio (ZFL), through different cellular assays for 1, 6, and 12 h. The 7 results showed that the effects of isolated goethite NPs on ZFL cells were 8 different from the effects produced in association with the selected herbicides: 9 co-exposure with glyphosate suppressed the effects of goethite NPs; and co- 10 exposure with the formulated product, intensified the effects of iron oxide NPs. 11 Regarding cell viability, the concentrations of 1 and 10 mg L-1 of goethite 12 NPs were cytotoxic to the ZFL line after 6 h of exposure, interfering in 13 mitochondrial metabolism and lysosomal integrity, without changes in the 14 plasma membrane. Several in vitro studies have already demonstrated the 15 ability of different iron oxide NPs to cause cytotoxic damage in lymphocyte cells 16 (Sonmez et al., 2016), lung cell lines (Bhattacharya et al., 2012; Karlsson et al., 17 2008), neural cells (Costa et al., 2016), and keratinocytes and dermal 18 microvascular endothelial human cells (Bayat et al., 2015) in different cellular 19 targets (plasma membrane, mitochondria, and lysosomes). 20 Iron oxide NPs are more likely to cross biological membranes (Singh et 21 al., 2010) due to their large surface area and high reactivity. Once internalized, 22 these NPs can undergo the acid dissolution process within lysosomes (Xia et 23 al., 2008), with subsequent liberation of free Fe ions (Fe2+) (Laskar et al., 2012; 24 Singh et al., 2010). As a result, the excess of these ions in cells can lead to iron 25 imbalance and cause direct damage to the mitochondria, such as morphological 26 alterations or decreases in mitochondrial membrane potential (Freyre-Fonseca 27 et al., 2011; Levi and Rovida, 2009; Teodoro et al., 2011). In the latter case, the 28 free iron can react with hydrogen peroxide and oxygen produced by the 29 mitochondria to produce highly reactive hydroxyl radicals and ferric ions (Fe3+) 30 via the Fenton reaction (Singh et al., 2010). Another important factor to be 31 considered is that depending on the surface charge of nanomaterials, it may 32 interact with the inner surface of the lysosomal membrane, which could result in 33 damage to this organelle (Asati et al., 2010; Cho et al., 2012). 15 1 From assessment of the concentrations of dissolved iron in the exposure 2 solutions, it can be inferred that the positive cytotoxicity results are not related 3 to the excess of free Fe ions. Although treatment N3 presented the highest 4 values of dissolved iron, no cytotoxic damage was observed in ZFL cells 5 exposed to 100 mg L-1 of goethite NPs. In contrast, the lowest dissolved iron 6 values found in goethite NPs exposure solutions were at concentrations that 7 resulted in damage to mitochondrial and lysosomal viability (N1 and N2). 8 To evaluate the possible DNA damage to ZFL cells, the comet assay 9 alkaline version was used. This technique is widely used due to its high 10 sensitivity to detect DNA molecule damage, such as single- and double-strand 11 breaks, alkali-labile sites, incomplete excision repair sites, and cross-links 12 (Singh et al., 1988; Tice et al., 2000). The results of this assay demonstrated 13 the genotoxic potential of goethite NPs at concentrations of 1 and 10 mg L-1 14 after 1 h and the concentration of 100 mg L-1 after 6 h. DNA damage caused by 15 iron oxide NPs has been reported in different cell lines and the concentrations 16 of these NPs are quite varied (Auffann et al., 2006; Bhattacharya et al., 2012; 17 Bayat et al., 2015; Karlsson et al., 2008; Sonmez et al., 2016). Moreover, the 18 majority of studies report that DNA damage caused by exposure to 19 nanoparticles is indirect, resulting from ROS formation by Fenton reaction (Rim 20 et al., 2013; Singh et al., 2012). However, in the present study, the 21 establishment of oxidative stress was not observed in any period of exposure 22 and, in this particular case, it is suggested that dissolved iron values found for 23 treatment N3 are related to the reported genotoxicity. In addition, it is possible 24 that the DNA damage of the ZFL cells is due to the direct action of goethite 25 NPs. According to Singh et al. (2010), super magnetic iron oxide nanoparticles 26 can cause direct damage in the DNA by mechanisms still unknown. 27 Although the toxic potential glyphosate has been demonstrated for 28 mussels Perna perna, and fishes D. rerio and Jenynsia multidentata in vitro 29 (Sandrini et al., 2013), in the present work the herbicide was not cytotoxic and it 30 did not promote oxidative stress or cause DNA damage to ZFL cells. In 31 agreement, Lopes et al. (2018) reported the absence of cytotoxic effects on the 32 same cell line exposed to glyphosate at concentrations of 0.65 and 3.25 mg L-1 33 for 24 and 48 h. 16 1 In turn, when associated with glyphosate, the observed toxicity effects of 2 goethite NPs in ZFL cells disappeared. Glyphosate forms inner-sphere surface 3 complexes on goethite by a ligand exchange mechanism, resulting in the 4 formation of inner-sphere surface complexes, where the phosphonate group of 5 Gly binds Fe+3 ions at the surface (Liu et al., 2014; Sheals et al., 2002). 6 Considering the great adsorption capacity of goethite to glyphosate, it is 7 possible that the formation of this complex resulted in a decrease in available 8 Fe+3 ions and, consequently, the suppression of toxic effects on ZFL cells 9 exposed only to goethite NPs. The type of contaminant and its interaction with 10 iron-based NPs has been previously reported by Fajardo et al. (2015) who 11 showed that different effects could be obtained depending on the chemical 12 properties of the pollutant. 13 On the other hand, several authors have reported that glyphosate-based 14 formulations may cause alterations in mitochondrial function, such as: inhibition 15 of succinate dehydrogenase enzyme activity, transmembrane reduction 16 capacity, and inhibition of ATP synthesis (Koller et al., 2012; Peixoto, 2005; 17 Pereira et al., 2018; Ugarte, 2014; Vanlaeys et al., 2018). In the present study, 18 the results demonstrated that Roundup® was cytotoxic, with a decrease in 19 mitochondrial viability at 6 h and no effect on lysosomal and plasma membrane 20 viability parameters. In agreement with the results found, Lopes et al. (2018), 21 using the ZFL lineage, observed a significant reduction in mitochondrial 22 metabolic activity after exposure to Roundup® (3.25 mg L-1) at 24 and 48 h, 23 whereas lysosomal integrity was reduced only after 24 h exposure, with no 24 alteration in membrane viability at the two experimental times tested. 25 Among the effects described for fish, the genotoxicity of Roundup® has 26 been pointed out as one of the most harmful (Cavalcante et al., 2008; Ghisi and 27 Cestari, 2013; Guilherme et al., 2014; Marques et al., 2014; Moreno et al., 28 2014). Cavalcante et al. (2008), using the fish P. lineatus, exposed to 10 mg L-1 29 of Roundup®, showed a significant increase in erythrocytes DNA damage after 30 6 and 96 h of exposure and 6 and 24 h for gills, corroborating its genotoxic 31 effect. In agreement with previous studies, the present work confirmed the 32 genotoxic potential of Roundup® with a significant increase in DNA damage 33 after 1 and 6 h of exposure. After 12 h, the DNA damage of the ZFL cells 34 returned to CTR levels, which could indicate possible activation of the DNA 17 1 repair system, in order to restore the breaks caused by exposure to the 2 formulated product. The DNA repair system is an important factor in the 3 prevention of serious genetic damage, such as mutations, DNA breaks, and 4 chromosomal aberrations (Marques et al., 2014), and although the comet assay 5 does not make it possible to reliably infer the repair process, this test is used to 6 detect genomic lesions that can be corrected. 7 According to the results found for isolated goethite NPs or isolated 8 Roundup®, we found that both caused cytotoxic and genotoxic damage to the 9 ZFL cells. When associated, the effects produced were more pronounced, even 10 for the highest concentration of goethite NPs (100 mg L-1), demonstrating 11 cytotoxic characteristics for mitochondrial and lysosomal metabolism, ROS 12 promotion, and ACAP activation, as well as genotoxic potential, even for the 13 longest exposure time. 14 In relation to this intensified toxicity demonstrated by the co-exposure, it 15 is possible to infer that both substances could act together, producing a 16 synergistic effect and intensifying the toxicity for the ZFL cells. The entry of toxic 17 molecules, such as Roundup®, into cells is facilitated due to the adsorbent 18 capacity of nanomaterial. Thus, NPs can serve as carriers, even for 19 contaminants, increasing the intracellular concentration of these compounds 20 and, consequently, their potential toxicity (Azevedo Costa et al., 2012; Choi et 21 al., 2007; Lei et al., 2018; Limbach et al., 2007). Moreover, the presence of the 22 other substances in the formulated product, such as surfactants, could favor the 23 entrance of NPs to the ZFL cells, causing the damage observed in this study. 24 Pisanic et al. (2007) reported that coordination between the surfactant and 25 nanostructures facilitates entry of both the nanostructures and surfactants into 26 the cells or interaction with the cells. 27 Finally, considering the absence of toxic effects produced by glyphosate 28 and the cytotoxic and genotoxic damage produced by Roundup® to ZFL cells in 29 the present work, it is possible to suggest that the presence of other compounds 30 in its composition may play an important role in the toxicity of this herbicide. The 31 different compounds present in the Roundup® formulation have already been 32 compared in several aquatic organisms and the order of toxicity of the chemical 33 agents found was: POEA ˃ Roundup® ˃ glyphosate ˃ glyphosate 34 isopropylamine salt (IPA), indicating that toxicity of the formulated product can 18 1 be attributed to the surfactant POEA (Tsui and Chu, 2003). In another study, it 2 was pointed out that the genotoxic potential of Roundup® for the fish Anguilla 3 anguilla is directly related to the genotoxicity of surfactant POEA (Guilherme et 4 al., 2012). Likewise, Navarro and Martinez (2014) evidenced a significant 5 increase in DNA damage of erythrocytes in the fish P. lineatus after exposure to 6 POEA, supporting the genotoxic character of this surfactant. Although the 7 toxicity of the surfactant POEA has been demonstrated in several studies, our 8 data cannot support that the increase in Roundup® toxicity for ZFL cells is 9 exclusively associated with POEA. 10 In summary, these data lead us to conclude that the concentrations of 11 goethite NPs used were not safe for the ZFL lineage. In addition, it was shown 12 that goethite NPs and Roundup®, both isolated, presented cytotoxic and 13 genotoxic effects and, when co-exposed, produced a synergistic effect, 14 intensifying the previously reported damage. On the other hand, glyphosate did 15 not promote cytotoxic, biochemical, or genotoxic damage to ZFL cells and, in 16 association, the toxic effects produced by isolated goethite NPs were 17 suppressed. In comparison, given the lack of toxic effects of glyphosate, it is 18 possible to suggest that the presence of other compounds in the formulated 19 product favors the toxicity of this herbicide when compared to the active 20 ingredient glyphosate. 21 Considering the lack of studies investigating the possible toxic effects of 22 goethite NPs, especially for aquatic organisms, our study is extremely 23 significant, as we evaluated the mechanism of action of nanomaterial at the 24 cellular level, contributing to knowledge of the toxic effects generated by iron 25 oxide NPs. In addition, the results may favor discussion and investigation of 26 strategies to determine environmentally safe NPs concentrations that result in 27 an effective tool for removing contaminants from polluted aquatic ecosystems. 28 29 Conflicts of interest 30 There are no conflicts of interest. 31 32 Acknowledgements 33 The authors would like to thank the Araucaria Foundation and CNPq (PRONEX: 34 24732/2012), the National Institute of Science and Technology in Aquatic 19 1 Toxicology (INCT-TA, CNPq: 573949 / 2008-5), and the São Paulo Research 2 Foundation (#2017/21004-5) for financial support. To CAPES for granting a 3 master's degree grant to N.D.G. da Silva and CNPq for the research grant to 4 C.B.R. Martinez. This work is part of the Master dissertation of N.D.G. da Silva 5 for the Post-graduate Program in Biological Sciences of the State University of 6 Londrina. 7 8 References 9 Amado, L.L., Garcia, M.L., Ramos, P.B., Freitas, R.F., Zafalorf, B., Ferreira, 10 L.R., Yunes, J.S., Monserrat, J.M., 2009. A method to measure total 11 antioxidant capacity against peroxyl radicals in aquatic organisms: 12 Application to evaluate microcystins toxicity. Sci. Total Environ. 407, 2115- 13 2123. 14 Amarante, O.P.J., Santos, T.C.R., Brito, N.M., Ribeiro, M.L., 2002. Glifosato: 15 Propriedades, toxicidade, usos e legislação. Química Nova 25, 589-593. 16 Ankamwar, B., Lai, T.C., Huang, J.H., Liu, R.S., Hsiao, M., Chen, C.H., Hwu, 17 Y.K., 2010. Biocompatibility of Fe3O4 nanoparticles evaluated by in vitro 18 cytotoxicity assays using normal, glia and breast cancer cells. 19 Nanotechnology 21, 75102. 20 Asati, A., Santra, S., Kaittanis, C., Perez, J.M., 2010. Surface-charge- 21 dependent cell localization and cytotoxicity of cerium oxide nanoparticles. 22 ACS Nano 4, 5321-5331. 23 Auffan, M., Decome, L., Rose, J., Orsiere, T., Meo, M., Briois, V., Chaneac, C., 24 Olivi, L., Berge-Lefranc, J., Botta, A., Wiesner, M., Bottero, J., 2006. In vitro 25 interactions between DMSA-coated maghemite nanoparticles and human 26 fibroblasts: a physicochemical and cyto-genotoxical study. Environ. Sci. 27 Technol. 20, 4367-4373. 28 Azevedo Costa, C.L., Chaves, I.S., Ventura-Lima, J., Ferreira, J.L.R., Ferraz, L., 29 Carvalho, L.M., Monserrat, J.M., 2012. In vitro evaluation of co-exposure of 30 arsenium and an organic nanomaterial (fullerene, C60) in Zebrafish 31 hepatocytes. Comp. Biochem. Physiol. Parte C 155, 206-212. 32 Barja, B.C. and dos Santos Afonso, M., 2005. Aminomethylphosphonic acid and 33 glyphosate onto goethite: a comparative study. Environ. Sci. Technol. 39, 34 585-592. 20 1 Bayat, N., Lopes, V.R., Sanchez-Dominguez, M., Lakshmanan, R., Rajarao, 2 G.K., Cristobal, S., 2015. Assessment of functionalized iron oxide 3 nanoparticles in vitro: introduction to integrated nanoimpact index. Environ. 4 Sci. Nano 2, 380-394. 5 Benbrook, C.M., 2016. Trends in glyphosate herbicide use in the United States 6 and globally. Environ. Sci. Eur. 28(3), 1-15. 7 Bhattacharya, K., Hoffmann, E., Schins, R.F.P., Boertz, J., Prantl, E., Alink, 8 G.M., Byrne, H.J., Kuhlbusch, T.A.J., Rahman, Q., Wiggers, H., Schulz, C., 9 Dopp, E., 2012. Comparison of micro- and nanoscale Fe+3-containing 10 (hematite) particles for their toxicological properties in human lung cells in 11 vitro. Toxicol. Sci. 126, 173-182. 12 Bols, N.C., Dayeh, V.R., Lee, L.E.J., Schirmer, K., 2005. Use of fish cell lines in 13 the toxicology and ecotoxicology of fish. Piscine cell lines in environmental 14 toxicology. J. Biochem. Mol. Biol. 6, 43-84. 15 Borggaard, L.K. and Gimsing, A.L., 2008. Fate of glyphosate in soil and the 16 possibility of leaching to ground and surface waters: a review. Pest Manag. 17 Sci. 64, 441-456. 18 Bridi, D., Altenhofen, S., Gonzalez, J.B., Reolon, G.K., Bonan, C.D., 2017. 19 Glyphosate and Roundup® alter morphology and behavior in zebrafish. 20 Toxicology 392, 32-39. 21 Bruggen, A.H.C., He, M.M., Shin, K., Mai, V., Jeong, K.C. Finckh, M.R., Morris, 22 J.G., 2018. Environmental and health effects of the herbicide glyphosate. 23 Sci. Total Environ. 616-617, 255-268. 24 Caballero-Gallardo, K., Olivero-Verbel, J. and Freeman, J.L., 2016. 25 Toxicogenomics to evaluate endocrine disrupting effects of environmental 26 chemicals using the zebrafish model. Curr. Genomics 17(6), 515-527. 27 Cavalcante, D.G.S.M., da Silva, N.D.G., Marcarini, J.C., Mantovani, M.S., 28 Marin-Morales, M.A., Martinez, C.B.R., 2014. Cytotoxic, biochemical and 29 genotoxic effects of biodiesel produced by different routes on ZFL cell line. 30 Toxicol. In Vitro 28, 1117-1125. 31 Cavalcante, D.G.S.M., Martinez, C.B.R., Sofia, S.H., 2008. Genotoxic effects of 32 Roundup® on the fish Prochilodus lineatus. Mutat. Res. 655, 41-46. 33 Chen, Y.Y and Chan, K.M., 2018. Transcriptional inhibition of TCDD-mediated 34 induction of cytochrome P450 1A1 and alteration of protein expression in a 21 1 zebrafish hepatic cell line following the administration of TCDD and Cd2+. 2 Toxicol. Lett. 282, 121-135. 3 Cho, W.S., Duffin, R., Thielbeer, F., Bradley, M., Megson, I.L., MacNee, W., 4 Poland, C.A., Tran, C.L., Donaldson, K., 2012. Zeta potential and solubility to 5 toxic ions as mechanisms of lung inflammation caused by metal/metal-oxide 6 nanoparticles. Toxicol. Sci. 12, 1-31. 7 Choi, M.R., Stanton-Maxey, K.J., Stanley, J.K., Levin, C.S., Bardhan, R., Akin, 8 D., Badve, S., Sturgis, J., Robinson, J.P., Bashir, R., Halas, N.J., Clare, S., 9 2007. A cellular Trojan horse for delivery of therapeutic nanoparticles into 10 tumors. Nano Letters 7, 3759-3765. 11 Collins, A.R., Oscoz, A.A., Brunborg, G., Gaivão, I., Giovannelli, L., Kruszewski, 12 M., Smith, C.C., Štětina, R., 2008. The comet assay: topical issues. 13 Mutagenesis 23, 143-151. 14 Cornell, R.M. and Schwertmann, U., 2003. The iron oxides: structure, 15 properties, reactions, occurrences and uses. Secondedition: Wiley. 16 Costa, C., Brandão, F., Bessa, M.J., Costa, S., Valdiglesias, V., Kiliç, G., 17 Fernández-Bertólez, N., Quaresma, P., Pereira, E., Pásaro, E., Laffon, B., 18 Teixeira, J.P., 2016. In vitro cytotoxicity of superparamagnetic iron oxide 19 nanoparticles on neuronal and glial cells. Evaluation of nanoparticle 20 interference with viability tests. J. Appl. Toxicol. 36, 361-372. 21 De Gerónimo, E., Aparicio, V.C., Costa, J.L., 2018. Glyphosate sorption to soils 22 of Argentina. Estimation of affinity coefficient by pedotransfer function. 23 Geoderma 322, 140-148. 24 Dultz, S., Steinke, H., Mikutta, R., Woche, S.K., Guggenberger, G., 2018. 25 Impact of organic matter types on surface charge and aggregation of 26 goethite. Colloids Surf. A 554, 156-168. 27 Fajardo, C., Gil-Díaz, M., Costa, G., Alonso, J., Guerrero, A.M., Nande, M., 28 Lobo, M.C., Martín, M., 2015. Residual impact of aged nZVI on heavy metal- 29 polluted soils. Sci. Total Environ. 535, 79-84. 30 Fent, K., 2007. Permanent fish cell cultures as important tools in ecotoxicology. 31 ALTEX 24, 26-28. 32 Freyre-Fonseca, V., Delgado-Buenrostro, N.L., Gutiérrez-Cirlos, E.B., Calderón- 33 Torres, C.M., Cabellos-Avelar, T., Sánchez-Pérez, Y., Pinzón, E., Torres, I., 34 Molina-Jijón, E., Zazueta, C., Pedraza-Chaverri, J., García-Cuélla, C.M., 22 1 Chirino, Y.I., 2011. Titanium dioxide nanoparticles impair lung mitochondrial 2 function. Toxicol. Lett. 202, 111-119 3 Gajski, G., Gerić M., Žegura, B., Novak, M., Nunić, J., Bajrektarević, D., Garaj- 4 Vrhovac, V., Filipič, M., 2016. Genotoxic potential of selected cytostatic drugs 5 in human and zebrafish cells. Environ. Sci. Pollut. Res. 23, 14739-14750. 6 Ghisi, N.C. and Cestari, M.M., 2013. Genotoxic effects of the herbicide 7 Roundup® in the fish Corydoraspaleatus (Jenyns 1842) after short-term, 8 environmentally low concentration exposure. Environ. Monit. Assess. 185, 9 3201-3207. 10 Ghosh, C., Zhou, Y.L., Collodi, P., 1994. Derivation and characterization of a 11 zebrafish liver cell line. Cell Biol. Toxicol. 10, 167-176. 12 Goulart, T.L.S., Boyle, R.T., Souza, M.M., 2015. Cytotoxicity of the association 13 of pesticides Roundup Transorb® and Furadan 350SC® on the zebrafish cell 14 line, ZF-L. Toxicol. In Vitro 29, 1377-1384. 15 Guilherme, S., Santos, M.A., Gaivão, I., Barroso, C., Pacheco, M., 2012. 16 Differential genotoxicity of Roundup formulation and its constituents in blood 17 cells of fish (Anguilla anguilla): considerations on chemical interactions and 18 DNA damaging mechanisms. Ecotoxicology21, 1381-1390. 19 Guilherme, S., Santos, M.A., Gaivão, I., Pacheco, M., 2014. Are DNA-damaging 20 effects induced by herbicide formulations (Roundup® and Garlon®) in fish 21 transient and reversible upon cessation of exposure? Aquat. Toxicol. 155, 22 213-221. 23 Hanke, I., Wittmer, I., Bischofberger, S., Stamm, C., Singer, H., 2010. 24 Relevance of urban glyphosate use for surface water quality. Chemosphere 25 81, 422-429. 26 IBAMA (Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais 27 Renováveis), 2017. Relatórios de Comercialização de Agrotóxicos – Boletim 28 de Produção, Importação, Exportação e Vendas de Agrotóxicos no Brasil. 29 Available at: . 31 Jennings, P., 2015. “The future of in vitro toxicology”. Toxicol. In Vitro 29, 1217- 32 1221. 33 Jonsson, C.M., Persson, P., Sjöberg, S., Loring, J.S., 2008. Adsorption of 34 glyphosate on goethite (α-FeOOH): Surface complexation modeling 23 1 combining spectroscopic and adsorption data. Environ. Sci. Technol. 42, 2 2464-2469. 3 Karlsson, H.L., Cronholm, P.,Gustafsson, J., Möller, L., 2008. Copper oxide 4 nanoparticles are highly toxic: a comparison between metal oxide 5 nanoparticles and carbon nanotubes. Chem. Res. Toxicol. 21, 1726-1732. 6 Kharisov, B.I., Dias, H.V.R., Kharissova, O.V., Jiménez-Pérez, V.M., Pérez, 7 B.O., Flores, B.M., 2012. Iron-containing nanomaterials: synthesis, properties 8 and environmental applications.RSC Adv.2(25), 9325-9358. 9 Koller, V.J., Fürhacker, M., Nersesyan, A., Mišík, M., Eisenbauer, M., 10 Knasmueller, S. 2012. Cytotoxic and DNA-damaging properties of glyphosate 11 and Roundup in human-derived buccal epithelial cells. Arch. Toxicol. 86, 805- 12 813. 13 Lachner, D., Oliveira L.F., Martinez, C.B.R., 2015. Effects of the water soluble 14 fraction of gasoline on ZFL cell line: cytotoxicity, genotoxicity and oxidative 15 stress. Toxicol. In Vitro 30, 225-230. 16 Laskar, A., Ghosh, M., Khattak, S.I., Li, W., Yuan, X.M., 2012. Degradation of 17 superparamagnetic iron oxide nanoparticle-induced ferritin by lysosomal 18 cathepsins and related immune response. Nanomedicine 7, 705-717. 19 Lei, C., Sun, Y., Tsang, D.C.W., Lin, D., 2018. Environmental transformations 20 and ecological effects of iron-based nanoparticles. Environ. Pollut. 232, 10- 21 30. 22 Levi, S. and Rovida, E., 2009. The role of iron in mitochondrial function. 23 Biochim. Biophys. Acta 1790, 629-636. 24 Limbach, L.K., Wick, P., Manser, P., Grass, R.N., Bruinink, A., Stark, W.J., 25 2007. Exposure of engineered nanoparticles to human lung epithelial cells: 26 influence of chemical composition and catalytic activity on oxidative stress. 27 Environ. Sci. Technol. 41, 4158-4163. 28 Liu, H., Chen, T., Frost, R.L., 2014. An overview of the role of goethite surfaces 29 in the environment. Chemosphere 103, 1-11. 30 Lopes, F.M., Varela Junior, A.S., Corcini, C.D., da Silva, A.C., Guazzelli, V.G., 31 Tavares, G., da Rosa, C.E., 2014. Effect of glyphosate on the sperm quality 32 of zebrafish Danio rerio. Aquat. Toxicol. 155, 322-326. 33 Lopes, F.M., Sandrini, J.Z., Souza, M.M., 2018. Toxicity induced by glyphosate 34 and glyphosate-based herbicides in the zebrafish hepatocyte cell line (ZF-L). 24 1 Ecotoxicol. Environ. Saf. 162, 201-207. 2 Lushchak, O.V., Kubrak, O.I., Storey, J.M., Storey, K.B., Lushchak, V.I., 2009. 3 Low toxic herbicide Roundup induces mild oxidative stress in goldfish 4 tissues. Chemosphere 76, 932-937. 5 Maqueda, C., Undabeytia, T., Villaverde, J., Morillo, E., 2017. Behaviour of 6 glyphosate in a reservoir and the surrounding agricultural soils. 7 Sci. Total Environ. 593-594, 787-795. 8 Marques, A., Guilherme, S., Gaivão, I., Santos, M.A., Pacheco, M., 2014. 9 Progression of DNA damage induced by a glyphosate-based herbicide in fish 10 (Anguilla Anguilla) upon exposure and post-exposure periods – Insights into 11 the mechanisms of genotoxicity and DNA repair. Comp. Biochem. Physiol. 12 Part C166, 126-133. 13 Modesto, K.A. and Martinez, C.B.R., 2010a. Effects of Roundup Transorb on 14 fish: hematology, antioxidant defenses and acetylcholinesterase activity. 15 Chemosphere 81, 781-787. 16 Modesto, K.A. and Martinez, C.B.R., 2010b. Roundup causes oxidative stress 17 in liver and inhibits acetylcholinesterase in muscle and brainof the fish 18 Prochilodus lineatus. Chemosphere 78, 294-299. 19 Moreno, N.C., Sofia, S.H., Martinez, C.B.R., 2014. Genotoxic effects of the 20 herbicide Roundup Transorb® and its active ingredient glyphosate on the fish 21 Prochilodus lineatus. Environ. Toxicol. Pharm. 37, 448-454. 22 Morozesk, M., Franqui, L.S., Mansano, A.S., Martinez, D.S.T., Fernandes, 23 M.N., 2018. Interactions of oxidized multiwalledcarbono nanotube with 24 cadmium on zebrafish cell line: The influence of two co-exposure protocols 25 on in vitro toxicity tests. Aquat. Toxicol. 200, 136-147. 26 Mosmann, T., 1983. Rapid colorimetric assay for cellular growth and survival. J. 27 Immunol. Methods 65, 55-63. 28 Navarro, C.D.C. and Martinez, C.B.R., 2014. Effects of the surfactant 29 polyoxyethylene amine (POEA) on genotoxic, biochemical and physiological 30 parameters of the freshwater teleost Prochilodus lineatus. Comp. Biochem. 31 Phys. Part C 165, 83-90. 32 Novak, M., Žegura, B., Modic, B., Heath, E., Filipič, M., 2017. Cytotoxicity and 33 genotoxicity of anticancer drug residues and their mixtures in experimental 34 model with zebrafish liver cells. Sci. Total Environ. 601-602, 293-300. 25 1 O’Carroll, D., Sleep, B., Krol, M., Boparai, H., Kocur, C., 2013. Nanoscale zero 2 valent iron and bimetallic particles for contaminated site remediation. Adv. 3 WaterResour. 51, 104-122. 4 Orcelli, T., di Mauro, E., Urbano, A., Valezi, D.F., da Costa A.C.S., Zaia, C.T.B., 5 Zaia, D.A.M., 2018. Study of interaction between glyphosate and goethite 6 using several methodologies: an environmental perspective. Water Air Soil 7 Pollut. 229, 1-18. 8 Park, J.H. and Park, E., 2011. Influence of iron-overload on DNA damage and 9 its repair in human leukocytes in vitro. Mutat. Res-Gen Tox. En. 718, 56-61. 10 Peixoto, F., 2005. Comparative effects of the Roundup and glyphosate on 11 mitochondrial oxidative phosphorylation. Chemosphere 61, 1115-1122. 12 Pereira, A.G., Jaramillo, M.L., Remor, A.P., Latini, A., Davico, C.E., Silva, M.L., 13 Müller, Y.M.R., Ammar, D., Nazari, E.M. 2018. Low-concentration exposure 14 to glyphosate-based herbicide modulates the complexes of the mitochondrial 15 respiratory chain and induces mitochondrial hyperpolarization in the brain 16 Danio rerio. Chemosphere 209, 353-362. 17 Pessagno, P.C., Sánchez, R.M.T., dos Santos Afonso, M., 2008. Glyphosate 18 behavior at soil and mineral – water interfaces. Environ. Pollut. 153, 53-59. 19 Pisanic, T.R., Blackwell, J.D., Shubayev, V.I., Fiñones, R.R., Jin, S., 2007. 20 Nanotoxicity of iron oxide nanoparticle internalization in growing neurons. 21 Biomaterials 28(16), 2572-2581. 22 Rim, K.T., Song, S.W., Kim, H.Y., 2013. Oxidative DNA damage from 23 nanoparticle exposure and its application to workers’ health: a literature 24 review. Saf. Health Work 4, 177-186. 25 Rodrigues, L.B., Oliveira, R., Abe, F.R., Brito, L.B., Moura, D.S., Valadares, 26 M.C., Grisolia, C.K., Oliveira, D.P., Oliveira, G.A.R., 2017. Ecotoxicological 27 assessment of glyphosate-based herbicides: effects on different organisms. 28 Environ. Toxicol. Chem. 36, 1755-1763. 29 Sandrini, J.Z., Bianchini, A., Trindade, G.S., Nery, L.E.M., Marins, L.F.F., 2009. 30 Reactive oxygen species generation and expression of DNA repair-related 31 genes after copper exposure in zebrafish (Danio rerio) ZFL cells. Aquat. 32 Toxicol. 95, 285-291. 33 Sandrini, J.Z., Rola, R.C., Lopes, F.M., Buffon, H.F., Freitas, M.M., Martins, 34 C.M.G., Rosa, C.E., 2013. Effects of glyphosate on cholinesterase activity of 26 1 the mussel Perna perna and the fish Danio rerio and Jenynsia multidentata: 2 in vitro studies. Aquat. Toxicol. 130-131, 171-173. 3 Schwertmann, U. and Cornell, R.M., 2000. Iron oxides in the laboratory. 4 Preparation and characterization. Second edition: Wiley. 5 Sheals, J., Sjöberg, S., Persson, P., 2002. Adsorption of glyphosate on 6 goethite: molecular characterization of surface complexes. Environ. Sci. 7 Technol. 36, 3090-3095. 8 Singh, N., Jenkins, G.J.S., Asadi, R., Doak, S.H., 2010. Potential toxicity of 9 superparamagnetic iron oxide nanoparticles (SPION). Nano reviews 1, 5358- 10 5373. 11 Singh, N., Jenkins, G.J.S., Nelson, B.C., Marquis, B.J., Maffeis, T.G.G., Brown, 12 A.P., Williams, P.M., Wright, C.J., Doak, S.H., 2012. The role of iron redox 13 state in the genotoxicity of ultrafine superparamagnetic iron oxide 14 nanoparticles. Biomaterials 33, 163-170. 15 Singh, N.P., Mccoy, M.T., Tice, R.R., Schneider, E.L., 1988. A single technique 16 for quantification of low levels of DNA damage in individual cells. Exp. Cell 17 Res. 175, 184-191. 18 Singh, S., Khanna, V.K., Pant, A.B., 2018. Development of in vitro toxicology: a 19 historic story. In: In Vitro Toxicology. Academic Press, Chapter 1, 1-19. 20 Sonmez, E., Aydin, E., Turkez, H., Özbek, E., Togar, B., Meral, K., Çetin, D., 21 Cacciatore, I., Di Stefano, A., 2016. Cytotoxicity and genotoxicity of iron 22 oxide nanoparticles: an in vitro biosafety study. Arch. Biol. Sci. 68, 41-50. 23 Srivastava, S., Mishra, S., Dewangan, J., Divakar, A., Pandey, P.K., Rath, S.K., 24 2018. Principles for in vitro toxicology. In: In Vitro Toxicology. Academic 25 Press, Chapter 2, 21-43. 26 Tang, S., Allagadda, V., Chibli, H., Nadeau, J.L., Mayer, G.D., 2013. 27 Comparison of cytotoxicity and expression of metal regulatory genes in 28 zebrafish (Danio rerio) liver cells exposed to cadmium sulfate, zinc sulfate 29 and quantum dots. Metallomics 5, 1411-1422. 30 Tang, T., Boënne, W., Desmet, N., Seuntjens, P., Bronders, J., van Griensven, 31 A., 2015. Quantification and characterization of glyphosate use and loss in a 32 residential area. Sci. Total Environ. 517, 207-214. 33 Teodoro, J.S., Simões, A.M., Duarte, F.V., Rolo, A.P., Murdoch, R.C., Hussain, 34 S.M., Palmeira, C.M., 2011. Assessment of the toxicity of silver nanoparticles 27 1 in vitro: a mitochondrial perspective. Toxicol. In Vitro 25, 664-670. 2 Thit, A., Skjolding, L.M., Selck, H., Sturve, J., 2017. Effects of copper oxide 3 nanoparticles and copper ions to zebrafish (Danio rerio) cells, embryos and 4 fry. Toxicol. In Vitro 45, 89-100. 5 Tice, R.R., Agurell, E., Anderson, D., Burlinson, B., Hartmann, A., Kobayashi, 6 H., Miyamae, Y., Rojas, E., Ryu, J.C., Sasaki, Y.F., 2000. Single cell 7 gel/comet assay: guidelines for in vitro and in vivo genetic toxicology testing. 8 Environ. Mol. Mutagen. 35, 206-221. 9 Tosco, T., Papini, M.P., Vigghi, C.C., Sethi, R., 2014. Nanoscale zerovalent iron 10 particles for groundwater remediation: a review. J. Clean. Prod. 77, 10-21. 11 Tsui, M.T.K. and Chu, L.M., 2003. Aquatic toxicity of glyphosate-based 12 formulations: comparison between different organisms and the effects of 13 environmental factors. Chemosphere 52, 1189-1197. 14 Ugarte, R. 2014. Interaction between glyphosate and mitochondrial succinate 15 dehydrogenase. Comput. Theor. Chem. 1043, 54-63. 16 Vanlaeys, A., Dubuisson, F., Seralini, G.E., Travert, C., 2018. Formulants of 17 glyphosate-based herbicides have more deleterious impact than glyphosate 18 on TM4 Sertoli cells. Toxicol. In Vitro 52, 14-22. 19 Waiman, C.V., Avena, M.J., Garrido, M., Band, B.F., Zanini, G.P., 2012. A 20 simple and rapid spectrophotometric method to quantify the herbicide 21 glyphosate in aqueous media. Application to adsorption isotherms on soils 22 and goethite. Geoderma 170, 154-158. 23 WHO (World Health Organization), 2005. Water sanitation hygiene. Chemical 24 hazards in drinking-water: Glyphosate and AMPA. Availableat: 25 . 27 Xia, T., Kovochich, M., Liong, M., Mädler, L., Gilbert, B., Shi, H., Yeh, J.I., Zink, 28 J.I., Nel, A.E., 2008. Comparison of the mechanisms of toxicity of zinc oxide 29 and cerium oxide nanoparticles based on dissolution and oxidative stress 30 properties. ACS Nano 2, 2121-2134. 31 Yang, Y., Deng, Q., Yan, W., Jing, C., Zhang, Y., 2018. Comparative study of 32 glyphosate removal on goethite and magnetite: Adsorption and photo- 33 degradation. Chem. Eng. J. 352, 581-589. 34 Žegura, Z. andFilipič, M., 2019. The application of the Comet assay in fish cell 28 1 lines. Mutat. Res. 2 Gen.Tox.En.https://doi.org/10.1016/j.mrgentox.2019.01.011 https://doi.org/10.1016/j.mrgentox.2019.01.011 29 1 Table 1. Mean concentrations of total and dissolved Fe (mg L-1) in the exposure 2 solutions of the different treatments. 3 Fe concentrations (mg L-1) Total Dissolved Total Dissolved Total Dissolved CTR 0.76 0,67 CTR 0.76 0.67 CTR 0.76 0.67 N1 0.73 0.52 Gly 0.77 0.76 Rd 0.56 0.58 N2 5.23 0.69 GlyN1 1.48 0.97 RdN1 1.08 0.54 N3 60.0 8.36 GlyN2 6.23 0.95 RdN2 6.99 0.62 GlyN3 62.2 7.10 RdN3 71.1 4.97 30 1 Figure captions 2 3 Fig. 1. Mean diameter (A), polydispersity (B), and zeta potential (C) of the goethite NPs 4 present in the exposure solutions of 1, 10, and 100 mg L-1 (N1, N2, and N3) for 0, 24, 5 and 48 h. Results are mean ± SE (n = 3). Different letters indicate significant 6 differences between treatments for the same experimental time (P < 0.05). 7 8 Fig. 2. Cell viability (%) based on plasma membrane integrity (A-C), mitochondrial 9 activity (D-F), and lysosomal integrity (G-I) for ZFL cells exposed to: 1, 10, and 100 mg 10 L-1 of goethite nanoparticles (N1, N2, and N3); 3.6 mg L-1 of glyphosate (Gly) and the 11 co-exposures (GlyN1, GlyN2, and GlyN3) and 10 mg L-1 of Roundup® (Rd) and the co- 12 exposures (RdN1, RdN2, and RdN3) or only to the Dulbecco´s PBS (CTR) for 1, 6, and 13 12 h. Results are mean ± SE (n = 4). Different letters indicate significant differences 14 between treatments for the same experimental time (P < 0.05). 15 16 Fig. 3. Production of reactive oxygen species (ROS) in ZFL cells exposed to: A) 1, 10, 17 and 100 mg L-1 of goethite nanoparticles (N1, N2, and N3); B) 3.6 mg L-1 of glyphosate 18 (Gly) and the co-exposures (GlyN1, GlyN2, and GlyN3); and C) 10 mg L-1 of Roundup® 19 (Rd) and the co-exposures (RdN1, RdN2, and RdN3) or only to the Dulbecco´s PBS 20 (CTR) for 1, 6, and 12 h. Results are mean ± SE (n = 4). Different letters indicate 21 significant differences between treatments for the same experimental time (P < 0.05). 22 23 Fig. 4. Antioxidant capacity against peroxyl radicals (ACAP) of ZFL cells exposed to: A) 24 1, 10, and 100 mg L-1 of goethite nanoparticles (N1, N2, and N3); B) 3.6 mg L-1 of 25 glyphosate (Gly) and the co-exposures (GlyN1, GlyN2, and GlyN3); and C) 10 mg L-1 of 26 Roundup® (Rd) and the co-exposures (RdN1, RdN2, and RdN3) or only to the 27 Dulbecco´s PBS (CTR) for 1, 6, and 12 h. Results are mean ± SE (n = 4). Different 28 letters indicate significant differences between treatments for the same experimental 29 time (P < 0.05). 30 31 Fig. 5. A) Score of DNA damage in ZFL cells exposed to 0.5 mM of MMS (Positive 32 Control or PC); B) 1, 10, and 100 mg L-1 of goethite nanoparticles (N1, N2, and N3); C) 33 3.6 mg L-1 of glyphosate (Gly) and the co-exposures (GlyN1, GlyN2, and GlyN3); and 34 D) 10 mg L-1 of Roundup® (Rd) and the co-exposures (RdN1, RdN2, and RdN3) or only 35 to the Dulbecco´s PBS (CTR) for 1, 6, and 12 h, quantified by the comet assay. Results 36 are mean ± SE (n = 4). Different letters indicate significant differences between 37 treatments for the same experimental time (P < 0.05). E) Photomicrographs of ZFL 31 1 cells processed for comet assay showing increasing degrees of DNA damage (original 2 magnification: 1000x). Fig. 1. Fig. 2. Fig. 3. Fig. 4. Fig. 5. Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: