Mitochondrial bioenergetics and locomotor activity are altered in zebrafish (Danio rerio) after exposure to the bipyridylium herbicide diquat

Accepted Manuscript Title: Mitochondrial bioenergetics and locomotor activity are altered in zebrafish (Danio rerio) after exposure to the bipyridylium herbicide diquat Authors: Xiao H. Wang, Christopher L. Souders II, Yuan H. Zhao, Christopher J. Martyniuk PII: DOI: Reference:

S0378-4274(17)31445-5 https://doi.org/10.1016/j.toxlet.2017.10.022 TOXLET 9988

To appear in:

Toxicology Letters

Received date: Revised date: Accepted date:

25-7-2017 6-10-2017 29-10-2017

Please cite this article as: Wang, Xiao H., Souders, Christopher L., Zhao, Yuan H., Martyniuk, Christopher J., Mitochondrial bioenergetics and locomotor activity are altered in zebrafish (Danio rerio) after exposure to the bipyridylium herbicide diquat.Toxicology Letters https://doi.org/10.1016/j.toxlet.2017.10.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

Mitochondrial bioenergetics and locomotor activity are altered in zebrafish (Danio rerio) after exposure to the bipyridylium herbicide diquat Xiao H. Wang 1, 2, Christopher L. Souders II 2, Yuan H. Zhao 1*, Christopher J. Martyniuk 2*

1 State Environmental Protection Key Laboratory of Wetland Ecology and Vegetation Restoration, School of Environment, Northeast Normal University, Changchun, Jilin, 130117, P. R. China 2 Center for Environmental and Human Toxicology, Department of Physiological Sciences, College of Veterinary Medicine, UF Genetics Institute, University of Florida, Gainesville, Florida, 32611 USA *Corresponding author Email address: [email protected] and [email protected] Tel: 352-294-4636 Fax: 352-392-4707

Highlights 

Diquat is a non-selective bipyridylium herbicide in the same class as paraquat.



Seven day old zebrafish treated with diquat 10-100 nM showed increased behavior activity.



Basal respiration and ATP production were decreased following a 24 hours diquat exposure at 100 µM.



Larvae exposed to 10 µM diquat showed higher transcript levels of catalase compared to control fish.



This study improves mechanistic understanding of diquat exposure in fish at early stages of development.

Abstract: Diquat is a non-selective bipyridylium herbicide which has replaced its sister compound paraquat, as paraquat is associated to an increased risk for Parkinson’s disease. However, the propensity of diquat to propagate reactive oxygen species ensures that diquat remains an exposure risk in non-target organisms. In this study, zebrafish (Danio rerio) embryos were exposed to diquat (1, 10, 100 µM) beginning at ~6 hours post fertilization for up to 7 days to learn more about the mechanisms underlying diquat toxicity during vertebrate development. Zebrafish embryos exposed to diquat for 96 hours did not show any significant mortality nor deformity compared to controls. Moreover, there was no difference in the timing of hatch, an indicator of stress, for fish exposed to diquat. To determine whether changes in mitochondrial bioenergetics occurred in early development as a response to diquat exposure, oxygen consumption rate was measured in whole embryos. Basal respiration and ATP production were decreased following a 24 hours diquat exposure at 100 µM, suggesting that diquat negatively affects oxidative phosphorylation. We also assessed locomotor behavior as a sensitive endpoint for impaired activity and neurotoxicity. Seven day old (7 dpf) zebrafish treated with diquat at the highest doses tested (10-100 nM) showed an increase (hyper-activity) in total distance travelled, velocity, movement cumulative duration, and overall activity compared to unexposed fish. Lastly, in 7d fish, we measured transcripts related to redox balance and apoptosis as diquat has been reported to induce oxidative stress and can affect mitochondrial bioenergetics. Larvae exposed to 10 µM diquat showed higher transcript levels of catalase compared to control fish, implying that reactive oxygen species are produced

following diquat exposure. Transcript levels of sod1, sod2, bcl2, bax and caspase 3 however did not vary in abundance among treatments with diquat. This study improves mechanistic understanding of diquat in fish at early stages of development and presents evidence that diquat disrupts mitochondrial bioenergetics and behavior.

Key words: herbicide, neurotoxicity, oxidative stress, mitochondrial dysfunction, ATP production, oxygen consumption rate 1. Introduction Diquat is a non-selective bipyridylium herbicide that is related to the neurotoxin paraquat. Paraquat has been widely used in agricultural since the 1960s (US EPA. 2009) and can enter aquatic environments through run-off. Diquat, structurally similar to paraquat, has replaced paraquat in many agricultural applications because paraquat has been restricted in use or prohibited due to an increased risk for neurodegenerative diseases such as Parkinson’s disease (McCormack et al., 2002). Diquat is used in agriculture as a plant growth regulator for the control of broadleaf and grassy weeds in terrestrial non-crop and aquatic areas, but it has been included on a priority list of herbicides of potential concern established by the European Union (EU) due to its widespread application (Ducrot et al., 2010). Diquat is highly soluble in water and is detected as residues in the environment (Siemering et al., 2008). Therefore, because of its solubility and a wide range of applications in aquatic and terrestrial environments, the potential impacts of diquat on fish remains an important question to be address. Studies investigating diquat toxicity and its effects on higher level endpoints in fish are relatively few. Williams et al. (2016) showed that 20 µM diquat induced pericardial edema and slight dorsal or ventral curvature in zebrafish larvae. In the same study, diquat was also demonstrated to disrupt swimming behavior by delaying inflation of the swim bladder at 20 µM. Acute toxicity of diquat in different fish species (Stizostedion vitreum, Micropterus salmoides, Micropterus dolomieu, Lepomis macrochirus, Gambusia affinis) at early life stages has also been reported, and can range between 0.75 mg/L to 289 mg/L (LC50 values) (Paul et al., 1994). In another study, Sanchez et al. (2006) found that

three-spined stickleback (Gasterosteus aculeatus L.) exposed to environmentally relevant (22, 44 µg/L) and higher concentration (222, 444 µg/L) of diquat did not show effects on hepatic oxidative stress biomarkers, but report that diquat did alter hepatic enzymes for xenobiotic metabolism compared to control fish. Specifically, Ethoxyresorufin-O-deethylase (EROD) activity was decreased at all concentrations tested while GST was increased at 222 and 444 µg/L. In another study, the 96-h median tolerance limit for goldfish was reported to be 85 mg/L for diquat and hypertrophied gill epithelial cells were detected in fish at 32 mg/L diquat (Charles and Berry, 1984). Although not a fish, diquat at ecologically relevant concentrations of 22.2, 44.4 and 222.2 µg/L also altered the expression of stress-related transcripts in the aquatic invertebrate L. stagnalis (Bouétard et al., 2013). Thus, there are a handful of studies on diquat that report responses for enzyme activity, gross morphology, and mortality in aquatic organisms Much of what we know about diquat in terms of its mechanisms of action (MOA) comes from cell-based studies. The MOA of diquat has been investigated in different biological systems, and similar to its sister compound paraquat, diquat has been shown to induce oxidative stress in cells (Suleiman et al., 1987). The high propensity of diquat to propagate reactive oxygen species (ROS) is a predominant mechanism underlying diquat-induced cellular damage in different cell models (Slaughter et al., 2002; Zhang et al., 2012; Nisar et al., 2015). Diquat is proposed to transfer an electron to molecular oxygen, facilitating the formation of a superoxide anion radical, in the presence of cytochrome P450 reductase and NADPH, thus leading to sub-lethal and lethal effects in cells and tissues (Sandy et al., 1986; Jones and Vale, 2000). Oxidative stress arising from the redox cycling of diquat has been documented in both in vivo and in vitro studies using mammalian models (Karuppagounder et al., 2012; Nisar et al., 2015), with some of these data generated in aquatic animal models (Sanchez et al., 2006; Bouétard et al., 2013). However, studies report that the oxidative stress response induced by diquat can result in either an increased or decreased antioxidant response. This depends on the experimental paradigm, duration of exposure, and drug concentration (Slaughter et al., 2002; Bouétard et al., 2013; Zheng et al., 2013; Mao et al., 2014).

While there have been a number of cell-based experiments conducted to characterize

diquat toxicity, there remains a need to include mechanistic investigations in whole animals.

It has been demonstrated that another bipyridilium herbicide paraquat acts through a mechanism of mitochondrial dysfunction (Nunes et al., 2016; Wang et al., 2016). Diquat, which structurally resembles paraquat, has also been reported to impair mitochondrial function (Slaughter et al., 2002; Drechsel and Patel 2009). For example, exposure to diquat can lead to a significant increase in the activity of the mitochondrial enzyme glutamate dehydrogenase (GDH) in human neuroblastoma SH-SY5Y cells (Slaughter et al., 2002), and in the same cell type, diquat is proposed to result in a significant loss of mitochondrial transmembrane potential (Nisar et al., 2015). Drechsel and Patel (2009) also reported that diquat impairs mitochondrial complexes I and III, resulting in an aberrant electron transport and overproduction of ROS. In addition to these in vitro studies, diquat is reported to disrupt mitochondrial function in in vivo animal model, and one study investigating yellow perch (Perca flavescens) exposed to diquat reported that the animals suffered from significant respiratory stress with diquat (Bimber et al., 1976). Another toxic effect observed with exposure to diquat is reduced cell viability (Slaughter et al., 2002; Ran et al., 2004; Zhang et al., 2012; Nisar et al., 2015). It has been reported that diquat can electrostatically bind with DNA, and can lead to cell shrinkage, nuclear fragmentation, chromatin condensation, and chromosomal fragmentation, further exacerbating in cell death (Zhang et al., 2012). However, there is some debate as to whether apoptosis or necrosis is the major contributor to diquat-induced cell death pathways. Zhang et al. (2012) reported that diquat can act through apoptotic pathways in rat adrenal pheochromocytoma cells (PC12). Another study also demonstrated that chronic exposure to diquat can induce granulosa cell apoptosis and increase transcriptional responses of apoptosis-related genes in female ICR mice (Zhang et al., 2016). However, Nisar et al. (2015) reported that this herbicide may cause cell death in SH-SY5Y cells by programmed necrosis rather than apoptosis. Thus, some debate remains about the mechanisms underlying diquat induced toxicity, although these mechanisms may not be mutually exclusive. The objective of this study was to evaluate the sub-lethal effects of diquat in zebrafish. Zebrafish are an exceptional model for screening toxicants, and are highly sensitive to environmental pollutants during early stages of development (Flinn et al., 2008; McGrath and Li, 2008). While exposure to diquat

can result in oxidative stress, mitochondrial impairment, and apoptosis in cells, no study to the best of our knowledge has been conducted for bioenergetics. Assessing mitochondrial bioenergetics in larval zebrafish may act to improve understanding as to the molecular mechanisms underlying diquat toxicity. In addition, diquat toxicity may also be related to movement disorders and neurodegeneration, due to its similar structure to paraquat, a dopaminergic neurotoxin that induces mitochondrial dysfunction in a number of animal models (Bortolotto et al., 2014; Nunes et al., 2016). In mammals, diquat has been observed to induce abnormal gait and akinesia in male C57/BL-6 mice, two phenotypes that are implicated in locomotor deficiencies that are associated with the neurotoxic properties of diquat (Karuppagounder et al., 2012). In larval zebrafish, diquat has also been shown to disrupt swimming behavior by delaying inflation of the swim bladder (Williams et al., 2016). As such, the aims of this study were to determine the effect of diquat on mitochondrial bioenergetics and oxidative stress (at the transcript level), as well as locomotor activity in zebrafish. 2. Materials and Methods 2.1. Breeding of zebrafish Adult wildtype zebrafish (AB strain, Danio rerio) were raised in the Cancer Genetics Research Center at the University of Florida. Adult fish were maintained in an automated recirculating Pentair Aquatic Eco-Systems with an mean water pH of ~7.2±1 and mean temperature of ~28±1°C. Fish of approximately 3-4 months of age were exposed to a photoperiod cycle of 14h of light and 10h of dark for breeding. In the morning, the embryos were collected according to the criteria of Kimmel et al. (1995). Experimental procedures were approved by the University of Florida Institutional Animal Care and Use Committee and experiments with embryos and larvae were carried out at the Center for Environmental and Human Toxicology (University of Florida). 2.2. Experimental paradigm for exposures to diquat Supplemental figure 1 is provided to clarify the experimental design and endpoints measured. The experimental groups consisted of one control and three treatment groups. Control embryos were incubated in embryo rearing medium (ERM: 0.1368 mol NaCl, 0.0054 mol KCl, 0.0002 mol Na2HPO4,

0.0013 mol CaCl2, 0.0044 mol KH2PO4, 0.0010 mol MgSO4, 0.0042 mol NaHCO3, 1 L distilled water) at a pH of 7.2. Treatment groups consisted of waterborne incubations in one dose of 1, 10, 100 µM (0.362 mg/L, 3.62 mg/L, 36.2 mg/L) of diquat dibromide monohydrate (CAS: 6385-62-2, > 95%, Sigma, USA). Based upon the aforementioned studies, we selected 1, 10, 100 µM (or 0.362 mg/L, 3.62 mg/L, 36.2 mg/L) for our exposure, as we aimed to capture sub-lethal effects of the herbicide. These concentrations spanned a range of biological effects (e.g. gene expression, behavior) reported by others (Sanchez et al. 2006, Williams et al. 2016) and our lowest dose was one that is comparable to environmental levels after spraying diquat in agricultural fields (Ritter et al., 2000; Emmett, 2002). Diquat stock solution was prepared in distilled water and diluted in ERM to the final nominal concentrations of 1, 10, 100 µM. Chemicals were prepared fresh every day. Ninety-six embryos at 6 hours post fertilized (hpf) were incubated in a 96-well plate (N = 24) at a mean temperature of ~26°C±0.5°C for 96 hours exposure. Development, mortality and hatch time of embryos were recorded using an EVOS FL Auto Imaging System (Life Technologies) to obtain bright field images every hour during a 96 hours exposure. In a separate experiment, five embryos at 6 hpf were incubated with diquat in a glass beaker (N = 5 replicate beakers for each group) at 26°C. After 48 hours exposure, one embryo from each beaker from each group was selected randomly for the assessment of mitochondrial respiration (N = 5). We measure single embryos in the assay as we can better address individual variability in the oxygen consumption rates among embryos. Additional experiments were conducted for locomotion, and embryos were placed in two 96-well plates and continuously exposed to diquat for 5 days post fertilization (dpf) and 7 dpf for locomotion (N = 24) at 26°C. Lastly, another independent experiment was conducted for gene expression analysis, and ten embryos at 6 hpf were exposed to diquat in 5 or 6 glass beakers each group for 96 hours at 26°C. At 96 hours, a single pool of hatched larvae from each beaker (biological replicate) were flash-frozen using liquid nitrogen and stored at –80 °C for genes expression (N = 5 or 6 pools/treatment). 2.3. Oxygen consumption rate and respirometry

After exposure to diquat for 48 hours, the oxygen consumption rate was measured by the XFe24 Extracellular Flux Analyzer (Seahorse Bioscience). A Utility Plate was filled with 1 mL of XFe Calibrant fluid in each well and incubated with the sensor cartridge overnight at 26 °C. Following plate calibration, one embryo was placed in each well of an Islet Capture Microplate with 525 µL of ERM, or one dose of either 1, 10 and 100 μM of diquat (N = 5/treatment). The calibration plate was replaced by the Islet Capture Microplate to measure the oxygen consumption rate of embryos. The following pharmacological agents were added sequentially to assess respiratory sources: oligomycin (75 μM), carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP, 54 μM), sodium azide (200 mM). The final concentrations per well were 9.4 μM oligomycin, 6 μM FCCP, and 20 mM sodium azide (Wang et al., 2017). This protocol consisted of the following time cycles: 2 minutes for mixing, 1 minute pause, and then 2 minutes oxygen levels determination. Ten measurement cycles were performed for basal respiration of embryos, following by 18 measurement cycles after oligomycin injection to inhibit ATP production. Maximum respiratory was measured at 8 measurement cycles following FCCP injection and non-mitochondrial respiration was measured at 24 measurement cycles after introducing sodium azide to inhibit mitochondrial respiration. Additional details on the method are provided in Liang et al. (2017). 2.4. Zebrafish locomotion Embryos were exposed in two 96-well plates separately until 5 dpf and 7 dpf, then placed into a DanioVision™ instrument (Noldus Information Technology, Leesburg, VA). The 96-well plate was placed in the DanioVision Observation Chamber overnight at 26°C before behavioral test, to ensure appropriate tracking and control for variability caused by time of day (MacPhail et al., 2009). The activities of larvae (N = 24) were simultaneously and individually tracked using an infrared analog camera that installed in the DanioVision™ Observation Chamber for 50 minutes, consisting of alternating 10 minutes periods of light and dark and analyzed using EthoVision® XT software (Noldus Information Technology) as outlined in Zhang et al. (2017). These experiments were performed independently twice for each time point and data were combined to strengthen the overall analysis by first normalizing each dataset to an arbitrary value (control set to 1 and treatments relative to control) prior to combining all the data into a single data set. The total distance travelled, swimming velocity,

activity, and movement cumulative duration of each larvae were recorded to evaluate larval behavior. 2.5. Real-Time PCR Analysis Pools of ten larvae were collected following 96 hours exposure of diquat for gene expression analysis. Six replicates were prepared for each group. TRIzol® Reagent (Life Technologies, Carlsbad, CA, USA) was used to extract total RNA from samples and RNA pellets were reconstituted in RNase-DNase water. The quality of RNA for all samples was verified using the 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). The averaged RNA integrity number (RIN) for all samples was (9.26 ± 0.10). RNA samples were purified using the RNeasy Mini Kit column (Qiagen, Valencia, CA, USA). The quantity of RNA samples was measured by the NanoDrop-1000 (Thermo Scientific, USA). The cDNA was synthesized from 2 µg of column purified RNA using iScript (BioRad) as per the manufacturer’s protocol. Real-time PCR assays followed suggestions outlined by MIQE standards for high quality real-time PCR experimentation (Taylor et al., 2010). Real-time PCR analysis was performed by the CFX Connect System (BioRad), and was conducted with 5 or 6 biological replicates/group in duplicate, including 4 no reverse transcriptase control (NRT) and 4 no template control (NTC) in duplicate on each plate. The two-step thermal cycling parameters were as follows: initial 1-cycle Taq activation at 95 °C for 30 s, followed by 95 °C for 5 s, and primer annealing at 58 °C for 5 s. After 40 cycles, a dissociation curve was generated starting at 65.0 - 95.0 °C with increments of 0.5 °C every 5 s. Primer sets for target genes were collected from the literature (Table S1). The transcripts measured in this study comprised of superoxide dismutase 1 (sod1), superoxide dismutase 2 (sod2), catalase (cat), heat shock protein 70 (hsp70), bax, B-cell CLL/lymphoma 2a (bcl2) and caspase 3 (casp3). Three housekeeping genes (ribosomal subunit 18 (rps18), beta-actin (β-actin), and elongation factor 1α (ef1α)) were tested, but two of them (rps18 and β-actin) were used to normalize all target genes expression. Elongation factor 1α (ef1α) was removed due to high variability across individuals and treatments, which worsened the stability value (M-value). Using two reference genes, we achieved a combined M-value of 0.93 and CV of 0.32 using the CFX Manager™ software to evaluate stability of reference genes. Normalized gene

expression values were extracted using CFX Manager™ software with the relative ΔΔCq method based on the method of Livak and Schmittgen, (2001). 2.6. Statistical analysis Statistical analysis was performed by using Graph-Pad Prism version 6.0 (GraphPad Sofware Inc., La Jolla, CA, USA). Data are reported as mean value ± standard error. For mitochondrial respiration and zebrafish locomotion, differences in mean values were analyzed with one-way ANOVA following by a Dunnett’s post hoc test to the control group. The data for locomotor behavior was normalized and combined to strengthen the analysis. For gene expression analysis, differences in mean values were analyzed with Kruskal-Wallis test following by a Dunn’s post hoc test. Statistical significance of difference was set at p < 0.05. All figures were prepared using Graph-Pad Prism version 6.0 (GraphPad Sofware Inc., La Jolla, CA, USA). 3. Results 3.1. Mortality, hatch time and morphology Development of fish embryos were observed over 96 hours. Significant mortality and deformity were not observed for zebrafish larvae exposed to diquat. All experimental fish hatched out successfully within the exposure time. The embryos hatched out at 79.79 ± 0.15 hpf, 80.25 ± 0.23 hpf, 79.63 ± 0.26 hpf, 79.75 ± 0.51 hpf, respectively, in control group and three treatment groups. The hatch time of the embryos exposed to diquat were not different from the control embryos (F

(3, 92)

= 0.74, p = 0.53)

(Supplemental Fig. S2). 3.2. Mitochondrial bioenergetics After 48 hours, whole zebrafish embryos were assessed for mitochondrial respiration (Fig. 1A). The basal respiration of fish significantly decreased in 100 μM diquat by ~50% (F

(3, 16)

= 4.05; p = 0.026)

(Fig. 1B). Oligomycin can inactivate the mitochondrial ATP synthase, thereby, the difference between basal oxygen consumption rate and oxygen consumption rate following oligomycin injection indicates the amount of respiration that is linked to mitochondrial ATP production (ATP-linked respiration).

ATP-linked respiration was reduced by diquat, and there was a significant difference observed between embryos exposed to 100 µM diquat and controls by ~40-45% (F

(3, 16)

= 3.71; p = 0.039) (Fig. 1C).

FCCP is a mitochondrial uncoupler and induces maximal oxygen consumption, thus, the maximal respiration can be calculated following FCCP injection. After 48 hours exposure, diquat did not affect maximal respiration (Fig. 1D). Sodium azide blocks mitochondrial respiration and enables calculation of non-mitochondrial respiration driven by processes outside the mitochondria. There were no differences significant differences among groups for non-mitochondrial respiration after 48 hours exposure (Fig. 1E). Taken together, there is evidence that diquat disturbs mitochondrial respiration in live embryos. 3.3. Zebrafish locomotion The activity of 5 dpf larval zebrafish were not significantly different from the control fish (Supplemental Figure S3 A-D). However, 7 dpf larvae showed higher locomotor activities at 10 µM diquat compared with the control fish (Fig. 2A-D), which corresponded to total distance travelled (F

(3,

188)

= 5.09; p = 0.0021), swimming velocity (F (3, 188) = 5.09; p = 0.0021), movement cumulative duration

(F

(3, 188)

= 5.76; p = 0.0009) and activity (F

(3, 188)

= 7.43; p < 0.0001). The activity of 7 dpf larval

zebrafish in 100 μM diquat was also significantly increased compared with the control fish. Larvae exposed to 1 and 100 µM diquat at 7 dpf did not show differences in total distance travelled, swimming velocity and movement cumulative duration compared to the control fish. Thus, there may be an inverted U-shape dose response for the locomotor activity of 7 dpf larvae. 3.4. Transcripts related to oxidative stress and apoptosis The transcript levels of sod1, sod2, cat and hsp70 were measured to determine whether or not diquat affected the expression of genes indicative of an oxidative stress in zebrafish larvae (Fig. 3 A-D). Cat expression was significantly increased for fish exposed in 10 μM diquat when compared with the control fish (K = 6.44, p = 0.073). Other transcripts such as sod1, sod2 and hsp70 in diquat-treated fish were not different in relative levels when compared to the control fish. The transcript levels of bax, bcl2 and casp3 were measured to determine whether or not diquat affected the expression of genes related to apoptosis in zebrafish larvae (Fig. 3 E-G). The data showed that the expression levels for these genes

were not differentially affected by diquat. Taken together, there is some evidence that embryos are up-regulating enzymes needed to mitigate ROS production (i.e. catalase). 4. Discussion Diquat has replaced paraquat for some agricultural applications because of the higher human risks associated with paraquat. Despite reports of lower toxicity of diquat compared to paraquat, the propensity of diquat to induce redox homeostasis disturbance ensures that diquat remains a concern for aquatic organisms. In the present study, significant mortalities and deformities were not detected in zebrafish after 96 hours exposure to diquat. It has been reported that 20 µM diquat induces pericardial edema and slight dorsal or ventral curvature in zebrafish larvae (Williams et al., 2016). A lack of response for the morphological abnormality in this study may be attributed to the different exposure periods. In the study by Williams et al. (2016), the exposure period started at 4 dpf and ended at 6 dpf, compared with 6 hpf to 4 dpf used in this study, thus there may be a critical window of sensitivity in embryos to diquat and deformities manifest between ~4-6 dpf.

In this case, the chorion may be acting

as a physical barrier to reduce the uptake of diquat (Mandrell et al., 2012), and this many also contribute to its relatively low toxicity to zebrafish embryos. In the present study, the hatch time of fish was not significantly different in animals exposed to diquat when compared to the control fish. The hatch time of embryos is an important stress response for fish larvae (Barton, 2002). In contrast, diquat’s sister compound paraquat significantly accelerated the hatch time of zebrafish embryos (Wang et al., 2017). Diquat may not elicit the same stress response as paraquat due to its lower toxicity. Conversely, diquat is proposed to arrest cell growth and increase cell cycle in mammalian cells (Tanaka and Amano, 1989; Shackelford et al., 2000; Freeman and Rayburn, 2006), and perhaps these are mechanisms that are involved in maintaining the timing of hatch for the fish. Conversely, zebrafish may not be experiencing significant stress following diquat exposure. Diquat exhibited very low toxicity to zebrafish larvae based upon survival and lack of any obvious deformities. In addition to the physical barrier of chorion, due to the property of hydrophilicity and ionization of diquat, the blood-brain barrier (BBB) may obstruct its uptake (Abbott et al., 2006) into the central

nervous system; thus, zebrafish exposed to lower doses of diquat may not exhibit adverse neuronal effects nor deformity. Klüver et al. (2015) recently suggested that the bioconcentration of diquat in developing zebrafish may not reach a steady state equilibrium state within a 96-hour period. Hence, longer exposure to the herbicide may be required to elicit adverse effects on zebrafish development and morphology. Despite no evidence for overt mortality or deformity due to diquat exposure, larvae did exhibit sub-lethal effects that included impairment of oxidative respiration and locomotor activity. We hypothesized that diquat would impair mitochondrial function in embryos based upon literature demonstrating that it can induce oxidative stress. We found that basal respiration and ATP-linked respiration were decreased in embryos following a 48 hour exposure to diquat compared to control fish. Other studies also show that chemical exposures can affect the bioenergetics of zebrafish embryos, for example, paraquat, polycyclic aromatic hydrocarbons and triclosan were shown to disrupt mitochondrial respiration in embryos (Knecht et al., 2013; Shim et al., 2016; Wang et al., 2017). Diquat toxicity resembles that of paraquat toxicity in that there is superoxide anion production through redox cycling (Jones and Vale, 2000). It is reported that diquat activates mitochondrial enzyme GDH and reduces mitochondrial membrane potential in SH-SY5Y cells (Slaughter et al., 2002; Nisar et al., 2015). A study by Drechsel and Patel, (2009) also supported the hypothesis that diquat affects mitochondrial bioenergetics directly. In the study, diquat was observed to attenuate the production of H2O2 after the dissipation of mitochondrial membrane potential by adding FCCP. This indicates that ATP production is involved in diquat-induced ROS. These data also suggest that diquat is interacting directly with mitochondria to regulate bioenergetics. However, unlike paraquat, less is known about the molecular targets of diquat. Hence, rigorous experiments are required to test how diquat interacts with different protein complexes in the electron transport chain. Behavioral responses act as sensitive indicators of environmental stressors in fish early life stages (Selderslaghs et al., 2010; Sloman and McNeil, 2012). Karuppagounder et al. (2012) reported that diquat induces behavior impairments in male C57/BL-6 mice. However, little is known about the effect of diquat on fish behavior. Paraquat, which has similar structure and physiochemical property to diquat, is well documented to induce behavior deficits in mammalian and fish models (Litteljohn et al., 2009;

Bortolotto et al., 2014; Nellore and Nandita, 2015; Wang et al., 2017). Additionally, locomotor responses of zebrafish larvae to neuroactive drugs resembles that of mammals (Irons et al., 2010). Therefore, we considered behavioral impairments as a putative consequence underlying diquat toxicity in zebrafish larvae. We report that the locomotor performance of 5 dpf zebrafish larvae after diquat exposure was not different from control fish. However, the locomotor activity of 7 dpf zebrafish larvae was higher in response to diquat at 10 µM compared to the control group for all endpoints measured, and total activity was also affected in 7dpf zebrafish with the highest dose of diquat (100 µM). Irons et al. (2010) reported that low doses of D-amphetamine can result in hyperactivity in zebrafish larvae, while higher doses can result in hypo-activity. Thus, chemical dose is an important variable for the observed behavioral responses that can occur in fish. Moreover, we report that the developmental time period is an important variable for diquat exposure, as 5 dpf larvae did not show the same behavioral responses as 7 dpf larvae. Chemical exposures may produce different behaviors later in life (Tierney, 2011), and this may reflect differences in central nervous system development between stages. One potential mechanism for the behavioral manifestations in larvae is the effects of diquat on mitochondrial oxidative phosphorylation and ATP production at an earlier stage of development (Allen et al., 2008). There is some evidence to suggest that energy deficits in zebrafish larvae are related to changes in locomotor performance (Zhang et al., 2017). Additionally, the vulnerability of neurons to mitochondrial damage due to their high metabolic activity and energetic requirement can also result in locomotor deficits of animals (Camilleri and Vassallo, 2014). An alternative hypothesis for the increase in locomotor activity may be that, when diquat inhibits mitochondrial ATP production at higher doses, then embryonic and larval fish may be required to use all nutrient stores more rapidly that unexposed animals. These depleted reserves may initiate higher foraging activity in search of food. We did not measure the onset of exogenous feeding in the diquat-exposed groups but it would be interesting to explore whether diquat-induced hyperactivity is linked with foraging for food. In addition to ATP production, mitochondria are also important organelles that regulate the apoptotic and redox signaling pathways in cells (Dumollard et al., 2007; DiMauro and Schon, 2008; Meyer et al., 2013). Hence, we measured transcriptional responses for stress- and apoptosis-related genes in larvae.

The expression proteins of sod1 and sod2 genes are two important enzymes that mitigate mitochondrial ROS overproduction and regulate apoptotic signaling (Pias et al., 2003; Kasahara et al., 2005; Sea et al., 2015). We determined that the mRNA levels of sod1 and sod2 for zebrafish larvae exposed to diquat were not different from untreated fish. Conversely, zebrafish larvae exposed to paraquat showed an increase of sod2 expression (Wang et al., 2017). This may be related to the higher toxicity of paraquat, as diquat is reported to be less toxic. In addition to sod1 and sod2, cat is also an important gene encoding catalase to protect the cell from oxidative damage. In the present study, the transcript levels for cat were increased in zebrafish exposed to 10 µM diquat. This also corresponded to the dose that was most associated with a change in locomotor activity. It has been reported that microsomes in rat liver cell have sufficient cat activity to prevent accumulation of hydrogen peroxide caused by diquat (Gage, 1968). We hypothesize that diquat induced ROS occurs not only in mitochondria but also in cytoplasm, and there are also evidences that diquat can produce ROS both inside and outside the mitochondria (Drechsel and Patel, 2008; Nisar et al., 2015). An increase expression level of cat, but not sod1 and sod2, may be attributed to ROS production that is independent of mitochondria. The mode of antioxidant response appears to be variable depending on experimental paradigm. An increased antioxidant response was found in SH-SY5Y cells treated with 25 µM diquat for 48 hours exposure (Slaughter et al., 2002), while there were reduced activities of antioxidant enzymes for weaned pigs treated with 10 mg/kg diquat at 96 hours after injection (Zheng et al., 2013; Mao et al., 2014). Additionally, transcript levels for bax, bcl2 and casp3 were not differentially regulated by diquat at the doses tested in this study, suggesting that mitochondrial dysfunction in larvae may not involve changes in these transcripts. Apoptotic process in cells can result from an increased bax/bcl2 ratio, which can cause loss of mitochondrial transmembrane potential (Jang et al., 2015; Hu et al., 2017). In this study, embryos maintain this ratio among groups, suggesting there may be a compensatory mechanism to regulate apoptosis. In summary, diquat did not result in significant mortality and deformity in larval zebrafish exposed to 1, 10 and 100 µM diquat. However, we observed significant differences in behavioral responses for 7d old larvae in 10 µM diquat, whereas this was not the case for 5d old larvae, suggesting that differential responses in behavior may be more apparent later in life.

One reason for observing effects

at later developmental stages may be related to detoxification of diquat in the liver, and once the liver matures, the larvae may begin to detoxify diquat, leading to daughter compound(s) that is (are) more toxic than the parent. Future studies should address this aspect, and determine whether detoxification of diquat and daughter compounds are more toxic at later stages. In terms of oxidative stress as a mechanism, the decrease in basal respiration, ATP production, and the up-regulation of catalase transcript levels suggests mitochondrial dysfunction and ROS generation with diquat. We also measured the expression levels of apoptosis-related genes, and they were not differentially regulated by diquat, suggesting that mitochondrial dysfunction may not significantly involve initiation of apoptotic pathways in zebrafish. However, additional methods such as apoptotic stains would better address this possibility. This study improves mechanistic understanding of diquat exposure in fish at early stages of development and presents evidence that diquat disrupts mitochondrial function and behavior.

Acknowledgments The authors have no conflict of interest to declare. We thank Edward Flynn for zebrafish husbandry and technical support. This research was funded by the University of Florida and the College of Veterinary Medicine (CJM) and supported by the National Natural Science Foundation of China (21377022； 21777022). The support provided by China Scholarship Council (CSC) for Xiaohong Wang (No. 201606620043) to visit the University of Florida is acknowledged.

References Abbott, N.J., Rönnbäck, L., Hansson, E., 2006. Astrocyte-endothelial interactions at the blood-brain barrier. Nat. Rev. Neurosci. 7, 41-53. Allen, D.G., Lamb, G.D., Westerblad, H., 2008. Skeletal muscle fatigue: cellular mechanisms. Physiol. Rev. 88, 287-332. Aramendía, M.A., Borau, V., Lafont, F., Marinas, A., Marinas, J.M., Moreno, J.M., Porras, J.M., Urbano, F.J.,

2006.

Determination

of

diquat

and

paraquat

in

olive

oil

by

ion-pair

liquid

chromatography-electrospray ionization mass spectrometry (MRM), Food Chem. 97, 181-188. Barton, B.A., 2002. Stress in fishes: a diversity of responses with particular reference to changes in circulating corticosteroids. Integr. Comp. Biol. 42, 517-525. Bouétard, A., Besnard, A.L., Vassaux, D., Lagadic, L., Coutellec, M.A., 2013. Impact of the redox-cycling herbicide diquat on transcript expression and antioxidant enzymatic activities of the freshwater snail Lymnaea stagnalis. Aquat. Toxicol. 126, 256-265. Bimber, D.L., Boenig, R.W., Sharma, M.L., 1976. Respiratory stress in yellow perch induced by subtoxic concentrations of diquat. Ohio J. Sci. 76, 87-90. Bortolotto, J.W., Cognato, G.P., Cristoff, R.R., Roesler, L.N., Leite, C.E., Kist, L.W., Bogo, M.R., Vianna, M.R., Bonan, C.D., 2014. Long-term exposure to paraquat alters behavioral parameters and dopamine levels in adult zebrafish (Danio rerio). Zebrafish. 11, 142-153. Camilleri, A., Vassallo, N., 2014. The centrality of mitochondria in the pathogenesis and treatment of Parkinson’s Disease. CNS Neurosci. Ther. 20, 591-602. Charles, R., Berry, Jr., 1984. Toxicity of the herbicides diquat and endothall to goldfish. Environ. Pollut. 34, 251-258. DiMauro, S., Schon, E.A., 2008. Mitochondrial disorders in the nervous system. Annu. Rev. Neurosci. 31, 91-123.

Drechsel, D. A., Patel, M., 2008. Role of reactive oxygen species in the neurotoxicity of environmental agents implicated in Parkinson’s disease. Free Radic. Biol. Med. 44, 1873-1886. Drechsel, D.A., Patel, M., 2009. Differential contribution of the mitochondrial respiratory chain complexes to reactive oxygen species production by redox cycling agents implicated in parkinsonism. Toxicol. Sci. 112, 427-434. Ducrot, V., Péry, A.R.R., Lagadic, L., 2010. Modelling effects of diquat under realistic exposure patterns in genetically differentiated populations of the gastropod Lymnaea stagnalis. Phil. Trans. R. Soc. B. 365, 3485-3494. Dumollard, R., Duchen, M., Carroll, J., 2007. The role of mitochondrial function in the oocyte and embryo. Curr. Top. Dev. Biol. 77, 21-49. Emmett, K., 2002. Final risk assessment for diquat bromide. Washington State Department of Ecology, Washington, DC. Flinn, L., Bretaud, S., Lo, C., Ingham, P.W., Bandmann, O., 2008. Zebrafish as a new animal model for movement disorders. J. Neurochem. 106, 1991-1997. Freeman, J.L., Rayburn, A.L., 2006. Aquatic herbicides and herbicide contaminants: in vitro cytotoxicity and cell-cycle analysis. Environ. Toxicol. 21, 256-263. Gage, 1968. The action of paraquat and diquat on the respiration of liver cell fractions. Biochem. J. 109, 757-761. Hu, X., Shen, H., Wang, Y., Zhao, M., 2017. Liver X receptor agonist TO901317 attenuates paraquat-induced acute lung injury through inhibition of NF-𝜅B and JNK/p38 MAPK signal pathways. Biomed. Res. Int. Article ID 4652695. Irons, T.D., MacPhail, R.C., Hunter, D.L., Padilla, S., 2010. Acute neuroactive drug exposures alter locomotor activity in larval zebrafish. Neurotoxicol. Teratol. 32, 84-90. Jang, Y.J., Won, J.H., Back, M.J., Fu, Z., Jang, J.M., Ha, H.C., Hong, S.B., Chang, M., Kim, D.K., 2015.

Paraquat induces apoptosis through a mitochondria-dependent pathway in RAW264.7 Cells. Biomol. Ther. 23, 407-413. Jones, G.M., Vale, J.A., 2000. Mechanisms of toxicity, clinical features, and management of diquat poisoning: a review. J. Toxicol. Clin. Toxicol. 38, 123-128. Karuppagounder, S.S., Ahuja, M., Buabeid, M., Parameshwaran, K., Abdel-Rehman, E., Suppiramaniam, V., Dhanasekaran, M., 2012. Investigate the chronic neurotoxic effects of diquat. Neurochem. Res. 37, 1102-1111. Kasahara, E., Lin, L.R., Ho, Y.S., Reddy, V.N., 2005. Sod2 protects against oxidation-induced apoptosis in mouse retinal pigment epithelium: Implications for age-related macular degeneration. Invest. Ophthalmol. Vis. Sci. 46, 3426-3434. Kimmel, C.B., Ballard, W.W., Kimmel, S.R., Ullmann, B., Schilling, T.F., 1995. Stages of embryonic development of the zebrafish. Dev. Dyn. 203, 253-310. Klüver, N., König, M., Ortmann, J., Massei, R., Paschke, A., Kühne, R., Scholz, S., 2015. Fish embryos toxicity test: Identification of compounds with weak toxicity and analysis of behavioral effects to improve prediction of acute toxicity for neurotoxic compounds. Environ. Sci. Technol. 49, 7002-7011. Knecht, A.L., Goodale, B.C., Truong, L., Simonich, M.T., Swanson, A.J., Matzke, M.M., Anderson, K.A., Waters, K.M., Tanguay, R.L., 2013. Comparative developmental toxicity of environmentally relevant oxygenated PAHs. Toxicol. Appl. Pharmacol. 271, 266-275. Liang, X., Souders, C.L. 2nd, Zhang, J., Martyniuk, C.J. 2017. Tributyltin induces premature hatching and reduces locomotor activity in zebrafish (Danio rerio) embryos/larvae at environmentally relevant levels. Chemosphere. 189:498-506. Litteljohn, D., Mangano, E., Shukla, N., Hayley, S., 2009. Interferon-gamma deficiency modifies the motor and co-morbid behavioral pathology and neurochemical changes provoked by the pesticide paraquat. Neuroscience. 164, 1894-1906.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2 -ΔΔCT method. Methods. 25, 402-408. MacPhail, R.C., Brooks, J., Hunter, D.L., Padnos, B., Irons, T.D., Padilla, S., 2009. Locomotion in larval zebrafish: Influence of time of day, lighting and ethanol. Neurotoxicology. 30, 52-58. Mandrell, D., Truong, L., Jephson, C., Sarker, M.R., Moore, A., Lang, C., Simonich, M.T., Tanguay, R.L., 2012. Automated zebrafish chorion removal and single embryo placement: optimizing throughput of zebrafish developmental toxicity screens. J. Lab. Autom. 17, 66-74. Mao, X., Lv, M., Yu, B., He, J., Zheng, P., Yu, J., Wang, Q.Y., Chen, D.W., 2014. The effect of dietary tryptophan levels on oxidative stress of liver induced by diquat in weaned piglets. J. Anim. Sci. Technol. 5, 49. McCormack, A.L., Thiruchelvam, M., Manning-Bog, A.B., Thiffault, C., Langston, J.W., Cory-Slechta, D.A., di Monte, D.A., 2002. Environmental risk factors and Parkinson’s disease: selective degeneration of nigral dopaminergic neurons caused by the herbicide paraquat. Neurobiol. Dis. 10, 119-127. McGrath, P., Li, C.Q., 2008. Zebrafish: a predictive model for assessing drug-induced toxicity. Drug Discov. Today. 13, 394-401. Meyer, J.N., Leung, M.C., Rooney, J.P., Sendoel, A., Hengartner, M.O., Kisby, G.E., Bess, A.S., 2013. Mitochondria as a target of environmental toxicants. Toxicol. Sci. 134, 1-17. Minnema, D.J., Travis, K.Z., Breckenridge, C.B., Sturgess, N.C., Butt, M., Wolf, J.C., Zadory, D., Herberth, M.T., Watson, S.L., Cook, A.R., Botham, P.A., 2016. Dietary administration of diquat for 13 weeks does not result in a loss of dopaminergic neurons in the substantia nigra of C57BL/6J mice. Regul. Toxicol. Phar. 75, 81-88. Nellore, J., Nandita, P., 2015. Paraquat exposure induces behavioral deficits in larval zebrafish during the window of dopamine neurogenesis. Toxicol. Rep. 2, 950-956. Nunes, M.E., Müller, T.E., Braga, M.M., Fontana, B.D., Quadros, V.A., Marins, A., Rodrigues, C.,

Menezes, C., Rosemberg, D.B., Loro, V.L., 2017. Chronic treatment with paraquat induces brain injury, changes in antioxidant defenses system, and modulates behavioral functions in zebrafish. Mol. Neurobiol. 54, 3925-3934. Nisar, R., Hanson, P.S., He, L., Taylor, R.W., Blain, P.G., Morris, C.M., 2015. Diquat causes caspase-independent cell death in SH-SY5Y cells by production of ROS independently of mitochondria. Arch. Toxicol. 89, 1811-1825. Paul, E.A., Simonin, H.A., Symula, J., 1994. The toxicity of diquat, endothall, and fluridone to the early life stages of fish. J. Freshwater Ecology. 9, 229-239. Pias, E.K., Ekshyyan, O.Y., Rhoads, C.A., Fuseler, J., Harrison, L., Aw, T.Y., 2003. Differential effects of superoxide dismutase isoform expression on hydroperoxide-induced apoptosis in PC-12 cells. J. Biol. Chem. 278, 13294-13301. Ran, Q., Liang, H., Gu, M., Qi, W., Walter, C.A., Roberts, L.J., Herman, B., Richardson, A., Van Remmen, H., 2004. Transgenic mice overexpressing glutathione peroxidase are protected against oxidative stress-induced apoptosis. J. Biol. Chem. 279, 55137-55146. Ritter, A.M., Shaw, J.L., William, W.M., Travis, K.Z., 2000. Characterizing aquatic ecological risks from pesticides using a diquat dibromide case study. I. Probabilistic exposure estimates. Environ. Toxicol. Chem. 19, 749-759. Sanchez, W., Palluel, O., Lagadic, L., Aït-Aïssa, S., 2006. Biochemical effects of nonylphenol polyethoxylate adjuvant, diquat herbicide and their mixture on the three-spined stickleback (Gasterosteus aculeatus L.). Mar. Environ. Res. 62, S29-S33. Sandy, M.S., Moldeus, P., Ross, D., Smith, M.T., 1986. Role of redox cycling and lipid peroxidation in bipyridyl herbicide cytotoxicity. Studies with a compromised isolated hepatocyte model system. Biochem. Pharmacol. 35, 3095-3101. Sea, K., Sohn, S.H., Durazo, A., Sheng, Y., Shaw, B.F., Cao, X., Taylor, A.B., Whitson, L.J., Holloway, S.P., Hart, P.J., Cabelli, D.E., Gralla, E.B., Valentine, J.S., 2015. Insights into the role of the unusual

disulfide bond in copper-zinc superoxide dismutase. J. Biol. Chem. 290, 2405-2418. Selderslaghs, I.W.T., Hooyberghs, J., De Coen, W., Witters, H.E., 2010. Locomotor activity in zebrafish embryos: A new method to assess developmental neurotoxicity. Neurotoxicol. Teratol. 32, 460-471. Shackelford, R. E., Kaufmann, W. K., Paules, R. S., 2000. Oxidative stress and cell cycle checkpoint function. Free Radic. Biol. Med. 28, 1387-1404. Shim, J., Weatherly, L.M., Luc, R.H., Dorman, M.T., Neilson, A., Ng, R., Kim, C.H., Millard, P.J., Gosse, J.A., 2016. Triclosan is a mitochondrial uncoupler in live zebrafish. J. Appl. Toxicol. 36, 1662-1667. Siemering, G.S., Hayworth, J.D., Greenfield, B.K., 2008. Assessment of potential aquatic herbicide impacts to California aquatic ecosystems. Arch. Environ. Contam. Toxicol. 55, 415-431. Slaughter, M.R., Thakkar, H., O’Brien, P.J., 2002. Effect of diquat on the antioxidant system and cell growth in human neuroblastoma cell. Toxicol. Appl. Pharmacol. 178, 63-70. Sloman, K.A., McNeil, P.L., 2012. Using physiology and behavior to understand the responses of fish early life stages to toxicants. J. Fish Biol. 81, 2175-2198. Tanaka. R., Amano. Y., 1989. Genotoxic effects of paraquat and diquat evaluated by sister-chromatid exchange, chromosomal aberration and cell-cycle rate. Toxicol. In Vitro. 3, 53-57. Taylor, S., Wakem, M., Dijkman, G., Alsarraj, M., Nguyen, M., 2010. A practical approach to RT-qPCR-publishing data that conform to the MIQE guidelines. Methods. 50, 1-5. Tierney, K.B., 2011. Behavioral assessment of neurotoxic effects and neurodegeneration in zebrafish. Biochim. Biophys. Acta. 1812, 381-389. US EPA. 2009. Risks of Paraquat Use to Federally Threatened California Red-legged Frog (Rana aurora draytonii). Pesticide Effects Determination. Environmental Fate and Effects Division, Office of Pesticide Programs, US Environmental Protection Agency, Washington, D.C. Vanholder, R., Colardyn, F., De Reuck, J., Praet, M., Lameire, N., Ringoir, S., 1981. Diquat intoxication:

report of two cases and review of the literature. Am. J. Med. 70, 1267-1271. Wang, Q., Liu, S., Hu, D.Y., Wang, Z., Wang, L., Wu, T.F., Wu, Z.B., Mohan, C., Peng, A., 2016. Identification of apoptosis and macrophage migration events in paraquat-induced oxidative stress using a zebrafish model. Life Sci. 157, 116-124. Wang, X.H., Souders, C.L., Zhao, Y.H., Martyniuk, C.J., 2017. Paraquat alters mitochondrial bioenergetics, locomotion, and the expression of the dopamine system in zebrafish larvae (Danio rerio). Toxicol. Appl. Pharmacol. In review. Williams, L.M., Lago, B.A., McArthur, A.G., Raphenya, A.R., Pray, N., Saleem, N., Salas, S., Paulson, K., Mangar, R.S., Liu, Y., Vo, A.H., Shavit, J.A., 2016. The transcription factor, nuclear factor, erythroid 2 (Nfe2), is a regulator of the oxidative stress response during Danio rerio development. Aquat. Toxicol. 180, 141-154. Zhang, J.L., Souders, C.L., Denslow, N.D., Martyniuk, C.J., 2017. Quercetin, a natural product supplement, impairs mitochondrial bioenergetics and locomotor behavior in larval zebrafish (Danio rerio). Toxicol. Appl. Pharmacol. 327, 30-38. Zhang, J.Q., Gao, B.W., Wang, J., Wang, X.W., Ren, Q.L., Chen, J.F., Ma, Q., Xing, B.S., 2016. Chronic exposure to diquat causes reproductive toxicity in female mice. PLoS ONE. 11, e0147075. Zheng, P., Yu, B., He, J., Tian, G., Luo, Y., Mao, X., Zhang, K., Che, L., Chen, D., 2013. Protective effects of dietary arginine supplementation against oxidative stress in weaned piglets. Br. J. Nutr. 109, 2253-2260. Zhang, Q., Wang, C., Liu, W.P., Zhang, X.F., Zhuang, S.L., 2012. Evidence for DNA-diquat interaction and cytotoxicity in in vitro rat cells. Environ. Chem. Lett. 10, 35-39.

Figure caption Fig. 1. Mitochondrial bioenergetics (oxygen consumption rate) of zebrafish embryos exposed to embryo rearing media (ERM), or one dose of 1, 10, or 100 μM of diquat from 6 hpf to 54 hpf (48 hour exposure). (A) Diagram depicting profile of oxygen consumption rates throughout the assay. (B) Basal respiration, (C) ATP-linked respiration, (D) Maximal respiration, (E) Spare capacity respiration, and (F) Non-mitochondrial respiration. Data are expressed as mean ± standard error (n = 5). Asterisks (*) indicates a significant difference between control fish and fish in a designated treatment at p < 0.05.

Fig. 2. Locomotor activity of zebrafish larvae exposed to embryo rearing media (ERM), or one dose of 1, 10, or 100 μM of diquat from 6 hpf to 7 dpf. All values are relative to the control group which has a mean value of one (relative units). (A) Total distance moved, (B) Velocity, (C) Moving cumulative duration, (D) Activity are presented in relation to the control group. Data are expressed as mean ± standard error (n = 24/group). Asterisks (*) indicates a significant difference between control fish and treated fish at p < 0.05.

Fig. 3. Transcript levels for genes related to oxidative stress and apoptosis in zebrafish embryos. The steady state mRNA levels of (A) sod1, (B) sod2, (C) catalase, (D) hsp70, (E) bax, (F) bcl2, and (G) caspase 3 in zebrafish embryos exposed to either embryo rearing media (ERM), or one dose of 1, 10, or 100 μM of diquat from 6 hpf up to 96 hours. Data are expressed as mean ± standard error (n = 5-6).

Asterisks (*) indicate a significant difference between control fish and treated fish at p < 0.05.