Acute toxicity and cholinesterase inhibition of the nematicide ethoprophos in larvae of gar Atractosteus tropicus ( Semionotiformes : Lepisosteidae )

Biomarkers are a widely applied approach in environmental studies. Analyses of cholinesterase (ChE), glutathione S-transferase (GST) and lipid peroxidation (LPO) are biomarkers that can provide information regarding early effects of pollutants at different biochemical levels on an organism. The aim of this study was to evaluate the biomarker approach on a Costa Rican native and relevant species. For this, larvae of gar (Atractosteus tropicus) were exposed to the organophosphorus nematicide, ethoprophos. Acute (96hr) exposure was conducted with pesticide concentrations ranging from 0.1μg/L to 1 500μg/L. The 96hr LC50 calculated was 859.7μg/L. After exposure, three biomarkers (ChE, GST and LPO) were analyzed in fish that survived the acute test. The lowest observed effect concentration (LOEC) regarding ChE activity inhibition was 50μg/L. This concentration produced a significant inhibition (p<0.05) of the enzyme by 20%. The highest concentration tested without showing any effect on ChE activity and therefore considered as no observed effect concentration (NOEC) was 10μg/L. Ethoprophos concentration of 400μg/L caused a ChE inhibition by 79%. In this study, no significant variations (p>0.05) in GST activity and LPO were observed in A. tropicus larvae after exposure to ethoprophos. Rev. Biol. Trop. 60 (1): 361-368. Epub 2012 March 01.

The application of large amounts of agricultural pesticides is a serious environmental problem in Costa Rica, that has increased its imports during the last 30 years (Ramírez et al. 2009).In this period, it was found that insecticides/nematicides make up about 10% of the total amount of active ingredients (a.i.) being imported.Of these, ethoprophos was among the top ten pesticides imported relative to quantity (Ramírez et al. 2009).The intensity and volume of pesticide use, related to the expansion of intensive agriculture, therefore represents a serious threat for biodiversity, especially in fragile and complex systems such as wetlands.
Tropical ecosystems harbor much of the world's biodiversity, and are subjected to very high land conversion rates due to agriculture.Conservation and management of biodiversity urgently needs information regarding impacts at the ecotoxicological level on native organisms, so that the sensitivity of individual species as well as of ecosystems can be characterized.More importantly, information regarding species-specific sensitivity to common pesticides and pollutants is an essential tool for establishing limits and regulations on pollutants released to the environment through agricultural activities.In this regard, most of the studies concerning species-specific ecotoxicology, and its use to establish limits and regulations have been carried out in North America and Europe (US-EPA 1986, OECD 2007).Some studies has been carried out in South America and Mexico (Akaishi et al. 2002, de la Torre et al. 2005, Azevedo et al. 2009, González-Mille et al. 2010), but information is scarce for Central America.
Atractosteu tropicus inhabit the San Juan River, its tributaries in Northern Costa Rica and the wetlands of the Caño Negro National Wildlife Refuge, where it is considered an ensign species.However, the intensive pesticide use in agriculture at the surroundings of the refuge, especially in orange and pineapple crops carry a potential hazard to the gar and other fish populations inhabiting the refuge (Soto & Ramírez 2002).Criticisms regarding the environmental damage caused by pesticide use in the Caño Negro National Wildlife Refuge have been published in the local press (Chacón 2009, Barquero 2010, Hernández 2010).However, until now, A. tropicus has not been considered a sentinel species for monitoring the effects of environmental pollution in the area.In this regard, studies developed with other species of gar fish have demonstrated the potential of these fish for monitoring the effects of pollutants, including pesticides, heavy metals and endocrine disruptors (Osborne & Rulifson 2010, Aguilera et al. 2010).These studies have evaluated the bioaccumulation of metals and the response of biomarkers.
Biomarker evaluation allows for early detection of pollutant impacts on organisms.Biomarkers, such as the inhibition of cholinesterase (ChE) activity are popular and widely used tools in the field of environmental toxicology, mainly with the goal of estimating the effects of pollutants at the sub-individual level (Triebskorn et al. 2001, Guilhermino 2006, Sanchez et al. 2007, Almeida et al. 2010).Inhibition of ChE activity is one of the best characterized biomarkers and has been intensively used in environmental studies, showing a specific response to organophosphate (OP) and carbamate pesticides (Gruber & Munn 1998, Thompson 1999, Pathiratne et al. 2008, Mdegela et al. 2010).The performance of ChE inhibition as a biomarker has rendered it as a potential substitute of chemical analysis used to detect exposure to OP and carbamate pesticides (Peakall 1992).Considering the intensive use of organophosphates in Costa Rican agriculture and the use of this kind of pesticides in the Caño Negro area (Soto & Ramírez 2002), ChE stands out as an ideal biomarker to evaluate agriculture-related pollution effects in the area.Further, it would be useful to trace a link between pesticide use, detection of residues in the environment and their toxic effects.
Biotransformation and oxidative stress biomarkers such as Glutathione S-transferase (GST) and lipid peroxidation (LPO) are also widely applied in environmental studies (Mdegela et al. 2006, Guilhermino 2006, Sanchez et al. 2007); these markers are more unspecific, but still useful signals of environmental stress as they might respond to a broad diversity of pollutants.
In Costa Rica, the process of induced breeding of A. tropicus has been achieved twice with promising results, and with the objective of producing fingerlings for aquaculture and for reintroduction in areas where wild populations has been considerably diminished (Protti et al. 2010).Considering the availability of A. tropicus larvae from controlled spawns and the ecological relevance of this species, the production of ecotoxicological information on this fish would facilitate, its use as a sentinel species in ecologically vulnerable areas.The objective of this study was to evaluate the acute toxicity and the effects at the sub-individual level of the insecticide/nematicide ethoprophos on the Costa Rican indigenous fish A. tropicus.

MATERIALS AND METHODS
Fish used in the assays were bred in the Laboratory of Natural Resources and Wildlife (LARNAVISI); toxicity tests and biomarker analyses were carried out at the laboratory of Ecotoxicology (ECOTOX), both laboratories are located at the Campus Omar Dengo, National University of Costa Rica.A batch of five-days old individuals (wet wight=0.031±0.0014g;total length=17.42±0.58mm;mean+standard deviation) fed only once before assays was used.Breeding of fish; toxicity testing and biomarker analyses were undertaken between May and August, 2009.
For acute 96-horus assay, fish were placed into glass containers (one individual/container) that held 100mL of filtered (MILLIPORE) and UV-treated (PURA) water.After a 24hr acclimation period, a volume of ethoprophos stock solution (79.84mg a.i./L) was added to each container to obtain the nominal concentrations of each assay.The concentrations used in the assays were 0.1, 1, 10, 50, 100, 200, 300, 400, 500, 600, 700, 900 and 1500µg/L.Ten replicates were used per concentration, assays were static and during the exposure fish were not fed.After 96hr, deaths were recorded and surviving fish were retrieved, placed in 1.5mL microtubes and stored at -20°C for biomarker analyses.
The range of concentrations of ethoprophos used as treatments were estimated from preliminary assays with other native freshwater species (not included).In this study, several treatments, in a wide range of concentrations, were used in order to obtain enough treatments with surviving fish for biomarker analyses.
Whole-fish samples were used for biomarker analyses due to the small size of organisms which made difficult to obtain specific tissues.Samples were prepared according to Monteiro et al. (2007).Briefly, for homogenization, 1mL of 0.1M potassium phosphate buffer (pH 7.2) was added to each microtube.Fish were then homogenized for 10s using an ultrasonic homogenizer (Branson Sonifier 450).Samples were centrifuged at 8 160g for 5min.Supernatants were transferred to new microtubes and used for analysis of biomarkers.Samples were kept on ice during the whole process.
Protein contents of all samples were determined according to the Bradford method (Bradford 1976) adapted to microplate (Bio-Rad Laboratories Inc. 2005).Bovine g-globulin (Bio-Rad) was used as the protein standard.
ChE activity was measured using the Ellman method (Ellman et al. 1961), adapted to microplate (Guilhermino et al. 1996).Because no differentiation between acetylcholinesterase (AChE) and butyrylcholinesterase (BChE) was determined, the activity measured was considered as total ChE.Ellman assay was conducted by exposing the samples to a reaction solution (75mM acetylthiocholine and 10mM 5,5'-dithiobis-(2-nitrobenzoic acid)).Change of absorbance at 412nm was followed in a microplate reader (MultiSkan Ascent, Thermo Electron Corporation) for 15min.ChE activity was calculated for a linear period of the reaction and then expressed as U/mg, U=Nmol/min.
Glutathione S-transferase (GST) activity was determined by the method of Habig et al. (1974), based on the conjugation of reduced glutathione (GSH) with 1-chloro-2,4-dinitrobenzene.The assay was conducted in a microplate as described by Booth et al. (2000), and the reaction was monitored at 340nm during three minutes.GST activity was expressed as U/mg, U=Nmol/min 1 .
Lipid peroxidation (LPO) was measured as described by Torres et al. (2002).Level of LPO was expressed as Nmol of thiobarbituric acidreactive substances (TBARS) per milligram of protein.
SPSS statistics 17.0 software was used for data analysis.Probit regression was used to calculate the 96hr LC 50 .Differences among treatments were tested using ANOVA followed by the Tukey HSD post-hoc test.Kruskal-Wallis non-parametric test was used when samples did not meet the normality and/or homogeneity of variance.In all cases, fish were analyzed individually and differences were considered significant at a probability level of p<0.05.

RESULTS
Acute toxicity is one of the most relevant endpoints in ecotoxicological studies, and is also an important parameter to characterize the sensitivity of species to xenobiotics.Ethoprophos has a high toxic effect on A. tropicus larvae.The determined 96hr LC 50 of ethoprophos for A. tropicus was 859.7µg/L (95% confidence limits=659.3-1170.6µg/L).
Organophosphorous pesticides affect the nervous system of organisms by inhibiting the enzyme ChE.In this study, ChE was significantly inhibited (p<0.05) in A. tropicus larvae exposed to ethoprophos at a concentration of 50µg/L, value representing the lowest observed effect concentration (LOEC).At this concentration, enzyme activity was inhibited by 20% whereas at a concentration of 400µg/L of the pesticide, inhibition of the enzyme reached 79% (Fig. 1).At concentrations below 50µg/L, no ChE inhibition was observed and 10µg/L was determined as the no observed effect concentration (NOEC).
Oxidative stress and biotransformation markers are known to respond to a diverse range of environmental stressors.In this study, GST and LPO measured in whole-organism samples did not revealed a significant response (p>0.05) after exposure to ethoprophos compared to the control groups.GST activity was low in the whole-organism samples (Fig. 2A).A similar result was observed for LPO determinations (Fig. 2B).Even considering the low levels of response measured, no significant difference was observed among treatments.

DISCUSSION
Evaluation of toxicological parameters in tropical species is essential if the purpose is to characterise the impact of pollutants on the environment.The 96hr LC 50 value calculated for ethoprophos on A. tropicus demonstrated a high toxicity of the pesticide for this species according to data from the Pesticide Action Network (PAN) of North America.An inhibition of 20% in ChE activity is considered a reliable indicator to diagnose exposure to organophosphates and carbamates, whereas enzyme inhibition of over 50% can be considered as an evidence that exposure to these pesticides caused the mortality in fish kill scenarios (Peakall 1992, Thompson 1999).At 50µg/L ChE activity was inhibited by 21%, but no mortality was observed at this concentration.These results show the suitability of the biomarker as an early warning signal; although other indigenous fish species should be tested regarding their sensitivity to anti-ChE pesticides, in order to produce a reliable species sensitivity distribution for the neotropical environment.
Determinations of oxidative stress and biotransformation markers were not able to distinguish effects of exposure, even at the highest test concentration of 1 500µg/L of the pesticide.Measurements should be done in specific tissues (liver or gills), as they are the primary organs involved with the pollutants entering the organism (Giaria et al. 2008).Also, gills and liver are frequently used in studies with more consistent results for these biomarkers (Li et al. 2008, Dinu et al. 2010).In this case, the small size of the fish made it impossible to sample these organs, but it should be considered in future experiments.
Furthermore, some contradicting results has been reported about the response of GST and LPO in fish towards ChE inhibiting pesticides.In the case of organophosphates, Almeida et al. (2010) found no significant response of these biomarkers in Dicentrarchus labrax after exposure to fenitrothion (OP).On the other hand, inhibition of GST was observed in Pseudorasbora parva exposed to methomyl (carbamate) (Li et al. 2008).It also should be considered that the activities of GST and LPO vary widely between fish species (Solé et al. 2006, Sanchez et al. 2007, Azevedo et al. 2009, Oliveira et al. 2010).Nevertheless, characterization of these non-specific biomarkers is necessary if the species is going to be used in field monitoring, considering that in the environment, organisms would be exposed to unknown mixtures of pollutants.
Our findings that A. tropicus larvae are highly sensitive to ethoprophos support the use of this species as a sentinel organism for the potential effects of pesticide exposure for wetlands embedded in a matrix of intensive agriculture.Cholinesterase activity in A. Tropicus larvae also resulted to be a sensitive biomarker of exposure to organophosphates.Future studies should be conducted with organisms of a larger size in order to characterize accurately the response of biomarkers in specific organs.The possibility of using blood samples in future biomarker studies should be evaluated as a non destructive technique which would be especially important working with valuable and endangered species.Further investigation should be conducted to evaluate the sensitivity of A. tropicus to other pollutants to increase the value of the species as a tool field monitoring.

ACKNOWLEDGMENTS
Three anonymous reviewers contributed with their corrections, comments and suggestions to this article.Sascha Pfennig is grateful with the German Academic Exchange Service (DAAD) for his Ph.D. scholarship.
Figures are presented as means with their respective standard errors.

Fig. 1 .
Fig. 1.Cholinesterase activity in A. tropicus larvae exposed to ethoprophos.Activity measured in whole-organism samples.Values are expressed as Mean+standard error.