This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

Author's personal copy

Toxic effects of pure anatoxin-a on biomarkers of rainbowtrout, Oncorhynchus mykiss

Joana Osswald a, António Paulo Carvalho a,b, Laura Guimarães a,*,Lúcia Guilhermino a,c

a Interdisciplinary Centre of Marine and Environmental Research (CIIMAR/CIMAR), University of Porto, Rua dos Bragas 289,4050-123 Porto, PortugalbDepartment of Biology, Faculty of Sciences, University of Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugalc ICBAS – Institute of Biomedical Sciences of Abel Salazar, University of Porto, Department of Population Studies,Laboratory of Ecotoxicology, Rua de Jorge Viterbo Ferreira, n 228, 4050-313 Porto, Portugal

a r t i c l e i n f o

Article history:Received 3 November 2012Received in revised form 11 April 2013Accepted 16 April 2013Available online 3 May 2013

Keywords:Anatoxin-aRainbow troutBiomarkersBiotransformationNeurotransmissionEnergy metabolism

a b s t r a c t

Anatoxin-a is a neurotoxin produced by various bloom-forming cyanobacteria. Although itshows widespread occurrence and is highly toxic to rodents, its mechanisms of action andbiotransformation, and effects in fish species are still poorly understood. The main aim ofthis study was, thus, to investigate sub-lethal effects of anatoxin-a on selected biochemicalmarkers in rainbow trout fry in order to get information about the mechanisms of toxicityand biotransformation of this toxin in fish. Trout fry were administered sub-lethal doses ofanatoxin-a (0.08–0.31 mg g�1) intraperitoneally. Livers and muscle tissue were collected72 h later for quantification of key enzyme activities as biochemical markers. Enzymesassessed in muscle tissues were related to cholinergic transmission (acetylcholinesterase[AChE]), energy metabolism (lactate dehydrogenase [LDH] and NADPþ-dependent iso-citrate dehydrogenase [IDH]). Enzymes assessed in the liver were involved in biotrans-formation (ethoxyresorufin-O-deethylase [EROD] and glutathione S-transferases [GST]).The results indicated a significant increasing trend for AChE activity with the dose ofanatoxin-a, possibly representing an attempt to cope with overstimulation of muscle ac-tivity by the toxin, which competes with acetylcholine for nicotinic receptors binding.Anatoxin-a was also found to significantly induce the activities of liver EROD and GST,indicating the involvement of phase I and II biotransformation in its detoxification. Like-wise, lactate dehydrogenase activity recorded in fry muscle increased significantly withthe dose of anatoxin-a, suggesting an induction of the anaerobic pathway of energy pro-duction to deal with toxic stress induced by the exposure. Altogether, the results suggestthat under continued exposure in the wild fish may experience motor difficulties, possiblybecoming vulnerable to predators, and be at increased metabolic demand to cope withenergetic requirements imposed by anatoxin-a biotransformation mechanisms.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Cyanobacteria are ubiquitous prokaryotes that canproliferate at high densities (cyanobacterial blooms) inaquatic environments and particularly in freshwater sys-tems. The main concern about these microorganisms istheir ability to produce secondary metabolites that are

* Corresponding author. Tel.: þ351 22 340 1819; fax: þ351 22 339 0608.E-mail addresses: joanaosswald@ciimar.up.pt (J. Osswald),

apcarval@fc.up.pt (A.P. Carvalho), lguimaraes@ciimar.up.pt(L. Guimarães), lguilher@icbas.up.pt (L. Guilhermino).

Contents lists available at SciVerse ScienceDirect

Toxicon

journal homepage: www.elsevier .com/locate/ toxicon

0041-0101/$ – see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.toxicon.2013.04.014

Toxicon 70 (2013) 162–169

Author's personal copy

toxic – cyanotoxins – and may cause death in humans andin aquatic and terrestrial organisms (Azevedo et al., 2002).Cyanotoxins have been classified according to their maintarget in mammals as neurotoxins, hepatotoxins, dermo-toxins and cytotoxins. Anatoxin-a is a neurotoxin producedby various bloom-forming cyanobacteria, including Ana-baena,Oscillatoria and Aphanizomenon (Osswald et al., 2007and references herein). Despite their widespread occur-rence and high toxicity (10 min LD50 of 250 mg kg�1,following intraperitoneal injection) (Rogers et al., 2005),little attention from researchers has been paid to anatoxin-a, probably because it has seldom been implicated intoxicosis and appears to be less frequent than hepatotoxins(Codd et al., 2005). Nevertheless, some studies havedescribed high incidences of anatoxin-a in different parts ofthe globe, some of them causing animal deaths (Puschneret al., 2008). Considering the current climate change sce-nario, with the expected increases in water temperatureand higher risk of formation of toxic blooms, moreknowledge on the effects of anatoxin-a particularly inaquatic organisms is urgently needed. In this context ourteam has been studying the effects of anatoxin-a in aquaticanimals, including fish that are a crucial link in aquatic foodwebs (Osswald et al., 2007, 2008, 2009, 2011).

Anatoxin-a is an alkaloid and a potent nicotinic agonist(Thomas et al., 1993) that acts by blocking the cholinergicneurotransmission. In the last decades its mode of actionhas been described and characterized, mainly in rodents(Campos et al., 2010; Fawell et al., 1999; Spivak et al., 1980).Anatoxin-a mimes the neurotransmitter acetylcholine(ACh) present in cholinergic synapses of the peripheral andcentral nervous systems of vertebrates. In the absence ofanatoxin-a, when a nervous impulse reaches the synapsesACh is released into the synaptic cleft and binds to the re-ceptors on the effector organs. After transmission of theelectrical impulse, ACh is hydrolyzed by acetylcholines-terase (AChE) allowing for signal termination and conse-quent re-establishment of the postsynaptic restingpotential. Since anatoxin-a is not degraded by esterases andit competes with ACh for the same cholinergic receptors,when anatoxin-a is present the effector organ remains in astate of activation. In case of exposure to lethal doses theanimal ends up dying by respiratory arrest (Carmichaelet al., 1979). Because fish are vertebrates possessing nico-tinic receptors and ACh, they are expected to show a re-action to anatoxin-a similar to that of rodents. However,despite its toxicity and interference with cholinergictransmission, the mode of action, biotransformation andeffects of anatoxin-a and/or its metabolites are still poorlydescribed, both in rodents and fish species.

The aim of this work was, therefore, to investigatewhether sub-lethal doses of anatoxin-a could induce alter-ations in relevant biochemical markers in rainbow trout(Oncorhynchus mykiss), an animal model widely used inaquatic toxicology. Changes in the activity of some key en-zymeswereused as biomarkers. The selected enzymeswereAChE, lactate dehydrogenase (LDH) and NADPþ- dependentisocitrate dehydrogenase (IDH) in muscle tissue, andethoxyresorufin-O-deethylase (EROD) and glutathione S-transferases (GST) in the liver. AChE is the cholinesterasepresent in rainbow trout brain and white muscle (Sturm

et al., 2007), being involved in the cholinergic neurotrans-mission as described above. LDH and IDH are involved incellular respiration, respectively in the glycolysis and thecitric acid cycle, and their activities may be used as in-dicators of themetabolic state of organisms under chemicalor natural stress (Castro et al., 2004; Polakof et al., 2006;Almeidaet al., 2010). ERODactivity is awidelyused indicatorof xenobiotics uptake infish, providingevidenceof receptor-mediated induction of cytochrome P450-dependant mon-ooxygenases, namely the CYP1A subfamily responsible forthe metabolization of various xenobiotics in liver phase Ibiotransformation (Van der Oost et al., 2003; Smith andWilson, 2010). GST play an important role in liver phase IIbiotransformation, catalyzing the conjugation of gluta-thione with a wide variety of xenobiotics, rendering themmore soluble and facilitating their excretion (Van der Oostet al., 2003). It is expected that this set of enzymes couldprovide novel information on the toxic effects andbiotransformation pathways of anatoxin-a in fish.

2. Materials and methods

2.1. Holding of fish

One month old O. mykiss of the same parenthood werekindly provided by an aquaculture facility (Castro & Gaberoat River Coura) in the north of Portugal. They were main-tained under controlled conditions in the bioterium (BOGA-CIIMAR) for two months, in a tank with recirculation ofaerated and dechlorinated tap water at 14 �C � 1 �C. Duringthis period fish were fed ad libitum with standard com-mercial pellets (A. Coelho & Castro, Portugal). No deaths ordisease symptoms were recorded during this period ofacclimatization. Therefore, fish were concluded to behealthy and proper for experimentation. Two weeks beforestarting the experiments, 3 months old fry (average totallength and freshweight of 4.8 cmand1g, respectively)wererandomly separated in groups according to the treatmentsdescribed below (sections 2.3.1–2.3.3). Each group wasplaced in their respective floating cage, inserted in the sametank under the same conditions as in the acclimatizationperiod. Feeding was halted 24 h before the experiments.

2.2. Injection solutions

Phosphate buffer saline (PBS) (pH 7.4) was preparedusing p.a. chemicals from Sigma–Aldrich�. After adding thechemicals and confirming the pH, this solution was steril-ized by autoclave and kept at 4 �C for 24 h before use in theexperiments.

A stock solution of anatoxin-a fumarate (Tocris Biosci-ence,U.K.)waspreparedbydilution inPBS to a concentrationof 0.5 mg mL�1 of pure anatoxin-a. Final anatoxin-a solutionswere prepared within 12 h prior to injection, by diluting thesterile stock solution in PBS to thedesired test concentration.

2.3. Bioassays

2.3.1. Range-finding bioassayBased on the results of previous studies (Osswald et al.,

2011), four doses (0.005, 0.05, 0.5 and 5 mg g�1 of fish fresh

J. Osswald et al. / Toxicon 70 (2013) 162–169 163

Author's personal copy

weight (f.w.)) and one vehicle control (PBS) were admin-istered intraperitoneally to the 3 months old rainbow troutfry. For ethical reasons, in this bioassay, each treatment wasapplied only to three fish in order to diminish the numberof experimented animals. Injectionwas administered in theventral midline, midway between the pelvic and pectoralfins, in a proportion of 2 mL per g of fish f.w. A 10 mL pre-cision syringe (Hamilton, U.S.A.) was used with a 25G/5/800,0.50 � 16 mm needle. Immediately after injection, fry weretransferred to an individual aerated container for obser-vation during the first 30 min. After 30 min, they were putback in the respective cage and observed every 8 h until72 h after injection. Death and poisoning symptoms wererecorded; fish were considered dead if touching of thecaudal peduncle produced no reaction (OECD, 1992).

2.3.2. Median lethal dose and median lethal time bioassayThis assay followed the OECD guidelines for acute

toxicity testing with fish (OECD, 1992). Based on the resultsof the range-finding bioassay, median lethal dose (LD50)and median lethal time (MLT) values were determinedusing five anatoxin-a doses ranging between 0.31 and0.50 mg g�1 f.w. (0.31, 0.35, 0.39, 0.44 and 0.5 mg g�1)injected to groups of 9 individuals. Two control groups of 9individuals each were used: one control group with noinjection (ø) and another group (PBS) injected with thevehicle. Injection and observation procedures were thesame as in the range-finding bioassay (section 2.3.1.). After72 h, surviving fish were euthanized by decapitation.

2.3.3. Biomarkers bioassayBased on the results of the bioassay described in section

2.3.2, four anatoxin-a doses (0.08, 0.12, 0.20 and 0.31 mg g�1

f.w.) were tested in order to analyze biomarkers indicativeof mechanisms of toxicity and biotransformation. Fishconditions, injection and observation procedures were thesame as in the previous assays. After 72 h, fish wereeuthanized by decapitation and dissected. In this assay allfry were weighed and measured after death in order todetermine the condition factor (CF, 100� total body weight[g]/total length3 [cm]) and the hepatosomatic index (HSI,100 � liver weight [g]/body weight [g]).

2.4. Biomarkers

2.4.1. Tissue/organ collectionDissection and collection of samples were carried out on

dry ice; collected samples were immediately stored at�80 �C until further processing within two days. Two typesof samples were excised from each fry: two samples ofdorsal muscle and the whole liver. Muscle samples werestored individually. The livers of three animals from thesame treatment and the same cage were pooled togetherin order to be able to determine biotransformationbiomarkers.

2.4.2. Preparation of biological materialHomogenization of muscle and liver samples was car-

ried out on ice, using a mechanical homogenizer (YstralGmbH d-7801 Dottigen). For AChE determinations, c.a. 0.1 gof muscle tissue was homogenized in 1 mL of ice-cold

phosphate buffer (0.1 M, pH 7.2) and centrifuged at3300 g for 6 min at 4 �C. The supernatant was recoveredand enzymatic analysis proceeded in this fraction. For IDHanalysis, each tissue sample (c.a. 0.1 g) was homogenized in1 ml of ice-cold Tris–NaCl buffer (0.1 M, pH 7.2) andcentrifuged (3300 g, 6 min, 4 �C). The supernatant was thencollected and used to determine the enzymatic activity. Asimilar procedurewas followed for LDH samples, but in thiscase the samples were first submitted to three freezing(�80 �C) thaw (4 �C) cycles to lyse the cells and release LDH,and then homogenized and centrifuged (3300 g, 6 min,4 �C). Enzymatic quantifications followed immediately asdescribed in section 2.4.3.

Liver samples were homogenized in ice-cold phosphatebuffer (0.1 M, pH 7.4) at the proportion of 10 ml per gramf.w. These homogenates were centrifuged (15,000 g,15 min, 4 �C) to recover the post-mitochondrial superna-tants that were divided in aliquots and immediately storedat �80 �C. The aliquots were used to determine EROD andGST enzymatic activities as described in the followingsection 2.4.3.

2.4.3. Enzymatic analysisAll chemicals used in these analytical procedures were

purchased from Sigma–Aldrich� except when stated. Allenzymatic activities were determined at 25 �C. Spectro-photometric readings were performed in a microplatespectrophotometer Powerwave 340 with KC-Junior soft-ware, v1.6 (BioTek Instruments, Inc.) using Costar� micro-plates (ref 3370). The fluorimetric EROD assaywas run in anFP-6200 spectrofluorimeter (Jasco, Inc.). Protein content ofsupernatants was quantified by the method of Bradfordadapted to microplate using the Bio-Rad Protein Assay DyeReagent Concentrate (Bio-Rad Laboratories, Inc.). In all as-says proper blanks were prepared by substituting the vol-ume of sample by the same amount of the respectivehomogenization buffer. All sample and blank reactionswere assessed in triplicate. Except for EROD, enzyme ac-tivities will be given in nmol min�1 mg�1 of total solubleprotein; EROD activity will be expressed in pmol min�1 mgprotein�1.

AChE was analyzed by the standard spectrophoto-metric method of Ellman et al. (1961). The increase inabsorbance due to the reaction of thiocholine with 5,50-dithio-bis-2-nitrobenzoate (DTNB) was followed at412 nm, as described previously (Guilhermino et al., 1996).LDH quantification was done by the standard spectro-photometric method of Vassault (1983), by following thedecrease in absorbance due to NADH oxidation resultingfrom the conversion of pyruvate to lactate at 340, asadapted by Diamantino et al. (2001). IDH activity wasmeasured following the spectrophotometric methoddeveloped by Ellis and Goldberg (1971); the increase inabsorbance due to the reduction of NADPþ, mediated byIDH, was monitored at 340 as adapted to microplate byLima et al. (2007). EROD was determined in a cuvette byfluorimetry (lexc ¼ 530 nm; lem ¼ 585 nm) following themethod described by Burke and Mayer (1974) where theproduction of resorufin at 25 �C in the presence of 7-ethoxyresorufin and NADPH was measured. GST quantifi-cation was done using the technique described by Habig

J. Osswald et al. / Toxicon 70 (2013) 162–169164

Author's personal copy

et al. (1974) by following the conjugation of GSH with1-chloro-2,4-dinitrobenzene (CDNB) at 340 nm, asdescribed by Frasco and Guilhermino (2002).

2.5. Statistical analysis

Linear regression was used to estimate the 10 (LD10), 50(LD50), and 90% (LD90) lethal doses, as well as theirrespective 95% confidence intervals (95% CI), according tothe equations presented in Zar (1999). MLT values weredetermined by Probit analysis (Finney, 1971). Concerningthe other endpoints, the two control groups (ø and PBS)were first comparedwith theMann–Whitney U test to seekfor possible alterations induced by the vehicle treatment.Because no significant differences between the two groupswere found for any of the parameters analyzed(Supplementary Table 1), subsequent comparisons toassess the effects of anatoxin-a were all performed againstthe PBS group. Data regarding the condition indexes (CFand HSI) and biochemical parameters determined in indi-vidual fish (i.e., activities of the enzymes AChE, LDH, andIDH) were then assessed for normality of data distributionsand homogeneity of variances using the Shapiro–Wilks andLevene tests, respectively. Possible effects of anatoxin-aexposure on CF, HSI, and IDH activity were then analyzedby one-way analysis of variance (Pavlova et al., 2010).ANOVAwith linear contrasts was used to test for a possibleeffect of anatoxin-a treatment on AChE activity. Dataregarding LDH activity, for which ANOVA assumptionscould not be met even after logarithmic transformation,and EROD and GST activities (determined in pooled liversamples) were analyzed by the Jonckheere–Terpstra test formonotonic trends. A step-down approach for this non-parametric test was used to determine the No ObservedEffect Level (NOEL) and the Lowest Observed Effect Level(LOEL) as recommended by OECD (OECD, 2006). The sig-nificance level was set at p < 0.05 for all statistical testsperformed. All statistical analyses were carried out in IBMSPSS v19.0.

3. Results

In all the bioassays, the variation of pH was alwaysbelow 1 unit, oxygen concentrationwas always higher than60% of air saturation, temperature did not change morethan 2 �C. No mortality was recorded in the control treat-ments. Therefore the assays were considered valid.

3.1. Range-finding bioassay

In the range-finding assay, all fry survived in the PBScontrol and in the lowest doses of anatoxin-a tested: 0.005and 0.05 mg g�1. In the two highest doses, average survivaltime was 30 and 17 min for 0.5 and 5 mg g�1, respectively.Considering these results and the OECD guidelines (OECD,1992), it was decided to determine the LD50 in five dosesbetween 0.50 and 0.31 mg g�1 with a geometric progression(step value¼ 1.27). Therefore, the test treatments in assay Bwere 0.31, 0.35, 0.39, 0.44 and 0.5 mg.g�1 of anatoxin-a plusthe two control groups (ø and PBS).

3.2. LD50 and MLT bioassay

The MLT was calculated for each tested dose and variedbetween 28 min in the lowest dose to 13.50 min in thehighest one (Fig.1). Lethal doses were determined based onthe linear relationship found between dose and cumulativedeath (Fig. 2): LD10 ¼ 0.25 mg g�1, 95% IC ¼ (0.15–0.31);LD50¼ 0.36 mg g�1, 95% IC¼ (0.30–0.39); LD90¼ 0.48 mg g�1,95% IC ¼ (0.42–0.72).

3.3. Biomarkers bioassay

Results obtained for the physiological indexes, CF andHSI, are shown in Fig. 3A and B, respectively. No significantvariation of these parameters was found across the exper-imental groups (One-way ANOVA, CF: F(4,38) ¼ 0.83,p ¼ 0.514; HSI: F(4,38) ¼ 1.19, p ¼ 0.333).

Muscle AChE activity showed a significant linear in-crease with anatoxin-a dose (ANOVA with linear contrasts,F(1,39) ¼ 4.39, p ¼ 0.032), indicating alterations on cholin-ergic motor transmission (Fig. 4). Concerning energeticmetabolism enzymes, the Jonckeere–Terpstra test revealeda significant trend in LDH data, with a tendency for signifi-cant increase in activity with the anatoxin-a dose (Jonck-eere–Terpstra test, JT ¼ 1.74, p ¼ 0.042) (Fig. 5A), indicatingan enhancement of the anaerobic pathway of energy pro-duction. The step-down approach indicated a NOEL value of0.12 mg g�1 and a LOEL of 0.20 mg g�1, (JT ¼ 1.73, p ¼ 0.046).For IDH, although anatoxin-a exposure appeared todecrease the enzyme activity, no significant differenceswere found among experimental groups (F(4,38) ¼ 1.49,p ¼ 0.224) (Fig. 5B). The two biotransformation enzymesdescribed in Fig. 6 were significantly affected by anatoxin-aexposure, indicating detoxification through phase I andphase II processes. Jonckeere–Terpstra test revealed a sig-nificant trend in the data, with the activity of both enzymesincreasing with the dose administered (EROD: JT ¼ 2.81,p ¼ 0.002; GST: JT ¼ 2.43, p ¼ 0.007). For EROD activity, thestep-down approach allowed determining NOEL and LOELvalues of 0.20 and 0.31 mg g�1, respectively. For GST activity,the NOEL was 0.12 mg g�1, and the LOEL was 0.20 mg g�1

(JT ¼ 2.27, p ¼ 0.012).

Fig. 1. Cumulative death (%) of O. mykiss fry as a function of the dose ofanatoxin-a, recorded for 25 min after intraperitoneal injection.

J. Osswald et al. / Toxicon 70 (2013) 162–169 165

Author's personal copy

Fig. 2. Median lethal times (MLTs) for O. mykiss fry exposed to differentdoses of anatoxin-a. Bars represent the 95% confidence intervals.

Fig. 3. Average condition indexes of O. mykiss fry at 72 h after intraperito-neal injection with vehicle (PBS) and with different doses of anatoxin-a. A:condition index (CF); B: hepatosomatic index (HSI). Vertical bars correspondto the respective standard errors. Average values were obtained by analyzingnine individuals per treatment.

Fig. 4. Average activity (nmol min�1 mg protein�1) of acetylcholinesterase(AChE) enzyme in muscle of O. mykiss fry at 72 h after intraperitoneal in-jection with vehicle (PBS) and with different doses of anatoxin-a. Verticalbars correspond to the respective standard errors. Average values wereobtained by analyzing nine individuals per treatment. Asterisk indicatessignificant differences relative to PBS (p < 0.05).

Fig. 5. Average activity (nmol min�1 mg protein�1) of enzymes related toenergetic metabolism in O. mykiss fry at 72 h after intraperitoneal injectionwith vehicle (PBS) and with different doses of anatoxin-a. A: lactate dehy-drogenase (LDH); B: NADPþ-dependent isocitrate dehydrogenase (IDH).Vertical bars correspond to the respective standard errors. Average valueswere obtained by analyzing nine individuals per treatment. Asterisk in-dicates significant trend for increase in activity with the doses of anatoxin-a(Jonckheere–Terpstra test, p < 0.05).

J. Osswald et al. / Toxicon 70 (2013) 162–169166

Author's personal copy

4. Discussion

The bioassays performed in this work revealed an acutemedian lethal dose of pure anatoxin-a of 0.360 mg g�1 (i.p.)for rainbow trout fry. Up to date no data were availableconcerning lethal doses of pure anatoxin-a for fish species.The i.p. LD50 value determined in the present work issimilar to the LD50 previously reported for mice exposed toanatoxin-a: 0.250 mg g�1 (form not reported, Rogers et al.,2005), 0.375 mg g�1 (form not reported, Fitzgeorge et al.,1994), and 0.386 mg g�1 ((þ) isoform of anatoxin-a,Valentine et al., 1991). It is, however, lower than that re-ported by Valentine et al. (1991) for mice exposed to theracemic mixture of anatoxin-a. The (þ) isoform is the nat-ural form produced by cyanobacteria. Interestingly, thisisoform was shown to be at least 2.5–6 times more potentthan the racemic mixture not only regarding lethality, butalso motor, cardiac, and respiratory activity (Adeyemo andSirén, 1992; MacPhail et al., 2007; Valentine et al., 1991). In

contrast the (�) isomer was shown to be much less potentthan either the (þ) isoform or the racemic mixture(Adeyemo and Sirén, 1992; Valentine et al., 1991). Becausethe racemic mixture was used in the present study, theresults obtained appear to suggest that the natural form ofanatoxin-a may be more toxic to fish than to rodents.

The low MLT values reported in this study are in goodagreement with the known high toxicity of anatoxin-a tovertebrates. To the best of our knowledge, no MLT value ofpure anatoxin-a has been reported previously for any otherorganism. Previous work with carp exposed to toxiclyophilized cells (Osswald et al., 2007) found an MLT of26:39 h. This represents a much slower effect whencompared to MLT values found in the present assay. How-ever, the exposure route, the toxic agent (toxic cells vs. puretoxin) and the test species employed in both studies weredifferent, preventing important inferences. Nevertheless,knowledge of MLT values obtained under defined experi-mental conditions is important for future investigations onfurther sub-lethal effects of anatoxin-a aiming at under-standing its mode of action in fish.

After 72 h of exposure to single doses of anatoxin-a nochanges were observed in either of the physiological con-dition indices assessed, relative to control fish. CF and HSIhave been largely used to monitor changes in fish health.Environmental stressmay cause alterations in these indicesdepending on pollutant exposure, abiotic conditions, sea-son, food availability, and physiological status of the or-ganisms. On the other hand, several field and laboratorystudies reported increased or decreased HSI values inresponse to contaminant exposure (reviewed in Van derOost et al., 2003). For example, according to Van der Oostet al. (2003), 38% of 24 laboratory studies showed signifi-cant increase of the HSI, and none showed significant CFalterations, upon exposure to various environmental con-taminants. In a controlled assay as in the present work,variation of abiotic and biotic factors is kept to a minimum.Hence, variation in the physiological condition indicesshould mainly be due to the exposure to the toxic sub-stance. However, in this assay, the observation period wasprobably too short to reveal putative alterations in the CFand the HSI. The induction of phase I and phase IIbiotransformation enzymes found in the exposed fry wereexpected to increase energy requirements in order to dealwith liver detoxification of anatoxin-a. This appears to befurther supported by the LDH induction found in the frymuscle. Nevertheless, these indices correspond to higher-level responses that are usually observed well afterlower-level effects (e.g., on biochemical biomarkers) havebeen detected.

Comparing to PBS treatment, a steady dose-dependentincrease in AChE activity was found in rainbow trout fryexposed to anatoxin-a. This increment could indicate aninduction of AChE synthesis triggered by anatoxin-a stim-ulation of muscle activity. In support of this hypothesis,studies in rats have shown that some AChE variants can beover expressed not only after external muscle fiber stimu-lation and physiological stress, but also by exposure toanticholinesterasic agents and acetylcholine analogs(Pregelj et al., 2007; Soreq and Seidman, 2001). Sinceknowledge about AChE function is still incomplete

Fig. 6. Average activity of biotransformation enzymes in the liver of O.mykiss fry at 72 h after intraperitoneal injection with vehicle (PBS) and withdifferent doses of anatoxin-a. A: glutathione S-transferases (GST) innmol min�1 mg protein�1; B: ethoxyresorufin-O-deethylase (EROD) inpmol min�1 mg protein�1. Vertical bars correspond to the respective stan-dard errors. Average values were obtained by analyzing three pools of threelivers per treatment. Asterisk indicates significant trend for increase in ac-tivity with the doses of anatoxin-a (Jonckheere–Terpstra test, p < 0.05).

J. Osswald et al. / Toxicon 70 (2013) 162–169 167

Author's personal copy

(Anglade and Larabi-Godinot, 2010; Soreq and Seidman,2001) it could be hypothesized that in real scenarios,when anatoxin-a is present at sub-lethal levels, AChEoverproduction could further contribute to alteration ofmuscle and neurological functions, plus other less well-characterized effects, with harmful consequences for theindividual or the community. In fact, actual available in-formation point out a role of AChE on several other bio-logical functions, such as neuritogenesis, cell adhesion,synaptogenesis, haematopoiesis and thrombopoiesis(Soreq and Seidman, 2001). Future studies using moleculartools, enabling real time quantification of mRNA expression(i.e., RT-qPCR) of AChE gene(s) after exposure to anatoxin-awill contribute to clarify this issue.

Because anatoxin-a is a neurotoxin that causes hyper-activity, muscle contraction and fasciculation, followed byconvulsions and respiratory arrest, in various vertebrates,including fish (Carmichael et al., 1975; Osswald et al., 2007;Rogers et al., 2005), we analyzed two enzymes related tothe muscle energy production pathways, LDH and IDH.Studying their activity is relevant to better understand howintoxicated organisms handle energy homeostasis re-quirements after a toxic challenge caused by anatoxin-a.Our results showed an increase in LDH activity in intoxi-cated animals, suggesting an effort to get additional energyto cope with the toxic stress caused by anatoxin-aexposure.

In this work, EROD activity showed a significant trendto increase with the dose of anatoxin-a administered, witha 48% induction found in fry exposed to the highest testtreatment. These results indicate that anatoxin-a ismetabolized in liver phase I biotransformation. EROD isresponsible for the biotransformation of various xenobi-otics in the fish liver, as for example polycyclic aromatichydrocarbons (PAH) and dioxin-like compounds, amongothers (Van der Oost et al., 2003; Smith and Wilson,2010). Also, this enzyme is the most dominant in oxida-tion processes involved in phase I biotransformation,sometimes leading to the conversion of certain substancesto more toxic intermediates such as epoxides (Schlenket al., 2008). In nature, anatoxin-a can undergo epoxida-tion and hydroxylation reactions caused by high pH andsunlight, being transformed to non-neurotoxic products(James, 1998). As far as we know there are no studiesevaluating other types of toxicity of these putativebiotransformation derivatives but we understand that itwould be important.

GST activity significantly increased with the dose ofanatoxin-a, suggesting its involvement in the biotransfor-mation of this neurotoxin. Relative to PBS controls, fryexposed to 0.2 and 0.31 mg g�1 showed a GST induction of86 and 60%, respectively. A previous study investigatinganatoxin-a effects in plants also found induction of GSTactivity after exposure of aquatic plants to 25 mg mL�1

during four days (Mitrovic et al., 2004). The GST inductionfound in our work could be due to the conjugation of themetabolite(s) formed after EROD action with GSH, and/orthe direct binding of part of the parental compound toglutathione, both reactions catalyzed by GST. To betterunderstand this, future studies should also focus onunraveling the mechanisms of biotransformation of this

toxin and its metabolites in the trout. The slight decrease ofGST activity recorded in fry treated with the highest testeddosemight result from a breakdown of GST function at highdoses. This pattern of variation is in agreement with a bell-shaped trend of the response, which usually shows aninitial activation of enzyme synthesis with increasingcontaminant exposure, thereafter decreasing progressivelydue to the enhanced catabolic rate and/or a direct inhibi-tory action of the toxicant on the enzyme molecules(Viarengo et al., 2007). Previous studies have found thisbell-shaped pattern of GST activity in other fish speciesexposed to mercury (Elia et al., 2003).

To our knowledge, this is the first study addressingenergy metabolism and biotransformation effects ofanatoxin-a in animal species. In summary, short-term i.p.exposure to anatoxin-a elicited increased AChE activity infry trout, and enhanced anaerobic metabolism and liverphase I and II biotransformation. These results raiseconcern about the effects of anatoxin-a under conditions ofchronic exposure in the wild. They suggest that in such asituation fish may experience motor difficulties, possiblybecoming vulnerable to predators, and be at increasedmetabolic demand to cope with energetic requirementsimposed by anatoxin-a detoxication mechanisms.

Acknowledgments

We thank colleagues Joana Almeida, Ph.D., Luís Vieira,Ph.D., and Carlos Gravato, Ph.D., for their kind technicaladvising in the enzymatic determinations. We also thankProfessor Vitor Vasconcelos and Professor Walter Osswaldfor their constructive review of the manuscript. Thisresearch was partially supported by the European RegionalDevelopment Fund (ERDF) through the COMPETE – Oper-ational Competitiveness Program and national fundsthrough FCT – Foundation for Science and Technology,under the project “PEst-C/MAR/LA0015/2011”. The workwas also supported by the Portuguese Foundation for Sci-ence and Technology (FCT), program QREN-POPH/FSE[SFRH/BPD/37804/2007].

Ethical statement

All fish experiments were performed in accordancewiththe current guidelines for the care of laboratory animalsand ethical guidelines for investigation in conscious ani-mals set by the General Directorate of Veterinary ofPortugal, and approved by the institutional committee. Allefforts were made to minimize the number of animals usedand their suffering.

Conflict of interest statement

The authors declare that there are no conflicts ofinterest.

Appendix A. Supplementary material

Supplementary material related to this article can befound at http://dx.doi.org/10.1016/j.toxicon.2013.04.014.

J. Osswald et al. / Toxicon 70 (2013) 162–169168

Author's personal copy

References

Adeyemo, O.M., Sirén, A.L., 1992. Cardio-respiratory changes and mor-tality in the conscious rat induced by (þ)- and (þ/�)-anatoxin-a.Toxicon 30, 899–905.

Almeida, J.R., Oliveira, C., Gravato, C., Guilhermino, L., 2010. Linkingbehavioural alterations with biomarkers responses in the Europeanseabass Dicentrarchus labrax L. exposed to the organophosphatepesticide fenitrothion. Ecotoxicology 19, 1369–1381.

Anglade, P., Larabi-Godinot, Y., 2010. Historical landmarks in the histo-chemistry of the cholinergic synapse: perspectives for future re-searches. Biomed. Res. 31, 1–12.

Azevedo, S., Carmichael, W., Jochimsen, E., Rinehart, K., Lau, S., Shaw, G.,Eaglesham, G., 2002. Human intoxication by microcystins duringrenal dialysis treatment in Caruaru–Brazil. Toxicol 181–182, 441–446.

Burke, M.D., Mayer, R.T., 1974. Ethoxyresorufin: direct fluorimetric assayof a microsomal-O-deethylation which is preferentially inducible by3- methylcholantrene. Drugs Metab. Dispos. 2, 583–588.

Campos, F., Alfonso, M., Durán, R., 2010. In vivo modulation of [alpha]7nicotinic receptors on striatal glutamate release induced by anatoxin-A. Neurochem. Int. 56, 850–855.

Carmichael, W.W., Biggs, D.F., Gorham, P.R., 1975. Toxicology andpharmacological action of Anabaena flos-aquae toxin. Science 187,542–544.

Carmichael, W.W., Biggs, D.F., Peterson, M.A., 1979. Pharmacology ofanatoxin-a, produced by the freshwater cyanophyte Anabaena flos-aquae NRC-44-1. Toxicon 17, 229–236.

Castro, B.B., Sobral, O., Guilhermino, L., Ribeiro, R., 2004. An in situbioassay integrating individual and biochemical responses usingsmall fish species. Ecotoxicology 13, 667–681.

Codd, G.A., Morrison, L.F., Metcalf, J.S., 2005. Cyanobacterial toxins: riskmanagement for health protection. Toxicol. Appl. Pharmacol. 203,264–272.

Diamantino, T.C., Almeida, E., Soares, A.M.V.M., Guilhermino, L., 2001.Lactate dehydrogenase activity as an effect criterion in toxicity testswith Daphnia magna straus. Chemosphere 45, 553–560.

Elia, A.C., Galarini, R., Taticchi, M.I., Dorr, A.J., Mantilacci, L., 2003. Anti-oxidant responses and bioaccumulation in Ictalurus melas undermercury exposure. Ecotoxicol. Environ. Saf. 55, 162–167.

Ellis, G., Goldberg, D.M., 1971. An improved manual and semiautomaticassay for NADP-dependent isocitrate dehydrogenase activity, with adescription of some kinetic properties of human liver and serumenzyme. Clin. Biochem. 2, 175–185.

Ellman, G.L., Courtney, K.D., Andres Jr., V., Featherstone, R.M., 1961. A newand rapid colorimetric determination of acetylcholinesterase activity.Biochem. Pharmacol. 7, 88–95.

Fawell, J.K., Mitchell, R.E., Hill, R.E., Everett, D.J., 1999. The toxicity ofcyanobacterial toxins in the mouse: II Anatoxin-a. Hum. Exp. Toxicol.18, 168–173.

Finney, D., 1971. Probit Analysis. Cambridge University Press, Cambridge.Fitzgeorge, R.B., Clark, S.A., Keevil, C.W., 1994. Routes of intoxication. In:

Codd, G.A., Jefferies, T.M., Keevil, C.W., Potter, E. (Eds.), DetectionMethods for Cyanobacterial Toxins. The Royal Society of Chemistry,London.

Frasco, M.F., Guilhermino, L., 2002. Effects of dimethoate and beta-naphthoflavone on selected biomarkers of Poecilia reticulata. Fish.Physiol. Biochem. 26, 149–156.

Guilhermino, L., Lopes, M., Carvalho, A., Soares, A., 1996. Acetylcholines-terase activity in juveniles of Daphnia magna Straus. Bull. Environ.Contam. Toxicol. 57, 979–985.

Habig, W.H., Pabst, M.J., Jakoby, W.B., 1974. Glutathione-S-transferases,the first enzymatic step in mercapturic acid formation. J. Biol. Chem.249, 7130–7139.

James, K., 1998. Sensitive determination of anatoxin-a, homoanatoxin-aand their degradation products by liquid chromatography withfluorimetric detection. J. Chromatogr. A 798, 147–157.

Lima, I., Moreira, S.M., Osten, J.R.V., Soares, A., Guilhermino, L., 2007.Biochemical responses of the marine mussel Mytilus galloprovincialisto petrochemical environmental contamination along the North-western coast of Portugal. Chemosphere 66, 1230–1242.

MacPhail, R.C., Farmer, J.D., Jarema, K.A., 2007. Effects of acute and weeklyepisodic exposures to anatoxin-a on the motor activity of rats:comparison with nicotine. Toxicology 234, 83–89.

Mitrovic, S.M., Pflugmacher, S., James, K.J., Furey, A., 2004. Anatoxin-aelicits an increase in peroxidase and glutathione S-transferase activityin aquatic plants. Aquat. Toxicol. 68, 185–192.

OECD, 1992. OECD Guideline for Testing of Chemicals. Fish, Acute ToxicityTest. OECD Environment Directorate, Paris.

OECD, 2006. Current Approaches in the Statistical Analysis of Ecotoxi-cology Data: A Guidance to Application. O.E.H.A.S. Publications, OECDEnvironment Directorate, Paris.

Osswald, J., Rellán, S., Carvalho, A., Gago, A., Vasconcelos, V., 2007. Acuteeffects of an anatoxin-a producing cyanobacterium on juvenile fish–Cyprinus carpio L. Toxicon 49, 693–698.

Osswald, J., Rellan, S., Gago, A., Vasconcelos, V., 2008. Uptake and depu-ration of anatoxin-a by the mussel Mytilus galloprovincialis (Lamarck,1819) under laboratory conditions. Chemosphere 72, 1235–1241.

Osswald, J., Carvalho, A.P., Claro, J., Vasconcelos, V., 2009. Effects ofcyanobacterial extracts containing anatoxin-a and of pure anatoxin-a on early developmental stages of carp. Ecotoxicol. Environ. Saf.72, 473–478.

Osswald, J., Azevedo, J., Vasconcelos, V., Guilhermino, L., 2011. Experi-mental determination of the bioconcentration factors for anatoxin-ain juvenile rainbow trout (Oncorhynchus mykiss). Proc. Int. Acad.Ecol. Environ. Sci. 1, 77–86.

Pavlova, V., Furnadzhieva, S., Rose, J., Andreeva, R., Bratanova, Z.,Nayak, A., 2010. Effect of temperature and light intensity on thegrowth, chlorophyll-a, concentration and microcystin production byMicrocystis aeruginosa. Gen. Appl. Plant Physiol. 36, 148–158.

Polakof, S., Arjona, F.J., Sangiao-Alvarellos, S., Martín del Río, M.P.,Mancera, J.M., Soengas, J.L., 2006. Food deprivation alters osmoreg-ulatory and metabolic responses to salinity acclimation in giltheadsea bream Sparus auratus. J. Comp. Physiol. B 176, 441–452.

Pregelj, P., Trinkaus, M., Zupan, D., Trontelj, J.J., Sketelj, J., 2007. The role ofmuscle activation pattern and calcineurin in acetylcholinesteraseregulation in rat skeletal muscles. J. Neurosci. 27, 1106–1113.

Puschner, B., Hoff, B., Tor, E.R., 2008. Diagnosis of anatoxin-a poisoning indogs from North America. J. Vet. Diagn. Invest. 20, 89–92.

Rogers, E.H., Hunter III, E.S., Moser, V.C., Phillips, P.M., Herkovitz, J.,Muñoz, L., Hall, L.L., Chernoff, N., 2005. Potential developmentaltoxicity of anatoxin-a, a cyanobacterial toxin. J. Appl. Toxicol. 25,527–534.

Schlenk, D., Celander, M., Gallagher, E., George, S., James, M.,Kullman, S.W., Hurk, P.V.D., Willet, K., 2008. Biotransformation infishes. In: Giulio, R.T., Hinton, D.E. (Eds.), The Toxicology of Fishes. CRCPress, Taylor & Francis Group, Boca Raton, pp. 153–234.

Smith, E.M., Wilson, J.Y., 2010. Assessment of cytochrome P450 fluoro-metric substrates with rainbow trout and killifish exposed to dexa-methasone, pregnenolone-16[alpha]-carbonitrile, rifampicin, and[beta]-naphthoflavone. Aquat. Toxicol. 97, 324–333.

Soreq, H., Seidman, S., 2001. Acetylcholinesterase – new roles for an oldactor. Nat. Rev. Neurosci. 2, 294–302.

Spivak, C.E., Wiktop, B., Albuquerque, E.X., 1980. Anatoxin-a: a novel,potent Agonist at the nicotinic receptor. Mol. Pharmacol. 18, 384–394.

Sturm, A., Radau, T.S., Hahn, T., Schulz, R., 2007. Inhibition of rainbow troutacetylcholinesterase by aqueous and suspended particle-associatedorganophosphorous insecticides. Chemosphere 68, 605–612.

Thomas, P., Stephens, M., Wilkie, G., Amar, M., Lunt, G., Whiting, P.,Gallagher, T., Pereira, E., Alkondon, M., Albuquerque, E., 1993.(þ)-Anatoxin-a is a potent agonist at neuronal nicotinic acetylcholinereceptors. J. Neurochem. 60, 2308–2311.

Valentine, W.M., Schaeffer, D.J., Beasley, V.R., 1991. Electromyographicassessment of the neuromuscular blockade produced in vivo byanatoxin-a in the rat. Toxicon 29, 347–357.

Vassault, A., 1983. Lactate dehydrogenase. In: Bergmeyer, H.O. (Ed.),Methods of Enzymatic Analysis. Enzymes: Oxireductases, Trans-ferases. Academic Press, New York, pp. 118–126.

Van der Oost, R., Beyer, J., Vermeulen, N.P.E., 2003. Fish bioaccumulationand biomarkers in environmental risk assessment: a review. Environ.Toxicol. Pharmacol. 13, 57–149.

Viarengo, A., Lowe, D., Bolognesi, C., Fabbri, E., Koehler, A., 2007. The useof biomarkers in biomonitoring: a 2-tier approach assessing the levelof pollutant-induced stress syndrome in sentinel organisms. Comp.Biochem. Physiol. C Toxicol. Pharmacol. 146, 281–300.

Zar, J., 1999. Biostatistical Analysis, fourth ed. Prentice Hall, Inc., UpperSaddle River.

J. Osswald et al. / Toxicon 70 (2013) 162–169 169