gy 28 (2006) 245–250www.elsevier.com/locate/neutera

Neurotoxicology and Teratolo

Antipsychotics produce locomotor impairment in larval zebrafish

Nicholas J. Giacomini 1, Brian Rose 1, Kayta Kobayashi 1, Su Guo ⁎

Department of Biopharmaceutical Sciences, Programs in Human Genetics and Biological Sciences, University of California, San Francisco,California, 94143-0446, USA

Received 22 August 2005; received in revised form 31 December 2005; accepted 26 January 2006Available online 9 March 2006

Abstract

Zebrafish has been a favored vertebrate genetic model organism for studying developmental processes. It also holds a great potential forunderstanding the genetic basis of behavior and associated behavioral disorders. Despite such potential, their use in the study of behavior is greatlyunder-explored. It is well known that multiple classes of drugs used to treat psychiatric diseases produce extrapyramidal side (EPS) effects andconsequent movement disorders in humans. The underlying molecular causes of these drug-induced movement disorders are poorly understood.Here we report that zebrafish treated with the antipsychotics fluphenazine and haloperidol (both of which can induce severe EPS in humans)develop movement defects. In contrast, another antipsychotic olanzapine, which produces mild to little EPS in humans, leads to minimalmovement defects in zebrafish. These results establish a rapid assay system in which the effects of EPS-inducing agents can be assessed. Thus,future genetic screening in zebrafish shall identify genes and pathways that elucidate drug-induced movement disorder in human as well asprovide insights into the brain control of locomotor activity. Future chemical screening in zebrafish may act as a preclinical test for the EPS effectof certain drugs, as well as a test used to researching drugs made to counteract the effects of EPS.© 2006 Elsevier Inc. All rights reserved.

Keywords: Drug-induced movement disorders; Extrapyramidal side effects; Antipsychotics; Dopamine; Locomotor behavior; Zebrafish; Genetics; Model organism

1. Introduction

The execution of any complex behavior in an animalinevitably requires regulation of the locomotor centers in thebrain. The importance of the brain in regulating locomotoractivity is well highlighted by the fact that destruction of braindopaminergic (DA) neurons causes Parkinson's disease, themost common movement disorder characterized by bradykine-sia, tremor, rigidity, and postural imbalance [16]. The projectionof DA neurons to the striatum is important for regulating motoroutput. How this pathway is regulated and how it regulatesmovement are largely unclear. This is in part due to the fact thatlocomotor activity can be complicated by many factors such asmuscle integrity, making it difficult to find the brain under-pinnings through directly assessing movement in a geneticallytractable animal [15].

⁎ Corresponding author. Tel.: +1 415 502 4949; fax: +1 415 502 8177.E-mail address: suguo@itsa.ucsf.edu (S. Guo).

1 These authors contributed equally to the work.

0892-0362/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.ntt.2006.01.013

It has been well established that both antipsychotics and anti-depressants can have extrapyramidal side effects (EPS) leadingto movement disorders in individuals who are treated with thesemedications [6,12]. These antipsychotics include haloperidoland fluphenazine, both of which are dopaminergic antagoniststhat primarily inhibit D2-family of receptors; the anti-depres-sants include the selective serotonin-reuptake inhibitors (SSRI)[18]. The EPS side effects of these drugs usually develop within1 month of the initiation of the offending medication inapproximately 60% patients and in approximately 90% within 3months [20]. The likely risk factors include prior history ofmovement defects, age, gender, and genetically determineddifferences in drug metabolism and possibly drug action [20]. Itis important to point out that the development of these drug-induced movement disorders is not restricted to aged indivi-duals and young people are equally susceptible; in addition, theconditions are generally reversible once the medications areremoved, suggesting that the drugs produce an interference withneuron function rather than killing the neurons.

Similar to our understanding of brain control of locomotoractivity, the molecular and cellular mechanisms underlying

246 N.J. Giacomini et al. / Neurotoxicology and Teratology 28 (2006) 245–250

drug-induced movement disorders are largely unclear. On-targettoxicity to undesired neural circuitry (e.g. blockade of striataldopamine D2 receptors) has been proposed [17]. A recent studysuggests that the alternatively spliced dopamine D2S and D2Lreceptors may differentially contribute to the actions ofantipsychotic and psychotic agents in mice [21]. Alternatively,the EPS side effects may be due to off-target toxicity to yetunknown molecular receptors. Association of dopamine D3receptor gene variants with neuroleptic-induced akathisia hasbeen reported [5]. Taken together, genetic factors are implicatedin mediating EPS effects of these antipsychotic agents, but themolecular identity of the majority of these factors remains to bedetermined.

If EPS-causing antipsychotics or anti-depressants can inducemovement defects in a genetically tractable animal, subsequentforward genetic analyses will allow the identification of genesand pathways that mediate locomotor impairment by anti-psychotics. These analyses will also shed important light on thebrain control of locomotor activity: for instance, they provide ameans to identify mutants that move normally in the presence ofthe drug, thus circumventing the lack-of-specificity problemencountered when screening directly for movement-defectivemutants. With these goals in mind, we set out to determinewhether movement defects can be induced with antipsychoticsin the zebrafish Danio rerio. Being a vertebrate, zebrafish shareextensive similarity to humans in cellular structure, organphysiology, and genetic blueprint, and have been used to modelhuman cancer and cardiovascular disorders [4,9,19]. Neurotox-in-induced Parkinson's disease models have also been de-scribed in zebrafish [1,3,13]. Most water-soluble compoundsdirectly administered in the tank water can be easily taken upinside zebrafish, making it feasible to carry out large-scalescreening [8]. The early development of zebrafish is rapid: inmerely five-day post-fertilization, zebrafish already possessmany patterns of behavior including free swimming, foodhunting, and escape from predators [7]. Their nervous system isalso well developed; for example, the brain DA neurons can bedetected as early as 24 h post-fertilization [10].

Here we report that upon treatment with the antipsychoticsfluphenazine and haloperidol that induce severe EPS in humans,larval zebrafish displayed movement defects that include bothreduced swimming speed and erratic movements. In contrast,treatment with olanzapine that has mild to minimal EPS effect inhumans only mildly reduced larval swimming speed and causedlittle erratic movements. These findings, together with theamenability of zebrafish to large-scale genetic and chemicalscreening, shall permit identification of genes and pathwaysinvolved in drug-induced movement disorders and the braincontrol of locomotor activity, as well as surveying a broad arrayof chemical compounds to identify those devoid of EPS.

2. Methods

2.1. Zebrafish maintenance and husbandry

Zebrafish were housed in a fish facility at the University ofCalifornia, San Francisco. ∼30 pairs of healthy adult

zebrafish were used for the study. For mating, a pair ofadult zebrafish was placed in a breeding cage overnight. Thefertilized eggs were then placed in petri dishes filled with eggwater (0.2 g Instant Ocean salts, 0.12 g CaSO4/l water).Instant Ocean salts was purchased from Aquatic Ecosystems,Apopka, FL, USA. Eggs were kept at 28 °C. Procedures forthe study of zebrafish conform to recommended UCSFAnimal Use guidelines.

2.2. Drug treatment

Two sets of experiments were undertaken to investigatethe effects of antipsychotics on larval zebrafish: A) Healthythirteen- or fourteen-day old larval zebrafish (also known asfry) derived from the laboratory strain (AB) were used. Weused 10 larvae in a group because less variability in basallocomotor activity was observed in this setting, aspreviously reported [14]. Groups of 10 larval zebrafishwere placed in the following solutions: 1) Fry water (100 gInstant Ocean/1 L distilled H2O) only (Control). 2) Differentconcentration of fluphenazine ranging from 0.1 to 50 μM(Sigma) dissolved in fry water for 40–45 h. 3) L-dopadissolved in fry water (31.6 mM) (Sigma) for 24 h. 4) Afterexposure to fluphenazine for 40–45 h, larval zebrafish weretransferred to fry water for 24 h. 5) After exposure tofluphenazine for 40–45 h, larval zebrafish were transferredto L-dopa (31.6 mM) for 24 h. Drug concentrations weredetermined base on knowledge of doses used in humans aswell as our experimental trials: the blood concentration offluphenazine in treated human patients is ∼4.5 μM, and thedoses of other two antipsychotics are in similar range tofluphenazine [2,11]; the daily dose of L-dopa compared tofluphenazine is ∼60 :1 in humans [11]. A minimum of 10groups was examined for each treatment condition. B) Totest whether responses to antipsychotics can be observed in7-day old larval zebrafish after shorter exposure time (forthe purpose of high throughput screening), 7-day old frywere exposed to the following solutions: 1) Fry water only(control). 2) 0.01% DMSO only (control, used to dissolvehaloperidol and olanzapine). 3) Fluphenazine dissolved inthe fry water. 4) Haloperidol dissolved in 0.01% DMSO andfry water. 5) Olanzapine dissolved in 0.01% DMSO and frywater. The treatment time in this experiment was 2 h. Aminimum of 20 groups was examined for each treatmentcondition.

2.3. Measuring locomotor activity and erratic movements inzebrafish

Quantitative determination of the locomotor activity of larvalzebrafish was as previously described [3]. Briefly, they wereplaced in a view tray on top of the light necessary for successfulvideo recording, allowed to habituate for 5 min, and theirlocomotor activity was recorded for 5 min using a digital videocamera linked to a computer. The data were analyzed withDigital Imaging Analysis Software (DIAS) (Solltech, Ohio),and followed with Microsoft Excel.

247N.J. Giacomini et al. / Neurotoxicology and Teratology 28 (2006) 245–250

To observe erratic movements (defined as rapid bouts ofvertical or sideway swimming, since larval zebrafish normallyswim in a horizontal zig-zag pattern), fish were observedimmediately after the administration of drugs at 5-min intervalsfor the first hour, and 15-min intervals after that. Each groupwas observed for 5 s, and the number of vertical or sidewaybouts of swimming were recorded.

2.4. Statistical analysis

We used SPSS 11.5 for Windows to conduct statisticalanalyses. In cases of multiple comparisons we first performedANOVA and subsequently conducted Dunnett's T3 pairwisecomparison post hoc test to determine significant differencesbetween individual treatments. The Dunnett's T3 post hoc testwas used because it is robust to the analysis of multiple unequalvariances, so that all treatments can be compared to the basallevel of locomotor activity.

3. Results

3.1. The antipsychotic fluphenazine impairs locomotor activityof larval zebrafish

Fluphenazine, a commonly used water-soluble antipsy-chotic, was tested on two-week old larval zebrafish. Aftertesting a range of concentrations, we found that 0.98 and1.57 μM of fluphenazine produced significant impairment ofswimming speed in these fish after 40–45 h exposure (Fig. 1)(P<0.001, n=20 groups), as compared to the controls thatwere identically handled except without the addition of thedrug. Their locomotion was recoverable upon the removal ofthe drug (Fig. 1). This observation suggests that theantipsychotic fluphenazine impairs locomotor activity oflarval zebrafish.

Fig. 1. The antipsychotic fluphenazine impairs locomotor activity of larval zebrafiswimming speed of larval zebrafish (P<0.001, compared to control). L-dopa aloneTransfer of larval zebrafish to L-dopa following fluphenazine treatment led to a signifwater only (#P<0.05). However, the swimming speed of both groups (recovered in

3.2. Levo-dopa treatment facilitates recovery of locomotorimpairment induced by fluphenazine

As described above, although zebrafish could spontaneouslyrecover from locomotor impairment as humans do afterstopping intake of the offending antipsychotics, we wonderedwhether addition of levo-dopa might facilitate the recovery.Since L-dopa has been successfully used to reverse thesymptoms of human Parkinsonistic patients, and the antipsy-chotic fluphenazine is known to interfere with dopaminesignaling in humans, a L-dopa facilitated locomotor recoveryfrom fluphenazine would suggest a similar dopaminergicimpairment by fluphenazine in larval zebrafish. Therefore, weadministered levo-dopa to post-fluphenazine-treated animals.As shown in Fig. 1, L-dopa treatment alone did not significantlyincrease locomotor activity (P>0.5, n=20). However, therecovery of swimming speed from fluphenazine treatment wassignificantly facilitated by the administration of L-dopa(P<0.02, n=20). This analysis suggests that the reduction oflocomotor activity by fluphenazine in larval zebrafish is likelydue to its interference with dopamine signaling.

3.3. Different antipsychotics differentially reduce the swimmingspeed in larval zebrafish

We next asked whether the locomotor impairment is ageneral feature of antipsychotics that induce movement defectsin humans. To this end, we first developed a high throughputassay for examining locomotor impairment caused by anti-psychotics. We used 7-day old larval zebrafish that do notrequire feeding and increased the drug concentration whilereducing the time of treatment. We found that 7-day old larvalzebrafish displayed significantly reduced swimming speed upontreatment with 9 μM fluphenazine for ∼ 2 h (Fig. 2). Thelocomotor impairment was reversible upon removal of the drug

sh. Treatment with fluphenazine (0.98 and 1.57 μM) significantly reduced thedid not significantly alter the swimming speed (P>0.5, compared to control).icant increase of swimming speed compared to larval zebrafish transferred to fryfry water or the L-dopa) was significantly lower than control (⁎P<0.05).

Fig. 2. Multiple different antipsychotics impair locomotor activity of larval zebrafish. The locomotor activity of fish treated with fry water, DMSO, olanzapine,fluphenazine, and haloperidol. Whereas olanzapine mildly affected the locomotor activity (⁎P<0.05), both fluphenazine and haloperidol more dramatically reducedlocomotor activity (⁎P<0.001). Compared to the effect of olanzapine, the effects of fluphenazine and Haloperidol are significantly more severe (#P<0.05).

248 N.J. Giacomini et al. / Neurotoxicology and Teratology 28 (2006) 245–250

(data not shown). We next determined the effects of 9 μMhaloperidol and 9 μM olanzapine on 7-day old larval zebrafish.Haloperidol, similar to fluphenazine, leads to severe movementdefects in humans, whereas olanzapine has mild to minimalsuch side effect when similar doses were administered [11]. Asshown in Fig. 2, whereas zebrafish treated with fluphenazineand haloperidol moved significantly slower than those treatedwith only fry water or DMSO (used to dissolve haloperidol andolanzapine), the effects of olanapine were significantly milder.This is consistent with the fact that olanzapine induces a rathermild movement defects in humans.

3.4. Fluphenazine and haloperidol induce erratic movementsin larval zebrafish

We noted other movement defects prior to the reduction ofswim speed in larval zebrafish exposed to antipsychotics.Antipsychotic-treated zebrafish exhibited unusual body posi-tioning during swimming (Fig. 3): normal zebrafish maintain aposition in parallel to the surface of the water duringmovement, whereas most zebrafish treated with antipsychoticswere not able to maintain such a position, indicative of a lackof postural balance. In addition, antipsychotic-treated zebrafishexhibited erratic swimming patterns, manifested by bouts ofvertical swimming or sideway swimming, suggesting aproblem with coordination (Fig. 3A). We quantified thisbehavior by counting the number of larval zebrafish exhibitingvertical and sideway swimming patterns during a 5-sobservation at different times of treatment. As shown in Fig.3B, we observed a significantly more erratic movements at 30and 60 min following the drug treatment. This quantitativeanalysis showed that only the fluphenazine- and haloperidol-

treated fish exhibited significant erratic movements, whereasolanzapine had minimal such effect (Fig. 3B), consistent withthe fact that it only mildly affected the locomotor activity oflarval zebrafish (Fig. 2). Taken together, we conclude thatsimilar to the findings in humans, fluphenazine and haloperidolcaused strong movement defects, whereas olanzapine hadminimal such effects in larval zebrafish.

4. Discussion

We report in this study that larval zebrafish displaymovement defects characterized by reduced average swimmingspeed and increased erratic movements upon treatment withantipsychotics that induce movement disorders in humans. Thisstudy provides a simple assay in a tractable vertebrate geneticmodel organism to identify candidate genes for understandingand treating antipsychotic-induced movement defects as well asto study the molecular genetic mechanisms underlying the brainregulation of locomotor activity.

We choose to use larval zebrafish in our study, because theyare highly permeable to chemicals administered in the tankwater, and they are readily available in large quantities.Moreover, development of an assay in larval zebrafish makesit convenient for subsequent genetic or chemical screening. As afirst test of the effects of antipsychotics on larval zebrafish, wechoose to use two-week old larval zebrafish, because they aremore mature, and swim faster than the 7-day old. Also, lowconcentrations of fluphenazine and long time treatment werechosen in this experiment, because the blood concentration offluphenazine in treated human patients is ∼4.5 μM [2]. In thisexperiment, we observe that fluphenazine impairs locomotoractivity in larval zebrafish. This result was subsequently

Fig. 3. Erratic movements of antipsychotic-treated larval zebrafish. (A) Photographs of control (left) and fluphenazine-treated (right) fish, showing an erraticswimming bout of the fluphenazine-treated fish that was not observed in control fish. (B) Average number of fish demonstrating erratic movement during the first hourof a time course immediately following administration of drugs. For each drug dose, fish were observed for 5 s at 30 and 60 min following drug treatment, duringwhich the number of vertical and sideway swimming patterns was recorded. Only fluphenazine and haloperidol induced significant erratic movement (P<0.001,compared to controls).

249N.J. Giacomini et al. / Neurotoxicology and Teratology 28 (2006) 245–250

confirmed in 7-day old larval zebrafish treated with higherconcentration of fluphenazine and in shorter exposure time. Thefact that fluphenazine produces reversible locomotor impair-ment in larval zebrafish from two different stages and by twodifferent treatment courses strongly suggests that the observedeffects are specific. The observation that administration of L-dopa facilitates the recovery of locomotor activity furtherimplies that the locomotor impairment action of fluphenazine isdue to its pharmacological interference with dopaminesignaling.

The observations that another antipsychotic haloperidol alsoimpairs movement further strengthen the conclusion thatantipsychotics produce locomotor impairment in larval zebra-fish. Moreover, we observed an interesting correlation betweenthe ability of antipsychotics to produce EPS effects in humanand their ability to impair locomotor activity in larval zebrafish.These results suggest that the pharmacological actions of thesedrugs may be well conserved in vertebrates, and studies inzebrafish could provide important insights into the understand-ing of the sides effects of these drugs as well as brain control oflocomotor activity.

Acknowledgment

We wish to thank Kathy Giacomini for her encouragement toinitiate this study. We would like to thank members of the Guo

laboratory for helpful discussions, Susie Lee and MichaelMuncha for fish maintenance. This work was supported bygrants from Burroughs Wellcome Fund, American Parkinson'sDisease Association, and NINDS.

References

[1] O.V. Anichtchik, J. Kaslin, N. Peitsaro, M. Scheinin, P. Panula,Neurochemical and behavioral changes in zebrafish Danio rerio aftersystemic administration of 6-hydroxydopamine and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine, J. Neurochem. 88 (2004) 443–453.

[2] R.J. Baldessarini, F.I. Tarazi, in: J.G. Hardman, L.E. Limbird, A.G. Gilman(Eds.), The Pharmacological Basis of Therapeutics, 2001.

[3] S. Bretaud, S. Lee, S. Guo, Sensitivity of zebrafish to environmental toxinsimplicated in Parkinson's disease, Neurotoxicol. Teratol. 26 (2004)857–864.

[4] K. Dooley, L.I. Zon, Zebrafish: a model system for the study of humandisease, Curr. Opin. Genet. Dev. 10 (2000) 252–256.

[5] P. Eichhammer, et al., Association of dopamine D3-receptor gene variantswith neuroleptic induced akathisia in schizophrenic patients: a general-ization of Steen's study on DRD3 and tardive dyskinesia, Am. J. MedGenet. 96 (2000) 187–191.

[6] J.M. Errea-Abad, J.R. Ara-Callizo, A.-R.C., Drug-induced parkisnonism:clinical aspects compared with Parkinson disease, Rev. Neurol. 27 (1998)35–39.

[7] J.R. Fetcho, K.S. Liu, Zebrafish as a model system for studying neuronalcircuits and behavior, Ann. N.Y. Acad. Sci. 860 (1998) 333–345.

[8] R. Gerlai, M. Lahav, S. Guo, A. Rosenthal, Drinks like a fish: zebrafish(Danio rerio) as a behavior genetic model to study alcohol effects,Pharmacol. Biochem. Behav. 67 (2000) 773–782.

250 N.J. Giacomini et al. / Neurotoxicology and Teratology 28 (2006) 245–250

[9] S. Guo, Linking genes to brain, behavior, and neurological diseases: whatcan we learn from zebrafish? Genes Brain Behav. 3 (2004) 63–74.

[10] S. Guo, et al., Mutations in the zebrafish unmask shared regulatorypathways controlling the development of catecholaminergic neurons, Dev.Biol. 208 (1999) 473–487.

[11] J.G. Hardman, L.E. Limbird, A.G. Gilman, The Pharmacological Basis ofTherapeutics, McGraw-Hill Medical Publishing Division, New York,2001.

[12] P.H. Kelly, R.J. Miller, The interaction of neuroleptic and muscarinicagents with central dopaminergic systems, Br. J. Pharmacol. 54 (1975)115–121.

[13] C.S. Lam, V. Korzh, U. Strahle, Zebrafish embryos are susceptible to thedopaminergic neurotoxin MPTP, Eur. J. Neurosci. 21 (2005) 1758–1762.

[14] B. Lockwood, S. Bjerke, K. Kobayashi, S. Guo, Acute effects of alcoholon larval zebrafish: a genetic system for large-scale screening, Pharmacol.Biochem. Behav. 77 (2004) 647–654.

[15] J. Martin, Locomotor activity: a complex behavioral trait to unravel,Behav. Processes 64 (2003) 145–160.

[16] C.W. Olanow, W.G. Tatton, Etiology and pathogenesis of Parkinson'sdisease, Annu. Rev. Neurosci. 22 (1999) 123–144.

[17] K. Ossowska, Neuronal basis of neuroleptic-induced extrapyramidal sideeffects, Pol. J. Pharmacol. 54 (2002) 299–312.

[18] Y. Sawada, R. Nasu, H. Matsuo, H. Ohtani, Relationship between structureand drug-induced parkinsonism, Ann. Pharmacother. 34 (2000) 668.

[19] J.T. Shin, M.C. Fishman, From zebrafish to human: modular medicalmodels, Annu. Rev. Genomics. Hum. Genet. 3 (2002) 311–340.

[20] J.A. Van Gerpen, Drug-induced parkinsonism, The Neurologist 8 (2002)363–370.

[21] R. Xu, D. Hranilovic, L.A. Fetsko, M. Bucan, Y. Wang, Dopamine D2Sand D2L receptors may differentially contribute to the actions ofantipsychotic and psychotic agents in mice, Mol. Psych. 7 (2002)1075–1082.