Developing zebrafish models of autism sp
Contents lists available at
Progress in Neuro-Psychopharmacology & Biological
Psychiatry
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / p n p
Review article Developing zebrafish models of autism spectrum disorder (ASD)
Adam Michael Stewart Keith Wong , Manoj K. Poudel , Allan V. Kalueff a b ZENEREI Institute and Zebrafish Neuroscience Research Consortium (ZNRC), 309 Palmer Court, Slidell, LA 70458, USA c Department of Neuroscience, University of Pittsburgh, A210 Langley Hall, Pittsburgh, PA 15260, USA d Department of Biomedical Engineering, University of Virginia, 415 Lane Road, Charlottesville, VA 22908, USA
University of California San Diego (UCSD) School of Medicine, 9500 Gilman Dr, La Jolla, CA 92093, USA a r t i c l e i n f o a b s t r a c t Article history:
Autism spectrum disorder (ASD) is a serious neurodevelopmental disorder with complex symptoms and unclear,
Received 30 September 2013
multi-factorial pathogenesis. Animal (rodent) models of ASD-like behavior are extensively used to study genet-
Received in revised form 22 November 2013
ics, circuitry and molecular mechanisms of ASD. The evolutionarily conserved nature of social behavior and its
Accepted 28 November 2013
molecular pathways suggests that alternative experimental models can be developed to complement and
Available online 6 December 2013
enhance the existing rodent ASD paradigms. The zebrafish (Danio rerio) is rapidly becoming a popular model organism in neuroscience and biological psychiatry to study brain function, model human brain disorders and
Keywords:
explore their genetic or pharmacological modulation. Representing highly social animals, zebrafish emerge as
Autism spectrum disorder
a strong potential model organism to study normal and pathological social phenotypes, as well as several
Behavioral tests
Social deficits other ASD-like symptoms. Here, we discuss the developing utility of zebrafish in modeling ASD as a new emerg-
Translational research ing field in translational neuroscience and drug discovery. Zebrafish © 2013 Elsevier Inc. All rights reserved.Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
27 2. Traditional experimental models relevant to ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28 2.1. Genetic rodent models relevant to ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28 2.2. Pharmacological rodent models relevant to ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29 3. Zebrafish models relevant to ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.1. Behavioral and pharmacological models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
30 3.2. Physiological correlates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
31 3.3. Genetic models relevant to ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
33 3.4. Environmental models potentially relevant to ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
33 4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
33 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
34 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
34
1. Introduction Autism spectrum disorder (ASD) is a serious debilitating mental ill-
ness affecting approximately 1–2% of the general population (
Recently revisited by the Abbreviations: ADHD, attention deficit hyperactivity disorder; ASD, autism spectrum
American Psychiatric Association (2013), ASD represents a neurode-
disorder; AVP, arginine vasopressin; DSM-5, Dianostic and Statistical Manual of mental
velopmental disorder characterized by impaired social communication,
disorders, 5th edition; FXS, fragile X syndrome; GABA, gamma-aminobutyric acid; MDMA, 3,4-methylenedioxy-N-methylamphetamine; NMDA, N-methyl-D-aspartate; for details
repetitive behavior and cognitive deficits (see
PCP, phencyclidine; SERT, serotonin transporter; SFARI, Simons Foundation Autism
of clinical phenotypes associated with ASD). In addition to these core Research Initiative; SSRIs, selective serotonin re-uptake inhibitors; V1aR, V1a receptor. symptoms, ASD shows high (~90%) heritability, representing one of the ⁎ Corresponding author at: ZENEREI Institute, 309 Palmer Court, Slidell, LA 70458, USA. most heritable brain disorders ).
Tel./fax: +1 240 328 2275.
E-mail address:
Notably, ASD is a polygenic disorder with multiple genetic determinants (A.V. Kalueff).
0278-5846/$ – see front matter © 2013 Elsevier Inc. All rights reserved.
28 A.M. Stewart et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 50 (2014) 27–36
While various experimental (e.g., genetic or pharmacological) ma- nipulations model certain symptoms and/or disordered pathways of ASD, they do not reflect the entire disease state. However, several social, motor and cognitive phenotypes ( ) are commonly assessed in rodent models of ASD, providing important mechanistic insights into its neurobiology also see discussion of this further in the text. The evolutionarily conserved nature of social behavior and its molecular pathways suggests that novel experimental models can be developed to complement and enhance the existing rodent ASD paradigms. Can we use zebrafish to study ASD-related path- ogenesis? Addressing this important question, here we discuss recent advances and outline future promising directions of research in the field of novel zebrafish models of ASD-like states.
Mood Sleep Anxiety ADHD Seizures Irritability (aggression, self-aggression) Social deficits Repetitive behaviors Language impairment Core clinical symptoms Cognitive deficits * * * * * * * * * Fig. 1. Summary of key clinical features of autism spectrum disorder (ASD) and related syndromes (based on Kas, Glennon, 2013). Core ASD symptoms in this diagram are marked with color (circle sizes are adjusted for illustration purposes and do not represent a particular aspect of notice). Asterisks denote domains (clusters of symptoms) that can be modeled in zebrafish (also see for details; ADHD—attention deficit and hyperactiv- ity). Note that majority of clinical ASD phenotypes can be modeled in zebrafish. (For inter- pretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
C. Symptoms' trajectory Symptoms must be present in the early developmental period (but may not become fully manifest until social demands exceed limited capacities, or may be masked by learned strategies in later life); these symptoms cause clinically significant impairment in social, occupational, or other important areas of current functioning.
B. Behavioral and cognitive perseverations Restricted, repetitive patterns of behavior, interests or activities, manifested by: i) stereotyped/repetitive motor movements, use of objects or speech (e.g., simple motor stereotypies, idiosyncratic phrases); ii) insistence on sameness, routines, ritualized patterns in verbal/nonverbal behavior (e.g., distress at small changes, rigid thinking patterns, stable rituals); iii) highly restricted interests abnormal in intensity or focus (e.g., strong attachment to objects, excessively circumscribed or perseverative interest); iv) hyper/hyporeactivity to sensory input or sensory aspects of the environment (e.g., indifference to pain/temperature, adverse response to specific sounds or textures, excessive smelling/touching objects, visual fascination with lights or movement).
A. Social deficits Persistent deficits in social communication and interaction, including: i) deficits in social–emotional reciprocity and social approach, reduced sharing of interests, failure to initiate/respond to interactions; ii) deficits in nonverbal communication, poorly integrated verbal/nonverbal communication, poor eye contact and body language, deficits in understanding/use of gestures; a deficit in facial expressions and recognizing facial affect; iii) deficits in developing, maintaining and understanding relationships, problems with adjusting behavior to various social contexts, making friends and developing interest in peers.
Table 1 Diagnostic criteria for autism spectrum disorder (ASD), according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) of the American Psychiatric Association (2013).
Loss-of-function mutations in the Nlgn4 gene in mice encoding for Neuroligin-4 impair social behavior and vocal communication, establishing Nlgn4 mice as a genetic model of ASD-related pathogenesis ). Shank proteins are also involved in modulation of synaptic communication and
With mouse and rat genomes now being fully characterized, various ge- netic models of ASD and related disorders have been developed. For ex- ample, the Fragile X Syndrome (FXS) is an inherited mental retardation disorder caused by a single mutation in the FMR1 gene in the X chromosome. FVB and C57BL/6 mice with FMR1 genetic knockout dis- play some FXS social and behavioral symptoms (
Laboratory rodents are highly sociable animals, and therefore are useful to study normal and pathological social behaviors (
2.1. Genetic rodent models relevant to ASD
2. Traditional experimental models relevant to ASD
and candidate genes. For example, the SFARI Gene database (
currently lists 546 genes associated with ASD. Thus, ASD
Taken together, this suggests that zebrafish models can be used exten- sively in translational neuroscience research
discussion). However, mounting recent experimental evidence shows that zebrafish possess high genetic and physiological homology to mam- mals and display complex affective, social and cognitive responses which are similar to those observed in rodents and humans
for
Can zebrafish be used to model complex brain dis- orders? For decades, zebrafish have been viewed as ‘simple’ organisms with relatively primitive, instinctively-driven behaviors suitable mainly for screening drugs, genetic mutations or developmental defects (see
ders, new experimental approaches using zebrafish (Danio rerio) are rap- idly gaining popularity in neuroscience research
). Complementing traditional rodent models of brain disor-
Animal (experimental) models of brain disorders are an indispens- able tool for drug discovery and dissecting the pathogenic mechanisms of brain disorders
Despite recent progress in dissecting the neural underpinnings of ASD , its pharmacological therapy is complicated by the lack of efficient, disorder-specific and safe medica- tion ). For example, the Food and Drug Administration approves two drugs (atypical neuroleptics ris- peridone and aripiprazole) for treating ASD-associated irritability Other agents, such as methylphenidate, selective serotonin re-uptake inhibitors (SSRIs), valproate, atomoxetine, α2 ad- renergic agonists and olanzapine, can also treat some ASD symptoms, but are not effective in alleviating others (
has a complex, poorly understood pathogenesis and associated genetic/ environmental risk factors, aberrant brain circuits and disordered molec- ular pathways (
- aspartate (NMDA) antagonist-induced be- haviors ( ).
- tf/J (BTBR) inbred mice display core behavioral phenotypes of autism, including social deficits and stereotypic behavior
for details). Emphasizing the importance of cross-domain model-
), whereas V1aR−/− mice display social deficits, and may there- fore be a useful model of ASD-like behavior and other disorders associat-
2.2. Pharmacological rodent models relevant to ASD
In addition to genetic models, rodents are valuable in studying drug- evoked responses relevant to ASD. For example, as social withdrawal is one of the symptoms of schizophrenia and ASD, the social approach test has long been used to study the psychopharmacology of social pheno- types ( ). D -Amphetamine disrupts rodent sociability without stimulating locomotor activity (this effect was not reversed by antipsychotics), whereas the gamma-aminobutyric acid (GABA)-A inverse agonist FG-7142 reduces sociability, reversed by the GABA-A antagonist, flumazenil (Hanks, Dlugolenski, 2013).
Antipsychotic drugs antagonize both dopaminergic-induced and glutamatergic N-methyl-
D
While rodent models appear to be a valuable tool to study ASD genetics and pharmacology ( ), there is a growing need for alternative, time/cost-efficient and high-throughput models of this disorder. Furthermore, domain-oriented approaches to complex human brain disorders focus on various domains (e.g., cogni- tive, affective, motor or social) that represent clusters of various individ- ual behaviors, and overlap during brain pathogenesis (see
ing of CNS pathobiology, these ‘integrative’ approaches assess multiple disordered phenotypes as a system, and are particularly useful in study- ing diseases such as ASD (which is polygenic and displays a wide range of phenotypes with different severity) ( ). In addition, this allows the researchers to analyze multiple co-expressed behaviors, simultaneously generating rich behavioral data in several distinct neurophenotypic domains ( recognize the importance of cross-species modeling of complex human brain disorders. Given evolutionarily conserved mechanisms of social behavior, it is therefore
to those exhibited by other mouse models with oxytocin deficits ). Moreover, the arginine vasopressin (AVP) system also plays an important role in social behavior, and the V1a receptor (V1aR) gene has been linked to social deficits
Table 2 Comparison of rodent phenotypes relevant to autism spectrum disorder (ASD; selected from for a comprehensive zebrafish behavioral catalog).
Mouse phenotypes Examples of relevant zebrafish phenotypes and models Lack of preference for social novelty
Social preference test (
A) Reduced social interactions Social interaction test, shoaling, social preference tests, mirror stimulation test ( A–C)
Reduced pup ultrasonic vocalizations Not available Reduced adult vocalizations Not available
Reduced vocalizations Not available Impaired learning and memory Various learning/memory tasks Impaired social recognition Social learning/recognition tasks Increased repetitive self- grooming Not available
Repetitive stereotyped circling behavior Stereotypic circling swimming
similar
complex with Neuroligin. Loss-of-function mutation in the Shank3 gene, encoding for Shank protein 3, inhibits synaptic transmission and im- pairs mouse social behaviors, modeling the 22q13.3 deletion syndrome associated clinically with delayed development and speech impairment given recent human genetic studies linking SHANK3 to this disorder ( ).
). Social environmental factors also affect behavioral
Serotonin transporter (SERT) gene knockout in C57BL/6J mice re- sults in increased grooming ( and causes social deficits in both SERT−/− mice and rats (
). SERT
variant Ala56 mice display differential serotonin neuron firing patterns with altered social behavior, communication and repetitive behavior and the ITGB3 knockout mice also show excessive grooming in novel environments (
In addition to mutant mice, several non-mutant mouse strains are relevant to ASD. For instance, the BTBR T
), which have been used in many studies relevant
to ASD (
or fluoxetine administration normalizes social activity in BTBR, but not in control C57BL/6 mice. Environmental factors, such as enriched housing in BTBR mice, reduce grooming duration, suggesting that enrichment normalizes certain repetitive behaviors (
outcomes in mouse ASD models. For example, adolescent BTBR mice reared in the same housing as ‘social’ C57BL/6 strain exhibit markedly increased sociability, compared to BTBR adolescents reared with other BTBR mice. In contrast, adult BTBR mice do not show improvement in repetitive (self-grooming) behavior, suggesting that behavioral recov- ery differentially affects the social but not the motor ASD-like pheno- type in this model ( ).
ments and deficits in social recognition
Likewise, the C58/J strain exhibits a behavioral phenotype re- sembling core ASD symptoms, including excess locomotor activity, social deficits, abnormal repetitive behaviors
),
poor T-maze performance and motor stereotypies (
). BALB/c mice also exhibit low sociability, exaggerated aggres-
sion, enlarged brain mass, low serotonin levels ( suggesting them as a genetic model of ASD-related states. Systemic treatment of BALB/c mice with MK-801, an allosteric inhibitor of the gluta- matergic (NMDA) receptors, elicits stereotypic circling behavior Together, this suggests the involvement of glutamatergic pathways in the pathogenesis of locomo- tor stereotypies and social impairment in the BALB/c mouse model of ASD-like behavior. Interestingly, acute administration of oxytocin does not alter social behavior, whereas its sub-chronic administration in- creases sociability 24 h post-treatment in BALC/cByJ mice and for 1–2 weeks in C58/J mice
Oxytocin receptor Oxtr
knockout mice exhibit decreased social ability but maintain normal locomotor activity Mice lacking the Cd38 gene (which encodes for CD38, an oxytocin secretion regulator;
) exhibit oxytocin signaling defects, behavioral impair-
29 A.M. Stewart et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 50 (2014) 27–36
30 A.M. Stewart et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 50 (2014) 27–36
Fig. 2. Social behavioral tests commonly used in zebrafish neurophenotyping research. A: The zebrafish social interaction test in which two unfamiliar zebrafish explore each other (target fish is shown in red). Note that in the model shown here, the separators are made from glass, providing visual cues sufficient to evoke strong social phenotypes in zebrafish. In various modifications of this test, holes can be made in the separators (to allow water to pass through), providing additional (olfactory) cues important for evoking social re- sponses in zebrafish. B: The zebrafish social preference test in which proximity of fish to each other in the shoal is measured as an index of social 'shoaling' behavior (also see photos in Fig. 3). C: The zebrafish mirror stimulation test, which quantifies social and ag- gressive behavior (also see ). D: The trace pattern analysis of zebrafish, which as- sesses various repetitive behaviors, such as stereotypic ‘thigmotaxic’ dwelling (bottom image) or ‘tight circling’ (top image).
Familiar Normal school Disrupted, loose school Mirror C B A D
Target fish Conspecific Empty zone Target fish Conspecific Group Target fish Kin Non-kin Target fish Unfamiliar
As mentioned earlier, there is a general lack of clinically efficient ASD-specific drugs, stressing the importance of developing novel com- pounds to treat various symptoms of ASD. For example, risperidone (clinically effective to treat irritability and motor/aggressive behaviors in ASD patients) does not reduce their social deficits and repetitive/ obsessive behaviors . While risperidone was not tested
scribed for these agents in rodents (see above). Collectively, this sup- ports the evolutionarily conserved nature of ASD-related social and motor phenotypes in various species, again emphasizing the transla- tional value of zebrafish models to study human ASD.
resembling the effects on ASD-related social behavior de-
Interestingly, zebrafish circling behavior can be induced by selected psychotropic drugs, such as glutamatergic antagonists MK-801, PCP or ketamine ( paralleling rodent drug-evoked circling locomotion already discussed above in relation to ASD. In line with this, various pharmacological agents have also been shown to potently modulate zebrafish social behaviors. For example, zebrafish shoaling is markedly disrupted by various psychoactive drugs
). Thus, zebrafish models are fully capable of displaying both hallmark behavior- al symptoms of ASD—social deficits and behavioral perseverations, indi- cating high translational potential of zebrafish models for ASD-related states.
also an important ASD-related phenotype ( Therefore, it may be critical to assess the availability of zebrafish behavioral stereotypies that can be measured in various tests. While zebrafish do not display patterned and complex grooming behavior, they do show several common behavioral stereotypies, including repetitive, stereotypic ‘thigmotaxic’ swimming near the walls, or specific circling behavior, involving rotational swimming in tight circles (
), behavioral perseverations are
In addition to social deficits
aggression, which will be robustly disrupted if the fish display social deficits. For example, increased head-butting or mirror biting in this test may represent elevated aggression and/or sociability in zebrafish, whereas abnormally low responses can be used to quantify ASD-like symptoms
for review). This situation can evoke both social behavior and
The mirror stimulation test, quite specific for aquatic models, is based on mirror presentation and the fact that fish perceive their own reflection as another zebrafish (
important to foster innovative modeling of ASD, expanding the spec- trum of model organisms for its translational research
, zebrafish spent most of their time swimming in dynamic groups (schools), characterized by short inter-
3. Zebrafish models relevant to ASD
3.1. Behavioral and pharmacological models
The zebrafish (Danio rerio) is a new popular model organism in bio- medical research. These fish are particularly useful for translational neu- roscience because of their high physiological and genetic homology to humans, ease of genetic manipulation, fully characterized genome, and rapid development
) and staying very close to conspecifics. It has recently been suggested that the zebrafish system may contribute to understanding ASD patho- biology and genetics ( Here, we discuss how zebrafish models can be applied practically for experimental studies of autism.
show that many of the clinical and rodent ASD-like phenotypes can be successfully modeled in zebrafish. For example, sim- ilar to a mouse social interaction protocol ), in the zebrafish social interaction test two unfamiliar zebrafish introduced together explore each other, demonstrating a wide range of quantifiable social behaviors ( which can easily be assessed in terms of dura- tion and frequency of various types of social contacts and approach.
In the shoaling test (
In the zebrafish social preference test, also adapted from rodent studies ), a target fish given a choice between staying close to the empty vs. conspecific zone, spends significantly more time near the conspecific area (
. In other modifications of this model, zebrafish typically spend more time near a group of zebrafish (vs. a single fish), also showing kin recognition/preference (
)
and spending more time during social investigation of novel (unfamil- iar) zebrafish (
A). Similarly, zebrafish models can also take advan- tage of the availability of various color variants, assessing the duration and frequency of social contacts with such phenotypically distinct strains of fish. For example, while the wild type zebrafish typically avoid shoals of white-skin stripeless nacre mutants, their social prefer- ence for nacre fish increases by systemic injections of oxytocin (and its fish analog isotocin), paralleling its similar pro-social effects in rodents ); see discussion further.
fish distance, smaller zebrafish group area size/diameter, as well as rel- ative polarizations. In contrast, disorganized social structure in zebrafish is characterized by reduced polarization of fish shoals, looser and larger schools, and higher percentage of fish leaving the group and spending time outside the shoal
A.M. Stewart et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 50 (2014) 27–36
31 PCP 5 mg/L Control MDMA 160 mg/L Control Ethanol 4% vol/vol Control Reserpine 10 mg/L 7 days Control
Fig. 3. Examples of pharmacological modulation of zebrafish social (shoaling) phenotypes. Note that zebrafish shoaling responses are markedly impaired by acute 20-min exposure to
phencyclidine (PCP, 5 mg/L), 3,4-methylenedioxymethamphetamine (MDMA, 160 mg/L), ethanol (4% vol/vol) and chronic reserpine (10 mg/L for 7 days) in representative 4-fish shoals
(Kalueff et al., 2009–2013 unpublished studies; also see
. Alcohol at mild-to-high doses is commonly known to cause social withdrawal in human and rodents, also showing a similar profile in zebrafish. Reserpine, a plant alkaloid which
depletes brain monoamines, is known to impair social behavior in rodents ), also causing social deficits in zebrafish in the shoaling test used here. Overall, these
data support sensitivity and face validity of zebrafish autism spectrum disorder (ASD)-related social phenotypes to pharmacological modulation by drugs known to evoke ASD-like symp-
toms clinically and in experimental (rodent) models. In addition to drugs evoking ASD-like behavior, it is logical to expect that zebrafish models can be used to study agents that modulate
zebrafish behavior in the opposite direction, i.e., for screening for novel potential anti-ASD compounds (see discussion of zebrafish sensitivity to risperidone, aripiprazole and fluoxetine in
the text).activity habituation to novelty (reflecting spatial working memory) can be reli-
Fluoxetine, also commonly used in clinical and rodent ASD
ably measured in zebrafish, and is affected by various pharmacological studies ( ). Memory performance in various mazes albeit not yet tested in fish social tests, also
strongly modulates zebrafish motor and emotional behavior can be assessed in zebrafish in a manner similar to rodent studies
). Since attention deficit hyperactivity disorder (ADHD) is frequently comorbid with ASD zebrafish to compounds clinically relevant to correcting some ASD symp- toms, suggesting a potential predictive validity of zebrafish screens to , and is related to both behavioral anti-ASD drugs. Clearly, future studies examining in-depth this applica- and cognitive clusters of ASD symptoms the possibility of
becomes
tion of zebrafish models to ASD research merit further scrutiny. modeling ADHD-like phenotypes in zebrafish ( important and translationally relevant to ASD (also see Cognitive deficits are widely recognized as an important clinical aspect of ASD, and are often observed in rodent models of this for details on recent progress in experimental disorder Zebrafish possess excellent cognitive abili- models of attention in zebrafish). ties, which can be comprehensively evaluated in various cognitive tasks ( ). In addition to
3.2. Physiological correlates zebrafish social cognitive models (e.g., social or kin recognition, Fig.
2a, various other paradigms can be used to assess non-social In addition to behavioral biomarkers, ASD is often accompanied clini- cally by altered physiological (e.g., neuroendocrine) functions. For
32 A.M. Stewart et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 50 (2014) 27–36
and analyzed (like other ASD-related
for details of social behavioral phenotypes in zebrafish).
Developmental nature of observed symptoms Aberrant developmental trajectories of zebrafish social behaviors (e.g., in shoaling test across lifespan) a Based on DSM-5; see
Presence of specific behavioral stereotypies (e.g., cycling behavior, figure 8 swimming, stereotyped thigmotaxis-like swimming and other motor stereotypies, such as repetitive jaw movements) and/or cognitive deficits (e.g., detected in various motor and cognitive tasks, such as impaired habituation in the novelty tests or aberrant patterns of alternation in Y or/T-mazes)
Clinical symptoms Social deficits Reduced social interaction in the shoaling test (e.g., less time spent in shoals, increased shoal size, increased average inter-fish distance and farthest neighbor distance). Reduced social interaction and/or investigation time and frequency in the shoaling, mirror stimulation and the social preference tests (e.g., reduced time spent near conspecific, reduced time investigating new (vs. familiar) conspecific, reduced time and contacts with mirror). Repetitive behaviors and cognitive inflexibility
Overview of zebrafish phenotypes relevant to autism spectrum disorder (ASD).
Social preference and recognition A natural tendency to spend time close to conspecifics; can be observed as part of shoaling behavior, kin recognition, social recognition, or preference of the ‘conspecific’ vs. ‘empty’ zones ). Zebrafish also recognize familiar from unfamiliar zebrafish (social recognition) and distinguish between own vs. phenotypically different other strain Stereotypic locomotion A pattern of rigid, repetitive behaviors (e.g., swimming from corner to corner, stereotypic jaw movements) evoked in zebrafish under some conditions (e.g., treatment with psychostimulants and PCP; Submissive behavior A social behavior following aggressive confrontations. Submissive fish stays immobile (with fins retracted), typically near the bottom or near the surface, with the caudal part of the body oriented downward ( ) Table 3
Fighting Agonistic confrontation between two individuals to establish social dominance ( ). During the first phase, the fish assess each other by exhibiting display, biting/nipping, which continues until the first chase/flee occurs. Next, the ‘winner’ (chaser) initiates all agonistic behaviors, while the ‘loser’ displays fleeing, submission behavior or freezing
Habituation Tendency to show a decreased response upon repeated exposure to a novel stimulus/environment (reflecting cognitive ability of zebrafish). Includes
inter-trial (inter-session) and intra-trial (intra-session) habituation, quantified by calculating the ratios of behavioral activity during the initial vs. latest trials, or by assessing the behavioral profile of fish across the trial(s) Kin preference and recognition The preference for kin vs. unrelated zebrafish, particularly robust in juvenile zebrafish. Zebrafish are able to seek/recognize kin from unrelated zebrafish; this behavior involves approach/attraction and leads to increased time spent near kin (kin preference) ) Mirror stimulation response Complex behaviors evoked in fish by mirror exposure; most likely linked to aggression and social investigation; typically includes approach, head-butting, biting the mirror and chasing own reflection Place preference The tendency to establish a preferred location in which the fish spends more time; can be induced by social reward (e.g., sight of the conspecific) Polarization Behavioral characteristic of adult zebrafish reflecting the degree to which members of the group are moving in the same direction; is high in established zebrafish groups Zebrafish shoaling can be quantified manually or using automated video-tracking systems, assessing several endpoints, including the average inter-fish distance; shoal area size; proximity (time each member of the shoal spent within a specified distance from each other); nearest and farthest neighbor distances; time spent in shoal; time spent away from shoal; number of animals leaving the shoal and polarization (heading) Social interaction Normal social behavior of zebrafish, a reciprocal change in zebrafish behavior influenced by the presence or actions of other conspecifics Some examples include fighting/aggression, shoaling, courtship, approach/boldness (social investigation), social recognition and social preferenceBehaviors Brief description Aggression Complex behaviors (including approach, fin raise, biting, charging, chasing and circling) directed at conspecifics in adult zebrafish; may appear in the context of establishing dominance (social interaction) Charge Movement towards a second fish with increasing acceleration; establishes social dominance and marks the resolution of a zebrafish ‘fight’ Circling Repetitive swimming in a circular direction; can be defined by their diameter (e.g., 2 body lengths/~5 cm) and quantified by the number of complete circles per trial, the number (%) of animals showing circles and the direction of circling (left- or right-rotations); automated methods may also quantify turn angle and angular velocity Lateral display involves two fish lined up parallel to each other head to tail, raising dorsal fins (fin raise), extending caudal fins and circling. Frontal display involves two fish approaching each other from the front with the attempt of biting
Table 4 Selected zebrafish behaviors potentially relevant to autism spectrum disorder (ASD; based on for a general framework).
symptoms, ) via high-throughput screening, further supporting the use of zebrafish as an efficient model for ASD and drug discovery.
example, a dysregulated oxytocinergic system has been strongly impli- cated in social deficits common for ASD
Therefore, continued investigation into the develop-
). Likewise, seizure be-
), zebrafish offer an excellent model for cortisol screening, as they not only possess a robust neuroendocrine stress axis, but also utilize cortisol (like humans), which can be easily and reliably quantified in fish (
Another potential physiological correlate of ASD may include altered cortisol levels, often reported in autistic patients. While some species differences may exist in mineralocorticoid vs. glucocorticoid action of this hormone (
), zebrafish may provide use- ful tools to investigate the pharmacological profiles of oxytocin-related compounds, as well as to test the potential of novel anti-ASD drugs targeting the oxytocinergic neuroendocrine system.
humans ( Since the oxytocinergic system is increasingly recognized as a potential therapeutic target in the treatment of ASD (
similar to
either oxytocin or isotocin (equally potent relative to each other), or an oxytocin antagonist, has been shown to modulate social preference and anxiety-related behavior in zebrafish
). Recently, peripheral administration of
mental and functional role of oxytocin is critical. The zebrafish homologue of oxytocin, isotocin, facilitates numerous facets of social behavior in tel- eosts, including social approach, fear, reproduction-related vocalizations and courtship behavior (
havior is commonly observed in ASD in humans ( ). Importantly, experimental sei- zures can be easily evoked in zebrafish
3.3. Genetic models relevant to ASD
In both humans and rodents, environmental factors play a signifi- cant role in the development of ASD
33 A.M. Stewart et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 50 (2014) 27–36
A B Fig. 4. Additional models for zebrafish social behaviors. A: An experimental setup utilizing standardized automated (vs. live) stimulus fish image presentation in zebrafish sociability models relevant to ASD. These approaches enable a better, standardized control of exper- imental conditions across all groups and eliminate potential confounding effects on the tested fish (e.g., resulting from aberrant stimulus fish behavior). In addition, they also en- able testing multiple additional experimental parameters (which can be easily adjusted using computers), such as size, shape, number of subjects, color, skin pattern, body orien- tation/angle, position in the tank, as well as swimming speed and trajectory of the present- ed stimulus fish. B: The potential application of robotic fish (e.g., ) for modeling zebrafish social behaviors, such as shoaling, relevant to ASD-like behavior (note that the same approaches can also be applied to zebrafish social preference tests shown in ).
Unlike rodent ASD Robotic fish Zebrafish
Notably, zebrafish behavior is strongly reliant on visual sensory infor- mation about their social environment, which offers a possibility of using standardized still images and ‘virtual reality’ videos of individual or group zebrafish rather than utilizing live fish as social stimuli (e.g.,
). A highly significant correlation for manual vs. automated analy- ses in such studies demonstrates the capacity of video-tracking technol- ogy to assess zebrafish social behaviors in a high-throughput and reliable manner, further fostering zebrafish models relevant to ASD ).
4. Conclusion In summary, zebrafish emerge as an important model species in translational neuroscience and the neurogenetics of ASD. The presence of key behavioral phenotypes which parallel human and rodent ASD, as well as their sensitivity to major groups of pharmacological agents used clinically to treat (or evoke) ASD, also supports their role in experimen- tal modeling of autism. The availability of recently developed video- tracking tools enables efficient, objective and data-dense quantification of zebrafish social phenotypes (
modulation of ASD-like behavior in zebrafish has not yet been thoroughly investigated, future studies of environmental effects on zebrafish phenotypes relevant to ASD may be important. Given the well-recognized history of zebrafish tests in developmental pharmacol- ogy and toxicology, the possibility of developing ‘environmental’ zebrafish models of ASD-like pathogenesis can represent an interesting direction for future research in this field.
). Thus, while the environmental
suggest that ASD-like state may also be modulated by environmental stimuli in zebrafish. For example, it has been suggested that zebrafish also have gene-by-environment epigenetic regulation of complex be- haviors, such as ASD-like phenotypes Consistent with this, exposure to a pesticide chlorpyrifos induces ASD-like behaviors in zebrafish
). Recent studies
Considering the high heritability of ASD and the ease of genetic manipulations in zebrafish, this species is also emerging as a useful tool to study ASD genetics. For example, the ASD-linked human 16p11.2 chromosomal area has a homologous region in zebrafish, consisting of several genes important for brain development
including KCTD13, which is responsible for head size and
confirm the potential of zebrafish for genetic mapping of complex be- havioral traits, but also establish genetic determinants of social behavior in this species (Wright et al., 2003, 2006), generally consistent with the notion of high heritability of ASD.
). Overall, these results not only
) strongly supports the value of zebrafish models to study CNS pathogenesis, including ASD. In addition to genetically modified fish, recent studies on strain differences in shoaling have revealed important insights into the genetics of zebrafish social behavior. For example, based on strong differences between a wild-derived strain of fish and the laboratory AB strain, quantitative trait loci (QTL) analysis identified a genomic region on chromosome 21 responsible for shoaling behavior (
The potential role of SHANK3 in human ASD ) has recently been paralleled by zebrafish models, where the genetic knock- down of zebrafish zshank3 orthologous genes (zs3.1 and zs3.2) by morpholino results in a reduction in the head size and markedly impaired swim responses to touch ( . The fact that this gene is strongly implicated in CNS function in zebrafish, rodents and humans
wise, mounting evidence has recently implicated neurexins and neuroligins in neurodevelopmental disorders, including ASD and psychoses ( These genes have been extensively characterized in zebrafish, demonstrating high sequence conservation with the human genes, therefore suggesting that zebrafish models (with their simpler circuitry and high-throughput capacity) may prove extremely useful in identifying therapeutic strategies to treat ASD ).
Like-
The AUTS2 locus has also been strongly implicated in ASD and other human brain disorders, such as ADHD, epilepsy, dyslexia, motor delay and language delay ). The knock-down of auts2 in zebrafish results in a smaller head size, neuronal reduction and de- creased mobility. Collectively, this suggests that AUTS2 plays an impor- tant role in neurodevelopment, consistent with both clinical and zebrafish phenotypes relevant to ASD
for a review).
). Reflecting a neurodevelopmental aspect of ASD pathogenesis this phenotype illustrates the potential of zebrafish in modeling ASD (
). Zebrafish met is expressed in the cerebellar primordium (later localizing in the ventricular zone), and its morpholino knockdown reduces the size of the cerebellum and affects its cellular morphology, thereby paralleling the correlation between altered MET regulation and ASD (
neurogenesis in zebrafish, and was identified as a major contributor in some cases of autism in humans ( Similarly, the ASD susceptibility gene met regulates zebrafish cerebellar development and motor neuron migration. Met encodes a tyrosine kinase receptor, im- plicated in autism and brain development in mice and zebrafish (
3.4. Environmental models potentially relevant to ASD
34 A.M. Stewart et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 50 (2014) 27–36
Egashira N, Tanoue A, Matsuda T, Koushi E, Harada S, Takano Y, et al.
Cachat J, Kyzar EJ, Collins C, Gaikwad S, Green J, Roth A, et al.
Capurro V.
Carneiro AM, Cook EH, Murphy DL, Blakely RD.
Carter CS. Carter CS, Grippo AJ, Pournajafi-Nazarloo H, Ruscio MG, Porges SW. Carter MD, Shah CR, Muller CL, Crawley JN, Carneiro AM, Veenstra-VanderWeele J.
Chadman KK. Corbett R, Camacho F, Woods AT, Kerman LL, Fishkin RJ, Brooks K, et al.
Crawley JN.
Ebstein RP, Knafo A, Mankuta D, Chew SH, Lai PS. Edvardson S, Ashikov A, Jalas C, Sturiale L, Shaag A, Fedick A, et al.
Egan RJ, Bergner CL, Hart PC, Cachat JM, Canavello PR, Elegante MF, et al.
El-Kordi A, Winkler D, Hammerschmidt K, Kastner A, Krueger D, Ronnenberg A, et al.
Brodkin ES. Burket JA, Cannon WR, Jacome LF, Deutsch SI.
Elsen GE, Choi LY, Prince VE, Ho RK.
Evans B. Fairless AH, Katz JM, Vijayvargiya N, Dow HC, Kreibich AS, Berrettini WH, et al.
Farmer C, Thurm A, Grant P. Gaikwad S, Stewart A, Hart P, Wong K, Piet V, Cachat J, et al.
Gao XM, Elmer GI, Adams-Huet B, Tamminga CA.
Gauthier J, Champagne N, Lafreniere RG, Xiong L, Spiegelman D, Brustein E, et al.
Gerlai R. Gerlai R. Gerlai R. Gerlai R, Fernandes Y, Pereira T.
Geschwind DH. Gilby KL.
Golzio C, Willer J, Talkowski ME, Oh EC, Taniguchi Y, Jacquemont S, et al.
Goodson JL, Bass AH.
social models based on automated stimulus presentation (
Braida D, Donzelli A, Martucci R, Capurro V, Busnelli M, Chini B, et al.
Bozdagi O, Sakurai T, Papapetrou D, Wang X, Dickstein DL, Takahashi N, et al.
) can therefore be better standardized and controlled experimentally, leading to improved reproducibility and reduced variance of data. Moreover, recent evidence suggests the utility of automated robotic fish (