Elsevier

Behavioural Brain Research

Volume 380, 17 February 2020, 112446
Behavioural Brain Research

Research report
Alarm cue-mediated response and learning in zebrafish larvae

https://doi.org/10.1016/j.bbr.2019.112446Get rights and content

Highlights

  • Adult zebrafish show alarm cue-mediated behavioural response and learning.

  • We investigated the presence of similar responses in larvae.

  • Zebrafish at 12 and 24 dpf decreased activity when exposed to alarm cue.

  • 24-dpf larvae learned to recognise a novel odour via pairing with alarm cue.

  • Alarm cue may allow developing paradigms to study anxiety, fear, stress and cognition.

Abstract

We investigated the behavioural and learning response of zebrafish larvae to chemicals released by injured conspecifics (the alarm cue). Many aquatic vertebrates and invertebrates exhibit an innate antipredator response to alarm cues because in nature, they reliably indicate the presence of predators. Likewise, when an individual simultaneously perceives a novel odour and alarm cue, it learns to recognise the novel odour as a predator odour. Alarm cue-mediated behavioural response and learning have been reported in some fish and amphibians during early ontogeny, but in zebrafish, they have been described only for adults. In this study, we demonstrated that zebrafish at 12 and 24 days post fertilization exhibited reduced activity when exposed to alarm cue obtained by homogenised larvae of the same age, with this response being greater for the older zebrafish. In addition, we showed that 24-dpf zebrafish conditioned with alarm cue plus a novel odour learned to recognise the novel odour as a threat and responded to it with antipredator behaviour. The innate behavioural response and the learned response after conditioning may be used to develop paradigms with which to study anxiety, fear, stress, learning and memory in zebrafish larvae.

Introduction

The zebrafish is gaining increasing importance as model for fear, stress and anxiety disorders [1,2]. For this reason, several behavioural paradigms have been developed (e.g. Hope et al. [3], Maximino et al. [4,5]), some of which exploit the innate response to alarm cues [6,7]. Alarm cues are substances contained in the skin and other tissues of many aquatic species [[8], [9], [10]]. When a prey is captured by a predator, the former’s damaged tissues release alarm cue in the water, which conspecifics can perceive via the olfactory system and use as a signal of danger to perform antipredator behaviours [[8], [9], [10]]. Adult zebrafish respond to alarm cues with typical antipredator behaviours, such as hiding or freezing, which cause a marked decrease in the locomotor activity [7]. Despite alarm cues having being studied for almost one century [11], their exact composition is unresolved. A line of evidence has suggested that hypoxanthine-3 N-oxide is the main substance causing antipredator response in fish [12,13], including in zebrafish [14]. However, this substance was not reliably detected in fish skin [15]. A recent biochemical fractionation study on zebrafish identified glycosaminoglycan chondroitin as the active substance [16]. Barreto and colleagues [17] reported that conspecific blood also evokes alarm response in Nile Tilapia. The alarm-cue effect of blood was detected also by studies on crustaceans [18], whereas a study on a cephalopod species revealed that even their ink evokes alarm response [19]. Alarm cues are considered by most authors a mixture of different substances released upon damage that fish and other aquatic organisms use both individually, and collectively to identify predation risk (reviewed in Chivers et al. [20]).

The zebrafish is also increasingly used for research on learning and memory [[21], [22], [23]], and alarm cues might be useful in this regard. Indeed, adult zebrafish, and other aquatic organisms, exploit alarm cue to learn about novel predators [24,25]. When a prey perceives alarm cue paired with a novel odour, it learns to recognise the novel odour as a threat. The conditioned response consists of antipredator behaviour (e.g., decreased activity) when exposed again to the conditioned cue (i.e., the novel odour), even without alarm cue [[8], [9], [10]]. This association between alarm cue and novel odour is extremely robust and occurs at the first presentation of the stimuli [10]. Intuitively, selection has favoured evolution of this one-trial learning mechanism to allow predator recognition after a single exposure, thereby increasing the prey’s chances of survival in future encounters [10].

Translational research is often interested in the use of larvae (i.e., age < 30 days post fertilisation, dpf) rather than of adult zebrafish [26]. Larvae can be collected in large numbers and tested early, given their quick development, making them particularly suitable to large-scale screenings of drugs and genotypes. The innate behavioural response to alarm cue and the learned response to stimuli conditioned with alarm cue could be ideal for studying anxiety, fear and stress responses, as well as learning and memory, in zebrafish larvae. However, to the best of our knowledge, larvae’s behavioural responses to alarm cue have not been used as a research model. A possible explanation is that an early study has reported the onset of zebrafish’s alarm cue response at the age of 48–52 dpf [27]. One may argue that this result seems somehow counterintuitive. Being able to respond to and recognise predators should substantially increase the chance of survival for zebrafish larvae, as reported in other species [28]. Therefore, there should be strong selection for early development of this ability. In line with this idea, response to alarm cues has been consistently documented in juvenile fish and amphibians, even at the embryonic stage [[29], [30], [31], [32], [33], [34]]. For example, Atherton and McCormick [29] monitored the heart beating of embryonic cinnamon clownfish, Amphiprion melanopus, exposed to conspecific alarm cue, finding significant increase in beating rates. This result was interpreted as evidence that cinnamon clownfish embryos can recognise alarm cue.

Detecting zebrafish larvae’s responses to alarm cue may be difficult because of methodological issues. Larvae might respond differently or more subtly to alarm cues compared to adults; therefore, most of the methodologies used for adults might not be adequate for larvae. Another potential confound is that, in the early study, zebrafish larvae were exposed to the alarm cues of older conspecifics [27]. In two fish species—brook char, Salvelinus fontinalis, and spiny chromis, Acanthochromis polyacanthus—individuals can distinguish the age of alarm cue donors and preferentially respond to the alarm cue of individuals of the same age [35,36]. Because size increases with age in most of fish species, alarm cue from larger or smaller conspecifics may indicate the presence of a predator that is not dangerous because it preys on individuals of a different age class. It is therefore possible that zebrafish larvae do not respond to the alarm cue of older conspecifics because it indicates the presence of a predator species that preys on older conspecifics and does not represent an immediate threat.

In experiment 1, we tested whether zebrafish larvae respond to alarm cue from same-age conspecifics. To detect the larvae’s response, we measured activity reduction and thigmotaxis using an automatic high-throughput tracking system, which was expected to improve the detection of the larvae’s behavioural responses. We compared three groups of larvae: one group exposed to alarm cue; one group exposed to water as a control, to ensure that the response in the alarm cue-exposed group was not due to the experimental manipulation; and one group exposed to fish odour as a further control for responses due to general fish odours or novel odours. We performed this experiment using subjects at 12 and 24 dpf.

In experiment 2, we investigated whether larvae can use alarm cue to learn to recognise a novel odour as a threat. First, we conditioned 24-dpf zebrafish using alarm cue paired with the odour of a novel, unfamiliar fish species to simulate a predator. After 6 h, we tested the larvae’s conditioned response to the fish odour alone; we also observed the response of a group of larvae exposed to water plus fish odour as control. Following a commonly-used approach [10], we assessed learning exploiting the antipredator response of larvae, which was identified in experiment 1 (i.e., activity reduction). In case of alarm cue-mediated learning, we expected that the group conditioned with alarm cues and fish odour would thereafter show a significantly stronger decrease in activity when exposed to fish odour, as compared to the control group pseudoconditioned with water. Because we suspected that a carryover effect had occurred in experiment 2, causing fish exposed to alarm cues to show reduced baseline activity, we performed experiment 3 with a larger interval (24 h) between the conditioning phase and the test phase.

Section snippets

Animal husbandry

We performed our experiments using larvae of wild-type zebrafish. The adults used for reproduction were maintained according to standard protocols in our laboratory at the University of Ferrara. Briefly, maintenance 200-L glass tanks housed mixed sex-groups with approximately 40 individuals. A heater kept the water temperature at 28 ± 1 °C, and illumination was provided with a 14:10 LD cycle. We fed the animals live brine shrimp nauplii (Ocean Nutrition, USA) and commercial flakes (Sera GmbH,

Distance moved

The model found significant main effects of type of cue (χ22 = 9.117, P =  0.010) and observation period (χ21 = 214.070, P <  0.0001), but not of larvae age (χ21 = 1.701, P =  0.192). Among the two-way interactions, the age × observation period and the type of cue × observation period were significant (χ21 = 35.432, P <  0.001 and χ22 = 134.738, P <  0.0001, respectively). The age × type of cue interaction was not significant (χ22 = 0.998, P =  0.607). The three-way interaction was significant (

Discussion

Behavioural responses induced by alarm cues may notably favour research on zebrafish anxiety, fear, stress and cognitive functions such as learning. This contribution might also be relevant for research on larvae, but their behavioural response to alarm cues has not yet been documented. Here, we provided evidence that zebrafish larvae exhibit innate response to alarm cues obtained by homogenising conspecific larvae. In addition, zebrafish larvae could be conditioned by pairing these alarm cues

CRediT authorship contribution statement

Tyrone Lucon-Xiccato: Conceptualization, Methodology, Formal analysis, Funding acquisition, Writing - original draft. Giuseppe Di Mauro: Conceptualization, Data curation, Writing - review & editing. Angelo Bisazza: Conceptualization, Funding acquisition, Writing - review & editing. Cristiano Bertolucci: Conceptualization, Funding acquisition, Writing - review & editing.

Acknowledgements

We have no competing interests. This work was supported by FAR2018, FAR2019 and FIR2018 from University of Ferrara to TLX and CB, by DOR grant from Università di Padova to AB, and by 31 st PhD program in Evolutionary Biology and Ecology at the University of Ferrara.

References (64)

  • R.E. Blaser et al.

    Experiments on learning in zebrafish (Danio rerio): a promising model of neurocognitive function

    Neurosci. Biobehav. Rev.

    (2014)
  • D. Hall et al.

    Visual and olfactory stimuli in learned release of alarm reactions by zebra danio fish (Brachydanio rerio)

    Neurobiol. Learn. Mem.

    (1995)
  • H. Richendrfer et al.

    On the edge: pharmacological evidence for anxiety-related behavior in zebrafish larvae

    Behav. Brain Res.

    (2012)
  • J.A. Atherton et al.

    Active in the sac: damselfish embryos use innate recognition of odours to learn predation risk before hatching

    Anim. Behav.

    (2015)
  • A. Poisson et al.

    Embryonic exposure to a conspecific alarm cue triggers behavioural plasticity in juvenile rainbow trout

    Anim. Behav.

    (2017)
  • M.C. Ferrari et al.

    The role of learning in the development of threat-sensitive predator avoidance by fathead minnows

    Anim. Behav.

    (2005)
  • C. Maximino et al.

    Fluoxetine and WAY 100,635 dissociate increases in scototaxis and analgesia induced by conspecific alarm substance in zebrafish (Danio rerio Hamilton 1822)

    Pharmacol. Biochem. Behav.

    (2014)
  • V.A. Quadros et al.

    Strain-and context-dependent behavioural responses of acute alarm substance exposure in zebrafish

    Behav. Processes

    (2016)
  • S.J. Schnörr et al.

    Measuring thigmotaxis in larval zebrafish

    Behav. Brain Res.

    (2012)
  • D.P. Chivers et al.

    Background risk influences learning but not generalization of predators

    Anim. Behav.

    (2016)
  • N. Miyasaka et al.

    Functional development of the olfactory system in zebrafish

    Mech. Dev.

    (2013)
  • P.R. Ziani et al.

    Nicotine increases fear responses and brain acetylcholinesterase activity in a context-dependent manner in zebrafish

    Pharmacol. Biochem. Behav.

    (2018)
  • R.M. Colwill et al.

    Visual discrimination learning in zebrafish (Danio rerio)

    Behav. Processes

    (2005)
  • M.C. Ferrari et al.

    The paradox of risk allocation: a review and prospectus

    Anim. Behav.

    (2009)
  • V.A. Quadros et al.

    Modulatory role of conspecific alarm substance on aggression and brain monoamine oxidase activity in two zebrafish populations

    Prog. Neuropsychopharmacol. Biol. Psychiatry

    (2018)
  • C. Maximino et al.

    Extending the analysis of zebrafish behavioral endophenotypes for modeling psychiatric disorders: Fear conditioning to conspecific alarm response

    Behav. Processes

    (2018)
  • V.A. Quadros et al.

    Involvement of anxiety-like behaviors and brain oxidative stress in the chronic effects of alarm reaction in zebrafish populations

    Neurochem. Int.

    (2019)
  • S. Jesuthasan

    Fear, anxiety, and control in the zebrafish

    Dev. Neurobiol.

    (2012)
  • S.J. Jesuthasan et al.

    The alarm response in zebrafish: innate fear in a vertebrate genetic model

    J. Neurogenet.

    (2008)
  • C. Brown

    Habitat–predator association and avoidance in rainbowfish (Melanotaenia spp.)

    Ecol. Freshw. Fish

    (2003)
  • G.E. Brown

    Learning about danger: chemical alarm cues and local risk assessment in prey fishes

    Fish Fish.

    (2003)
  • M.C. Ferrari et al.

    Chemical ecology of predator–prey interactions in aquatic ecosystems: a review and prospectus

    Can. J. Zool.

    (2010)
  • Cited by (12)

    • Innate visual discrimination abilities of zebrafish larvae

      2021, Behavioural Processes
      Citation Excerpt :

      The larvae’s cognitive abilities span perceptive modalities other than the visual ones. Zebrafish larvae can indeed associate tactile disturbances with a visual stimulation (Aizenberg and Schuman, 2011) or learn to recognise the odour of a predator after a single encounter (Lucon-Xiccato et al., 2020b). The need to respond appropriately to environmental stimuli appears to be critical for survival in the larval stage and consequently has driven the evolution of cognitive skills that develop early.

    • Thermal stress induces a positive phenotypic and molecular feedback loop in zebrafish embryos

      2021, Journal of Thermal Biology
      Citation Excerpt :

      Second, stress medium triggered behavioural hyperactivity compared to control embryos reaching an equivalent stage (prim-6) and the control medium containing regular metabolites, but not the fresh medium control. These results are in agreement with higher activities in rainbow trout embryos facing alarm cues (Poisson et al., 2017) but depart from lower behaviour activities in 24 hpf zebrafish embryos exposed to conspecific alarm cues (Lucon-Xiccato et al., 2020), which suggest that the response depends on the type of the cue and the tested model. Behavioural alteration was also observed in adult zebrafish in response to low pH and fasting stress-induced metabolites (Abreu et al., 2016), and adult marine invertebrates experiencing metabolites from low pH-stressed conspecifics and heterospecifics (Feugere et al., 2021).

    • Effects of clove oil on behavioral reactivity and motivation in Nile tilapia

      2021, Aquaculture
      Citation Excerpt :

      In this sense, fish recognize, through multiple sensory cues, potentially harmful stimuli (such as predators or predation risk) and trigger defense/stress responses (Kats and Dill, 1998; Barreto et al., 2003; Giaquinto and Volpato, 2001; Barcellos et al., 2007; Oliveira et al., 2014; Sanches et al., 2015; Zanuzzo et al., 2019). Aversive stimuli, such as predators (Lucon-Xiccato et al., 2020), exposure to new environments (Abreu et al., 2016; Kysil et al., 2017) and several social contexts (Soares et al., 2020) can generate typical reactions of fear and anxiety-like behaviors in fish. In addition, fish have structural-functional biological apparatus for nociception and pain perception (Sneddon, 2003; Sneddon et al., 2003; Sneddon, 2019).

    View all citing articles on Scopus
    View full text