To address the question of the neural mechanisms of defence against an external threat our lab uses the fruit fly as a model system. It’s amenable to the search for the neural mechanism of behaviour, and it allows the study of the behaviour of large groups of individuals. This is the ideal model system due to its large collection of powerful genetic tools, a rapidly increasing number of approaches to study neural circuits, and an expanding set of behavioural paradigms. Our team focuses on a pervasive defensive behavior that avoids detection by predators who mostly rely on movement for prey detectetion, characterized by complete immobility – FREEZING.

To investigate the flexible nature of freezing in flies, we searched for experimentally tractable environmental factors, inspired by the ones modulating vertebrate freezing, namely the social and spatial environment. In addition, we are insterested in the mechanisms by which freezing is instanciated. Knowing the sensory input for looming triggered freezing, we are focusing on the descending output that controls its somatic instatiation.

Main Interests

Defensive Behaviours    Social Behaviours      Spatial learning              Neural circuits                Cardiac activity                Muscle activity


Development of behavioural tasks, Optogenetics, Genetics and Physiology


Fruit fly


Our strategies for unravelling mechanisms of defence against external threats span diverse areas, studying the interactions between the brain, body and defensive behaviour.

Starting from a broad perspective, our lab undertakes detailed behavioural characterization, identifying specific elements of freezing behaviour responding to threatening stimuli. With this strong behavioural foundation, we then develop behavioural tasks, alter the social and environmental context, perform brain and body manipulations, and quantitatively analyse behaviour.

At the neurological level, we are interested in mapping the circuits underlying defensive behaviour and we also use electrophysiology and calcium imaging techniques to understand how these circuits work. We also work from descending outputs to investigate the functional role of muscles involved in freezing behaviour. 



Instantiation of defensive behaviours and internal state 

A successful defensive behaviour needs to be coordinated, directed and extremely fast. Achieving this requires a simultaneous response throughout the body, from the neural, muscular, and endocrine systems to the heart and circulatory system. We are interested in understanding how bodies enact defensive behaviours such as freezing, and conversely how different parameters of the body such as heart rate and internal state can affect the decision and the ability to freeze.


Dissecting the neural mechanisms underlying looming-triggered freezing behavior

Drosophila melanogaster when faced with visual threats displays a variety of complex and stereotypical behaviors that depend on the features of the threat, the context in which it occurs and physiological traits. Most studies focus on immediate responses to a single threatening stimulus presentation, overlooking the possibility that threats can lead to sustained  changes  in the flies’ behavior and internal state which may impact subsequent responses to the same stimulus mimicking a predator repeating its attack after a failed attempt at catching its prey. 

Leveraging the arsenal of genetic tools and new advances in electron microscopy, we use a combination of loss and gain of function experiments together with single-cell recordings, aiming to identify the underlying mechanisms and the dynamics within them that ultimately leads to the implementation of looming-triggered freezing behavior.

Behavioural/internal state and selection of defensive behaviour

Previous work in the lab established walking speed as a key factor in the choice between different defensive responses. Increased motor activity, and in particular walking speed, is one of the hallmarks of a state of increased arousal. Therefore, we hypothesize that an important factor modulating the selection of threat responses is the arousal state of the fly at the time of stimulus presentation. A growing number of studies have implicated biogenic amines in arousal in different behavioural contexts. Thus, we believe neuromodulation plays an important role in the selection of defensive strategy. Our goal is to uncover the molecules, circuitry and mechanisms that allow the fly to integrate a particular threat stimulus with its behavioural state to generate a specific defensive response.

Motor control of freezing behaviour

Absolute stillness during freezing is crucial to avoid detection by predators. Further, the need to respond fast implies the ability to sustain freezing in any posture. We aim to dissect the somatic parameters underlying the execution of sustained, stable freezing, including the sensory feedback required in local circuits and ascending to the brain, and muscle activation in the legs, using behavioural tracking, in vivo imaging and genetic manipulation

Image: Dr. Anna Hobbiss

Brain-heart interactions upon threat

Cardiac activity changes are a well characterized part of the physiological response during defensive behaviour. From insects to mammals, freezing behaviour is associated with a reduction in heart rate, bradycardia, while escape behaviours are associated with increased heart rate, tachycardia. We are interested in cardiac response to threat, its function and how it changes depending on the defensive behaviour chosen. We also investigate the neuronal mechanisms underlying cardiac regulation during behaviour. 

Barrios, N., Farias, M., & Moita, M. A. (2021). Threat induces changes in cardiac activity and metabolism negatively impacting survival in flies. Current Biology. (https://doi.org/10.1016/j.cub.2021.10.013)

Social modulation of defense behaviours

Flies, like many other animals, live in groups, which confer advantages in the context of predation. This group setting allows them to use cues from others to detect threats or as a safety signal, and thus deploy appropriate behaviours. Using detailed behavioural analysis, genetic manipulations and a setup with manipulable magnets, we identified these cues emanated from the movement of others. Through neuronal inactivation experiments we also uncovered lobula columnar visual projection neurons involved in their mediation. We are currently further exploring the detection of these cues at the sensory level and uncovering how the underlying sensory-motor transformations are instantiated, using optogenetic tools and imaging. In addition, we are studying the interplay between direct threat detection and reliance on social cues in different settings, and developing a computational model of the group behaviour.

Ferreira, C. H., & Moita, M. A. (2020). Behavioral and neuronal underpinnings of safety in numbers in fruit fliesNat. Comm. 11(1), 1-10. (https://doi.org/10.1038/s41467-020-17856-4)

How do predator-prey interactions modulate prey defensive behaviour?

Using behavioral analysis, in this project we aim to characterize how a fruit fly modulates its defensive behaviour depending on social interactions with its natural predator, the jumping spider. Indeed, social behaviour indeed, defined as the interaction between agents is not limited to members of the same species, therefore we can think of a prey-predator interaction as a particular type of social interaction. Depending on the social interaction the fly exhibits a vast array of defensive behaviours. However the precise characteristic of this interaction, namely which cues are used by the fly to select between the different defensive strategies, is unknown. Here we propose to shed light on this unexplored and intricate interaction


Photo: Jeff Burcher

How does context familiarity modulate defensive behaviours?

When facing a predatory threat, animals display different defensive behaviors: freeze, flight and fight. The selection of the appropriate defensive response depends on several factors such as the nature of the threat, the context and the animal’s internal state. However, very little is known about how the animal’s surroundings influence this choice. Combining genetic manipulations, sophisticated behavioral assays and imaging approaches, we aimed to provide insight on how learning about the spatial context impacts the selection of defensive behaviors in flies, and the underlying neuronal mechanisms of this process.

Neural mechanisms of sheltering behaviour 

What are the characteristics that are relevant for animals to identify a “place” as a shelter – a safe place to escape to in the face of danger?  Sheltering behaviour is ubiquitously expressed throughout the animal kingdom. However, its mechanistic underpinnings, namely how animals assess  protective features of a shelter, are poorly understood. Here, we propose to elucidate this process and the neuronal circuits supporting it, assessing the contribution of different sensory modalities. To do so, we will use behavioural tasks, neural manipulations, neuronal circuit mapping and in vivo imaging. Given that sheltering behaviour is conserved across species, we expect to find generalisable principles of organisation governing this process. And maybe shedding light on what are the features that make us feel safer in challenging situations! 

The impact of genetic background in the activity of single neurons

Due to their connectivity and organizational principles, neuronal circuits and gene networks share several common features. Whereas gene knockouts and gain-of-functions help us to dissect genetic pathways, perturbing neuronal activity is helping us to decipher the organization of neuronal circuits. Recent advancements in genetic engineering allow targeting the activity of single neurons, making it, for analogy purposes, comparable to mutating single genes in genetic pathways. While we continue with this endeavour, it is important to retain an important lesson from genetic studies, that is, the phenotypic effect of a mutation is highly dependent on the genetic background where it is introduced. In this project we are interested to know how the genetic background influences the phenotypes caused by the manipulation of single neurons.