The groups main research interest is the integrative study of social behaviour, which combines the study of proximate causes (gene modules, hormones, neural circuits, cognitive processes) and ultimate effects (evolutionary consequences). In particular, the research group aims to understand how brain and behaviour can be shaped by the social environment and how the cognitive, neural and genetic mechanisms underlying plasticity in the expression of social behaviour have evolved. For this purpose our research combines the study of model organisms (zebrafish, Drosophila) in the lab with the study of ecologically relevant models in the field (social evolution in African cichlids; cooperative behaviour in the cleaning mutualism between cleaner wrasse and their client reef fish). We also colaborate with colleagues in Human Psychology and Psychiatry to study the impact of social interactions and relationships (or their lack, i.e. loneliness) on human health.

Research Line 1 – Social Health

Topic 1: Social homeostasis

In social species individuals express a motivation to engage in social interactions and their lack is a major stressor with detrimental effects on Darwinian fitness. In Humans perceived social isolation (aka loneliness) has an effect size as a mortality risk as large as smoking, alcohol consumption, physical exercise or obesity. Despite these large effects, the mechanisms underlying social isolation stress at different levels (i.e. behavior, neural mechanisms, physiology, genetics) are still poorly understood. We hypothesize that the mechanisms that make individuals social must also be implicated in the detrimental effects of social isolation or impaired social relationships. Therefore, we propose to investigate causally the role of oxytocin, a neuropeptide with an evolutionary conserved role in the regulation of social behaviour, on social isolation and social bonding. However, as a complex trait, social isolation/ bonding is expected to be influenced by the interplay of many genetic and environmental factors. Therefore, it is important to search for genetic associations of susceptibility to social isolation at the whole genome level (e.g. GWAS) and to take into consideration gene x environment interactions. We are using a dual research strategy to address this research topic using: (1) an experimental approach with zebrafish as a model organism, since it is a social species whose social behaviour is well characterized and is easily manipulated in the lab, for which there is a large set of genetic tools (e.g. mutant and transgenic lines) available that allow the visualization and manipulation of neural circuits, and its genome has been fully sequenced and is well annotated; and (2) an observational (correlational) approach with human subjects, to study both (direct and indirect) genetic effects on the inter-individual variation in the susceptibility/ resilience to social isolation and how the establishment of social relationships changes the epigenome with potential health outcomes.

If successful this research line will provide significant advances in the understanding of the mechanisms underlying the physiological response to social isolation and to social interactions which has potential translational implications for loneliness health impacts in humans as well as for the understanding of the causal link between genetic variants and variation in a complex organismal trait, which will represent a major breakthrough for organismal biology.

Topic 2: Genetic models of Autism Spectrum Disorder (ASD)

ASD is a neurodevelopmental disorder that is characterized by deficits in social behavior. Animal models have been used to target autism-associated genes and assess their impact on behavioural phenotypes and underlying brain mechanisms. Zebrafish offer the opportunity of large-scale social phenotyping for drug screeening and for the assessment of putative genetic factors. However, this is often limited to simple behaviours, which prevents the identification of the specific cognitive impairments relevant for social interaction deficits. In our lab we have developed methodologies for phenotyping complex social phenotypes (e.g. social contagion) which have allowed us to identify specific cognitive deficits in different genetic models of ASD, namely a deficit in social attention in the Shank3a mutant and a deficit in social decision-making in oxytocin receptor mutants, that lead to a similar deficit in social behavior (Kareklas et al 2023 Molecular Autism 14). We aim to scale up this behavioral phenotyping approach to test a large set of ASD genetic mutants, already available, to disentangle how different genes contribute to different cognitive components of ASD social phenotypes. This approach will help to address the heterogeneity of ASD phenotypes and to identify specific biomarkers. Once relevant genetic mutants are identified we aim to screen drug libraries to identify specific drug effects on these different cognitive components.

Topic 3: Cognitive bias and stress resilience

The stress response is an adaptive mechanism for the organism to cope with challenging situations (stressors). The stress-regulatory circuit that is activated during a stressful event is highly dependent on the characteristics of the stressor. In general, stress responses to physiological threats (e.g. altered temperatures or hypoxia) are not modulated by the limbic system. However, stressors involving higher-order sensory processing (e.g. fear or exposure to novel environments) are limbic-sensitive and require a sequential stimulus assembly to obtain physiological meaning. Limbic circuits are therefore actively involved in the modulation of the resultant stress response, which is consequently dependent on the individual’s prior experiences. This fact implies that some kind of appraisal mechanism must allow organisms to subjectively evaluate a potential threat and to determine whether a stress response should be activated or not. The involvement of cognitive appraisal in the activation of the stress response opens the potential for evaluation biases, such that some individuals will consistently judge ambiguous stimuli as negative (pessimists) and others as positive (optimists). Consequently, optimistic/ pessimistic phenotypes are expected to vary in their resilience/susceptibility to stress-related disorders.

We aim to assess the neural mechanisms that underly the subjective assessment of ambiguous stimuli, to test the role of cognitive bias in the vulnerability to stress dysregulation and its specific detrimental consequences, and to assess cognitive bias in an ecological and evolutionary framework (i.e. function, evolutionary history) using both zebrafish and Drosophila as model organisms. We have already developed a judgement bias assay to measure decision-making under uncertainty in zebrafish (Espigares et al 2021 Biol Lett 17) and showed that judgment bias is casually linked with variation in stress resilience/ susceptibility and associated health consequences, with optimists being more resilient to stress-related disease (e.g. cancer onset and progression in a zebrafish melanoma line) than pessimists.

Research Line 2 – Social Evolution

Topic 4: The neuromolecular basis for the evolution of sociality

Understanding the evolution of sociality remains a central question in evolutionary behavioural ecology. Yet, genotype-to-phenotype mapping of social behaviour, which is needed to understand the causal mechanisms that underlie evolutionary trade-offs and how genetic and phenotypic architectures (neural, behavioural) evolve, poses great challenges. Behaviour is grounded in the brain, which raises a dual coding problem, since genes encode brain structure and functioning through gene regulatory networks (GRNs), whereas brains encode behaviour through neuronal electrochemical signalling. Moreover, the brain is a highly heterogeneous and complex organ. The first problem has been overcome with the finding that neuronal firing is paralleled by changes in brain gene expression, such that neurogenomic states can be used as a “common language” to link genes, brain and behaviour. The second problem can now be addressed with recent developments in single-cell and spatial transcriptomics, which allow to go beyond bulk RNA sequencing and to ground gene-behaviour relationships in specific neural networks with cellular resolution. We will seek to unravel the integrated genetic and neural architectures of elementary forms of social behaviour (social affiliation, social recognition) by combining experimental work (artificial selection for sociality) on a model organism (zebrafish) with comparative work on naturally occurring variation in social behaviour (cichlids). On one hand, zebrafish is an established teleost model organism that offers a set of genetic tools for the visualization and dissection of neural circuits. On the other hand, cichlids exhibit a vast array of divergent social phenotypes constituting a collection of natural mutants screened by natural selection. So far very few studies have linked behaviour associated genetic variants to causal mechanisms, the studies that have identified behavior associated gene networks lack spatial resolution (i.e. brain region or whole brain bulk RNAseq), and comparative work on vertebrate systems is mostly lacking. Only recently, advances in functional genomics allowed transcriptome analysis with temporal and spatial resolution in organismal tissues, and mathematical tools have been developed for analysis of complex systems, such as multi-layered networks, making it possible to study such complex datasets. Thus, the time is right to disentangle the molecular foundations of complex phenotypes, such as social behaviours, in an evolutionary framework using single cell and spatial transcriptomic approaches.