Chair: Bartosz Zglinicki (Łukasiewicz – PORT, Wroclaw, Poland)
Symposium 6: Advancing neuroscience with unbiased methods of automatization of behavioral studies
The subject of the session will place a strong focus on emerging need to introduce and refine automatic assessment of behavior (both individual and group behaviors) of animal model of choice. We need to acknowledge that in vivo experiments must be carried in parallel to the in vitro studies in general for the better understanding of tackled scientific problem and for future translation of results from animals to humans.
Aleksandra Badura, PhD, Associate Professor at Erasmus Medical Center, Rotterdam, Netherlands
The cerebellum as a driver of cortical maturation and cognitive flexibility
Understanding how the brain develops and how deviations from typical neurodevelopment are linked to health and disease remains a top priority in clinical neuroscience. Research to date has disproportionately focused on the development of the cerebrum, thereby omitting the so-called ‘small brain’, the cerebellum. While the cerebellum’s involvement in motor control is well documented, recent studies have made clear that it also plays a crucial role in higher cognitive function and that disrupted cerebellar development distorts cortical maturation. However, little is known about mechanisms underlying this complex cerebellar-cortical interplay during development. Here we have addressed this knowledge gap by studying the influence of disrupted cerebellar development on cortical maturation and behavioural phenotypes. Specifically, we examined the effects of disrupted crus 1 development in mice, using targeted cerebellar ablations at distinct developmental stages to elucidate the timing-dependent nature of behavioral and anatomical outcomes. Our findings reveal that lesions in crus I result in impairments in social and flexible behaviors without affecting motor skills, with the phenotypic impact varying by the timing of the lesion. Further, using an ultra-high field 7T MR imaging we have investigated anatomical differences across experimental groups. Through deformation-based morphometry, images were processed and analysed at the voxel level. Volume changes at the voxel level were then used to accurately retrieve changes in structure volumes through atlas registration. This approach allowed for patterns of deformation to be linked to patterns in behavioural defects at voxel and structural levels, as well as between individuals and experimental groups. Together, our results elucidate a developmental stage-specific effects of early cerebellar injury on behavioral phenotype and whole-brain anatomy. Such insights highlight the existence of critical periods that influence the cerebello-cortical development, potentially providing predictive value for neurodevelopmental deficits’ severity following cerebellar disruption.
Andrew Holmes, PhD, Senior Investigator at National Institute on Alcohol Abuse and Alcoholism, National Institute of Health, Rockville MD ,USA
Machine learning discovery of homecage behavior predictors of mouse fear and alcohol drinking
A major goal in psychiatry is to understand individual differences in risk for mental illness in order to identify disease markers that facilitate targeting of treatments to at-risk individuals. Artificial intelligence (AI) and machine learning has been leveraged to predict the long-term course of posttraumatic stress disorder (PTSD) and alcohol use disorder (AUD). However, whether AI can be exploited to predict how mice vary in phenotypes of relevance to these disorders has not been fully examined. In the current study, a computer vision-based system was trained to continuously and chronically classify homecage behaviors in male inbred (C57BL/6J) mice undergoing testing for fear-related behaviors and alcohol binge-drinking. Homecage behaviors were grouped into three macro feature-categories: Exploratory (EXP: Rear, Walk, Hang), Low activity (LOW: Rest, Sniff) and Consumption-related (CON: Eat, Paw, Drink, Groom). Assessment of freezing responses to a conditioned cue indicated marked inter individual variation during fear retrieval, extinction retrieval and fear renewal, such that the population was divisible into Low and High Fear, Extinction Intact and Impair and Low and High Renewal subgroups. The proportion of time spent in EXP behavior was greater in Extinction Intact than Impair mice during the extinction training period and greater in Low than High Renewal animals during the pre-conditioning baseline period. More time in EXP behavior during pre-alcohol baseline was also predictive of higher binge-drinking after three weeks of alcohol-access. These findings reveal distinct signatures of homecage behavior that represent predictive, a priori, markers of future performance on measures of mouse fear and alcohol binge drinking.
Adam Brosnan, PhD, Postdoc at Nencki Institute of Experimental Biology, Warsaw, Poland
Cage of Thrones: Automated Analysis of Social Behaviour and Power Dynamics in Mouse Hierarchies
Social hierarchies are thought to regulate social structures. In humans, social hierarchies manifest based on factors such as wealth, knowledge, age, and strength, which determine resource allocation. For example, in capitalist societies, wealth determines how much of a resource a person is able to obtain. In wild mice, a dominant mouse often has more access to territory, food, and mates. In this experiment, mouse social dominance is determined by chasing behaviour in a fully automated, semi-naturalistic, home cage monitoring system called ‘Eco HAB.’ Using this apparatus, we aim to model social dominance in socially housed (n = 30) C57BL/6 mice. We focus on three questions: (1) How quickly do mice form social hierarchies in Eco-HAB? (2) Once formed, are they stable? (3) Are they flexibly reformed across different social contexts? The automation of data collection and analysis in Eco-HAB offers significant advantages in studying social dynamics. By enabling continuous, non-invasive monitoring of behavior, the system provides high-resolution data that reveal intricate patterns of dominance and territoriality. Results demonstrate that mice form a hierarchy within 2 to 3 days, which, once established, is stable. In contexts exclusively composed of dominant animals, instead of regulating their society via chasing behaviour, animals switch to territoriality. The automated analysis streamlines the identification of these behaviors, reduces observer bias, and allows for the scalable study of multiple groups, making it a powerful tool for uncovering the complexities of social dominance.
Bartosz Zglinicki, PhD, Postdoc at Polish Centre for Technology Development, Łukasiewicz-PORT, Wroclaw, Poland
Establishing unified, automatic, unbiased platform for characterization of social behavior in mice
Validation of any biological intervention developed for tackling the root or symptoms of neuropsychiatric disorders requires proper tools. Disturbances of social behavior is a hallmark of many psychiatric conditions, such as depression, anxiety and schizophrenia. Capturing and quantifying broad range of social behavior simultaneously within a single tool is therefore essential, as it would allow for registration and clustering of groups of behavior specific for particular disorders. Such method would be of great advantage for better translational studies. Here, within Same-NeuroID project, we implemented a pipeline for efficient characterization of complex social behavior in mice. In the setup animals are housed together in a semi naturalistic environment with proper bedding, food and water access, and with night-day cycle. Animals are recorded for long hours to ensure a capture of diverse behavior. SLEAP.ai is used for pose estimation of recorded animals and extracted coordinates are processed by deepOF software to create set of features. These features are then used to “feed” models of both supervised and unsupervised learning for complex behavior classification.
Funding: This work was carrying out by the Horizon Europe Research and innovation funding programme under Grant Agreement 101079181 – SAME – NeuroID and the national science centre based on decision no. Dec-2021/41/b/nz3/04099 entitled „do astrocytes control synaptic connections in neural networks relevant to psychiatric diseases?”. Grant agreement: umo-2021/41/b/nz3/04099 – AstroSyCo