Mesoscale calcium imaging of freely moving mice engaged in social interaction reveals widespread inter-brain synchrony

The natural inclination to engage or associate with others is referred social interaction. The mechanisms involved in social behavior are extremely complex but are important for the health and well-being of every individual. When social interactions are compromised, they are a clear sign in several psychiatric disorders such as autism, schizophrenia, depression, and social anxiety disorder. Identifying the neural and behavioral mechanisms underlying these conditions would be important to improve their treatment. In mice, studies of social interaction are primarily based on examining behavior in individual animals by analyzing a single brain region at a time. In addition, monitoring neural activity using optical imaging techniques requires keeping the subject in a head-fixed condition, greatly limiting the behavioral repertoire. In humans, conversely, the neural activity of two or more subjects is acquired simultaneously using non-invasive techniques such as EEG, fMRI, or fNIRS during the execution of the behavioral paradigm. This type of study called Hyperscanning provides information about brain organization and the level of synchrony between the two participants' brain signals during the performance of various types of social activities. Recently, with the advent of miniaturized optical devices, it is possible to perform optical imaging in freely moving rodents, however, currently developed systems have fields of view limited to a few mm2 or require implantation of GRIN lenses in the brain site to be investigated. All of this precludes the study of neural dynamics distributed over large cortical areas associated with complex behaviors. In this Ph.D. project to overcome these problems, a new miniaturized wide-field optical system "Miniscope" was developed to perform calcium imaging, distributed over both hemispheres, in freely moving mice. Given the importance of social interactions in individual well-being, an innovative experimental paradigm was developed to study the behavioral repertoire exhibited by rodents. By combining this novel miniaturized optical system and the developed behavioral paradigm, a study of Hyperscanning in mice engaged in social interaction was conducted. This allowed simultaneous investigation of both the behavioral repertoire and neuronal activity distributed over the entire cortex. The results obtained show that the developed system allows the monitoring of calcium activity with a large wide field of view and sufficient optical resolution to perform mesoscale neuronal imaging. Design improvements made it possible to develop a system light enough to perform free-moving imaging without hindering animal behavior. Using the Miniscope, it was possible to record the neural activity distributed over the entire cortex mantle while two subjects are interacting in freely moving conditions. Wavelet Coherence Transform analysis was used to analyze the neural signal since this technique is widely employed in Hyperscanning studies in humans and provides information on how similar a pair of signals are as a function of frequency and time. Wavelet analysis conducted over the whole cortex revealed inter-brain coupling in two frequency bands (infra-slow and slow) modulated by the type of behavior. In detail, when two subjects simultaneously participate in the interaction (Gap Interacting), two coherence peaks in two frequency bands at 1/16 Hz (infra-slow) and about 4 Hz (slow) are evident. However, when only one of the two subjects in the dyad goes to the gap to interact (Gap Non-Interacting) a single coherence peak emerged in the infra-slow frequencies of lower intensity than in the previous behavioral condition, while the peak in the slow frequencies disappears. These results show that inter-brain coupling in slow frequencies is a characteristic event of the Gap Interacting condition, however, the Gap Non-Interacting is a sufficient behavioral condition to determine inter-brain coupling although there is no physical contact between dyad members. Moreover, when the synchrony of parceled brain regions is examined, it is found that not all cortical regions exhibit the same inter-brain coupling in the two frequency bands. Analysis of parcellated coherence maps by area highlights inter-brain coherence coupling of all cortical regions in the infra-slow band of the interacting dyad, while slow frequency analysis showed coupling between the two neural signals mainly in the somatosensory and visual area of both hemispheres and the associative area of the right hemisphere only. In conclusion similar to the results obtained in Hyperscanning experiments, we found that social interaction in mice modulates the coupling among the brains of interacting subjects and that this coupling is dependent on both the cortical areas and the frequency band. This matches the observations done in humans and therefore this platform provides an experimental framework to study social interactions in an animal model.