Resources
Authors & Affiliations
Archi Garg, Tanmai Dhani Reddy, Yoni Couderc, Daria Ricci, Tina Habchi, Anna Beyeler
Abstract
Anxiety disorders are marked by intense and chronic worrying along other symptoms, including an attentional bias towards negative valence. Preclinical models have linked the activity of the anterior insular cortex (or insula) to both anxiety and negative valence[1]. We recently showed an increased dopamine release onto anterior insula neurons expressing type-1 dopamine receptors (D1R) during anxiogenic events and during aversive stimuli[2]. We also revealed that both systemic and intra-anterior insula D1R activation heighten anxiety-like behaviors[2]. However, the impact of D1R on anterior insula population coding remained unknown. To reveal the dynamics of anterior insula neuronal activity, we performed single-unit recordings in mice during a classical anxiety assay, and during tastant consumption of positive or negative valence. We applied linear (principal component analysis, PCA) and non-linear (deep neural network[3]) dimensionality reduction to decipher neuronal activity in the anterior insula during those assays. In control conditions, PCA revealed higher spiking variability during anxiogenic compared to safe events, as well as during positive compared to negative tastant consumption. Similarly, the deep-learning analysis uncovered differential neuronal patterns when mice were in anxiogenic compared to protected areas, and during consumption of positive compared to negative tastants. Remarkably, along with increasing anxiety, systemic D1R activation altered neural structure in the anterior insula. Specifically, D1R activation increased spiking variability at the entrance of safe areas, and when mice explored safe spaces, while it decreased spiking variability during anxiogenic events and when mice were in exposed spaces. These results suggest that D1R activation induces a more reliable coding of anxiogenic events and spaces within the anterior insula. Interestingly, D1R systemic activation also altered coding of positive and negative tastants by anterior insula neurons. Our results provide a novel model of how neural structures in the anterior insula encode anxiety and taste valence, uncovering D1R-dependent coding mechanisms.