Cognitive dysfunctions, which are a hallmark of many psychiatric conditions including schizophrenia and attention-deficit hyperactivity disorder (ADHD), represent a major burden in the daily life of patients. It is particularly true since current medications poorly alleviate this class of symptoms. Therefore, a better understanding of the brain circuits involved in cognitive functions and/or dysfunctions is critical to better handle psychiatric pathologies and develop new therapies. Prefrontal cortex (PFC) regions have long been recognized as a neuronal hub contributing to various cognitive functions such as planning, attentional processes and working memory. The current study investigates how the claustrum (CLA) may participate in these cognitive insufficiencies.
The CLA remains one of the most enigmatic neuronal network in the brain. Although it was shown anatomically to share dense reciprocal connections with the neocortex, this network has rarely been studied because of its location and the difficulty to specifically manipulate its neurons without altering the activity of neighboring brain areas.
In this study, we molecularly distinguished CLA neurons from neighboring striatal and insular cortical neurons from adult brains using single cell RNA sequencing (scRNAseq). We identified a single neuronal population expressing all previously identified CLA marker genes. We used a transgenic mouse in which a Cre recombinase was specifically expressed in CLA glutamatergic neurons. Using conditional viral tracing and optogenetics, we studied the axonal projections originating from CLA neurons and demonstrated that the frontal cortex receives strong claustral inputs, which increase the excitability of pyramidal neurons both in vitro and in vivo.
Frontal regions being implicated in executive functions, we tested whether CLA neurons might be involved in cognitive flexibility and attention. Using microendoscopic calcium imaging, we monitored the activity evoked either in the CLA or in the medial prefrontal cortex (mPFC) during an attentional set-shifting task. In both networks, specific ensembles of neurons were activated during the task and displayed learning-dependent remapping. Specific CLA chemogenetic inhibition prevented the formation of reliable mPFC assemblies specifically during attentional set-shifting. These mice were concomitantly unable to shift their attention, phenocopying the consequence of mPFC lesions described in the literature, and a particular phenotype observed in schizophrenia (as well as in mouse models of this pathology). Mice were also unable to shift their attention when the formation of specific CLA cell assemblies was perturbed by optogenetic stimulation.
