Cell-type specific transcriptomic signatures of neocortical circuit organization and their relevance to autism

Abstract

A prevailing challenge in neuroscience is understanding how diverse neuronal cell types select their synaptic partners to form circuits. In the neocortex, major subclasses of excitatory projection neurons and inhibitory interneurons are conserved across functionally distinct regions. There is evidence these subclasses form circuits that depend primarily on their identity; however, regional cues likely also influence their choice of synaptic partners. We mined the Allen Brain Institute's single-cell RNA-sequencing database of mouse cortical neurons to study the expression of cellular adhesion molecules (CAMs) necessary for synapse formation in two regions: the anterior lateral motor cortex (ALM) and the primary visual cortex (VISp). We used the Allen's metadata to parse cells by clusters representing major excitatory and inhibitory subclasses that are common to both ALM and VISp. We then performed two types of pairwise differential gene expression analysis: 1) between different neuronal subclasses within the same brain region (ALM or VISp), and 2) between the same neuronal subclass in ALM and VISp. We filtered our results for differentially expressed genes encoding CAMs and developed a novel bioinformatic approach to determine the sets uniquely enriched in each neuronal subclass in ALM, VISp, or both. This analysis provides an organized set of genes that may regulate circuit formation in a cell-type specific manner. Furthermore, it identifies candidate mechanisms for the formation of circuits that are conserved across functionally distinct cortical regions or that are region dependent. Finally, we used the SFARI Human Gene Module to identify CAMs from our analysis that are related to risk for autism spectrum disorder (ASD). From over 3,000 differentially expressed genes, we found 40 ASD-associated CAMs that are enriched in specific neuronal subclasses in both ALM and VISp. Our analysis provides clear molecular targets for future studies to understand neocortical circuit organization and abnormalities that underly autistic phenotypes.