Neuron Identity Problem: From Genes to Function

Abstract

Cellular specialization is one of the key features of the vertebrate nervous system, which is composed of thousands to billions of neurons that that have distinct molecular, morphological, electrophysiological and connectivity properties. Deconstructing the nervous system into its constituent parts is one of the fundamental steps to understanding how it generates behavior and how its dysfunction might lead to neuropsychiatric disorders. To tackle this problem, we generated comprehensive cell type profiles of two regions of the zebrafish brain: 1) the habenula and 2) the telencephalon and sought to connect these molecular profiles to functional properties of said brain regions. To generate comprehensive cell type maps of these regions, we produced and analyzed single-cell RNAseq data from the habenula and telencephalon. Using a two-time point single cell analysis of each region combined with machine learning approaches, we investigated questions on the maintenance and divergence of cell type identities between a developing animal and a mature adult. We found that the habenular cell types were largely maintained between larva and adult. On the other hand, cell types in the telencephalon displayed significant divergences between developing larva and mature adult. To interrogate the functional roles of cell types, we combined molecular profiling approaches with in situ hybridization and computational image morphing to generate detailed spatial atlases for both regions of the brain. Using the spatial atlas of the habenula, we identified a population in the ventrolateral habenula that is activated by inescapable aversive environmental stimuli. To further understand the functional role of the cell types and marker genes, we generated mutants of cell type specific marker genes and assayed them for brain activity and behavior. We discovered that all the habenular mutants displayed aberrant locomotor activity and one displayed a reduction in arousal threshold, establishing the habenula as an important regulator of activity and perhaps even sleep. Finally, we utilized the cell type landscape of the forebrain to understand the molecular basis of a brain activity and behavioral phenotype in animals with loss of a schizophrenia associated gene znf536. Together, these observations demonstrate how the deconstruction of the brain into its constituent cell types can spur further discoveries of the role of cell types in generating animal behavior in health and disease.