The molecular gradients underlying wiring plasticity in the visual cortex

About This Grant

Project Summary Sensory experience during early postnatal life profoundly influences the development of neural circuits. The primary visual cortex (V1), as part of the neocortex, is a well-established site of developmental plasticity, where visual experience plays a key role in the proper development of its binocular circuitry. However, how vision regulates the organization of individual neurons in V1, the cell types and connections they form, and the molecules they express, remain poorly understood. I address this knowledge gap by combining single-cell transcriptomics, epigenomics, connectomics and computation to link genes, cell types, regulatory networks and circuit structure in layers 2 and 3 (L2/3) glutamatergic neurons in V1. My recent work found that L2/3 glutamatergic neurons in V1 form continuous rather than discrete cell types. These cells differentially express hundreds of molecules known to be important for neuronal cell-type identity in a graded fashion, and are spatially enriched in different sublayers of L2/3. In mice that are deprived of visual experience by dark rearing, this cell-type organization is disrupted in a specific way. Based on these findings, we hypothesis that the continuous and vision-dependent cell types are key structure underlying cortical developmental plasticity. Aim 1 of my research will determine the developmental origin of L2/3 cell-type continuum. I will generate a developmental atlas of the L2/3 neurons using three existing single-cell RNA-seq datasets from mice with and without visual deprivations. I will delineate the vision-dependent developmental programs of L2/3 neurons, and identify gene programs that are temporally-regulated, type-specific and/or vision- dependent, which will serve as candidates for follow-up experiments to examine their causal link in L2/3 wiring. Aim 2 will examine the axonal-projection patterns of L2/3 neurons in V1. These neurons send divergent projections from V1 to many higher visual areas (HVAs), and transcriptomic differences in L2/3 neurons are thought to be associated with differences in projection targets. We will use barcoded connectomics to systematically determine the relationship between L2/3 cell-type continuum and projection targeting specificity to HVAs. Aim 3, to be pursued during the R00 phase, will investigate the gene regulatory networks underlying L2/3 neurons. Using single-cell multiomics data, we have identified tens of thousands of genomic regions with graded differential accessibility along the L2/3 cell-type continuum. To understand the regulatory logic linking gene expression and genomic elements often known as enhancers, we will use both experimental and computational methods. Experimentally, we will perturb the expression levels of key transcription factors and examine changes in L2/3 cell-type organizations and projection patterns. Computationally, we will train a DNA language model to predict the graded chromatin profiles of L2/3 neurons, and use in silico perturbations to screen candidate genes. Ultimately, the two efforts could be combined into a closed-loop iteration where experiments improve the model, and the model prioritizes experiments. My work will investigate the influence of visual experience on genes, cell types and circuits in a subclass of cortical neurons, thereby providing insights into the molecular logic of wiring plasticity in the mammalian neocortex.