Project Summary In neuronal circuits, the relationship between gene expression and long-range neuronal projections defines the structure of circuits and constrains computational models of their functions. In the past few years, systematic studies on the gene expression and projections of cortical neurons across the whole cortex have revealed tremendous diversity in cell types defined by gene expression, i.e. transcriptomic types, and in long-range projections. These previous studies have laid the foundation for understanding the complex relationship among gene expression, projections, and cortical areas at cellular resolution. Unraveling this relationship, however, is challenging, because it requires associating gene expression and projections across large areas of the cortex at cellular resolution. Furthermore, gene expression that are associated with the development of projections may be transiently expressed in development. Here we overcome these challenges by applying two high-throughput in situ sequencing-based neuroanatomical approaches, BARseq and retro-BARseq, to comprehensively interrogate both gene expression and projections in the cortex. These techniques rely on in situ sequencing of RNA barcodes to determine the projections of thousands of neurons together at cellular resolution. Because these techniques use in situ sequencing, they can also simultaneously interrogate the expression of hundreds of genes in the same cells. Thus, the highly multiplexed nature and the ability to associate projections with gene expression in the same cells make BARseq and retro-BARseq uniquely suited for unraveling how gene expression associate with projections. We will combine these high-throughput techniques with robust meta-analysis to take advantage of existing systematic single-cell RNAseq datasets in the cortex to identify association between gene expression and projections with unprecedented details. We will focus on the corticothalamic circuit because its projections are highly diverse and organized across both cortical areas and transcriptomic types. We will start by building a whole-cortex to whole-thalamus projection map of transcriptomic types. We will match transcriptomic types in our data to cell types in reference transcriptomic datasets to broaden our impact. We will then focus on a portion of the cortex encompassing multiple neighboring cortical areas to identify gene correlates of projections within the same transcriptomic types across neighboring areas. Finally, we will apply the same approach to the developing brain to assess the developmental association between gene expression and projections. The datasets generated in this study will not only provide a foundational resource that enables future structural and functional studies at the cell type resolution, but also provide a unique view of the relationship between cellular programs encoded by gene expression and wiring diagrams encoded by projections.

Project Narrative Long-range projections of neuronal types underlie the structure of neuronal circuits and constrain computations, but these relationship between projections and neuronal types remain largely unclear. We will unravel the relationship between gene expression and projections in the mouse corticothalamic circuit by combining state-of-the-art high-throughput in situ sequencing-based neuroanatomical techniques and robust meta-analysis. We will build a comprehensive spatiotemporal map of the corticothalamic projections in the context of neuronal gene expression and link our data with rich external reference single-cell RNAseq data to extract gene-projection relationship with unprecedented details.