An anatomical cell type atlas of gene expression, morphology and connectivity

Zeng, Hongkui

Abstract

(Research Segment 2) The brain circuit is an intricately interconnected network of numerous cell types. To understand the principles of information processing in the brain circuit, it is essential to determine a catalog of cell types, how they are distributed throughout the brain, and how they are connected to each other. A cell's precise location within the brain (where is it?), its specific dendritic and axonal morphology (what does it look like?), and its structural connectivity with other cells in circuits and networks (who does it connect to?) are all critical anatomical factors which contribute to the definition and accounting of different cell types. We will produce comprehensive anatomical cell census data from brains of adult male and female mice, leveraging our existing mouse brain reference and connectivity atlases, and employing three major scalable approaches. We will map the spatial organization of transcriptomic cell types identified in Research Segment 1 using multiplexed error-robust fluorescence in situ hybridization (MERFISH) with combinatorial marker gene sets identified in the single-cell RNA-seq experiments. Furthermore, we will use MERFISH to determine the microenvironment, tissue composition and ratio of various cell types. We will generate full neuronal morphologies of representative cell types in major brain regions, using two different high-throughput and high-resolution whole-brain fluorescent imaging approaches and semi-automated morphology reconstruction methods. These systematically collected data will be used to discover rules underlying cell types as defined by their full dendritic and axonal morphologies. We will use an optimized rabies tracing system to do monosynaptic, retrograde trans-synaptic tracing to map whole-brain inputs to genetically-identified cell populations brain-wide. By combining this with the Allen Institute's already created anterograde projectome, we will be able to generate a first iteration of the mesoscale, input/output circuit wiring diagram. These different types of anatomical data will be integrated with each other and with other cell type characterization data modalities in a variety of ways, including coupling rabies tracing, full neuronal morphology, or in vivo functional imaging with multiplexed FISH, to derive an integrated cell type classification scheme.

Public Health Relevance

(Research Segment 2) A cell's precise location within the brain (where is it?), its specific dendritic and axonal morphology (what does it look like?), and its structural connectivity with other cells in circuits and networks (who does it connect to?) are all critical anatomical factors which contribute to the definition and accounting of different cell types. We will produce comprehensive anatomical cell census datasets from the mouse brain, combining gene expression with neurons' full dendritic and axonal morphologies and their connectivity patterns. These datasets are essential for the generation of an integrated cell type classification scheme.