Brain-wide Correspondence Between Neuronal Epigenomics and Long-Distance Projections [1]. Brain-wide Correspondence Between Neuronal Epigenomics and Long-Distance Projections [1]

Single-cell genetic and epigenetic analyses parse the brain’s billions of neurons into thousands of “cell-type” clusters, each residing in different brain structures. Many of these cell types mediate their unique functions by virtue of targeted long-distance axonal projections to allow interactions between specific cell types. Here we have used Epi-Retro-Seq to link single cell epigenomes and associated cell types to their long-distance projections for 33,034 neurons dissected from 32 different source regions projecting to 24 different targets (225 source→target combinations) across the whole mouse brain. We highlight uses of this large data set for interrogating both overarching principles relating projection cell types to their transcriptomic and epigenomic properties and for addressing and developing specific hypotheses about cell types and connections as they relate to genetics. We provide an overall synthesis of the data set with 926 statistical comparisons of the discriminability of neurons projecting to each target for every dissected source region. We integrate this dataset into the larger, annotated BICCN cell type atlas composed of millions of neurons to link projection cell types to consensus clusters. Integration with spatial transcriptomic data further assigns projection-enriched clusters to much smaller source regions than afforded by the original dissections. We exemplify these capabilities by presenting in-depth analyses of neurons with identified projections from the hypothalamus, thalamus, hindbrain, amygdala, and midbrain to provide new insights into the properties of those cell types, including differentially expressed genes, their associated cis-regulatory elements and transcription factor binding motifs, and neurotransmitter usage. Overall design: In Epi-Retro-Seq, the retrograde viral tracer rAAV2-retro-Cre is injected in the target region in an INTACT mouse, turning on Cre-dependent nuclear-GFP expression in neurons that project to the injected target, throughout the mouse brain. The brain is then sectioned into eighteen 600-micron coronal slices, and the brain regions of interest are mannually dissected from each slice. Nuclei are sampled from 4 mice (2 male and 2 female) for each projection target. Nuclei from each of the dissected source regions are prepared, from which GFP+/NeuN+ nuclei (the GFP-labeled projection neurons) are isolated as single nuclei using fluorescence activated nuclei sorting (FANS) and assayed using snmC-Seq2/3 to profile their genome-wide DNA methylation signatures. Low-quality nuclei were removed.

单细胞遗传与表观遗传分析（single-cell genetic and epigenetic analyses）可将大脑数十亿神经元划分为数千个“细胞类型（cell-type）”簇，每个簇均分布于不同的脑区结构中。
此类细胞类型中的多数，通过靶向性长距离轴突投射实现特定细胞类型间的相互作用，从而介导其独特的生理功能。
本研究采用Epi-Retro-Seq技术，对全小鼠脑中源自32个不同起源脑区、投射至24个不同靶区（共225种起源→靶区组合）的33034个神经元进行分析，将单细胞表观基因组、对应细胞类型与其长距离投射特征相关联。
本研究展示了该大型数据集的多重应用场景：既可用于探究投射神经元类型与其转录组、表观基因组特征之间的通用规律，也可用于针对与遗传学相关的细胞类型及神经连接提出并验证特定科学假说。
我们对该数据集进行了全面整合，并针对每个解剖起源脑区中投射至各靶区的神经元的可区分性开展了926次统计学比较。
我们将本数据集整合至由数百万神经元构成的已注释BICCN细胞类型图谱（BICCN cell type atlas）中，从而将投射神经元类型与共识细胞簇进行关联。
结合空间转录组数据的进一步分析，可将富集投射神经元的细胞簇定位至比原始解剖取材更小的起源脑区中。
我们通过对源自下丘脑（hypothalamus）、丘脑（thalamus）、后脑（hindbrain）、杏仁核（amygdala）和中脑（midbrain）的已鉴定投射神经元进行深入分析，展示了该数据集的上述应用能力，为这些细胞类型的特征提供了新见解，包括差异表达基因、相关顺式调控元件、转录因子结合基序以及神经递质使用模式等。
总体实验设计：在Epi-Retro-Seq技术流程中，我们向INTACT小鼠的靶脑区注射逆行病毒示踪剂rAAV2-retro-Cre，使全脑内投射至该注射靶区的神经元中启动Cre依赖的核GFP表达。
随后将小鼠大脑切成18片厚度为600微米的冠状脑片，并从每片脑片中手动分离目标脑区。
针对每个投射靶区，我们从4只小鼠（2雄2雌）中取样细胞核。
制备每个解剖起源脑区的细胞核样本，利用荧光激活细胞核分选技术（fluorescence activated nuclei sorting, FANS）分离出GFP+/NeuN+的细胞核（即GFP标记的投射神经元），并采用snmC-Seq2/3技术对其全基因组DNA甲基化特征进行检测分析。
我们剔除了质量不达标的细胞核样本。