Neuron types in the developing mouse brain are divided into partially matching epigenomic and transcriptomic classes

Abstract

We hypothesize that the differential activity of transcription factors (TFs) regulating patterning and cell fate determination manifests in a differential genome activation in different cell lineages and cell types. Here, we have mapped mRNA expression and accessible chromatin regions in single cells derived from the developing mouse CNS. We first defined a reference set of features for scATAC-seq read quantitation across samples, allowing comparison of chromatin accessibility between brain regions and cell types directly. Second, we integrated the scRNA and scATAC data to form a unified resource of transcriptome and chromatin accessibility landscape for the cell types in diencephalon to anterior hindbrain in E14.5 mouse embryo. Importantly, we implemented methods to achieve high level of statistically supported resolution of clustering for single-cell -omics. We show high level of concordance between the cell clustering based on the chromatin accessibility and transcriptome in analyzed neuronal lineages, indicating that both genome and transcriptome can be used for cell type definition. Hierarchical clustering by the similarity in genome activation reveals complex interrelationships between the cell types derived from different anteroposterior domains of the E14.5 mouse CNS. Further work is required to describe the fraction of genome activation resulting from the activity of patterning genes and the fraction activated by the cell fate selection regulators.