Single-cell Analysis Reveals the Molecular Roadmap of Mouse Embryogenesis

Abstract

Mammalian embryogenesis is a rapid and complex process that involves the proliferation and diversification of cells. Within a few weeks, a single-cell zygote gives rise to hundreds of millions of cells that express a wide range of molecular programs. These cells eventually give rise to all of the tissues and organs in the adult body. There are two major questions in mammalian embryogenesis: how cells transition from one stage of development to the next, and what the molecular factors that control this process are. Our current understanding of these questions, particularly in the in vivo context, is still incomplete. Although intensively studied, a systematically ascertained delineation of the major cell lineages and trajectories that comprise in vivo mammalian development, as well as the regulatory transcription factors (TFs) that drive it, remains elusive. Recently, we and other researchers have used a variety of single-cell methods to profile millions of cells from a whole mouse embryo. This allowed us to characterize the transcriptional profiles (single-cell RNA-seq, or scRNA-seq), chromatin accessibility (single-cell ATAC-seq, or scATAC-seq), and other modalities of individual cells, from a series of “snap-shots” of the embryo at different timepoints, spanning the pre-gastrulation, gastrulation, and organogenesis periods. For example, Pijuan-Sala et al. captured the transcriptional profiles of over 100,000 individual cells from the mouse embryo during gastrulation, from embryonic day (E) 6.5 to E8.5. Cao et al., on the other hand, developed a new single-cell combinatorial indexing technology (sci-RNA-seq3) to profile the transcriptomes of around 2 million cells derived from mouse embryos staged between stages E9.5 and E13.5. I hypothesized that these and other single-cell datasets, especially if they were combined, would offer significant possibilities for obtaining a comprehensive understanding of mouse embryogenesis. Therefore, this thesis will introduce three different projects that leverage scRNA-seq datasets with comprehensive computational analysis to address this question. In the first project, we integrated multiple published scRNA-seq datasets to define cell states at 19 stages of mouse gastrulation and early organogenesis, from E3.5 to E13.5. We then heuristically connected these cell states to their pseudo-ancestors and pseudo-descendants based on their transcriptional similarity, creating a graph of cellular trajectories. We then leveraged this graph to identify TFs and TF motifs that regulate the emergence of new cell types. In the second project, we applied sci-RNA-seq3 to fill the data gap between late gastrulation and birth, by profiling the transcriptional states of 12.4 million nuclei from 83 precisely staged embryos spanning late gastrulation (E8) to birth (postnatal day 0 or P0), with a temporal resolution of at least 6 hours. We identified hundreds of cell types and performed deeper analyses of the development of the posterior embryo during somitogenesis, as well as the kidney, mesenchyme, retina, and early neurons. We leveraged these data, together with other published datasets, to construct a rooted tree of cell type relationships that spans mouse development from zygote to pup, followed by identifying sets of TFs and other genes as candidate drivers of cell type differentiation. In the third project, we aimed to establish scRNA-seq of the whole embryo as a scalable platform for the systematic phenotyping of mouse genetic models. We used sci-RNA-seq3 to profile and analyze the gene expression of 1.6 million nuclei from 101 embryos of 22 mutants and 4 wild-type genotypes at E13.5. We then developed and applied several analytical frameworks to detect differences in the composition and/or gene expression of 52 cell types or trajectories across genotypes. This study takes a comprehensive approach to understanding mammalian embryogenesis. By examining the cellular trajectories that cells take during development and the molecules that drive these trajectories, we aim to answer fundamental questions about this process. We believe that this work will provide a foundation for a deeper understanding of mammalian development.