Scalable methods for genomic analysis of in vitro models of mammalian embryogenesis

Abstract

Mammalian development, from the single-celled zygote to a multicellular individual, is an incredible dynamic journey that is marked by many milestones measurable across many scales. In fact, by the end of the first two weeks of human embryogenesis, most precursors of major tissues and organs required for life are already present. This developmental milestone is known as gastrulation. Here the embryo or gastrula undergoes invagination, creating the blastopore and three major layers. For example, the outermost layer, known as ectoderm, gives rise to the nervous system and skin; the middle layer, known as the mesoderm, gives rise to the musculoskeletal system and the heart; the innermost layer, known as the endoderm, gives rise to internal organs such as the lungs and liver. Collectively, these developmental cell types constitute the three major germ layers. Thus it is at the stage of gastrulation that cells of the embryo are specified toward distinct fates, leaving behind their relatively indistinct transcriptional states as pluripotent precursors. The advent of large consortia efforts, like the Human Genome Project or ENCODE, has ushered in new sequencing technologies, e.g. single-cell molecular phenotyping modalities like scRNA-seq, that are capable of uncovering the individual components or features of the genome that support the blueprint for multicellularity. For example, we now know that the genome can be partitioned into two categories: the coding genome and the non-coding genome. The coding genome is largely made up of genes, including cell-type specifying transcription factors (TFs). While approximately ~22,000 protein-coding genes have been decoded and cataloged, of which ~1600 or so are thought to be TFs, the overall coding proportion only makes up 1-2% of the mammalian genome. The other 98% of the genome is defined by the non-coding genome, where ~1 million non-coding regulatory elements, namely enhancers, are thought to reside. Despite our ever-growing knowledge of the genome, we know very little about the transcription factors or enhancers that are required for the myriad of cell types required for mammalian development. How this remarkable process unfolds at the molecular level is a timely question that remains elusive. The focus of my PhD has been to elucidate how the process of early development works, particularly when cells undergo cell fate specification during gastrulation. More specifically, I have been intensely focused on understanding the dynamics of germ layer formation through 1) functional characterization of non-coding DNA elements or enhancers, 2) defining key developmental transcription factors, and 3) tracing histories of cell lineages as they are emerging within a multicellular system. To tackle these complex areas of investigation, I have developed scalable methods applied to multicellular in vitro embryoid model systems of early development. In the first chapter, I describe current strategies to understand early development and cell fate specification. In the second chapter, I describe efforts to perturb and record lineages using a novel platform for clonal organoid generation. In the third chapter, I describe a highly multiplexed method with single-cell resolution for measuring autonomous activity of non-coding regulatory DNA in a multicellular context. Finally, in the last chapter, I conclude with my thoughts on the future of in vitro models alongside multi-modal measurements.