The single-cell transcriptional landscape of hematopoiesis

Abstract

Hematopoietic stem cells (HSCs) are responsible for the production of all mature hematopoietic cells during the entire life of an organism. It is now known that HSCs are generated from a specific subset of endothelial cells in the main arteries of the developing mouse and zebrafish embryos. In the aorta of the mouse embryo, HSCs are first detected as part of intra-aortic hematopoietic clusters, transient small aggregates of cells in contact with the blood circulation. In the zebrafish embryo, HSCs undergo a sub-aortic ingression before being released into the blood stream. In both species, HSCs are later amplified in secondary sites before migrating to the bone or kidney marrow where they will reside during adulthood. In the past ten years, technological advances such as single-cell mRNA Sequencing, have deepened our understanding of the molecular mechanisms underlying the endothelial to hematopoietic transition. Additionally, novel sophisticated lineage-tracing methods now allow us to track the progeny of single hematopoietic cells. This thesis aim is to characterize the mechanisms underlying HSC production and function in the mouse and zebrafish. In addition, this thesis contains a variety of novel methods derived from single-cell mRNA-Sequencing technologies that have been applied in the field of hematopoiesis. In chapter 2, we present a single-cell mRNA-Sequencing dataset of various endothelial and hematopoietic cell populations isolated from the mouse embryonic aorta. Additionally, we developed a novel method to mechanically isolate intra-aortic hematopoietic clusters based on their physical location inside the aorta. We show that transcriptome differences between dorsal or ventral intra-aortic hematopoietic clusters are minimal. Our dataset is a resource for the scientific community to guide further functional studies of key regulators of HSC production. In chapter 3, we introduce ScarTrace, a method to track the clonal origin and characterize the transcriptome of thousands of zebrafish single cells, including zebrafish hematopoietic cells. We find that a small amount of mesodermal ancestors are responsible for the production of all hematopoietic cells in the adult zebrafish. In addition, we use ScarTrace to study the clonal history of the zebrafish brain, eyes and caudal fin during regeneration. Our results allowed to refine the time window of right-left specification of head organs and to discover clonally distinct populations of myeloid cells in the caudal fin. In chapter 4, we present GateID, a novel algorithm to purify cell types by combining single cell FACS and mRNA-Sequencing data. GateID allows unbiased cell type identification of single cells based on their transcriptome and prediction of non-intuitive FACS gates based on FACS index data. Importantly, GateID allows enriching for cell types without using specific antibodies or fluorescent transgenes. We demonstrate the power of GateID by enriching several hematopoietic cell types from the zebrafish whole kidney marrow (including HSCs) and alpha and beta cells from the human pancreas. Ultimately, GateID can be used to enrich any cell type of interest after generation of a single-cell mRNA Sequencing and FACS index dataset necessary to predict FACS gates.