Single-cell measurement of dynamic cellular behaviors in development, regeneration, and malignancy

Abstract

While we are now often able to precisely measure the current molecular state of individual cells in mammalian tissues using single cell omics technologies, it can still be difficult to study dynamic behaviors such as differentiation, proliferation, and cell death in situ at a similar level of detail. One general strategy for characterizing these behaviors within the native tissue context in multicellular organisms is to measure both current transcriptomic cell state as well as additional information about shared cell ancestry, past molecular states, etc in the same set of individual cells. During my PhD, I worked on multiple projects in which we augmented static single-cell phenotype data with additional information on past cell state in order to study biological systems undergoing rapid cellular changes, such as the bone marrow and the developing mammalian embryo.

Chapter 1 describes our work investigating the effects of the most common driver mutation (JAK2-V617F) in myeloproliferative neoplasms on patient hematopoietic stem and progenitor cells. We measured the transcriptome and JAK2 genotype in single cells from bone marrow from these patients and determined that the mutation had a direct effect on hematopoietic cells, affecting their differentiation behavior and increasing their expression of inflammation-related genes. Additionally, we observed that bone marrow monocytes with the mutation also had higher expression of some pro-inflammatory genes and expressed a cell surface marker associated with the development of fibrosis.

In Chapters 2 and 3, we set out to measure hematopoietic stem and progenitor cell differentiation and division kinetics in vivo using two different experimental strategies. In Chapter 2, we developed a new system that temporarily labels cells at the top of the differentiation hierarchy using multiple fluorescent proteins. These proteins are gradually lost due to cell division and degradation after cells differentiate out of the hematopoietic stem cell (HSC) state, allowing us to estimate the amount of time and number of cell divisions that have passed since these cells were HSCs. Unfortunately, we were not able to achieve high enough initial fluorescent protein expression levels for the experimental system to be useful. Instead, we used a complementary approach, described in Chapter 3, to measure HSC differentiation rates using permanent, inducible fluorescent labeling of long-term HSCs. We measured the rate of fluorescent label propagation in downstream hematopoietic progenitor cell populations and used mathematical modeling to conclude that HSCs differentiate slowly into downstream cell types.

Finally, in Chapter 4 we investigated the process of cell fate specification during early mouse embryogenesis by measuring both transcriptional state and lineage history information in individual cells just after gastrulation. To do this, we used a CRISPR barcoding system in which the barcode sequences encoding lineage information are transcribed and therefore able to be read using scRNA-seq. We found that fate restriction occurs gradually during gastrulation and investigated the origins of endothelial cells in different regions of the mouse embryo.