PDF, English - main document Download (5MB) | Terms of use

Abstract

Understanding the development of tissues and organs at a single cell level remains a challenge. In this thesis I present a novel barcoding system Polylox, recently developed by Hans-Reimer Rodewald and colleagues. Based on the loxP-Cre recombination system, Polylox allows endogenous barcoding of single cells in vivo. Using a Markov chain model for the recombination process, I find that n = 1,866,890 individual barcodes can be generated. Due to the structure of Polylox, barcodes have different generation probabilities. The mathematical model presented in this thesis, calibrated against experimental Polylox data, allows the assignment of generation probabilities to each observed barcode and the selection of informative barcodes based on their generation probabilities for clonal analyses. Experimental collaborators induced barcodes in hematopoietic stem cells (HSC) and I analysed the clone size distributions, finding large clone sizes of up to 3.8% in young mice (< 1 year old) and 21.5% in old mice (2 years old) of HSC in the adult bone marrow. I show that the appearance of large HSC clones in older mice is explained by a neutral drift model. Sampling from mature populations of the hematopoietic system revealed that a very large proportion of HSC contributes to adult hematopoiesis (85.7%). Additionally, many HSC realize multipotency in vivo, yet clustering analysis of barcode frequency distributions revealed a fundamental split between myelo- erythroid and common lymphoid lineage development. These findings support the longheld, but currently contested, view of a tree-like hematopoietic structure with few major branches. The description of the potential of common myeloid progenitors (CMP) as myeloerythroid restricted is largely dependent on transplantation and colony assays. Analysis of Polylox data places the CMP compartment downstream of the split inside the myelo-erythroid branch and shows the myelo-erythroid potential of CMP. By building a mathematical framework that allows the computing of the time evolution of the moments of barcode clone sizes, I show that Polylox data is consistent with previous work by Busch et al. further supporting the tree model of hematopoiesis. In addition to the analyses of single barcodes, I use network analysis techniques on observed barcode sets. The connectivity of barcode sets reflects the proliferative state of the system during labelling. I found evidence for a strong proliferative burst of at least three divisions a day in HSC progenitor cells at the time point of fetal liver formation at embryonic day 9.5 in the mouse embryo. Polylox proves to be a valuable technique for fate mapping that is not only applicable to hematopoiesis but a multitude of systems.