Abstract

In mammalian cell nuclei chromosome territories (CTs) occupy positions correlating with their gene-density and chromosome size. While this global radial order has been well documented, the question of whether a global neighborhood order is also maintained has remained a controversial matter. To answer this question I grew clones (of HeLa, HMEC and human diploid fibroblast cells) for up to 5 divisions (32 cells) and performed 3D FISH experiments to visualize the nuclear positions of 3 different CT pairs. Using different landmark-based registration approaches I assessed the similarity of CT arrangements in daughter cells and cousins. As expected from a symmetrical chromatid movement during mitotic anaphase and telophase, I was able to confirm previous findings of a pronounced similarity of CT arrangements between daughter cells. However, already after two cell cycles the neighborhood order in cousins was nearly completely lost. This loss indicates that a global neighborhood order is not maintained. Further, I could show in the present thesis that a gene density correlated distribution of CTs, which has already been shown in different cell types of various species appears to be independent of the cell cycle. Moreover I could provide evidence that the nuclear shape plays a major role in defining the extent of this gene-density correlated distribution, as nuclei of human, old world monkey and bovine fibroblasts showed an increased difference in the radial distribution of gene poor/dense CTs when their nuclei were artificially reshaped from a flat ellipsoid to a nearly spherical nucleus. The observation that a gene-density correlated distribution of CTs has been found in nuclei from birds to humans argues for a significant, yet undiscovered functional impact. So far CTs have been investigated mainly in cultured cells and to some extent in tissues, yet little is known about the origin and fate of CTs during early development. To gain insights into the very early organization of CTs in preimplantation embryos I have developed a fluorescence in situ hybridization (FISH) protocol, which enables the visualization of CTs in three dimensionally preserved embryos. Using this protocol I have investigated CTs of bovine chromosomes 19 and 20, representing the most gene-rich and gene-poor chromosomes, respectively. Equivalent to the distributions described in other species I could confirm a gene density related spatial CT arrangement in bovine fibroblasts and lymphocytes with CT 19 being localized more internally and CT 20 more peripherally. Importantly, I did not find a gene density related distribution of CTs 19 and 20 in early embryos up to the 8-cell stage. Only in embryos with more than 8 cells a significant difference in the distribution of both chromosomes became apparent that increased upon progression to the blastocyst stage. Since major genome activation in bovine embryos occurs during the 8- to 16-cell stage, my findings suggest an interrelation between higher order chromatin arrangements and transcriptional activation of the embryonic genome. Using another experimental set up I analyzed the topology of a developmentally regulated transgene utilizing bovine nuclear transfer (NT) embryos derived from fetal fibroblasts, which harbored a mouse Oct4/GFP reporter construct integrated at a single insertion site on bovine chromosome 13. I analyzed the intranuclear distribution of the transgene as well as its position in relation to its harboring chromosome in donor cell nuclei and day 2 NT embryos, where the transgene is still inactive as well as in day 4 NT embryos, where transgene expression starts, and day 7 NT embryos, where expression is highly increased. Compared to donor cell nuclei I found a more peripheral location of both BTA 13 CTs and the Oct4/GFP transgene in day 2, day 4 and day 7 NT embryos, although there was a trend of the transgene and both BTA 13 CTs to re-localize towards the nuclear interior from d2 to d7 embryos. Moreover, I found the transgene located at the surface of its harboring CT 13 in donor fibroblasts, whereas during preimplantation development of NT embryos it became increasingly internalized into the chromosome 13 territory, reaching a maximum in d7 NT embryos, i.e. at the developmental stage when its transcription levels are highest. These latter experiments show that the transfer of a somatic nucleus into a chromosome depleted oocyte triggers a large scale positional change of CTs 13 and of an Oct4/GFP transgene and indicate a redistribution of this developmentally regulated Oct4/GFP transgene during activation and upregulation in developing NT embryos.