A refined model of nucleolar formation derived from the analysis of UBF variants and the hierarchical silencing of human nucleolar organizer regions

The nucleolus, the largest sub-nuclear structure, forms around arrays of ribosomal RNA genes, termed nucleolar organizer regions (NORs). As the site of ribosome biogenesis, nucleolar function is intrinsically linked to cellular growth and proliferation control. Nucleolar morphology has long been used as a prognostic marker of cancer. This is one of the many examples of how aberrant nucleolar function is related to disease. The biochemistry of ribosome biogenesis has been extensively studied, leading to a comprehensive understanding of the process. However, we know relatively little about how the nucleolus is formed, and its relationship with NORs. Each diploid human cell contains around 600 copies of the human ribosomal RNA gene (rDNA) arranged in tandem repeats. The rDNA repeat is composed of the pre-ribosomal RNA coding sequences separated by non-transcribed intergenic spacer regions (IGS). NORs are located on the short arms of all five acrocentric chromosomes. Upstream Binding Factor (UBF), a nucleolar specific HMG-box protein, is essential for establishing and maintaining the activity status of ribosomal RNA genes. It binds constitutively across the whole of the rDNA repeat and generates a decondensed, transcriptionally permissive chromatin state. NORs may either exist in a UBF bound, transcriptionally active state, or a silent state. During mitosis, active, UBF bound NORs, are achromatic and therefore are referred to as secondary constrictions. In this thesis, I characterise the rDNA distribution across NORs for several cell lines, along with their activity status. This was accomplished using ImmunoFISH techniques. I determined that the majority of NORs having a detectable number of rDNA genes are UBF loaded and therfore active. To study how NORs form nucleoli, and the role of UBF in this process, I depleted UBF in these cells. I examined the effects on the activity status of NORs and their ability to participate in nucleolar biogenesis. Interestingly, I found that NORs were inactivated in a hierarchical manner, with NORs having smaller numbers of rDNA repeats being preferentially inactivated and dissociated from nucleoli. Similar results were observed in growth inhibited cells, illustrating for the first time that the activity status of NORs can be modulated in response to changes in cell state. In order to better describe the nature of the localization of UBF to the rDNA in nucleoli, I performed FRAP experiments to study UBF dynamics at these loci. This was done using a cell line expressing an endogenously tagged UBF_GFP gene, thereby ensuring the biological relevance of results obtained. A population of cells in which UBF dynamics was reduced was identified. This population has not been previously described. I then illustrated that driving cells into a quiescent state by either serum deprivation, or contact inhibition resulted in a similar reduction in UBF dynamics. The driving force for UBF localization and interaction with the rDNA is poorly understood. I analysed the N-terminus of UBF to determine if it plays some role in these functions. The first ~100 aa of this protein have been shown to contain both an extensively characterised dimerization domain and a putative SANT-like domain. Mutation of three conserved residues within the SANT-like domain abrogates localization. My results indicate that dimerization is crucial for correct localization of UBF, and furthermore suggest that the N-terminus has histone tail binding capabilities. In the final discussion I combine my findings about the behaviour of active NORs with recently described research regarding the biophysical properties of nucleoli and propose a refined model of nucleolar formation