Organization of chromatin and chromosomes in relationship to nuclear architecture
MetadataShow full item record
The overall hypothesis driving this thesis is that nuclear architecture and genomic function in the interphase cell nucleus are interdependent and thus alterations in structural arrangements in the nucleus could lead to corresponding alterations in nuclear function. To test this hypothesis, this thesis has performed novel research involving several complementary projects that investigate: (1) the role of gene amplification and chromatin organization to function in the cell nucleus; (2) the three dimensional arrangement of a large subset of human chromosomes in the cell nucleus; (3) the structural associations of a known nuclear matrix protein with chromosome territories in the cell nucleus and initial definition of proteins that potentially interact with matrin 3 in the cell nucleus. Comparison of the type I interferon (IFN) gene cluster in normal WI38 fibroblasts and the highly disrupted MG63 osteosarcoma nuclei offers a system to probe how alterations in the nuclear landscape affect function (Chapter 2). Fluorescent in situ hybridization (FISH) experiments revealed a 5 fold amplification of the IFN gene cluster in MG63 nuclei with the majority of foci gathered together in a region proximal to the nuclear periphery. The amplified IFN gene cluster was found located on a chromosome territory containing regions of chromosomes 4, 8 and 9 and interspersed within the IFN gene cluster foci were corresponding foci derived from amplified centromere 4 and 9 sequences. Although replication timing of the IFN gene cluster was found to be conserved in both cell lines transcriptional activity was observed to be altered. Quantitative RT-PCR demonstrated only half the expected increase in IFNβ transcripts in MG63 compared with WI38 based on the degree of gene amplification. RNA/DNA FISH experiments revealed 1-5 foci of IFNβ transcripts per cell with only approximately 5% of the cells showing RNA foci within the highly amplified IFN gene cluster indicating the susceptibility of gene regulation to aberrations in chromatin structure. To dissect the normal genome arrangement in fibroblast nuclei, the chromosome territory organization of eight of the large human chromosomes were studied in G0 synchronized WI38 diploid fibroblasts (Chapter 3). Radial positioning of these 8 chromosomes along with the gene poor chromosome 18 and gene rich chromosome 19 correlated well with a size-based distribution in which the larger chromosomes are closer to the nuclear periphery. Analysis of chromosome positions indicated preferential associations between many chromosome combinations. Computer imaging and modeling of the data confirmed that these associations were not a result of radial positioning or chromosome volumes. A geometric optimization method termed the Generalized Median Graph enabled identification of the most probable topological arrangement of these 8 chromosome pairs. High and low associating chromosome pairs from WI38 fibroblasts were then tested individually in fibroblast cell lines derived from lung, colon and skin tissue origins. While the percentage associations of chromosome neighbors were quite similar in all the lung fibroblasts, there were considerable differences when compared with the skin and colon derived fibroblast cells. We conclude that a large subset of human chromosomes has a preferred probabilistic spatial arrangement in WI38 cells and that the resulting chromosome-to-chromosome associations show tissue origin specificity. Together, the data presented here is consistent with a highly structured nucleus and suggests functional implications to genome organization due to the variability between cells from different tissues. It also demonstrates the importance of proper nuclear organization in regulating nuclear function. The abundant inner nuclear matrix protein, matrin 3 (matr3), was studied to determine its spatial organization with regards to chromosome territories and its interactions with other nuclear localized proteins (Appendix). Matr3 was found located throughout all autosomal chromosomes analyzed yet was observed to be excluded from the inactive X chromosome. There was also an absence of matr3 labeling from within both perinuclear and perinucleolar heterochromatin. Yeast two hybrid analysis frequently found matr3 to interact with itself in addition to 33 unique nuclear localized proteins. The vast majority of these proteins had previously identified roles in RNA metabolism. Further analysis of a selection of these proteins and SAFA by co-localization and co-immunoprecipitation experiments confirmed the interactions with matr3. Down regulation of matr3 by RNA interference was found to have little effect on global nuclear function.