The transcriptional activity of genes as well as the conformation of transcriptionally silenced genomic regions is defined by the epigenetic state of the associated chromatin fragment. Chromatin is composed of repetitive units called nucleosomes. Each nucleosome consists of a core complex of four histones and the fragment of DNA that is wrapped around it. The distinct residues of N-terminal tails of the histones that extend out from the nucleosomal core are subject to post-translational epigenetic modifications. The composition of these modifications defines the electric charge of the histone tails and, therefore, its coupling strength with DNA. Thus, the epigenetic mark composition of each individual nucleosome governs the overall conformation of the chromatin filament. In mammals, the tight conformation of chromatin in pericentric genomic regions, called pericentric heterochromatin (PCH), is important for the stability and the proper segregation of chromosomes. The epigenetic hallmark signature of PCH is di- or trimethylation of histone H3 at lysine 9 (H3K9me2/3) enriched over the entire centromere region. This epigenetic state is constitutively maintened through cell cycle progression and throughout multiple cell generations thereby preventing chromosomal breakages, missegregation and perturbed chromosomal interactions. This work investigates the specificity, propagation and the long-term autonomous memory effect for H3K9me2/3 silencing marker in PCH in mouse fibroblasts on the single cell level. We apply fluorescent microscopy techniques, high-throughput image-processing method and mathematical modeling to test the stability of the system upon changes in the expression of the chromatin-modifying enzymes. The network operating H3K9me2/3 in PCH was constructed and translated into a deterministic system of ordinary differential equations as well as formulated stochastically using the Gillespie simulation algorithm. The model incorporates the contribution of H3K9me2/3 binding protein HP1 together with H3K9 specific methyltransferase Suv39h, the H3K9 specific demethylase JMJD2 and cell cycle dependent kinase Aurora B as well as nucleosome collision processes via DNA looping. The realization that most of these chromatin-modifying processes depend on each other and also appear to be regulated by multiple positive and negative feedback loops has lead to the proposal of nonlinear stationary and dynamical features of PCH network function. The modeling simulations have revealed an increased variability in methylation degree upon increases in JMJD2 expression in the cell. This prediction was qualitatively supported by the heterogeneity observed experimentally on the single PCH foci level. However, in the experiment the response of the whole cell population remains monostable, in spite of the presence of a bistable memory element in the network. The model explains this property confirming the significant impact of the persistent silencing origins organized by the high residence time binding of HP1-Suv39h complexes that initiate the spread of the H3K9me2/3 mark. Therefore, on the population level the bistable mechanism of silencing propagation in PCH is hidden, appearing only as severe fluctuations in the H3K9me2/3 level. In summary, a consistent model of the silencing propagation in PCH was developed. Based on that model a scenario of PCH epigenetic state maintenance and robustness towards transient perturbation and intrinsic noise is established.
|Supervisor:||Kummer, Prof. Dr. Ursula|
|Place of Publication:||Heidelberg University|
|Date of thesis defense:||20 March 2014|
|Date Deposited:||03 Apr 2014 07:27|
|Faculties / Institutes:||The Faculty of Bio Sciences > Dean's Office of the Faculty of Bio Sciences|
|Subjects:||500 Natural sciences and mathematics
570 Life sciences