Structure and Function of Chromatin Domains at the Centromere Loci

The centromere is a specialized chromosomal domain forming on the highly repetitive alpha satellite DNA, upon which the kinetochore—the protein complex that links the centromere to spindle microtubules—assembles and directs the equal segregation of replicated sister chromatids to daughter cells during cell division. Centromere protein A (CENPA), the human centromere histone H3 variant that is found at all active kinetochores, is suggested to provide the key epigenetic information to mark the centromere identity independent of DNA sequence. In addition to CENPA, centromeres assemble large domains of pericentric heterochromatin that is thought to be important for maintaining their structure and function. Regarded as one of the most astonishing examples of epigenetic changes within the genome, neocentromeres are fully functional centromeres that form ectopically on non-repetitive DNA sequences and therefore, serve as an excellent model system for the study of the structure and function of different centromeric chromatin domains.;In this thesis I describe a unique case of centromere repositioning in humans, with a neocentromere forming at band 8q21 in an otherwise unrearranged chromosome 8, in which the endogenous centromere was inactivated. CENPA ChIP-seq demonstrated that this neocentromere is unique as it is located on a large variable tandemly repeated DNA (all other mapped neocentromeres are found in single copy non-repetitive DNA). Using this neocentromere as a model system to study the structure and function of endogenous centromeres, I have characterized its chromatin organization through ChIP-Seq of histone PTMs, and histone variants, in order to further understand the linear and higher order chromatin structures of the functional centromere. My observations suggest the presence of alternating CENPA and H3K9me3/H4K20me3 containing chromatin domains on this tandem array repeat.;I have further shown that pericentric heterochromatin is important for proper sister chromatid segregation and for the correction of monastrol-induced spindle connection errors. In addition, I observed the absence of the H3K4me2 and H3K9me2 histone PTMs and the H2A.Z variant, suggesting that they do not play a higher-order structural role at neocentromeres, contrary to what was suggested for endogenous centromeres. Based on my results, I further propose a model for the folding of the linear chromatin fiber into the centromeric specific higher-order chromatin organization.;Using the neocentromeric model system developed in our lab, I further investigate the physical properties of the CENPA containing particle. Recent observations both in vitro and in vivo have put forth conflicting models regarding the stoichiometry of the CENPA containing particle. Using paired-end sequencing of CENPA ChIP DNA acquired from MNase digested chromatin, I have shown for the first time in-vivo that the CENPA-containing particle is an octamer with loose termini at the DNA entry/exit site. I show that the CENPA containing octamer protects only the central ∼110bp of DNA unlike the 147bp protected by the canonical H3 containing-nucleosome, possibly altering the CENPA chromatin environment and making it unique compared to the chromatin along chromosome arms. I further demonstrated the same DNA fragment size distribution at endogenous centromeres suggesting it is not dependent on specific DNA sequences. Finally, I show that the CENPA octameric structure is not cell cycle dependent.;Collectively, I have characterized a neocentromeric model system that is well suited for the study of centromere biology. Using molecular cytogenetic and ChIP-seq approaches I have gained insights into the potential linear and high order organization of the centromeric chromatin. In addition, I have determined some of the epigenetic factors that are essential for proper centromere function. Finally, I have demonstrated in vivo the unique characteristic of the CENPA-containing nucleosome important for its role as the epigenetic mark that potentially provides a basis for how centromere identity is stably inherited.