The centromere is the site of kinetochore assembly that ensures proper chromosome segregation. Active centromeres are formed at chromosomal locations that do not appear to share homology between different species; this and other analyses has lead to the conclusion that centromeres are epigenetically determined. In all organisms, centromere location is specified by the assembly of unusual nucleosomes containing the histone H3 variant CENP-A in place of H3. However, an apparent paradox is that CENP-A in most organisms generally occurs on certain preferred sequences. The analyses presented focuses on the influence of DNA sequence on the selection of the locus where CENP-A chromatin are formed and whether there are any particular DNA features that promote CENP-ACnp1 assembly at centromeres in Schizosaccharomyces pombe. S. pombe provides an excellent model to study the structure and function of a complex eukaryotic centromere as it possesses epigenetically regulated centromeres that are structurally related to those of metazoa. Furthermore, plasmid-based circular minichromosomes provide a useful tool for studying these centromeres and the inactivation of minichromosome centromeres does not affect cell viability. The main aim of this study was to identify those features that enable centromeric sequences to assemble CENP-A chromatin. Manipulated circular minichromosomes were utilised to investigate the minimal central core sequence requirement for establishment of CENP-ACnp1 chromatin. These analyses showed that a minimal 2kb region from the central core of cen2 could form a functional centromere. A second aim was to analyse the DNA sequence requirements for centromere function on this minimal 2 kb region. To facilitate this, the endogenous central core region of endogenous cen2 was replaced with the central core region of cen1. This modified strain allows the structural and biological properties associated with plasmid borne central core 2 sequences to be analysed. Transcription of central domain sequences has been proposed to play a role in CENP-A establishment and/or maintenance. To explore the contribution of transcription potential promoters were mapped within the minimal 2 kb sequence and their regulatory elements investigated. Mutation of the minimal DNA element impedes its ability to assemble CENP-A chromatin. Therefore the primary DNA sequence of fission yeast centromeres is important for establishing functional centromeres and thus centromere location not entirely epigenetically regulated. It remains to be determined if the characteristics associated with theses sequences, and their mode of action, are conserved at other centromeres.

Human post-translational protein modifier protein SUMO-1/2/3 genes code for proteins homologous to yeast SMT3 protein, which is encoded by a suppressor 3 of MIF2 mutation in centromere protein gene. The yeast MIF2 protein shares at least two regions of similarity with mammalian centromere protein CENP-C. It would be of interest to investigate the possible interaction(s) between human CENP-C and SUMO-1/2/3 proteins. A CENP-C cDNA fragment was cloned using RT-PCR with total RNAs form Hela cells. This cDNA fragment encoding CENP-C amino acids 342-764 (MW 38 kDa designated C38) was tagged with EGFP. The sub-cellular localization and in vivo sumoylation in HeLa cells were carried out. The EGFP-C38 protein was shown to co-localize with active forms of Flag-SUMO-1/2/3GG proteins in nucleus of Hela cells. The EGFP -C38 protein was also shown co-immunoprecipitated with antibodies against SUMO proteins. The protein conjugates were analyzed on SDS-PAGE and their western blots were probed with either anti-GFP or anti- Flag antibodies. The molecular weight of EGFP-C38 protein was found to be higher than the expected MW, indicating that EGFP-C38 protein was sumoylized. This part ( 333 amino acids) of CENP-C protein (943 amino acids) was expressed and purified. The in vitro sumoylized His-C38 protein fragment was analyzed on SDS-PAGE, and the western blot was probed with either anti-SUMO-1 or anti-SUMO-2 antibodies. The C38-His protein fragment appeared to be sumoylized, and the isopeptide bond between the C-terminal glycine of SUMO and lysine of His-C38 was analyzed by MALDI-TOF-TOF. C38 cDNA was sub-fragmented into C28 and C10 fragments transformed to BL21 strain for expression protein and purified protein. S-tagged SUMO-1/2GG modify C28-His and C10-His fragments, the isopeptide bond between the C-terminal glycine of S-tagged SUMO-2GG and lysine of C10-His was identified analyzed by MALDI-TOF-TOF. The isopeptide bonds between either S-tagged SUMO-1GG and C28-His or S-tagged SUMO-1/2GG and C28-His /C10-His are being analyzed.

UA62784 is a novel fluorenone identified in a biologic screen of compounds that are selectively cytotoxic in DPC4 (deleted in pancreatic cancer)-deleted pancreatic cancer cells. We sought to determine the mechanism of action of UA62784, and discovered it to be a potent mitotic inhibitor. UA62784 affects the ATPase activity of the mitotic kinesin centromere protein E (CENP-E), but does not affect other known mitotic kinesins. This inhibition of ATPase activity is not caused by an inhibition of microtubule binding nor is it caused by a failure of the kinesin to translocate to the nucleus during mitosis. Despite the anti-cancer properties of this drug, UA62784 is relatively insoluble and is not suitable as a lead compound for further development.Once we determined the mechanism of action of UA62784, we sought to determine if analogs would demonstrate the same potent mitotic inhibition while also offering properties such as increased solubility. A small library of chemical analogs was generated wherein each compound was a slight variation of UA62784 (termed the DPC series). Several potential leads were identified which exhibited increased solubility and/or increased cytotoxic activity. When tested for CENP-E ATPase inhibition, some compounds were noted to inhibit other kinesins as well. We therefore created a screen where each of the DPC compounds was tested for activity in Eg5, CENP-E, MKLP-1, MCAK, and KIF3C kinesins. Within these data, there is a correlation between cellular IC50 and kinesin ATPase inhibition for CENP-E and MKLP-1. A few compounds emerged from these studies, including DPC046, which has a low cellular IC50 and inhibits all five kinesins to some degree. DPC046 was used in a mouse xenograft study to determine in vivo efficacy, but no significant tumor shrinkage was seen, likely due to solubility limitations affecting the amount of bioavailable compound.From these studies we conclude that the cytotoxic effects seen in UA62784 and its analogs are due, at least in part, to their inhibition of kinesin proteins. We demonstrate that compounds that inhibit CENP-E and other kinesin proteins hold promise in cytotoxically targeting pancreatic cancer cells. Further development is needed to optimize DPC046 compound solubility in order to increase in vivo efficacy.

As a centromere-specific protein complex in direct contact with the DNA, CenH3-containing nucleosomes are generally thought to act as the distinguishing epigenetic mark of active centromere location. Confusingly, seemingly disparate models have been proposed for the structure of CenH3 nucleosomes. The most widely supported model is an octameric structure that, like histone H3 nucleosomes, contains two subunits of each histone. Another more contentious, yet persistent model is the hemisome model proposed for fly and human CenH3 nucleosomes. In this case it is suggested that CenH3 nucleosomes contain only single subunit of each histone. One reason for this lack of consensus is that seemingly contradicting models are often proposed, even with material from the same organism, with little overlap in experimental approaches. For example, the proposed hemisome model for fly and human CenH3 nucleosomes is predominantly based on atomic force microscopy (AFM) imaging where the height of nucleosomes on a surface is measured. These AFM measurements are the main data used by protagonists for the hemisome model. However, data supporting an octameric model for human, and other, CenH3 nucleosomes is largely based on biochemical analysis of nucleosomes prepared in vitro, with little cross-over in the methodology used to generate data to support either model. In order to reach a consensus the same analyses needs to be applied to CenH3 nucleosomes assembled in vitro or extracted from cells. Here, recombinant Schizosaccharomyces pombe CENP-ACnp1 and H3 histones expressed and purified from E. coli have been assembled into nucleosomes. To our knowledge this is the first time that recombinant S. pombe nucleosomes have been produced, allowing the stoichiometry and composition of these nucleosomes to be examined in detail by a variety of biochemical and biophysical assays. The application of AFM has enabled the height of these recombinant nucleosomes to be measured and tests the ability of AFM to infer stoichiometry using defined material. The intriguing conclusion is that octameric CenH3 nucleosomes uniquely behave as tetrameric “hemisomes” as defined by AFM. In recent years the contribution of DNA sequence to directing H3 nucleosome location has received a great deal of interest. Since CENP-ACnp1 nucleosomes wrap DNA differently to H3 nucleosomes their preference for sequences that produce a stable nucleosome is expected to be altered. The development of protocols to assemble recombinant CENP-ACnp1 nucleosomes in vitro has also been used here to assess the contribution of primary DNA sequence to CENP-ACnp1 nucleosome positioning. CENP-ACnp1 and H3 nucleosomes were reconstituted on genomic DNA at low density and the resulting nucleosomal DNA from CENP-ACnp1 and H3 particles compared by Illumina sequencing. The stability of CENP-ACnp1 and H3 nucleosomes on specific ‘H3’ and ‘CENP-ACnp1’ sequences was cross-checked. Comparing these data with in vivo CENP-ACnp1 nucleosome positions has allowed the contribution of primary DNA sequence to CENP-ACnp1 nucleosome positioning to be explored.

<p>Centromeres, or sites of chromosomal spindle attachment during mitosis and meiosis, are non-randomly distributed in complex genomes and are largely associated with expansive, near-identical satellite DNA arrays. While the sequence basis of centromere identity remains a subject of considerable debate, one approach is to examine the genomic organization of satellite DNA arrays and their potential function. Current genome assembly and sequence annotation strategies, however, are dependent on robust sequence variation, and, as a result, these regions of near sequence identity remain absent from current genome reference sequences and thus are detached from explorations of centromere biology. This dissertation is designed as a foundational study for centromere genomics, providing the initial steps to characterize those sequences at endogenous centromeres, while further classifying `functional' sequences that directly interact with, or are capable of recruiting proteins involved in, centromere function. These studies build on and take advantage of the limited sequence variation in centromeric satellite DNA, providing the necessary genomic scope to promote biologically meaningful characterization of endogenous centromere sequences in both human and non-human genomes. As a result, this thesis demonstrates possible genomic standards for future studies in the emerging field of satellite biology, which is now positioned to address functional centromere sequence variation across evolutionary time.</p> / Dissertation

The kinetochore is a highly complex proteinaceous structure located at the primary constriction of mitotic chromosomes. Here, it performs an essential role in accurate chromosome segregation. Recently, much interest has been directed towards the Constitutive Centromere Associated Network (CCAN) components, as they participate in the formation of a scaffold involved in kinetochore assembly. It is therefore important to fully understand their role, and their distribution, at the kinetochore. Although many kinetochore proteins have already been identified, it is still unclear how centromeric chromatin folds to form the structure of the inner kinetochore. This is an interesting yet still open field of study, where the literature reports are still quite divided. In this study we take advantage of the high homologous recombination efficiency in DT40 B-lymphoma chicken cell lines, allowing the generation of conditional knockouts and deletion cell lines of several centromere proteins, subsequently engineered to stably express GFP:CENP-A. In the parental cell line the unfolding properties of the centromeric region were investigated by using TEEN buffer. Using fluorescence microscopy we were able to measure the length of many unfolded centromeric chromatin fibres, from both interphase and mitotic samples, based on the signal of GFP:CENP-A. A multi-peak analysis revealed the presence of discrete populations of fibres, recognised as peaks, in both interphase and mitotic samples. Compared with interphase, mitotic centromeres showed a greater level of compaction. Next, mutants for CCAN components, blocked in mitosis, were subjected to centromere chromatin unfolding. Results revealed that mitotic kinetochores depleted of CENP-C and CENP-S behaved similarly to the parental interphase samples, suggesting a role of those proteins in maintaining kinetochore structure. In contrast, CENP-O, CENP-H and CENP-I depletion did not seem to weaken the structure of the kinetochore. Additionally, we tested a hypothesis revealed by the multi-peak analyses, that chromatin layers exist in the inner kinetochore. Our data, when combined with published electron microscopy and crystallography measurements of centromere/kinetochore components, allowed us to assemble a robust and mathematically viable model that supports a boustrophedon organisation of the kinetochore chromatin. Finally, characterization studies of the novel kinetochore protein CENP-Z were performed. An involvement of CENP-Z in controlling the levels of di-methylation on lysine 4 of histone H3 was shown. This work represents an advance in our understanding of kinetochore structure in vertebrates.

Chinese hamster ovary cells (CHO) were arrested with hydoxyurea at the beginning of DNA synthesis. Subsequent treatment with caffeine induced cells to bypass S-phase and undergo mitosis with unreplicated genomes (MUG). Treated cells built a normal spindle and distributed unattached kinetochores to daughter cells. To determine if MUG cells obey the metaphase checkpoint, we used immunoflourescence to detect and localize known metaphase checkpoint and motor proteins. In addition, the drug taxol was used to stabilize microtubules in MUG cells. The localization of CENP- E, the presence of anaphase A, taxol arrest, and taxol release acted in a similar manner as in controls. The localization of kinesin differed from the controls and that of MAD2 was inconclusive. These results imply that MUG kinetochores behave similarly to controls and probably have an operational metaphase checkpoint.

Human centromeres contain large arrays of α-satellite DNA that are thought to provide centromere function. These arrays show size and sequence variations. However, the lower limit of the sizes of these DNA arrays in normal centromeres is unknown. Using a set of chromosome-specific α-satellite probes for each of the human chromosomes, interphase Fluorescence In Situ Hybridisation (FISH) was performed in a population screening study. This study demonstrated that extreme reduction of chromosome-specific α-satellite is unusually common in chromosome 21 (screened with the αRI probe), with a prevalence of 3.70%, compared to <=.12 % for each of chromosomes 13 and 17, and 0 % for the other chromosomes. No analphoid centromere was identified in over 17,000 morphologically normal chromosomes studied. All the low-alphoid centromeres are fully functional as indicated by their mitotic stability and binding to centromere proteins including CENtromere Protein-A (CENP-A), CENtromere Protein-B (CENP-B), CENtromere Protein-C (CENP-C), and CENtromere Protein-E (CENP-E). Sensitive metaphase FISH analysis of the low-alphoid chromosome 21 centromeres established the presence of residual αRI as well as other non-αRI α-satellite DNA suggesting that centromere function may be provided by (i) the residual αRI DNA, (ii) other non-αRI a-satellite sequences, (iii) a combination of i and ii, or (iv) an activated neocentromere DNA. (For complete abstract open document)

Centromere identity is integral for proper kinetochore formation and chromosome segregation. In most species chromosomes have a centromere at a defined locus that is propagated across generations. The histone H3 variant CENP-A acts as an epigenetic mark for centromere identity in most species studied. CENP-A is absent from the inactivated centromere on dicentric chromosomes and present at neocentromeres that form on non-centromeric sequences. Thus, the canonical centromere sequence is neither necessary nor sufficient for centromere function. Nevertheless, centromeres are generally associated with particular sequences. Understanding the organisation of centromeric sequence features will provide insight into centromere function and identity. In this study I use the fission yeast Schizosaccharomyces pombe model system to address the relationship between CENP-ACnp1 and centromeric sequence features. These analyses reveal that CENP-ACnp1 nucleosomes are highly positioned within the central domain by large asymmetric AT-rich gaps. The same sequence features underlying CENP-ACnp1 positioning are conserved in the related species S. octosporus, but are not found at neocentromeres, suggesting that they are important but non-essential for centromere function. CENP-ACnp1 over-expression leads to ectopic CENP-ACnp1 incorporation primarily at sites associated with heterochromatin, including the sites where stable neocentromeres form. Ectopic CENP-ACnp1 also occupies additional sites within the central domain that are not occupied in cells with wild-type CENP-ACnp1 levels. In wild-type cells CENP-ACnp1 occupied sites are likely also occupied by H3 nucleosomes or the CENP-T/W/S/X nucleosome-like complex in a mixed population. Several candidate proteins were investigated to determine a protein residing in the large gaps between CENP-ACnp1 nucleosomes could be identified. No proteins could be localised to the AT-rich gaps between CENP-ACnp1 nucleosomes, but the origin recognition complex in a promising candidate. The results presented in this thesis demonstrate that nucleosomes within the fission yeast centromere central domain are highly positioned by sequence features in a conserved manner. This positioning also allows for another complex, possibly the origin recognition complex, to bind to DNA. Nucleosome positioning, DNA replication, and transcription could individually and collectively influence CENP-ACnp1 assembly and centromere function. Further experiments in fission yeast will continue to provide insight into the general properties of centromere function and identity.

CENP-A est le variant d’histone qui remplace spécifiquement l’histone H3 au niveau des centromères et confère ses propriétés uniques à la chromatine centromérique. La cristallographie aux rayon X, ainsi que la digestion à la MNase des nucléosomes contenant CENP-A suggèrent une flexibilité de l’ADN entrant et sortant de ce nucléosome. Néanmoins ces déductions restent aujourd’hui au stade hypothétique, en particulier, rien n’est connu sur le rôle éventuelle de cette particularité dans la fonction du nucléosome CENP-A. L’utilisation de la cryo-électromicroscopie nous a permis de déterminer les caractéristiques de la dynamique de l’ADN sortant du nucléosome CENP-A. Nos analyses biochimiques, de protéomiques et de pseudo-génétiques révèlent que la flexibilité élevée de l’ADN du nucléosome CENP-A ne permet pas l’interaction avec l’histone de liaison H1. In vitro, remplacer les 2 tours de l’hélice aN de CENP-A avec les 3 tours de l’hélice aN de H3 permet de restaurer l’interaction de l’histone H1. In vivo, le replacement des nucléosomes CENP-A par des nucléosomes contenant ce même nucléosome hybride aN-CENP-A permet également le recrutement de H1, mais cela conduit également à la délocalisation d’un certain nombre de protéines du kinétochore. Ce kinétochore ne permet pas une ségrégation correcte des chromosomes et il conduit à des phases de mitose et de cytokinèse défectueuses. L’ensemble de ces données montre que la conservation au cours de l’évolution de la flexibilité de l’ADN dans le nucléosome CENP-A est essentielle pour l’accomplissement de la division cellulaire. / CENP-A is a histone variant, which replaces histone H3 at centromeres and confers unique properties to centromeric chromatin. The crystal structure and MNase digestion of CENP-A nucleosome suggests flexible nucleosomal DNA ends but their dynamics in solution remains elusive and their implication in centromere function is unknown. Using electron cryo-microscopy we determined the dynamic solution properties of the CENP-A nucleosome. Our biochemical, proteomic and genetic data reveal that the high flexibility of the DNA ends impairs histone H1 binding to the CENP-A nucleosome. Substituting the 2-turn aN-helix of CENP-A with the 3-turn N-helix of H3 results in particles able to bind histone H1. In vivo replacement of CENP-A nucleosomes with the same NH3-CENP-A hybrid nucleosomes leads to H1 recruitment, delocalization of kinetochore proteins and significant mitotic and cytokinesis defects. Put together, ourdata reveal that the evolutionarily conserved flexible ends of the CENP-A nucleosomes are essential to ensure the fidelity of the mitotic pathway.