Browsing by Author "Sullivan, Beth A"

Centromeres are essential chromosomal loci at which kinetochore formation occurs for spindle fiber attachment during mitosis and meiosis, guiding proper segregation of chromosomes. In humans, centromeres are located at large arrays of alpha satellite DNA, contributing to but not defining centromere function. The histone variant CENP-A assembles at alpha satellite DNA, epigenetically defining the centromere. CENP-A containing chromatin exists as an essential domain composed of blocks of CENP-A nucleosomes interspersed with blocks of H3 nucleosomes, and is surrounded by pericentromeric heterochromatin. In order to maintain genomic stability, the CENP-A domain is propagated epigenetically over each cell division; disruption of propagation is associated with chromosome instabilities such as aneuploidy, found in birth defects and in cancer.

The CENP-A chromatin domain occupies 30-45% of the alpha satellite array, varying in genomic distance according to the underlying array size. However, the molecular mechanisms that control assembly and organization of CENP-A chromatin within its genomic context remain unclear. The domain may shift, expand, or contract, as CENP-A is loaded and dispersed each cell cycle. We hypothesized that in order to maintain genome stability, the centromere is inherited as static chromatin domains, maintaining size and position within the pericentric heterochromatin. Utilizing stretched chromatin fibers, I found that CENP-A chromatin is limited to a sub-region of the alpha satellite array that is fixed in size and location through the cell cycle and across populations.

The average amount of CENP-A at human centromeres is largely consistent, implying that the variation in size of CENP-A domains reflects variations in the number of CENP-A subdomains and/or the density of CENP-A nucleosomes. Multi-color nascent protein labeling experiments were utilized to examine the distribution and incorporation of distinct pools of CENP-A over several cell cycles. I found that in each cell cycle there is independent CENP-A distribution, occurring equally between sister centromeres across all chromosomes, in similar quantities. Furthermore, centromere inheritance is achieved through specific placement of CENP-A, following an oscillating pattern that fixes the location and size of the CENP-A domain. These results suggest that spatial and temporal dynamics of CENP-A are important for maintaining centromere and genome stability.

The centromere is essential for chromosome segregation and genome stability. It is the site of kinetochore assembly and chromosome attachment to the spindle microtubules, and it is important for chromosome movement during mitosis and meiosis. Normal human chromosomes have one centromere, but genome rearrangements that occur with instability, aging, and disease often result in chromosomes with two centromeres, called dicentrics. Nearly seventy-five years ago, Barbara McClintock demonstrated that dicentric chromosomes in plants are associated with instability through mitotic "breakage-fusion-bridge" cycles. However, human dicentrics are unusually stable due to the poorly understood phenomenon of centromere inactivation. Centromere inactivation has been primarily studied in patient-derived dicentrics, limiting the derivation of a molecular pathway. Key centromere and kinetochore proteins are not present at inactive centromeres, but beyond these observations, the process of centromere inactivation is unclear. Epigenetic and sequence-dependent factors are known to contribute to centromere specification, but requirements for centromere assembly, maintenance, and suppression remain obscure. The aims of this research were to (1) determine the mechanism(s) by which de novo dicentric chromosomes are stabilized, (2) ascertain the factors influencing the involvement of specific chromosomes in de novo fusions, and (3) establish the epigenomic, temporal, and mechanistic basis of centromere inactivation. To uncover the mechanistic foundations of these processes, we developed in vitro cell culture systems to study the formation and stabilization of de novo dicentrics. We demonstrate that transient disruption of human telomere structure non-randomly produces dicentric fusions involving acrocentric chromosomes. This finding is notable since the most prevalent rearrangement in humans involves the acrocentrics and is called Robertsonian translocation (ROB). In some cases, centromere inactivation occurs by an apparently epigenetic mechanism. In other dicentrics, the size of the centromeric DNA array is reduced compared to the same array before dicentric formation. Many functional dicentrics persist for months after formation. Our results indicate that dicentric human chromosomes undergo alternative fates after formation across a broad temporal window. During transient telomere disruption, we observed a dramatic change in nucleolar appearance. Nucleolar proteins did not coalesce into condensed structures, but appeared dispersed throughout the nucleus. This surprising alteration in nucleolar organization and nuclear architecture suggests remodeling of the nucleolus and subsequent effects on nucleolar-associated chromosomes, such as the acrocentrics, could contribute to the high incidence of ROB formation. Further studies and development of additional cell culture systems will allow us to evaluate current models of centromere assembly and disassembly and the importance of chromatin organization to centromere function and genome architecture.

The centromere serves as the foundation for the kinetochore and attachment point for spindle microtubules during metaphase. The proper function of this locus is required to ensure chromosome segregation and genomic stability. In humans, repetitive alpha satellite DNA underlies the human centromere region and is organized into specific chromatin domains that are maintained by a complex combination of factors. Although the centromere region is generally thought to be specified epigenetically, some evidence suggests that the underlying DNA sequence is also involved in centromere function. To better define links between alpha satellite and function within the human centromere region, we investigated two attributes of alpha satellite DNA: its transcription into noncoding alpha satellite RNAs and genomic variation within the alpha satellite array. Noncoding transcripts produced from alpha satellite DNA are associated with normal centromere and pericentromere function and evidence from other organisms suggests RNAs from this region are pivotal in the centromere and kinetochore assembly cascade and in maintaining the chromatin environments of the centromere region. However, alpha satellite RNAs have not yet been fully characterized and data reflecting the chromosome-specific nature of alpha satellite arrays is lacking. Additionally, genomic variation within alpha satellite arrays has been linked to reduced centromere protein recruitment and chromosome instability, yet the molecular basis for this is unknown. These gaps in knowledge have stymied our understanding of the role of genomic and epigenetic attributes of alpha satellite that affect function within the human centromere region. Thus, this work aims to functionally characterize the role of alpha satellite transcripts and to determine how genomic variation impacts chromosome stability. Utilizing cytological and molecular techniques that allow the differentiation of alpha satellite RNAs from individual chromosomes and arrays, we have demonstrated that each chromosome produces unique noncoding RNAs that localize in cis to their site of production. Both centromeric and pericentromeric alpha satellite arrays produce noncoding RNAs, but these transcripts are spatially and functionally distinct. Alpha satellite RNAs from the centromere bind at least two key centromere proteins: CENP-A and CENP-C, while alpha satellite RNAs from the pericentromere colocalize with SUV39H1. Centromeric alpha satellite RNAs are required for complete loading of new CENP-A-containing nucleosomes, as well as maintenance of CENP-C levels. Genomic variation affects the origin of alpha satellite transcripts, such that highly variant arrays produce a different set of transcripts than wild type arrays. Further, the long-range organization of variation across the alpha satellite array in unstable chromosomes suggests certain spatial organizations of variation are poor platforms for building a stable centromere and kinetochore. Collectively, these findings implicate alpha satellite RNA and genomic variation and/or the interplay of these two elements as essential factors in the function of the human centromere region.

Centromeres are chromosomal loci essential for genome stability. Their malfunction can cause chromosome instability associated with cancer, infertility, and birth defects. This study focused on an intriguing centromere on human chromosome 17, which displays normal functional variation. Centromere identity can be found on either of two large arrays of repetitive DNA. We investigated inter-individual sequence variation on these two arrays and found association between array size, array variation, and centromere function. Our data suggest a functional influence of DNA sequence at this critical epigenetic locus.