Beth Sullivan, PhD

Associate Dean for Research Training

Headshot of Beth Sullivan, PhD361 CARL Building
Box 3054 DUMC
Durham, N.C. 27710
Phone: 919.684.9038
Fax: 919.684.2790


research •  biography • lab members • alumni • publications • lab websitetwitter 

Research in the Sullivan Lab involves studying how chromosomes are organized into inherited chromatin domains and understanding mechanisms of formation and behavior of chromosome abnormalities that are associated with birth defects, reproductive abnormalities, and cancer.

One focus of the lab’s research is the centromere, a specialized chromosomal site involved in chromosome architecture and movement, kinetochore function, heterochromatin assembly, and sister chromatid cohesion. Our experiments have uncovered a unique type of chromatin (CEN chromatin) formed exclusively at the centromere by replacement of core histone H3 by the centromeric histone variant CENP-A. Using high-resolution fluorescence microscopy and optical mapping, we have shown that centromeric chromatin (CEN chromatin) contains interspersed subdomains of CENP-A and H3 nucleosomes. Since CEN chromatin also contains H3, we are interested in determining if modifications of core histones functionally distinguish centromeres from other regions of the genome.

Centromeres have historically been considered heterochromatic, however, surprisingly, we found that CEN chromatin in humans and flies contain “euchromatic” modifications of H3, signifying an open or flexible chromatin conformation. These studies were the foundations for our current investigations into how centromeric chromatin affects transcription (and vice versa) and the identification of structural or functional elements that define centromere identity. Specifically, we study endogenous and engineered human chromosomes that contain two similar, adjacent blocks of centromeric satellite DNA arrays. On these chromosomes, usually only one satellite array is assembled into a functional centromere. We are testing how and why two physically linked satellite regions are functionally distinct.

The lab also studies genome stability, specifically the formation and behavior of human chromosomal abnormalities. During meiosis and mitosis, chromosome rearrangements often occur that produce chromosomes that have two (or more) centromeres. These chromosomes are dicentric and, in humans, occur as frequently as 1 in 1000 individuals. Barbara McClintock, a famous cytogeneticist and Noble prizewinner, studied dicentric chromosomes in maize (corn) in the 1930s. She described dicentrics as inherently unstable chromosomes because the two centromeres often segregated to opposite spindle poles in anaphase, leading to chromosome breakage. In humans, however, dicentric chromosomes exhibit unprecedented stability, and sometimes are even preferentially segregated to the ooycte in female meiosis. Human dicentrics are thought to be so stable because either both centromeres work together to make a super-centromere that has a functional advantage during meiotic segregation, or one of the centromeres is shut off, stabilizing the dicentric chromosome so that it behaves as if it has only a single centromere. 

We are particularly interested in understanding the mechanism of centromere inactivation – how and when does it occur? This phenomenon is currently understood as the loss of centromeric proteins and a cytological change in chromosome appearance. Beyond these observations, gathered mainly from studying patient-derived dicentrics, the molecular mechanism of inactivation is unclear. We have developed two novel strategies to generate human dicentric chromosomes in the laboratory. We are using several assays, including live cell studies and SNAP-tagging of centromere proteins, to dissect the process of centromere inactivation, and to understand how chromatin remodeling and chromosome structure affect centromeric fate.