Although genetic information is encoded in a one-dimensional array of DNA bases, all major DNA processes (replication, transcription, and recombination) are controlled by changes in the three-dimensional structure of DNA. Large-scale structural features of chromosomes including the arrangement of important chromosomal sites (origins, termini, and centromeres) and overall chromosome compactness change dramatically and predictably during the cell cycle.
One major limitation in studying chromosome dynamics is our inability to observe large-scale chromosome structure using standard fluorescence microscopy methods, which only label a discrete genetic position (focus). We are developing a novel chromosome painting technology to image individual domains within the entire chromosome in single cells. This method, inspired by in situ hybridization-based human karyotyping techniques, utilizes multi-color combinatorial labeling and high-resolution three-dimensional photography to generate whole genome maps of the chromosome. Our goal is to define the normal program of chromosome movement in E. coli using a cell cycle synchronization apparatus we designed called the “baby cell machine” (Bates et al, 2005).
In eukaryotic DNA replication, sister chromosomes are attached to each other (cohesion) until they are separated at nuclear division. By aligning sister chromosomes side-by-side along homologous sequences, chromosome cohesion enables precise attachment and segregation by the mitotic spindle apparatus and aids efficient repair of DNA damage by homologous recombination. Eukaryotic chromosome cohesion involves both a proteinaceous component, the cohesin complex, and physical entanglement of sister chromosomes, but the relative roles of these components in maintaining and regulating cohesion is unknown.
We previously showed that DNA in E. coli also has a cohesion period, but it is much shorter, lasting on average about 7 minutes (Joshi et al., 2011). Thus, there is a 300-400 kb window of cohesion behind each replication fork, after which sister duplexes separate toward opposite sides of the cell. Evidence suggests that bacterial cohesion is based solely DNA entanglements, without a cohesin-like protein component (Joshi et al., 2013). Because cohesion can be directly modulated by changing the cellular abundance the decatenating enzyme Topo IV (homologous to eukaryotic Topo II), we believe that "cohered" segments are highly catenated. Interestingly, cohesion persists much longer at a cluster of sites called "Snaps", and extended cohesion at this location was important for efficient chromosome segregation. We are currently exploring models of how these centromere-like Snaps are generated and how they promote chromosome segregation.
Dealing with Topology at the Replication Fork
Catenation occurs in DNA molecules that are under high helical tension (i.e., overwound). The presence of extensive DNA catenation behind the replication fork in E. coli implies that tension generated ahead of the fork by strand unwinding outpaces the relaxing ability of topoisomerase by a wide margin (forks can travel up to an astounding 1000 bp/sec!). In theory, if ahead-of-the-fork tension exceeds the energy required to rotate the ~800 kDa replisome, the entire fork will rotate in the opposite direction, wrapping nascent DNA duplexes around each other (catenation). Our lab is currently investigating factors that promote tension, including fork velocity and topological insulators (like DNA-bound proteins). We are also investigating how tension and catenation affect overall replisome progression, including fork stalling and reversal.