The bacterium Escherichia coli has a single, circular chromosome that is replicated and segregated with great precision to daughter cells during cell division. Replication proceeds bi-directionally from a single origin and terminates on the opposite side of the chromosome. The relative simplicity of this system and the limited number of cell components required for its propagation make it a model system for DNA replication and segregation in general. We have developed a P1 parS GFP-ParB system for localization by fluorescent microscopy of any desired locus on the E. coli chromosome. The technique works well in living cells and allows us to follow the fate of chromosomal sequences through several generations by time-lapse microscopy. In addition, we have used the technique, in combination with flow cytometry, to determine the spatial distributions of given loci at defined points in the cell cycle in a cell population. We currently have data for the dynamic behavior of multipleloci distributed around the chromosome in cells growing with a simple cell cycle at moderate growth rates. Our principle conclusions are as follows:1. Chromosome segregation is primarily accomplished while replication is still ongoing.2. The terminus is replicated at the cell center and the daughter termini of all cells remain attached there until cell division, at which time they rapidly segregate away from each other as the cell divides. The terminal foci are at the true cell center as the cell approaches division : the foci co-localize with a segment of the FtsZ ring. 3. The capture of the termini at the cell center is independent of the xerCD site-specific recombination system. It also appears to be independent of the C-terminal domain of FtsK, a protein implicated in DNA hadling at the cell center, although it is disrupted in many of the ftsK mutant cells due to aberrant or absent cell division events. 4. A terminus domain of 160kb., centered on the dif recombination site, segregates as a unit at cell division. Positions flanking this region segregate prior to cell division. 5. Origins of replication segregate fairly early in the cell cycle. On average, there is a delay of about 1/4 of a generation between origin initiation and segregation. However, the delay varies widely from cell to cell, with some origins segregating immediately after initiation. 6. Origins segregate from the cell center toward the poles and are free to move about for some time before becoming attached to the new cell centers. Daughter origins sometimes re-associate after initial segregation and dissociate again. Cohesion of origins, although it often occurs, is not a necessary or invariant feature of the cell cycle. 7. Positions around the chromosome that are intermediate between the terminal domain and the origin segregate before cell division, at times roughly corresponding to their map positions. There are some positions whose average segregation time appears earlier that predicted by an orderly segregation of markers timed by replication order. These may be regions that are handled in some special way by the segregation machinery. This year, we have put special emphasis on refining our analysis of E. coli chromosome segregation. In collaboration with the laboratory of Flemming Hansen., the technical University of Denmark, we have improved the GFP-ParB labelling method for the localization of DNA sequences, and have developed an automated method for the detection of cells and the measurement of the fluorescent foci. Our results clearly show that the recently published model of Bates and Kleckner (Cell, 121, 899-911) in which sister chromosomes cohere and are segregated as a unit, is incorrect. Rather, most of the chromosome is segregated smoothly as it is replicated.The overall mechanism for chromosome segregation in bacteria is clearly quite distinct from that of mitosis in higher organisms.
