DNA replication can be divided into three distinct steps: initiation, elongation, and termination. The bidirectional replication of a circular chromosome of bacteria terminates at a position where the two replication forks meet. Bacteria have developed a system that ensures termination will occur within a restricted terminus region. This is achieved by a combination of a DNA motif of 20 to 30 bp, called the ter sequence, and a cognate termination protein that recognizes ter sites and binds to them tightly.
Such systems were discovered in Escherichia coli and Bacillus subtilis, through the identification of the accumulation of Y-shaped replication intermediates at specific sites in the terminus regions (1), and they have been extensively characterized genetically and biochemically (2-4). The E. coli termination protein is encoded by the tus (terminus utilization substance) gene and has a molecular weight of 36 kDa (309 amino acid residues). The Tus protein specifically binds to the ter sites containing the consensus sequence of about 20 bp, and the Tus-Ter complex arrests a replication fork approaching from one direction but not from the other. This arrest is thought to be due to the orientation-dependent inhibition of unwinding of the DNA duplex by the DnaB DNA helicase at the apex of the replication fork (Fig. 1). Seven ter sites have been identified in the terminus region of the E. coli chromosome, as shown schematically in Figure 2. The clockwise replication fork can pass through the group1 ter sites, but if it arrives at the group2 ter sites before it meets the counterclockwise replication fork, it will stall there. Similarly, the counterclockwise replication fork will stall at the group1 ter sites if it has not met the clockwise fork. Thus, the termination event is regulated to occur in the terminus region opposite the oriC replication origin.
Figure 1. Polar inhibition of the DNA duplex unwinding by DNA helicases.
Figure 2. Location and orientation of seven ter sites on the E. coli chromosome (6).
A similar system to that of E. coli is also present in B. subtilis. The B. subtilis termination protein, encoded by the rtp (replication termination protein) gene, binds to several ter sites on the chromosome arranged similarly to those found in E. coli (5). The Rtp-Ter complex can either arrest or permit the passage of the replication fork, depending on the direction of its approach. Furthermore, the B. subtilis Rtp-Ter system efficiently arrests the progression of the replication fork in E. coli. However, the size (14.5 kDa, 122 residues) and amino acid sequence of Rtp protein are quite different from those of the E. coli Tus. Also, the functional B. subtilis arrest complex requires Rtp dimers to be bound to a longer ter sequence of about 30 bp in length.
The most intriguing aspect of the terminator-Ter complex is its polarity of the replication fork arrest, and two models have been proposed (6). One involves specific protein-protein interactions between the terminator and DNA helicase, and the other suggests a simple physical block of the complex (polar clump) against the action of the helicase. The findings that the Tus-Ter complex acts as a polar barrier to helicases of both prokaryotic and eukaryotic origins, and that the B. subtilis Rtp-Ter system efficiently arrests the progression of the replication fork in E. coli, support the latter model. The X-ray crystallography structures of the Tus and Rtp proteins have provided information about the molecular mechanism of the polar arrest (7, 8).
From an evolutionary point of view, E. coli and B. subtilis might have independently developed similar replication termination systems by convergent evolution, suggesting the advantage of the system for cell growth. However, the biological significance of the polar termination system is unclear. The terminator gene can be deleted without any apparent effect on cell growth in both B. subtilis and E. coli, indicating that chromosome replication can terminate wherever the two replication forks happen to meet. A similar system of polar replication fork block is characterized in the autonomously replicating plasmid R6K (9). Furthermore, fork-blocking sites have been reported in the yeast, pea, frog, and human genomes (10-13).