In the circular Escherichia coli chromosome, two replisomes are assembled at the unique origin of replication and drive DNA synthesis in opposite directions until they meet in the terminus region across from the origin. Despite the difference in rates of the two replisomes, their arrival at the terminus is synchronized through a highly specialized system consisting of the terminator protein (Tus) bound to the termination sites (Ter). This synchronicity is mediated by the polarity of the Tus−Ter complex that stops replisomes from one direction (non-permissive face) but not the other (permissive face). Two oppositely oriented clusters of five Tus–Ters that each block one of the two replisomes create a “replication fork trap” for the first arriving replisome while waiting for the late arriving one. Despite extensive biochemical and structural studies, the molecular mechanism behind Tus−Ter polar arrest activity remained controversial. Moreover, none of the previous work provided answers for the long-standing discrepancy between the ability of Tus−Ter to permanently stop replisomes in vitro and its low efficiency in vivo. Here, I spearheaded a collaborative project that combined single-molecule DNA replication assays, X-ray crystallography and binding studies to provide a true molecular-level understanding of the underlying mechanism of Tus−Ter polar arrest activity. We showed that efficiency of Tus−Ter is determined by a head-to-head kinetic competition between rate of strand separation by the replisome and rate of rearrangement of Tus−Ter interactions during the melting of the first 6 base pairs of Ter. This rearrangement maintains Tus’s strong grip on the DNA and stops the advancing replisome from breaking into Tus−Ter central interactions, but only transiently. We further showed how this kinetic competition functions within the context of two mechanisms to impose permanent fork stoppage. The rate-dependent fork arrest activity of Tus−Ter explains its low efficiency in vivo and why contradictory in vitro results from previous studies have led to controversial elucidations of the mechanism. It also provides the first example where the intrinsic heterogeneity in rate of individual replisomes could have different biological outcomes in its communication with double-stranded DNA-binding protein barriers.

The final stages of replication fork termination requires the timely and orderly orchestration of catalytic and enzymatic activities. Given the complexity of this process, it is conceivable that the final stages of fork termination is susceptible to problems that could trigger recombination, which could lead to deleterious genomic rearrangements if ectopic homologous sequences are recombined. Using the site-specific RTS1 barrier in fission yeast, I demonstrated that fork termination is generally not a recombinogenic process, and that hyper-recombination-induced by fork blockage at RTS1 is largely a result of replication fork restart. To investigate the actual mechanisms and proteins, which drive and influence recombination at a replication barrier, I studied the MHF proteins, which assist Fml1 in limiting crossovers during double-strand break (DSB) repair and promoting Rad51-mediated recombination at impeded replication forks, and are also components of the constitutive centromere-associated network (CCAN). Intriguingly, structural studies revealed that the MHF can exist as an octamer in vitro. I examined the biological significance of octameric MHF by employing three mutations that disrupt the octamer configuration in vitro. In fission yeast, these mutations cause hypersensitivity to methyl methanesulfonate (MMS), suggesting that the MHF octamer may have a role in DNA repair. One of the “octamerisation” mutants, exhibits greater hypersensitivity to MMS than the other two, and biochemical experiments indicated that this is because it confers an additional defect in MHF’s interaction with Fml1. Further genetic experiments on this mutant suggest that the ability of Fml1 to unwind D-loops depends more critically on its interaction with MHF than fork reversal. Additionally, I showed a synergistic interaction between Dcr1 and MHF, and demonstrated that in the absence of Dcr1, there is a greater need for recombination to tolerate/repair DNA damage. Lastly, I uncovered a novel function for the MHF in controlling the initiation of septation.