|Home | About | Journals | Submit | Contact Us | Français|
In this issue of Molecular Cell, Fachinetti et al. provide the first comprehensive map of replication termination sites (TERs) in Saccharomyces cerevisiae (Fachinetti et al., 2010). Strikingly, the majority of TERs is occupied by topoisomerase 2, which shields these regions against genomic instability.
There is arguably little to rival the most famous “terminator” known to mankind, but topoisomerase 2 (Top2) is making its own claim to fame in a beautiful study led by Marco Foiani’s laboratory that investigated termination of DNA replication in eukaryotic cells. Almost a decade after all putative replication initiation sites were determined in the budding yeast genome (Wyrick et al., 2001), Fachinetti and coworkers present us with a complementary map of 71 replication termination regions (TERs) that span approximately 5 kb each. The majority (67/71) of these sites was associated with Top2 and all of them exhibited one or more DNA elements that have previously been described in the context of replication fork pausing (Fachinetti et al., 2010).
Why has it taken so long to systematically identify eukaryotic TERs? Partly, because grasping the concept of termination has always been more difficult than envisioning DNA synthesis initiation. The DNA acrobatics that must occur when two replication forks converge are complex. It has been presumed that fork fusion occurred – somehow – over relatively broad regions. Precedent for this idea came from studies in bacterial systems. In E. coli, ten termination (Ter) sites are dispersed across a 2.6 Mb stretch of DNA. Actual fork resolution takes place approximately opposite from the replication origin oriC between TerC and TerA, which encompasses 260 kb (Duggin et al., 2008; Fig. 1A). Ter sites are polar and block replication forks only on one side, the non-permissive end, but not at the opposite, permissive end. The terminator protein (termination utilization substance) Tus binds to Ter sites in a sequence-specific manner and is “trapped” by a highly conserved cytosine when unwinding occurs at the non-permissive end (Mulcair et al., 2006). In contrast, when the DNA is unwound at the permissive end, Tus is easily displaced from Ter, allowing the replication fork to pass freely through the termination site. This mechanism explains how a simple monomeric clamp, such as Tus, can provide polarity as a replication fork barrier (RFB). Interestingly, unlike replication initiation, termination does not appear to be evolutionarily conserved. Even among prokaryotes, terminator binding sites and proteins do not exhibit any strong similarities, although they likely function in a common way (Duggin et al., 2008). These evolutionary disparities have certainly not helped the cause of deciphering the mechanism of replication termination in eukaryotic cells.
Several key advances in our understanding of replication fork dynamics in the context of other DNA metabolic processes, such as transcription, have paved the way for elucidating the principles of replication fork fusion in eukaryotes. One such cornerstone was the discovery of a RFB in the rDNA locus of budding yeast (Brewer and Fangman, 1988). This early report underscored the polar nature of this RFB, which promoted the selective pausing of the fork that moved opposite to the direction of transcription. This concept was further extended to tRNA genes (Deshpande and Newlon, 1996) and more recently to RNA polymerase II-dependent transcription units (Azvolinsky et al., 2009). Indeed, 64 of the 71 newly mapped TERs belonged to these two latter categories, whereas seven encompassed centromeres (CENs; Fig. 1B) either as their single characteristic feature or in combination with another element. Although CENs have no intrinsic polarity, like transcription units, their specific location showed a clear preference in halting one of the two converging forks. Tracking replication termination at such high resolution was accomplished by genome-wide mapping of bromodeoxyuridine incorporation into nascent DNA of synchronized cell cultures. Cells were grown at low temperatures or in the presence of hydroxyurea to slow replication (Fachinetti et al., 2010). In parallel, the authors monitored the occupancy of a specific replication fork component, DNA polymerase-ε (pol-ε). 47 of 71 TERs displayed an increase in pol-ε binding, further suggesting that replication forks came to a halt in these regions.
In a tour-de-force of two-dimensional replication gel analyses, the authors characterized 20 of the 71 TERs in more detail. These 20 regions represented all three classes of TERs (Fig. 1B), and uniformly showed a more persistent X-structure, the signature of converging forks, in strains in which RRM3 was deleted. The DNA helicase Rrm3 has been implicated in facilitating replication through natural pause sites (Ivessa et al., 2003). Strikingly, pausing at the vast majority of Pol II-associated sites did not appear to be enhanced in the absence of Rrm3 (Azvolinsky et al., 2009). However, the half-life of termination intermediates was increased, arguing that Rrm3 may play a more active role in termination than just regulating passage of the replication fork.
So how does Top2 fit into all of this? Top2 is known to be required for the decatenation of replicated DNA at the end of S phase. It is associated with migrating replication forks to counteract the accumulation of negative supercoils behind the fork. Top2 functions in concert with topoisomerase IA (Top3), which resolves positive supercoils ahead of the fork. Surprisingly, when Franchetti et al. examined TERs for Top2 and Top3 occupancy, only Top2 was associated with these sites. More importantly, Top2 resided at 67 TERs in S phase, G2/M or both. Specifically, Top2 was bound to 55 TERs in early S phase, long before replication forks invaded the area. This strongly suggested that Top2 is a hallmark of eukaryotic TERs (Fig. 1B). How Top2 is recruited to these particular DNA elements and when exactly it facilitates topological changes that aid in replication termination will have to await future studies. In the same vein, it cannot be fully excluded that Top2 has an additional structural role at TERs. For instance, it could have anti-helical activity similar to Tus in E. coli (Neylon et al., 2005)
Regardless of its precise function, loss of Top2 caused genome instability at a total of 13 sites in the genome during a single S phase. The authors conducted comparative genome hybridization in top2-1 mutants to map specific regions that exhibited altered DNA content under non-permissive conditions. Four of the 13 sites happened to be TERs. Given the documented accumulation of DNA damage in top2 mutants, this may not seem all that surprising. Nevertheless, these results convincingly argue that regulated replication termination is required to maintain genomic integrity. It will be interesting to further explore whether these four TERs have common characteristics that make them prone to become unstable. Moreover, two of the four TERs displayed local DNA amplification, reminiscent of the re-replication phenotype of tus mutants (Duggin et al., 2008).
In the end, we are left with questions about the similarities and differences between prokaryotic and eukaryotic strategies utilized in the termination of replication. Is Top2 the eukaryotic Tus? Maybe not, but it clearly distinguishes a regular pause site from a TER. Why does E. coli not employ transcription units as polar pausing elements? It turns out that the bacterial genome is organized in a way that maximizes replication efficiency. In other words, the direction of transcription is co-oriented with the leading strand (Rocha, 2008). Such a mechanism would be unfeasible for complex eukaryotic genomes. Whereas the 71 TERs likely do not constitute a complete termination map, they illuminate what has been until now a black box. The authors provide a framework that should allow not only for the systematic identification of TERs in higher eukaryotes, but also for mechanistic insight into the process of termination. It looks as if the terminator is back.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.