Sister chromatid decatenation: bridging the gaps in our knowledge

Abstract

Faithful chromosome segregation is critical in preventing genome loss or damage during cell division. Failure to properly disentangle catenated sister chromatids can lead to the formation of bulky or ultrafine anaphase bridges, and ultimately genome instability. In this review we present an overview of the current state of knowledge of how sister chromatid decatenation is carried out, with particular focus on the role of TOP2A and TOPBP1 in this process.

Introduction

During each cell cycle the entire genome is copied and then accurately separated between daughter cells during mitosis. Faithful chromosome segregation is critical in preventing genome loss or damage during cell division. This is underscored by the fact that the mechanisms governing chromosome separation are highly conserved throughout evolution. Furthermore, there is a much stronger imperative to ensure the fidelity of chromosome disjunction in complex multicellular organisms, such as humans, who rely on the co-operation between cells to form tissues and functional organs. At the most extreme, a single error in chromosome segregation made during early stages of development could be compounded many fold, disrupting and killing the unborn organism. The astounding task of segregating the genome is further complicated by the fact that replicated sister chromatids can remain knotted with each other due to topological constrains, such as catenation. Cells unable to resolve DNA entanglements in a timely manner end up with the so-called chromatin bridges formed between separating sister chromatids in mitotic cells. If unresolved, these bridges can lead to gross chromosome breakage and genome rearrangements,^1,2 both hallmarks of cancer cells.³ Accordingly, multiple genes involved in regulating sister chromatid segregation are potent tumor suppressors.^4-6

Initial studies identified so called “bulky” anaphase bridges as DAPI-positive stretches of DNA connecting sister chromatids during segregation in anaphase.⁷ In 2007, 2 groups identified a novel class of anaphase bridge termed “ultrafine anaphase bridges” (UFBs), which were defined as DNA bridges connecting the separating DNA masses in anaphase cells which were not readily stained with DNA-intercalating dyes such as DAPI, but could be identified by immunofluorescence analyses as being coated with certain proteins, namely the Bloom Syndrome helicase (BLM) and the Polo-like kinase-interacting checkpoint helicase (PICH).^7,8 The origin and processing of UFBs has been recently extensively reviewed by Liu and colleagues.⁹

Over the last decade it has become clear that a subset of UFBs arise from common genomic fragile sites, termed fragile site UFBs (FS-UFBs) which can be induced by replication stress, for example treatment with aphidicolin.¹⁰ These UFBs most likely result from sister chromatid interlinks generated by unresolved replication intermediates, and are characterized by the presence of the Fanconi Anemia (FA) protein FANCD2 forming foci that localize at either termini of these bridges.¹⁰ In addition, γH2AX foci are detected at the extremities of these bridges as mitosis progresses, suggesting induction of a DNA damage response.¹⁰ Recently, elegant work from 2 groups has shown that resolution of these structures requires the MUS81-EME1 endonuclease.^11,12 Some UFBs also originate from telomeres, as assayed by immunofluorescence experiments using DNA-FISH probes for telomeric loci.^10,13 These bridges, termed telomeric UFBs (T-UFBs) are more numerous upon treatment of cells with the replication inhibitor aphidicolin, or in cells lacking the Werner helicase (WRN). They are thought to arise from incomplete replication of telomeric loci, which consist of tandem repeat sequences prone to forming G-quadruplex structures, and are hence difficult to replicate.^9,10,13

Interestingly, the majority of UFBs arise from centromeric loci in the genome (centromeric UFBs or C-UFBs). These are thought to result from catenations of fully replicated dsDNA arising due to the persistence of sister chromatid cohesion at the centromere until the onset of anaphase.^9,14 Centromeres represent the genomic loci on which the kinetochore forms, serving as the binding platform for the mitotic microtubules.^14,15 They consist of tandem repeats of α-satellite DNA, at which sister chromatid cohesion persists until entry into anaphase, facilitated by the continuing presence of the cohesion complex at these loci.^14,15 Cohesin acts as a barrier to enzymes that can facilitate decatenation of centromeric DNA.¹⁴ The major enzyme involved in this process is Topoisomerase IIα (TOP2A).¹⁶

TOP2A is a type II topoisomerase, a family of enzymes required to manage DNA superhelicity and chromosome segregation.^16,17 TOP2A itself forms a dimer which can encircle dsDNA and facilitate decatenation via its ATP-dependent DNA strand passing activity.^16,17 TOP2A accumulates at centromeres during prometaphase (as assayed by immunofluorescence and epifluorescence microscopy) and remains there until anaphase, presumably facilitating decatenation of sister chromatids (reviewed extensively in¹⁶).

Several lines of evidence suggest that C-UFBs are resolved by the action of Topoisomerase IIα. For example, treatment of cells with the TOP2A inhibitors ICRF-193 or razoxane, leads to an increase in centromeric UFBs.^7,8 Moreover, depletion of TOP2A leads to the abnormal persistence of PICH coated UFBs emanating from centromere loci.¹⁸ These C-UFBs are found in a large percentage of anaphase cells during normal, unchallenged cell division,^8,19 decreasing in number as mitosis progresses, with no DNA double-strand breaks or induction of DNA damage response detectable in the resultant daughter cells, suggesting that they do not arise from aberrant DNA structures, but are a general feature of a normal cell division.¹⁹ How TOP2A is recruited to chromatin to facilitate sister chromatid decatenation has long been a key question in the field. Recent work from our and other groups has shed light on this process, with several non-mutually exclusive mechanisms described for recruitment of TOP2A to chromosomes, depending on the genomic locus and chromatin context.^4,20-22

Recruitment of TOP2A to Various Genomic Loci Through the Cell Cycle

Recent elegant work from Dykhuizen and colleagues has identified the BAF chromatin-remodeling complex (a member of the SWI/SNF complex family) as a major regulator of TOP2A recruitment to chromatin in mouse embryonic stem cells.⁴ The BAF250 subunit of this complex binds directly to TOP2A and facilitates its recruitment to over 12,000 sites across the genome.⁴ This localization is dependent on the ATPase activity of the BRG1 subunit of this complex, with defective recruitment of TOP2A to these genomic loci leading to the induction of DAPI-positive DNA bridges in anaphase cells, indicating defective sister-chromatid decatenation.⁴ It has not yet been established if BAF depletion or ablation of its TOP2A binding or ATPase activities impact upon the resolution of UFBs, but this would seem very likely. Interestingly, the authors identified many areas of the genome that TOP2A localizes to independently of the BAF complex. These are likely to include also centromeric loci, as recruitment of TOP2A in BAF mutant cell lines to these structures was not affected as assayed by immunofluorescence microscopy.⁴ This suggests that the BAF complex is most likely not required for the recruitment of TOP2A to centromeres and consequently C-UFB resolution. Moreover, this also implies that multiple mechanisms for TOP2A recruitment may exist to allow sister chromatid decatenation, depending on the genomic locus and local chromatin context.

Accordingly, d'Alcontres and colleagues have identified a role for the TRF1 subunit of the shelterin complex, which binds and protects telomere ends, in facilitating the recruitment of TOP2A to these loci to prevent the formation of both DAPI-positive anaphase bridges, and UFBs.²⁰ These data suggest a role for TOP2A in the resolution of T-UFBs, facilitated by its TRF-1 dependent recruitment to telomeres²⁰ and for the first time implies a role for TOP2A in the resolution of UFBs originating not from dsDNA catenenes, but from regions of unreplicated DNA/replication intermediates. However, the mechanisms governing TOP2A recruitment to other genomic loci, in particular to UFBs emanating from centromeres still remained enigmatic.

Recently, Germann and colleagues and our group have observed that the Topoisomerase IIβ-binding protein 1 (TOPBP1) (and its yeast orthologue Dpb11) localized to UFBs in yeast and transformed chicken cells²¹ and on UFBs in human cell lines.²²

TOPBP1, a highly conserved BRCT domain-containing protein, was first identified in a yeast 2-hybrid screen as an interacting partner of DNA topoisomerase-IIβ.²³ TOPBP1 promotes efficient initiation of DNA replication as well as cellular responses to DNA replication stress.²⁴ TOPBP1 is also necessary for a normal mitotic progression.²⁵ Yeast orthologues of TOPBP1 and TOP2A (Dpb11 and Top2) co-localize in nuclear tubes containing Dpb11-coated UFBs,²¹ suggesting that TOPBP1 may facilitate recruitment of type II topoisomerases to aid UFB resolution. Indeed this appears to be the case, as our group identified TOPBP1 as a novel TOP2A interacting protein, and demonstrated that TOPBP1 recruits TOP2A to UFBs thereby promoting their resolution in human cells.²² Despite this insight some important questions remain unanswered, namely, what the molecular basis of TOPBP1 recruitment to UFBs is, how exactly the TOPBP1-TOP2A interaction promotes the resolution of UFBs, and how this process is regulated.

The role of TOPBP1 in Controlling TOP2A Recruitment to UFBs

TOPBP1 localization to UFBs requires the highly conserved Lys704 in its BRCT domain 5, which also mediates TOPBP1s phospho-dependent interactions with BLM and 53BP1.^26,27 However, it seems that neither of these partners are required for its localization to UFBs,²² suggesting that this interaction is mediated via an as yet unidentified protein and is most likely phospho-dependent. Thus, a kinase/phosphatase pair could regulate TOPBP1 recruitment to UFBs. A good candidate would be one of the M-phase specific kinases such as PLK1 or another PLK family kinase, MPS1 or one of the Aurora kinases (reviewed in²⁸). Future studies aiming to identify these key regulators should clarify this question. It is clear, however, that this localization is required for the efficient resolution of UFBs.^21,22 In support of this idea, we observed that mutation of the highly conserved Lys704 to alanine leads to the increased prevalence of BLM-coated UFBs, the majority of which arise from centromeric loci. Likewise, treatment with TOP2A inhibitors or depletion of TOP2A lead to increased incidence of TOPBP1-coated UFBs.²² The requirement for TOPBP1 in UFB resolution appears to be evolutionarily conserved, as in yeast depletion of Dpb11 leads to the accumulation of UFBs, and a temperature sensitive Top2 mutant stain shows an increase in Dpb11-coated UFBs.²¹ Interestingly, treatment with the replication inhibitor aphidicolin does not increase the frequency of TOPBP1 UFBs,^21,22 suggesting that TOPBP1 is unlikely to participate in the resolution of UFBs originating from aphidicolin-induced fragile sites or telomeres.^10,13

As previously alluded to, we hypothesized that TOPBP1 may promote the access of TOP2A to UFBs in order to aid their resolution. Indeed, our data shows that TOP2A interacts with TOPBP1 in G2/M phase and this interaction requires TOPBP1s C-terminus encompassing the BRCT 7 and 8 motifs.²² However, these domains are not conserved down to yeast. Perhaps not surprisingly therefore, mutation of the highly conserved lysine residue (K1317A) in TOPBP1s BRCT 7, which mediates its interaction with FANCJ,²⁹ does not abrogate TOPBP1-TOP2A interaction (Fig. 1). Thus, this interaction may be mediated by the linker region bridging BRCT 7 and 8, as linker regions between tandem BRCT domains have been shown to mediate interactions with other proteins independently of phosphorylation.³⁰ These interactions rely on the presence of a β-hairpin motif within the inter-BRCT linker, supporting extensive contacts at the protein binding interface.³¹ Interestingly, the crystal structure of the BRCT 7/8 of TOPBP1 reveals an unusually long linker region compared with other phospho-peptide binding tandem BRCT domains,²⁹ indicating that this could play a role in the TOPBP1-TOP2A interaction.

Figure 1.

Mutation of highly conserved lysine K1317 within BRCT7 does not disrupt TOPBP1-TOP2A interaction. HEK 293FT control cells, or cells transiently expressing WT GFP-TOPBP1 and the indicated point mutant were nocodazole treated (0.1 µg/ml for 16h) and subjected to GFP-nanotrap immunoprecipitation. Western blotting for FANCJ, TOP2A and TOPBP1 confirm co-immunoprecipitation; FANCJ acts as an internal control.

Another possibility is that the most extreme part of the C-terminus of TOPBP1 promotes this interaction, which contains conserved residues between yeast Dpb11 and human TOPBP1. Moreover, it is unclear at present whether this interaction is direct or indirect. In vitro binding experiments with purified proteins and further deletion mutants mapping this interaction will address these issues.

It would appear that the majority of the UFBs bound by TOPBP1 originate from centromeric loci; however, only a subset of the BLM-positive UFBs induced by TOP2A depletion/inhibition stain for TOPBP1. This indicates that TOPBP1 may only bind to and aid the resolution of a subset of C-UFBs.²² Precisely what defines this subset remains an interesting question. It is possible that TOPBP1 promotes the resolution of a subset of canonical C-UFBs arising due to unresolved catenenes of dsDNA. Alternatively, TOPBP1 could also coat UFBs originating from unreplicated DNA and/or replication intermediate structures induced by the inhibition or depletion of TOP2A. In support of this notion, the α-satellite DNA at the centromere is late replicating and is comprised of large numbers of repetitive DNA sequences, which are canonically difficult to replicate.^32,33 These two scenarios are not mutually exclusive, and TOPBP1 may be required for the resolution of both types of bridge. Further studies should shed light on what exactly defines the TOPBP1-bound UFBs, and delineate the molecular mechanism of its recruitment to these structures.

Conclusions

Our work provides evidence for how TOP2A is recruited to a subset of genomic location to aid in sister chromatid decatenation and the role of TOPBP1 in this process. The key questions in this regard are whether TOPBP1-TOP2A interaction is direct, and if so which part of TOP2A is required for TOPBP1 binding. Moreover, the molecular mechanism of TOPBP1 recruitment to UFBs remains elusive. It is clear that TOPBP1 mediates TOP2A's recruitment to a subset of UFBs, and that TRF1 and the BAF complex promote TOP2A recruitment to telomeric loci and other discrete genomic loci, respectively. It remains to be understood, however, how TOP2A is recruited to other genomic loci to promote the resolution of sister chromatid and genome stability. These remain key questions in the field and their clarification will bring about a more complete understanding of sister chromatid decatenation.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Drs A. Blackford and M. Kliszczak for critical reading of the manuscript.

Funding

This work was supported by a WIMM/Medical Research Council Senior Non-Clinical Fellowship to WN.