Topoisomerase IIα prevents ultrafine anaphase bridges by two mechanisms

Topoisomerase IIα (Topo IIα), a well-conserved double-stranded DNA (dsDNA)-specific decatenase, processes dsDNA catenanes resulting from DNA replication during mitosis. Topo IIα defects lead to an accumulation of ultrafine anaphase bridges (UFBs), a type of chromosome non-disjunction. Topo IIα has been reported to resolve DNA anaphase threads, possibly accounting for the increase in UFB frequency upon Topo IIα inhibition. We hypothesized that the excess UFBs might also result, at least in part, from an impairment of the prevention of UFB formation by Topo IIα. We found that Topo IIα inhibition promotes UFB formation without affecting the global disappearance of UFBs during mitosis, but leads to an aberrant UFB resolution generating DNA damage within the next G1. Moreover, we demonstrated that Topo IIα inhibition promotes the formation of two types of UFBs depending on cell cycle phase. Topo IIα inhibition during S-phase compromises complete DNA replication, leading to the formation of UFB-containing unreplicated DNA, whereas Topo IIα inhibition during mitosis impedes DNA decatenation at metaphase–anaphase transition, leading to the formation of UFB-containing DNA catenanes. Thus, Topo IIα activity is essential to prevent UFB formation in a cell-cycle-dependent manner and to promote DNA damage-free resolution of UFBs.


Introduction
Genome stability requires accurate DNA replication during S-phase and correct chromosome segregation during mitosis. Errors impairing these two crucial steps are particularly prone to induce genetic instability [1,2]. DNA replication leads to the formation of intertwines between two DNA strands, referred to as DNA catenanes, the resolution of which requires the introduction of transitory breaks. Topoisomerases play a key role in DNA catenane processing. Topoisomerase IIα (Topo IIα) is a well-conserved double-stranded DNA (dsDNA)-specific decatenase enzyme [2][3][4][5][6]. Topo IIα activity leads to double-strand breakage followed by intramolecular strand passage and DNA re-ligation [2]. The decatenating activity of Topo IIα plays a major role in several aspects of chromosome dynamics, including DNA replication and chromosome segregation [7].
Topoisomerase activity ahead of the replication fork cannot resolve all dsDNA catenanes. Moreover, convergence of two replisomes leads to the steric hindrance of topoisomerase activity [2,8]. Consequently, some dsDNA catenanes are not resolved before the onset of mitosis. They form physical links between the sister chromatids and must therefore be processed by Topo IIα before chromosome segregation in anaphase [9]. Indeed, the disruption of Topo IIα activity leads to incomplete sister chromatid disjunction [10,11]. Sister chromatid anaphase bridges are of two types: chromatin anaphase bridges that can be stained with conventional dyes, such as DAPI, and ultrafine anaphase bridges (UFBs) that cannot be stained with conventional dyes or antibodies against histones. Both chromatin and ultrafine anaphase bridges result from a defect in sister chromatid segregation. During mitosis, PICH (Plk1-interacting checkpoint helicase), an SNF2-family DNA translocase involved in chromosome segregation [11][12][13][14][15], is recruited on both chromatin bridges and UFBs [10][11][12][13][14]. UFBs were discovered in 2007 [12,14] and are present in all cell lines tested. They are thus considered to be physiological structures [14,16]. Most UFBs are of centromeric origin, but some UFBs induced by replication stress originate from common fragile sites and are associated with FANCD2/FANCI proteins, whereas some other UFBs originate from telomeres or ribosomal DNA repeats [14,15,17,18]. UFBs were reported to contain either unresolved DNA catenations or replication intermediates [12][13][14]16,19]. Importantly, the total UFB population can be revealed only by PICH staining [12]. In a previous study, we reported that the intracellular accumulation of dCTP, due to cytidine deaminase (CDA) deficiency, leads to an excess of UFB-containing unreplicated DNA, due to a decrease in the basal activity of poly(ADP-ribose) polymerase 1 (PARP-1), which promotes the premature entry of cells into mitosis, before the completion of DNA replication has been completed [16,19].
Topo IIα inhibition leads to a large increase in the frequency of centromeric UFBs [11,12,14,[20][21][22]. In this study, we investigated the molecular origin of the increase in UFB frequency following Topo IIα inhibition. We showed that Topo IIα inhibition had no effect on global disappearance of UFBs during mitotic progression. However, we observed an aberrant UFB resolution leading to DNA damage within the next G1 as revealed by the increase in the frequency of 53BP1 foci. We also found that Topo IIα inhibition led to two types of UFBs, the type of UFB formed depending on the phase of the cell cycle. Topo IIα inhibition during S-phase impairs DNA replication, leading to the formation of UFB-containing unreplicated DNA during mitosis, whereas Topo IIα inhibition during mitosis prevents DNA decatenation, resulting in UFB-containing dsDNA catenanes. Thus, Topo IIα inhibition impairs both DNA replication during S-phase and DNA decatenation during mitosis, leading to the formation of two types of UFB with different molecular origins. Our results therefore demonstrate that Topo IIα activity is required to prevent the formation of UFBs through replication defects or a lack of resolution of DNA catenanes when cells enter mitosis.

Results and discussion
2.1. Topo IIα inhibition promotes ultrafine anaphase bridge formation before and during mitosis Topo IIα inhibition leads to a large increase in UFB frequency, and thus it has been proposed that Topo IIα activity is required for UFB resolution, accounting for the increase in UFB frequency upon Topo IIα inhibition [11,12,14,[20][21][22]. However, the increase in UFB frequency upon Topo IIα inhibition could also reflect, at least in part, an accumulation of newly formed UFB. We therefore first investigated whether Topo IIα inhibition compromised the resolution or the formation of UFBs. Topo IIα inhibition in HeLa cells with ICRF-159 (1 or 10 µM for 8 h), a catalytic Topo IIα specific inhibitor [23,24], led to an increase in UFB frequency in anaphase cells in a dose-dependent manner, as expected (figure 1a-c). We investigated whether Topo IIα inhibition affected UFB formation or resolution, by treating cells with ICRF-159 from S-phase until the end of mitosis. HeLa cell cycle duration is well described [25]: HeLa cells take about 8-10 h between S-phase and mitosis. Thus, mitotic cells after 8 h of ICRF-159 treatment correspond to cells that were in early S-phase when we added ICRF-159 to the cell culture medium (figure 1a). We then quantified PICH-positive UFBs from metaphase (the first step in mitosis, during which the distance between sister chromatids is sufficiently large for the visualization of UFBs) to telophase. Using this approach, we were able to assess UFB formation (by determining the increase in UFB frequency over time) and the global UFB disappearance (visualized as a decrease in UFB frequency during mitosis) (figure 1d).
ICRF-159 treatment led to an increase in UFB frequency at metaphase (red arrow, figure 1d). Thus, cells entered mitosis with a higher frequency of UFBs when Topo IIα was inhibited. Interestingly, the frequency of UFBs was also much higher at the metaphase-anaphase transition (green arrow, figure 1d), reflecting the formation of new UFBs early in mitosis. UFB frequency then decreased over time until the end of mitosis, in a similar manner in both untreated and ICRF-159treated cells. UFBs are therefore resolved even if Topo IIα is inhibited. These results indicate that either Topo IIα activity is dispensable for the resolution of pre-existing UFBs during mitosis, or UFB dissolution is aberrant in the absence of Topo IIα activity, possibly through DNA breakage. However, our results demonstrate that Topo IIα activity is strictly necessary to prevent the formation of new UFBs. Our observations also indicate that Topo IIα inhibition promotes UFB formation in two different ways: before the onset of mitosis, as revealed by the increase in UFB frequency at metaphase, and during mitosis, leading to an increase in UFB frequency at the metaphase-anaphase transition.

Topoisomerase IIα inhibition impairs complete DNA replication
We previously reported that delaying entry into mitosis allows the completion of DNA replication and prevents the formation of UFBs, strongly suggesting that, in unchallenging condition, these structures result from the accumulation of unreplicated DNA during mitosis [16,19]. Topo IIα inhibition leads to an increase in UFB frequency at metaphase (figure 1d). We therefore first investigated whether Topo IIα inhibition prevented the completion of DNA replication, leading to the formation of new UFB-containing unreplicated DNA on entry into mitosis. We therefore determined whether centromere replication was impaired upon Topo IIα inhibition. Cells were left untreated or were treated with ICRF-159 for 8 h. We used CREST staining to quantify double-dotted (yellow arrows, figure 2a) and single-dotted (white arrow, figure 2a) foci in prometaphase corresponding to fully replicated and unreplicated centromeres, respectively (figure 2a), as previously described [16]. The frequency of unreplicated centromeres was significantly higher in cells treated with 1 or 10 µM ICRF-159 for 8 h than in control cells ( figure 2a and b), demonstrating that Topo IIα inhibition impaired the replication of centromeric DNA. Interestingly, the frequency of unreplicated centromeres did not differ between cells treated with 1 and 10 µM ICRF-159, contrasting with the dose-dependent effect of ICRF-159 on UFB formation (figure 1c).
royalsocietypublishing.org/journal/rsob Open Biol. 10: 190259 For confirmation of the effect of Topo IIα inhibition on DNA replication, we then evaluated the levels of mitotic DNA synthesis (MiDAS). MiDAS contributes to the processing of unreplicated DNA sequences during mitosis and can therefore be used to detect problems leading to incomplete DNA replication during the previous S-phase [16,19,[26][27][28]. MiDAS can be visualized by 5-ethynyl-2 0 -deoxyuridine (EdU) incorporation, leading to the formation of foci on condensed chromosomes (yellow arrow, figure 2d). We found that treatment with 1 or 10 µM ICRF-159 for 8 h led to a significant increase in the percentage of prometaphase cells presenting MiDAS, with no dose dependence (figure 2c-e). These data confirm that Topo IIα inhibition results in an accumulation of unreplicated DNA during mitosis, reflecting incomplete DNA replication in the previous S-phase. These observations are consistent with several studies in yeast or in vitro, showing that Topo IIα facilitates DNA replication [2,[29][30][31][32]. They also suggest that Topo IIα activity is essential to promote complete DNA replication in mammalian cells.
Our data demonstrate that Topo IIα inhibition impairs the completion of DNA replication, probably leading to the formation of UFB-containing unreplicated DNA on entry into mitosis.

Topoisomerase IIα inhibition promotes the formation of two different types of ultrafine anaphase bridges depending on the phase of the cell cycle
Topo IIα inhibition increases total UFB frequency in a dosedependent manner, but impairs DNA replication independently of ICRF-159 concentration, suggesting that Topo IIα inhibition affects another process of UFB formation, in addition to DNA replication. Topo IIα activity is required for both the completion of DNA replication during S-phase (figure 2) [2,29-32] and the DNA decatenation during mitosis [9]. We royalsocietypublishing.org/journal/rsob Open Biol. 10: 190259 therefore investigated the respective contributions of these processes to the increase in UFB formation in response to Topo IIα inhibition. We analysed UFB frequency in HeLa anaphase cells after treatment either during S-phase (addition of ICRF-159 for 6 h followed by a release period of 3 h), or during mitosis (addition of ICRF-159 1 h before UFB analysis) (figure 3a). We confirmed the cell cycle phase specificity of our treatment by treating cells only during S-phase or only during mitosis (figure 3a), with EdU. As expected, all mitotic cells treated with ICRF-159 and EdU during S-phase were positive for EdU, whereas cells treated only during mitosis were EdUnegative (figure 3b). Consistent with these results, we found that Topo IIα inhibition during S-phase led to an increase in the percentage of unreplicated centromeres during mitosis and to an increase in the level of MiDAS (figure 3c and d), whereas inhibition during mitosis did not. These data confirm our previous findings ( figure 2) and demonstrate the cell cycle specificity of the treatment. These results indicate that Topo IIα activity is required during S-phase, to promote complete DNA replication.
In the same experimental conditions, Topo IIα inhibition during S-phase led to a slight, but significant, dose-independent increase in the mean number of UFBs per cell (figure 3b), probably due to the effect of Topo IIα inhibition on the completion of DNA replication (figures 2 and 3c,d). However, in cells treated with Topo IIα inhibitor only during mitosis, we observed a much higher dose-dependent increase in UFB frequency (figure 3b). These data suggest that most of the UFBs observed upon Topo IIα inhibition result from the loss of Topo IIα activity during mitosis.
We then investigated the effect of Topo IIα inhibition on the formation of different types of UFB as a function of the phase of the cell cycle. We treated cells with ICRF-159 (1 and 10 µM) during S-phase or during mitosis, and we analysed UFB   UFB frequency decreased during the course of mitosis in both treated and untreated conditions. We therefore hypothesized that Topo IIα inhibition in S-phase would impair complete DNA replication, leading to the formation of UFBcontaining unreplicated DNA on entry into mitosis. By contrast, the restriction of Topo IIα inhibition to mitosis had no effect on UFB frequency at metaphase (red arrow, figure 3f ). However, UFB frequency was much higher at the metaphaseanaphase transition, particularly in response to 10 µM ICRF-159 (green arrow, figure 3f ), reflecting the formation of new UFBs during mitosis. UFB frequency subsequently decreased during anaphase (figure 3f ). Interestingly, the increase in UFB frequency at metaphase-anaphase transition was not observed in cells treated only during S-phase (green arrow, figure 3e). Topo IIα activity is required to resolve centromeric DNA catenations at the metaphase-anaphase transition [7]. We therefore suggest that DNA decatenation is compromised when Topo IIα is inhibited during mitosis, promoting the formation of UFB-containing DNA catenanes in anaphase. Consistent with this hypothesis, we observed that 66% of UFBs were of centromeric origin when Topo IIα was inhibited during mitosis (figure 3b; electronic supplementary material, figure S1A and B). This region of the chromosome has already been shown to be associated with UFB-containing DNA catenanes [11,14,21,22]. More importantly, UFB frequency decreased from early anaphase to late anaphase in cells treated with ICRF-159 during S-phase or during mitosis, despite the maintenance of Topo IIα inhibition during mitosis. These data indicate that UFBs are either resolved in absence of Topo IIα activity or probably broken due to aberrant resolution.

Topoisomerase IIα activity prevents DNA damageassociated resolution of ultrafine anaphase bridge during mitosis
To determine if Topo IIα activity is dispensable or not for UFB resolution during mitosis, we addressed the specific fate of UFBs. It has been reported that aberrant UFB resolution at the end of mitosis causes DNA damage leading to the formation of 53BP1 bodies in the next G1 phase, to protect broken DNA ends until repair [33]. We investigated whether UFB resolution in cells treated with Topo IIα inhibitors was associated with DNA damage, by analysing the number of 53BP1 foci in the next G1 phase. Cells were synchronized by double thymidine block at the G1/S boundary and then released into cell cycle. Cells were left untreated or treated with ICRF-159 during S-phase or during mitosis (cell cycle distribution is shown in electronic supplementary material, figure S2A). First, to ensure that double thymidine block did not interfere with UFB formation when Topo IIα was inhibited, we analysed UFB frequency in synchronized cells treated with Topo IIα inhibitor during either S phase or mitosis, and we analysed the percentage of FANCD2-associated UFBs in these cells (electronic supplementary material, figure S2B and C). We confirmed an increase in UFB frequency in both cells treated during S-phase and during mitosis and found, as expected, that cells present a significant increase in the percentage of FANCD2-associated UFBs only when treated with Topo IIα inhibitor during S phase (electronic supplementary material, figure S2C). These results further support that the inhibition of Topo IIα during S-phase leads to unreplicated DNA marked by sister FANCD2 foci, but not when Topo IIα is inhibited in mitosis. Then, we analysed 53BP1 foci in the next G1 phase of synchronized cells, left untreated or treated with 1 or 10 μM ICRF-159 during S-phase or during mitosis (figure 4a-c; electronic supplementary material, figure S2A). Topo IIα inhibition during S-phase led to a slight, but significant, dose-independent increase in the number of 53BP1 foci (figure 4a-c). However, in cells treated with Topo IIα inhibitor only during mitosis, we observed a much higher dose-dependent increase in 53BP1 foci. These results demonstrate that DNA damage in G1 was correlated to UFB frequency in the previous mitosis, meaning that more UFBs are formed in the previous mitosis, and more 53BP1 foci are generated in the next G1. These observations indicate aberrant UFB resolution during anaphase when Topo IIα is inhibited, probably by DNA breakage. These results confirm that Topo IIα activity is necessary for DNA damage-free resolution of UFB during mitosis.
Overall, our data shed light on the molecular origin of the supernumerary UFBs observed following Topo IIα inhibition, showing that they correspond to newly formed UFBs and probably not, or to a lesser extent, to unresolved pre-existing UFBs. Indeed, our data demonstrate that maintaining Topo IIα inhibition during mitosis does not affect the global UFB disappearance after metaphase-anaphase transition but lead to an accumulation of DNA damage in the next G1, reflecting aberrant UFB resolution. We concluded that Topo IIα activity royalsocietypublishing.org/journal/rsob Open Biol. 10: 190259 in mitosis is necessary for the correct resolution of UFBs, as previously demonstrated [20].
We also found that Topo IIα inhibition during S-phase compromised the completion of DNA replication, leading to the accumulation of unreplicated DNA during mitosis, probably leading to the formation of UFB-containing unreplicated DNA. The restriction of Topo IIα inhibition to mitosis resulted in a much higher frequency of UFBs at the metaphaseanaphase transition, particularly in the presence of 10 µM ICRF-159, reflecting the formation of new UFBs during mitosis.  royalsocietypublishing.org/journal/rsob Open Biol. 10: 190259 These UFBs probably result from impaired DNA decatenation at the metaphase-anaphase transition and correspond to newly formed UFB-containing DNA catenanes (figure 4d). Thus, our results indicate that the excess UFB observed upon Topo IIα inhibition results mainly from newly formed UFBs, in a replication-or decatenation-dependent manner, rather than from a delayed UFB resolution.
In conclusion, our findings show that Topo IIα is necessary to promote DNA damage-free resolution of UFBs and further extend the role of Topo IIα activity during the cell cycle, by showing that Topo IIα is required for complete DNA replication.

Cell culture and treatments
HeLa cells were cultured in DMEM supplemented with 10% FCS as previously described [16].
ICRF-159 (Razoxane) was provided by Sigma Aldrich (R8657) and was added to the cell culture medium at a final concentration of 1 or 10 µM following the protocol described in figures 1a, 2c, 3a, 4a and electronic supplementary material, figure S2A. Thymidine was provided by Sigma Aldrich (T9250) and was added to the cell culture medium at a final concentration of 2 mM.
All cells were routinely checked for mycoplasma infection.

EdU staining
EdU incorporation into DNA was visualized with the Click-it EdU imaging kit (C10338 from Life Technologies), according to the manufacturer's instructions. EdU was used at a concentration of 2 µM for the indicated time. Cells were incubated with the click-it reaction cocktail for 15 min. Cell images were acquired with a three-dimensional deconvolution imaging system consisting of a Leica DM RXA microscope equipped with a piezoelectric translator (PIFOC; PI) placed at the base of a 63x PlanApo N.A. 1.4 objective and a Cool-SNAP HQ interline CCD camera (Photometrics). Stacks of conventional fluorescence images were collected automatically at a Z-distance of 0.2 μm (Metamorph software; Molecular Devices). Images are presented as maximumintensity projections generated with ImageJ software, from stacks deconvolved with an extension of Metamorph software.

Flow cytometry analysis
Cells were synchronized using double thymidine block: cells were incubated with 2 mM thymidine during 16 h and then released during 10 h in fresh medium and incubated again with 2 mM thymidine during 16 h. After ICRF-159 treatment, cells were detached by treatment with Accutase (Sigma), immediately washed in 1x PBS, fixed in 70% ethanol and stored at −20°C overnight. Cells were then washed twice with ice-cold 1x PBS and incubated with Vindelov solution (Tris HCl, pH 7.6 3,5 mM; NaCl 10 mM, propidium iodide 50 µg ml −1 ; NP40 0.1%; RNAse 20 µg ml −1 ) during 30 min in the dark. Finally, cell cycle analysis was analysed using FACSCanto II from BD Biosciences.

Statistical analysis
At least three independent experiments were carried out to generate each dataset and the statistical significance of differences was calculated with Student's t-test or two-way ANOVA, as indicated in figure legends.
Data accessibility. This article has no additional data.