ORIGINAL ARTICLE

Topoisomerase IIa maintains genomic stability through decatenation

G2 checkpoint signaling

JJ Bower1,2, GF Karaca1, Y Zhou1, DA Simpson1, M Cordeiro-Stone1,2,3 and WK Kaufmann1,2,3

1Department of Pathology and Laboratory Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA;2Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA and 3Center forEnvironmental Health and Susceptibility, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

Topoisomerase IIa (topoIIa) is an essential mammalianenzyme that topologically modifies DNA and is requiredfor chromosome segregation during mitosis. Previousresearch suggests that inhibition of topoII decatenatoryactivity triggers a G2 checkpoint response, which delaysmitotic entry because of insufficient decatenation ofdaughter chromatids. Here we examine the effects ofboth topoIIa and topoIIb on decatenatory activity in cellextracts, DNA damage and decatenation G2 checkpointfunction, and the frequencies of p16INK4A allele loss andgain. In diploid human fibroblast lines, depletion oftopoIIa by small-interfering RNA was associated withseverely reduced decatenatory activity, delayed progres-sion from G2 into mitosis and insensitivity to G2 arrestinduced by the topoII catalytic inhibitor ICRF-193.Furthermore, interphase nuclei of topoIIa-depleted cellsshowed increased frequencies of losses and gains of thetumor suppressor genetic locus p16INK4A. This study showsthat the topoIIa protein is required for decatenationG2 checkpoint function, and inactivation of decatenationand the decatenation G2 checkpoint leads to abnormalchromosome segregation and genomic instability.Oncogene (2010) 29, 4787–4799; doi:10.1038/onc.2010.232;published online 21 June 2010

Keywords: decatenation; G2 checkpoint; topoisomeraseIIa; topoisomerase IIb; chromosomal instability

Introduction

DNA topoisomerases are evolutionarily conservednuclear enzymes that have major functions in the cellcycle, including DNA replication, recombination andchromosome segregation (Larsen et al., 1996; Wang,2002). The two major families of topoisomerases aredifferentiated by the type of enzymatic reactions each ofthem performs. Type-I topoisomerases produce protein-associated single-strand breaks in DNA and relievesupercoiling tension by free rotation of the cut strand

around the intact strand. Type-II topoisomerasesproduce protein-associated DNA double-strand breaks(DSBs) and are capable of passing an intact DNAduplex through the protein-associated DSB. Thus, onlytype-II topoisomerases can separate knotted and inter-twined DNA molecules. Mammalian species express twotopoII isoforms: a and b. Each isoform is encoded by aseparate gene located on different human chromosomesand can be distinguished by their mass (Capranico et al.,1992; Austin and Marsh, 1998; Burden and Osheroff,1998). The two isoforms share B70% homology at theamino-acid level, and exist as homodimeric enzymaticcomplexes with similar catalytic activities, although a/bheterodimers have been observed (Biersack et al., 1996;Gromova et al., 1998).

TopoIIb is non-essential and constitutively expressed,whereas topoIIa is an essential gene maximally ex-pressed in G2- and M-phase (Heck et al., 1988; Goswamiet al., 1996; Akimitsu et al., 2003). TopoII activity isrequired for chromosome condensation, decatenation ofintertwined daughter DNA duplexes and centromereresolution (Maeshima et al., 2005; Coelho et al., 2008;Wang et al., 2008). Catenations between chromatidarms are removed before mitosis, while centromericcatenations persist up to the metaphase/anaphasetransition (Diaz-Martinez et al., 2006; Santamariaet al., 2007).

DNA and topoIIa form a reversible, covalentcomplex, often referred to as the cleavage complex(Burden and Osheroff, 1998; Cortes et al., 2003).TopoII-inhibiting drugs interfere with various stages ofthe catalytic cycle, and are therefore divided into twoclasses: poisons and catalytic inhibitors. TopoII poisonssuch as doxorubicin and etoposide stabilize the cleavagecomplex, which may block DNA replication forks ortranscriptional machinery and create DSBs (Xiao et al.,2003). On proteolysis of the stabilized cleavage complex,the DSB is exposed; thus, these drugs are used clinicallyto treat cancers that typically overexpress the topoIIenzyme (Jarvinen et al., 1996; Mao et al., 2001; Lorussoet al., 2007).

In contrast, catalytic inhibitors prevent the formationof the cleavage complex by intercalating into DNA andinhibiting topoII binding, or by stabilizing topoII in aclosed-clamp conformation after the ligation step ofthe catalytic cycle (Roca et al., 1994; Sehested andJensen, 1996). Therefore, topoII catalytic inhibitors are

Received 6 February 2010; revised 5 May 2010; accepted 13 May 2010;published online 21 June 2010

Correspondence: Dr WK Kaufmann, Department of Pathology andLaboratory Medicine, University of North Carolina at Chapel Hill,Rm 31-345 Lineberger Comprehensive Cancer Center, CB#7295,Chapel Hill, NC 27599, USA.E-mail: bill_kaufmann@med.unc.edu

Oncogene (2010) 29, 4787–4799& 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10

www.nature.com/onc

primarily used in the clinical setting as an adjunct toreduce the cardiotoxicity of topoII poisons (Lyu et al.,2007). The ability of topoII catalytic inhibitors toprotect against the toxicity of topoII poisons impliesthat poisons and catalytic inhibitors have very differenteffects on topoII activity, DNA binding and chromatinstructure.

Studies using topoII catalytic inhibitors suggest thatG2 cells monitor the catenation state of intertwinedsister chromatids following DNA replication andactively delay progression into mitosis pending sufficientchromatid decatenation (Downes et al., 1994; Deminget al., 2001). A subset of chromatid catenations appearto be organized in the centromere, and these catenationsare not separated until the onset of anaphase (Diaz-Martinez et al., 2006). Thus, the decatenation G2

checkpoint appears to monitor the sufficiency ofdecatenation along the chromosomal arms, and not itscompletion. The expression of an active decatenation G2

checkpoint was supported by evidence that the G2 delayin cells with catalytic inhibition of topoII was dependenton BRCA1 expression and could be over-ridden bycaffeine (Deming et al., 2001). BRCA1 could be requiredto activate topoIIa by ubiquitylation (Lou et al., 2005).Other studies have implicated ATR, ATM, Plk1, WRN,MDC1 and Chk1 in the decatenation G2 checkpoint(Deming et al., 2002; Franchitto et al., 2003; Robinsonet al., 2007; Luo et al., 2009; Bower et al., 2010).Although the presence of a decatenation G2 checkpointthat is independent of DNA damage has been supportedby a variety of studies (Doherty et al., 2003; Nakagawaet al., 2004; Damelin et al., 2005), the concept is stillcontroversial as ICRF-193 has been shown to activateDNA damage signaling in some cancer cell lines(Nakagawa et al., 2004; Park and Avraham, 2006).

The study described herein utilizes normal humandiploid fibroblast (NHDF) lines isolated from threedifferent individuals and depleted of either topoIIa ortopoIIb. These data support a distinction between thedecatenation and DNA damage-G2 checkpoints, andsuggest that topoIIa has a role in maintaining genomicstability through chromatid decatenation and decatena-tion G2 checkpoint signaling.

Results

TopoIIa is required for decatenatory activityAs there are two forms of topoII in mammals, it wasnecessary to determine the individual contribution ofeach isoform to cellular decatenatory activity. NHDFswere depleted of topoIIa or topoIIb using small-interfering RNA (siRNA) and assayed 48 h afterelectroporation. Representative western blots for threeNHDF lines are shown (Figure 1a, SupplementaryFigure S1a). Nuclear extracts were prepared andassayed for in vitro decatenatory activity on catenatedcircular DNA molecules isolated from trypanosomekinetoplasts (kDNA). Incubation of kDNA withNHDF nuclear extracts released free mini-circles and

intermediate mobility DNA species that likely representcatenated dimers or trimers, while topoII decatenatoryactivity was inhibited by the topoII catalytic inhibitor,ICRF-193 (Figure 1b). Furthermore, decatenatoryactivity was severely reduced in nuclear extracts isolatedfrom topoIIa-depleted cells relative to cells thatwere electroporated with non-targeting control (NTC)siRNA (Figure 1b, Supplementary Figure S1b). To-poIIb-depleted nuclear extracts displayed similar activityto that of the NTC extracts. Collectively, these resultssuggest that topoIIa accounts for the majority ofdecatenatory activity in actively growing NHDFs.

Catalytic inhibition of topoII with ICRF-193 inhibitsmitotic entryCatalytic inhibitors of topoII enzymes are thought toinduce a decatenation G2 checkpoint and prevent entryinto mitosis. In order to verify that the decatenation G2

checkpoint was effective in NHDFs, live-imagingbright-field microscopy was used. Mitotic NHDFs wereclearly distinguishable from interphase fibroblasts underbright-field microscopy because of their round morpho-logy, whereas interphase fibroblasts were flat, thin andelongated. Time-lapse images were collected every 2min

NHF3-hTERTNuclear Extracts

Catenated kDNANTC T2α T2β

siRNA

siRNA

Topo II β (T2β) Protein

Topo II α (T2α) Protein

T2α T2β T2αT2α T2βT2β

NHF1-hTERT NHF3-hTERT NHF10-hTERTCell Line

Control ICRF

NHF1-hTERTNuclear Extracts

% Depletion 98% 79% >99% 73% 85% 74%

Decatenated kDNA

% Decatenated kDNA 42842651

Figure 1 TopoIIa enzymatic activity accounts for the majority ofdecatenatory activity in normal human fibroblasts. (a) Westernimmunoblot analysis showing topoIIa depletion 48 h after electro-poration in three different normal human fibroblast lines with anaverage reduction in protein levels of 94%. TopoIIb depletionaveraged 75%. (b) The catalytic topoII inhibitor ICRF-193decreased the in vitro decatenatory activity of normal humandiploid fibroblast (NHDF) nuclear extracts. Small-interferingRNA (siRNA) depletion of topoIIa ablated the decatenatoryactivity of nuclear extracts, whereas topoIIb depletion had no effecton decatenatory activity. Results are representative of threeindependent experiments in three different NHDF cell lines.

TopoIIa and genomic stabilityJJ Bower et al

4788

Oncogene

and full-length movies can be viewed at our laboratorywebsite (http://top2a.med.unc.edu/jackie/home.html). Re-presentative images at 0, 1, 4 and 12 h after addition ofdimethyl sulfoxide (DMSO) or ICRF-193 are shown,with black arrows designating the mitotic cells(Figure 2). NHDFs exposed to DMSO entered andexited mitosis throughout the length of the movie,whereas catalytic inhibition of topoII with 4 mM ICRF-193 allowed mitotic exit, but not mitotic entry. Flow-cytometric analysis of cellular DNA content furthershowed that greater than 50% of fibroblasts were inG2 phase after 28 h of incubation with ICRF-193(Supplementary Figure S2). These results indicated thatmitotic entry was arrested in NHDFs on catalyticinhibition of topoII enzyme.

ICRF-193 does not induce detectable DNA damagein NHDFsA few studies have detected DNA damage (particularlyDSBs) on inhibition of topoII with ICRF-193. In orderto verify that catalytic inhibition of topoII with ICRF-193 did not induce a DNA damage response in NHDFs,immunofluorescence and flow cytometric studies wereperformed to determine the levels of a common DNAdamage marker, gH2AX. In Figure 3a, representativeimages are shown for NHF1-hTERTs treated for 15minwith DMSO, ICRF-193 or etoposide and stained forgH2AX by immunofluorescence. Etoposide-treatedNHDFs displayed many gH2AX foci after 15min,whereas no detectable gH2AX foci were present afterICRF-193 treatment. Flow cytometric studies wereperformed in parallel to quantify gH2AX positiveNHDFs (Figure 3b). All NHDFs showed an increasein gH2AX positive cells after etoposide treatment,whereas ICRF-193 treatment induced no detectablegH2AX staining above background levels. Similar

results have been observed by western immunoblotanalysis in lymphoblast and fibroblast cell lines (Boweret al., 2010). The slight increase in gH2AX staining seenat the 6 h time point was determined to be statisticallyinsignificant by a Student’s t-test, and was the result ofan idiosyncratic response by the NHF3-hTERT cell line.NHF3-hTERTs also displayed higher backgroundgH2AX staining and hypersensitivity to both ICRF-193 and topoIIa depletion (Zhou Y and Kaufmann WK,unpublished data and Figure 7c), suggesting that thesecells may display higher levels of cellular stress inculture. Overall, these results indicate that ICRF-193does not induce DSB formation in NHDFs.

TopoIIa depletion ablates the decatenation G2 checkpointresponse, but has no effect on the DNA damage G2

checkpoint responseICRF-193 inhibits topoII function by holding theenzyme as a closed clamp that is tethered to DNA butincapable of DNA strand passage, thus rendering cellsunable to remove DNA catenations (Roca et al., 1994;Jensen et al., 2000; Xiao et al., 2003). Previous studieshave suggested that G2 cells sense the presence ofentangled and/or catenated chromatids and activelydelay mitosis until decatenation is sufficiently advancedto permit mitotic cell division (Downes et al., 1994).Therefore, conditions that slow the rate of chromatiddecatenation should reduce the rate of mitotic entry bycausing a G2 delay, leading to an increase in the G2 cellfraction.

To test this hypothesis, fibroblasts were electropo-rated with NTC or topoIIa siRNA and analyzed fortheir ability to arrest in G2 and exit mitosis onDNA damage (ionizing radiation (IR)) or catalyticinhibition (ICRF-193) of topoII enzyme. Representativeflow cytometry plots of NHF1-hTERTs are shown

0 Hr 1 Hr 4 Hr 12 Hr

NHF1-hTERT0.1 % DMSO

NHF1-hTERT4 μM ICRF-193

Figure 2 Catalytic inhibition of topoII prevents entry into mitosis. Normal human diploid fibroblasts (NHDFs) were observedthrough time-lapse bright-field microscopy in the presence of dimethyl sulfoxide (vehicle control) or the catalytic topoIIinhibitor ICRF-193. Control cells continued to enter and exit mitosis throughout the 24 h period of observation. Blackarrows indicate several mitotic cells in each frame. After addition of ICRF-193, all initial mitotic cells exited mitosis within2–3 h of topoII inhibition, and no new mitotic figures were observed up to 24 h after treatment (see full length movies athttp://top2a.med.unc.edu/jackie/home.html.).

TopoIIa and genomic stabilityJJ Bower et al

4789

Oncogene

in Figure 4a. In both NTC and topoIIa-depletedfibroblasts, ionizing radiation produced a 75–80%average reduction in mitotic index 2 h after treatment(Figure 4b). This reduction of the mitotic index haspreviously been thought to reflect both an arrest of G2

cells and the exit of mitotic cells into interphase afterirradiation. The NTC-depleted fibroblasts retained afunctional decatenation G2 checkpoint; they displayed amean 60% reduction in mitotic index on ICRF-193treatment. In contrast, topoIIa-depleted fibroblasts

0

5

10

15

20

25

30

35

40

45

0.25 2 6

% P

ho

sph

o-H

2AX

+ C

ells

Hours After Treatment

0.1% DMSO 4 μM ICRF-193 4 μM Etoposide

No Treatment 0.1% DMSO - 15 Minutes

12 μM Etoposide - 15 Minutes 4 μM ICRF-193 - 15 Minutes

200X

1000X 1000X

200X 200X

200X

*

**

Figure 3 Catalytic inhibition of topoII with ICRF-193 does not induce phospho-H2AX foci in normal human diploid fibroblasts(NHDFs). (a) NHF1-hTERTs were exposed to 0.1% dimethyl sulfoxide, 4mM ICRF-193 or 12mM etoposide. Phospho-Ser 139 H2AX(gH2AX) foci were observed after a 15-min treatment with the topoII poison etoposide, which produces DNA DSBs. Catalyticinhibition of topoII with ICRF-193 did not induce detectable gH2AX foci in NHDFs. Representative images are shown at200�magnification and 1000�magnfication (inset). (b) Quantification of the percentage of cells expressing gH2AX in three NHDFlines treated with 4 mM etoposide or 4mM ICRF-193 by flow cytometry. gH2AX staining increased over time in the etoposide-treatedcells. ICRF-193 did not produce any detectable gH2AX staining until 6 h after treatment, and the signal was variable among thedifferent NHDF lines. NHF3-hTERT cells produced substantially more gH2AX staining than the other two fibroblast lines andshowed higher background levels of gH2AX. Results are presented as an average of three different NHDF cell lines±s.e.m. *Po0.05.

TopoIIa and genomic stabilityJJ Bower et al

4790

Oncogene

displayed a mean 257% increase in mitotic index ontreatment with ICRF-193. These results suggested thatwhile topoIIa depletion did not significantly alter theDNA damage-G2 checkpoint response, it ablated thedecatenation G2 checkpoint response.

One explanation for the increase in topoIIa-depletedmitotic cells after ICRF-193 treatment was that NHDFswere delaying mitotic exit because of the persistence ofcentromeric DNA catenations (Diaz-Martinez et al.,2006; Santamaria et al., 2007; Coelho et al., 2008;Wang et al., 2008). This effect might be enhanced onfurther inhibition of residual topoIIa and topoIIb withICRF-193, thus prolonging the amount of timespent in mitosis. If the amount of time NHDFs spentin mitosis was altered by a decrease in topoIIa activity,it was necessary to monitor the G2/M transition byexamining the mitotic entry rate (MER) of NHDFs inorder to discriminate between the G2/M and spindleassembly checkpoint mechanisms. NHDFs were col-lected over time in colcemid, a microtubule polymeriza-tion inhibitor that arrests cells with an intact spindleassembly checkpoint at metaphase. This assay allows fora strict examination of the G2 to M transition, andcalculation of the MER by linear regression provides aquantitative measure of decatenation G2 checkpointfunction.

NTC-, topoIIa-, and topoIIb-depleted fibroblast linesdisplayed an increase in the percentage of mitotic cellsover time in the presence of DMSO and colcemid(Figure 4c, Supplementary Figure S3). In contrast, whentopoII activity was inhibited with ICRF-193 in NTC- ortopoIIb-depleted fibroblasts, MERs were approximatelyzero. Surprisingly, the topoIIa-depleted fibroblastscontinued to enter mitosis, regardless of the presenceof ICRF-193. Comparison of MERs indicated thattopoIIa-depleted cells entered mitosis at a rate that wasapproximately half that of NTC-treated fibroblasts(Figure 4d). Mitotic entry in the topoIIa-depletedNHF1-hTERTs was independently verified by time-lapse bright-field microscopy and full-length videos canbe viewed at the laboratory’s website (http://top2a.med.unc.edu/jackie/home.html). Quantification of G2

fractions in this assay also revealed that a significant G2

accumulation occurred on topoIIa depletion or catalyticinhibition (Supplementary Figure S4).

Several attempts to correct the phenotype ofNHDFs by over-expressing a siRNA-resistant form oftopoIIa were unsuccessful because of the toxicityof topoIIa overexpression. Therefore, a second siRNAdirected towards topoIIa was tested in the mitoticentry assay. The results confirmed that the observedphenotype was not because of the off-target effects ofeither siRNA (Supplementary Figure S5). Taken to-gether, these data suggest that a catalytically inactivetopoIIa protein is required to arrest cells in G2 duringincubation with ICRF-193 and prevent mitotic entry inthe presence of entangled/catenated chromosomes.Furthermore, these results indicate that the decatenationG2 checkpoint does not sense DNA catenations per se,but rather senses the catalytically inactive form oftopoIIa.

TopoIIa depletion does not affect the ability of NHDFsto synthesize DNAAs topoIIa protein expression is cell cycle-dependentand an earlier report suggested that type II topoisome-rases are required for the completion of S phase inbudding yeast (Baxter and Diffley, 2008), it wasimportant to determine whether topoIIa has a role inDNA synthesis and/or the completion of S phase inNHDFs. If DNA replication or completion of S phasewere impaired, this could slow down the MERs. Flowcytometry was utilized to determine the capacity oftopoIIa-depleted cells to incorporate the thymidineanalog bromodeoxyuridine during S phase and repre-sentative flow cytometry profiles are shown (Figure 5a).While topoIIa-depleted cells displayed normal levels ofbromodeoxyuridine incorporation, there were slightlyfewer fibroblasts present in S phase when compared withNTC-treated fibroblasts (Figure 5b). This difference wasdetermined to be insignificant by Student’s t-test.However, the percentage of G2 cells increased intopoIIa-depleted fibroblasts, suggesting that cells wereindeed progressing through S phase and into G2 phase.Similar results were obtained on ICRF-193 treatment(Supplementary Figure S6) and by using synchronizedfibroblasts (not shown), suggesting that topoIIa deple-tion or inhibition does not interfere with DNAreplication or completion of S phase in NHDFs. Inaddition, recently published data from our laboratoryhas suggested that neither ATR nor Chk1 depletionaffected the MERs or decatenation G2 checkpointfunction of NHDFs (Bower et al., 2010). Similarly, noChk1 activation was observed in response to ICRF-193treatment. These data further support the conclusionthat topoII catalytic inhibition or depletion does notinduce a replicative stress response in NHDFs.

TopoIIa-depleted fibroblasts display abnormal metaphasechromosome structureInhibition of topoII with ICRF-193 produces character-istic chromosomal abnormalities in cells that evadethe decatenation G2 checkpoint. Metaphase prepara-tions of ICRF-193-treated cells often display under-condensed and entangled chromosomes (Deming et al.,2002). As the results in Figures 1b and 4c indicated thatNHDFs enter mitosis in the absence of significanttopoIIa activity, cytogenetic studies were performed toexamine chromosome morphology. TopoIIa-depletionsignificantly increased the frequencies of metaphasepreparations with under-condensed or entangled chromo-somes, whereas topoIIb-depleted fibroblasts displayedno significant increase in chromosomal aberrations(Figures 6a and b). Thus, depletion of topoIIa producedsimilar effects on chromosome structure as catalyticinhibition of topoII with ICRF-193.

Depletion of topoIIa increases the frequency of abnormalinterphase nuclei and promotes losses and gains of thep16INK4A tumor suppressor alleleAlthough the cytogenetic experiments in Figures 6a andb established that NHDFs would enter mitosis with

TopoIIa and genomic stabilityJJ Bower et al

4791

Oncogene

entangled chromosomes, it was unclear whether thesecells would exit mitosis, re-enter interphase, andpotentially duplicate a set of damaged or aneuploidchromosomes during subsequent rounds of DNAreplication and cell division. As entangled chromosomesmay not segregate properly at anaphase, failures of

segregation (non-disjunction) can generate anaphasebridges, lagging chromosomes and/or chromosomalbreakage. After restitution of interphase nuclear struc-ture, thin bridges connecting two nuclei may beapparent or one nucleus may have two lobes. Laggardor broken chromosomes are often manifested as

Pho

spho

- H

isto

ne H

3 (F

ITC

)

NTCsiRNA

Topo II αsiRNA

100

101

102

103

104

100

101

102

103

104

64 128 64 128 64 128 64 128

0.08 %1.70 % 1.48 % 0.41%

0.21 %1.03 % 0.90 % 1.92 %

80 % Inhibition

95 % Inhibition 72 % Inhibition

0 % Inhibition

DNA Content (PI)

Sham 1.5 Gy 0.1% DMSO 4μM ICRF-193

0%

50%

100%

150%

200%

250%

300%

350%

NTC siRNA Topo II α siRNA

% o

f C

on

tro

l (M

ito

tic

Ind

ex)

1.5 Gy IR 4μM ICRF-193

*

0

2

4

6

8

10

12

0 1 2 3 4 5 6 7% M

ito

tics

(30

,000

Cel

ls)

Time (Hrs)

NHF3-hTERT NTC siRNA

0123456789

10

0 1 2 3 4 5 6 7% M

ito

tics

(30

,000

Cel

ls)

Time (Hrs)

NHF3-hTERT Topo IIα siRNA

02468

10121416

0 1 2 3 4 5 6 7% M

ito

tics

(30

,000

Cel

ls)

Time (Hrs)DMSO ICRF-193

NHF3-hTERT Topo IIβ siRNA

NTC siRNACell Line 0.1% DMSO % Inhibition of Mitotic Entry

NHF1-hTERT 2.43 0.12 98%

NHF3-hTERT 1.54 -0.11 100%

NHF10-hTERT 1.61 0.01 100%

0.1% DMSO % Inhibition of Mitotic Entry

NHF1-hTERT 0.80 1.60 0%

NHF3-hTERT 1.10 1.25 0%

NHF10-hTERT 0.62 0.76 0%

0.1% DMSO 4μM ICRF-193

4μ μ M ICRF-193

4μM ICRF-193

% Inhibition of Mitotic Entry

NHF1-hTERT 0.16 88%

NHF3-hTERT 2.07 -0.11 100%

NHF10-hTERT 0.61 -0.06 100%

1.33

*

Topo IIα siRNACell Line

Topo IIβ siRNACell Line

***

TopoIIa and genomic stabilityJJ Bower et al

4792

Oncogene

micronuclei. The frequencies of interphase nuclei withstructural alterations such as bridges, blebs and micro-nuclei were therefore examined (Figure 6c). While NTC-treated fibroblasts showed a low frequency of nuclear

abnormalities (o5%), the frequencies of abnormalitieswere increased to B33 or B14% after depletion oftopoIIa or topoIIb, respectively (Figure 6d). Theseresults suggested that NHDFs could exit mitosis with

0

10

20

30

40

50

60

70

80

NTC siRNA Topo IIα siRNA Topo IIβ siRNA

Ave

rag

e %

of

Fib

rob

last

s(3

0,00

0 C

ells

)

G1 S G2/M

NHF10-hTERT NTC siRNA NHF10-hTERT Topo IIα siRNA

100

101

102

103

104

100

101

102

103

104

100

101

102

103

104

64 128 192 64 128 192

DNA Content (PI)

Brd

U In

corp

orat

ion

(FIT

C)

NHF10-hTERT Topo IIβ siRNA

64 128 192

*

*

Figure 5 Depletion of topoIIa or topoIIb does not affect DNA replication or completion of S phase in Normal human diploidfibroblasts (NHDFs). (a) Representative flow cytometry images of NHF10-hTERTs analyze for the ability to incorporatebromodeoxyuridine (BrdU) and plotted against DNA content. Bromodeoxyuridine-incorporating cells between 2N and 4N DNAcontent were designated as S phase cells. (b) Total percentages of fibroblasts in G0/G1, S or G2/M compartments in the presence orabsence of topoIIa or topoIIb. There was no statistically significant change in the percentage of cells found in G1 among the threegroups of NHDFs. TopoIIa-depleted NHDFs exhibited an increase in the percentage of G2/M fibroblasts compared to the NTCtreated NHDFs. However, neither topoIIa- nor topoIIb-depletion affected bromodeoxyuridine incorporation or initiated replicationfork collapse because of replication stress (cells containing S phase DNA content but no longer incorporating bromodeoxyuridine).Results are presented as the average of three independent experiments, each using a different NHDF lines. *Po0.05.

Figure 4 TopoIIa is required for the decatenation G2 checkpoint, but not for the DNA damage G2 checkpoint response. (a) Dualcolor flow cytometry using a mitosis-specific antibody directed towards phospho-Ser10 histone H3 and propidium iodide to analyzeDNA content was employed to examine the G2 checkpoint response to catalytic inhibition of topoIIa. Normal human diploidfibroblasts (NHDFs) depleted of topoIIa emptied mitosis 2 h after treatment with 1.5 Gy of ionizing radiation. In contrast, topoIIa-depleted fibroblasts failed to empty mitosis 2 h after ICRF-193 treatment, and instead appeared to accumulate in mitosis. Results arerepresentative of three independent experiments with NHF1-hTERTs. (b) Quantification of flow cytometry results illustrated inFigure 4a. Results are the average of three independent experiments with NHF1-hTERTs. *Po0.0025. (c) Depletion of topoIIa allowsNHDFs to continuously enter mitosis in the presence of ICRF-193. The rate of mitotic entry of topoIIa-depleted NHDFs wasapproximately half the rate of those cells electroporated with NTC small interfering RNA on average. TopoIIb-depleted fibroblastsdisplayed similar mitotic entry rates and a similar response to ICRF-193 as the NTC-treated fibroblasts. The results depicted in theillustration are the average of three independent experiments with NHF3-hTERTs, and were robustly conserved in all the threeNHDFs (shown in Supplementary Figure S3). (d) Rates of mitotic entry for NTC-, topoIIa- and topoIIb-depleted NHDFs. Mitoticentry rates were calculated as linear regression slopes from Figure 4c. The ratios of the slopes (4 mM ICRF-193:0.1% dimethylsulfoxide) were used to determine the percentage of mitotic entry inhibition (1-ratio� 100%). Maximum and minimum values ofinhibition were set at 100 and 0%, respectively. Results are an average of three independent experiments for each NHDF line.*Po0.005.

TopoIIa and genomic stabilityJJ Bower et al

4793

Oncogene

chromosomal mis-segregation and/or non-disjunctionerrors.

Abnormal chromosome segregation can lead to theloss and gain of genetic material and promote grosschromosomal rearrangements, thereby generating chro-mosomal instability (reviewed in Holland and Cleve-land, 2009). To determine whether decreased topoIIalevels and/or activity could generate genomic instabilityin NHDFs, fluorescence in situ hybridization probesdirected towards the tumor suppressor p16INK4A geneticlocus (red) and the centromere of chromosome 9 (green)were hybridized to cytogenetic preparations (Figure 7a).NHDFs with 2–4 copies of each locus were consideredto retain a normal allelic copy number to account forDNA replication in the log phase cells. At least 100interphase cells were analyzed from each NHDF line.The topoIIb-depleted fibroblasts failed to show anysignificant allelic copy number changes when comparedwith NTC-treated fibroblasts (Figure 7b). However, the

topoIIa-depleted NHDFs displayed increased frequen-cies of gains and deletions of the p16INK4A allele. Theseresults indicate that topoIIa is required for properchromatid segregation during mitosis, and that de-creased topoIIa levels and/or activity may generate orenhance the losses/gains of the tumor suppressorp16INK4A genetic locus. Collectively, these results suggestthat topoIIa is required to maintain a barrier againsttumor initiation and/or progression.

Owing to the large increase in the number ofabnormal interphase nuclei and the changes in the copynumber of the p16INK4A genetic locus after depletion intopoIIa level, topoIIa-depleted NHDFs were examinedfor their ability to undergo clonal expansion. Aclonogenic survival assay was performed in NHDFsafter transient topoIIa depletion (Figure 7c). Theseresults confirm that topoIIa is an essential gene withtransient depletion of the protein reducing colonyformation by an average of 68%. The colonies that

NTC siRNA

Topo IIα siRNA

Entangled

Undercondensed

0%10%20%30%40%50%60%70%80%90%

100%

% A

ber

ran

t M

etap

has

es

Entangled Undercondensed

Normal InterphaseNucleus

Bleb

Micronucleus Bridge

0%

5%

10%

15%

20%

25%

30%

35%

40%

% A

bn

orm

al N

ucl

ei

siRNA

*

*

*

*

NTC siRNA

Topo IIα siRNA

Topo IIβ siRNA

NTC siRNA

Topo IIα siRNA

Topo IIβ siRNA

Figure 6 Normal human diploid fibroblasts (NHDFs) entering mitosis without topoIIa activity have abnormal metaphasechromosome morphology and exit mitosis with abnormal nuclear morphology. (a) Examples of NHDF metaphase spreads after NTCsmall-interfering RNA (siRNA) treatment and topoIIa-depletion. Two phenotypes were observed in the topoIIa-depleted fibroblasts:under-condensation and entanglement. (b) Quantification of metaphase spreads from NTC-, topoIIa-, and topoIIb-depletedfibroblasts. Results are an average of three independent experiments, each using a different NHDF lines. At least 50 metaphase spreadsper siRNA-depletion were analyzed in a blinded manner for each line. *Po0.0025. (c) Examples of NHDF interphase nuclei and thetypes of abnormalities scored. (d) Quantification of the abnormal interphase nuclei in cells treated with NTC, topoIIa, or topoIIbsiRNA. Approximately 33% and 14% of interphase nuclei were abnormal after topoIIa or topoIIb depletion, respectively. Results arethe average of three independent experiments in the NHF1-hTERT line. Greater than one thousand interphase nuclei per treatmentwere analyzed. *Po0.01.

TopoIIa and genomic stabilityJJ Bower et al

4794

Oncogene

escaped clonal inactivation after depletion of topoIIamay have arisen from cells that recovered topoIIaexpression before lethality was manifested.

Discussion

The decatenation G2 checkpoint was originally pro-posed to delay entry into mitosis when chromosomeswere insufficiently decatenated by topoII. Earlier

evidence for this checkpoint was largely derived frompharmacological analyses using topoII catalytic inhibi-tors, such as ICRF-193 (Downes et al., 1994; Deminget al., 2001). This study provides biological evidence forthe existence of a decatenation G2 checkpoint that ismolecularly distinct from the DNA damage-G2 check-point in three genetically diverse NHDFs (Figures 3 and4). The earlier studies in which ICRF-193 triggered aDNA damage response in some cancer lines could haveused cell lines with a lax decatenation G2 checkpoint

0%10%20%30%40%50%60%70%80%90%

100%

Normal p16 loss p16 gain Cen 9 Loss Cen 9 Gain

% o

f In

terp

has

e N

ucl

ei

NTC siRNA Topo IIα siRNA Topo IIβ siRNA

NHF1-hTERT NTC siRNA NHF1-hTERT Topo IIα siRNA

Topo IIα siRNACell Line % Colony Forming Efficiency Average % of NTC siRNANHF1-hTERT 4% ± 0.7%NHF3-hTERT 1% ± 0.2% 32% ± 9%NHF10-hTERT 5% ± 1.2%

NTC siRNACell Line % Colony Forming Efficiency Average % of NTC siRNANHF1-hTERT 10% ± 0.6%NHF3-hTERT 8% ± 1.1% 100 %NHF10-hTERT 11% ± 2.4%

Topo IIβ siRNACell Line % Colony Forming Efficiency Average % of NTC siRNANHF1-hTERT 14% ± 0.8%NHF3-hTERT 17% ± 0.4% 160% ± 21%NHF10-hTERT 16% ± 1.8%

*

*

*

Figure 7 TopoIIa-depleted Normal human diploid fibroblasts (NHDFs) display allelic gains and losses of the tumor suppressorp16INK4A. (a) Examples of NHDF metaphase spreads and interphase nuclei after NTC and topoIIa depletion. The example of topoIIa-depletion depicts a bridged nucleus in which four copies of the p16INK4A allele and centromere 9 appear to undergo improper segregationto the spindle poles. (b) Quantification of the number of p16INK4A and centromere 9 alleles in interphase nuclei from NTC-, topoIIa-,and topoIIb-depleted fibroblasts. Results are an average of three independent blind experiments, one for each NHDF line. At least 100interphase nuclei per small-interfering RNA (siRNA)-depletion were analyzed for each line. *Po0.05. (c) NHDF lines were treatedwith NTC, topoIIa, or topoIIb siRNA and assayed for clonogenic survival. Results are presented as an average of three differentNHDF cell lines, assayed in triplicate, and expressed as a percentage of NTC siRNA. Individual colony forming efficiencies are alsodisplayed.

TopoIIa and genomic stabilityJJ Bower et al

4795

Oncogene

response, such as some HeLa cell lines (Bower JJ andKaufmann WK, unpublished data). HeLa cells wouldtherefore enter mitosis in the presence of catenatedchromosomes, acquire DSBs on mitotic exit, and thustrigger a gH2AX and DNA damage response.

TopoIIa-depleted mitotic cells displayed highfrequencies of entangled and under-condensed chromo-somes (Figures 6a and b) similar to biological effectsrecognized in cells treated with ICRF-193 that escapethe decatenation G2 checkpoint. Abnormal nuclearstructures that arise on the failure of chromatiddecatenation and subsequent mal-segregation were alsoobserved in topoIIa-depleted fibroblasts (Figures 6c andd), suggesting that the decatenation G2 checkpoint isrequired to prevent the breakage of sister chromatidsduring the metaphase/anaphase transition. Further-more, ICRF-193 treatment failed to prevent mitoticentry in topoIIa-depleted cells, indicating thatthe topoIIa protein has an important role during theG2/M transition. NHDFs that enter mitosis in theabsence of topoIIa, and thus escape decatenation G2

checkpoint activation, displayed losses and gains of thep16INK4A tumor suppressor gene. Collectively, theseobservations suggest that attenuation of the decatena-tion G2 checkpoint is one mechanism by which genomicinstability may be generated, leading to tumor initiationand/or progression.

Although topoIIa and topoIIb have similar in vitrobiochemical activities, the results herein imply that thephysiological roles of these proteins are very different.This is not unexpected, given that earlier studies havesuggested an independent role for topoIIb in repairingDNA cross-links and DSB-induced transcription factorrecruitment (Emmons et al., 2006; Ju et al., 2006).Furthermore, the data above suggest that topoIIbactivity is sufficient for S phase progression in NHDFs,but insufficient for decatenation G2 checkpoint activa-tion (Figures 4 and 5).

Our results are comparable to other studies in thattopoIIa appears to be required for condensation andsegregation of daughter chromosomes, with little con-tribution by topoIIb to this process. Targeted disruptionof the TOP2A gene in the HT1080 fibrosarcoma line,siRNA depletion of topoIIa in HeLa cells, and DT40conditional topoIIa-depleted avian B-cells producedsimilar results (Carpenter and Porter, 2004; Li et al.,2008; Johnson et al., 2009). Recent reports also suggestthat topoIIa may interact with mediator of DNAdamage checkpoint protein 1 (MDC1) to enforce thedecatenation G2 checkpoint (Luo et al., 2009), and thattopoIIa can be SUMOylated to assist in centromerelocalization and proper chromosome segregation(Diaz-Martinez et al., 2006; Dawlaty et al., 2008).However, none of these studies have addressed the roleof the decatenation G2 checkpoint in genomic stability.

As many dietary components are known topoIIcatalytic inhibitors/poisons (Bandele and Osheroff,2007, 2008; Barjesteh van Waalwijk van Doorn-Khos-rovani et al., 2007), it is possible that successive roundsof bypassing the decatenation G2 checkpoint could bean important consideration for the study of tumor

initiation and progression. As TOP2A allelic deletionshave been observed in breast cancer patients, exposureto these dietary compounds may increase genomicinstability in these cancers and accelerate tumorprogression (Beser et al., 2007). A further understandingof the signaling mechanisms involved in the decatena-tion G2 checkpoint could potentially lead to theidentification of early stage biomarkers and newtherapeutic targets for a subset of cancers.

Materials and methods

Experimental design and cell cultureNHDFs were described previously (Simpson et al., 2005). Eachfibroblast line was isolated from a different individual andunderwent o100 population doublings. Experiments wereperformed in all three NHDF lines to assess biologicalreproducibility and inter-individual genetic variation. NHDFswere grown in Dulbecco’s modified Eagle medium/10% fetalbovine serum/2mM L-glutamine and maintained at 5% CO2

and 37 1C. Periodic tests for mycoplasma contamination usinga commercial kit (Gen-Probe, San Diego, CA, USA) werenegative.

In vitro decatenation assay107 fibroblasts were fractionated with a ProtoJET NuclearExtraction Kit (Fermentas, Glen Burnie, MD, USA). Nuclearextracts were assayed for decatenatory activity using theTopoGEN Eukaryotic TopoII Assay kit (TopoGEN, PortOrange, FL, USA). kDNA was electrophoresed in a 1%agarose gel to separate topoisomers, stained with SYBRGold (Invitrogen, Carlsbad, CA, USA), analyzed using aTyphoon 9400 (GE Healthcare, Piscataway, NJ, USA),and quantified by ImageQuant TL version 2005 Software(GE Healthcare).

Time-lapse microscopyThe NHDFs were treated with 0.1% DMSO or 4mM ICRF-193 (Sigma, St Louis, MO, USA) at time 0. Bright-field time-lapse images were captured every 2min at the Michael HookerMicroscopy Facility and the Microscopy Services LaboratoryCore Facility at UNC-CH.

Immunofluorescence and gH2AX flow cytometryThe NHDFs plated on coverslips were treated for 0, 0.25, 2 or6 h with 4–12mM Etoposide, 4mM ICRF-193 or 0.1% DMSO.Coverslips were removed and fixed with 95% ethanol: 5%acetic acid. The remaining cells were analyzed by flowcytometry. Samples were stained with a Phospho-Ser139H2AX antibody (Millipore, Billerica, MA, USA) and ananti-mouse Texas Red secondary antibody (Santa CruzBiotechnology, Santa Cruz, CA, USA). Nuclei were counter-stained with 40,6-diamidino-2-phenylindole. Images wereobtained on a Leica DMIRB inverted fluorescence microscopeat the Michael Hooker Microscopy Facility . Flow cytometrysamples were measured on a Dako CyAN ADP instrument atthe Flow Cytometry Core Facility at UNC-CH and plots wereanalyzed using Summit 4.3 software. Percentages of gH2AXpositive cells were plotted against time and linear regressionslopes were calculated for each treatment group. Statisticalanalysis was performed using a Student’s t-test.

TopoIIa and genomic stabilityJJ Bower et al

4796

Oncogene

Ionizing radiationA gamma cell 40 Cs137 irradiation source was used for ionizingradiation treatment. Mitotic index was determined by the flowcytometry protocol described below 2h after irradiation.

Electroporation of fibroblastsThe NTC (50-UGGUUUACAU GUCGACUAA-30), topoIIa(50-CUGCCUGUUUAGUCGCUUUC A-30), and topoIIb(50-GAAGUUGUCUGUUGAGAGAUU-30) siRNA wereobtained from Dharmacon (Dharmacon, Lafayette, CO,USA). A second topoIIa siRNA was obtained from Invitrogen(50-GAGAUGUCACUAAUGAUGACCAUUA-30). TheNHDFs were electroporated using an NHDF NucleofectorKit (Lonza, Basel, Switzerland), and assayed for biologicaleffects 48 h post-electroporation. Protein depletion for eachindividual experiment was confirmed by western immunoblotat time of assay.

Western immunoblotStandard immunoblotting techniques were used to detectexpression of topoIIa and topoIIb. Briefly, equal amounts oftotal cellular protein were electrophoresed on 7% SDS–PAGEgels, transferred to nitrocellulose membranes and blocked in4% bovine serum albumin/tris-buffered saline with tween-20 .TopoIIa and topoII b antibodies were obtained from BDBiosciences (San Jose, CA, USA). Secondary anti-mouse Cy3/Cy5 antibodies were obtained from GE Healthcare andfluorescence was measured on a Typhoon 9400. Quantificationof protein expression was performed using ImageQuant TL.

Flow cytometric mitotic entry assayElectroporated fibroblasts were plated at a density of 106 cells(day 0), and fed with fresh medium on day 1. On day 2, cellswere treated with 100 ng/ml colcemid and either 0.1% DMSOor 4mM ICRF-193 for 2, 4 and 6 h. Fibroblasts were fixed with95% ethanol: 5% acetic acid. A phospho-Ser10 histone H3primary antibody (Millipore) and a fluorescein isothiocyanate-labeled secondary (Santa Cruz Biotechnology) were used toidentify mitotic cells. Propidium iodide was used to measureDNA content. Summit 4.3 plots were used to quantify thepercentage of NHDFs with 4N DNA content and phospho-Ser10 histone H3. The percentage of mitotic cells for eachsample was plotted against time and the resulting slope of theline was used to measure the MER (% entering mitosis/hour).

Flow cytometry bromodeoxyuridine incorporationFibroblasts were incubated with 10 mM bromodeoxyuridine for2 h and fixed with 70% ethanol. An anti-bromodeoxyuridine-fluorescein isothiocyanate antibody (BD Biosciences) was usedto identify cells with actively replicating DNA. Propidiumiodide was used to measure DNA content. Summit 4.3 plotswere used to quantify the percentage of fibroblasts in G0/G1, Sand G2/M phase.

Cytogenetic studiesFibroblasts were incubated with 100 ng/ml colcemid for 1 h,incubated in hypotonic KCl (75mM), and fixed with 3:1methanol:acetic acid. Nuclei were dropped onto microscope

slides and Geimsa-stained for visualization (Kaufmann et al.,1997). Metaphases were examined using an Olympus BH2microscope with a SPOT RT camera (Olympus, Center Valley,PA, USA). 50þ metaphases were scored for each treatmentgroup/cell line. All analyses were performed blinded to sampleidentity.

Nuclear abnormality assessmentThe NHF1-hTERT cells were electroporated with siRNA asdescribed above and allowed to adhere to coverslips. After48 h, the coverslips were fixed with 3:1 methanol:acetic acidand stained with 40,6-diamidino-2-phenylindole. Coverslipswere mounted onto slides and 1000þ interphase nuclei wereanalyzed in a blind manner for structural abnormalities (blebs,bridges, and micronuclei) at the Michael Hooker MicroscopyFacility .

Fluorescence in situ hybridization (FISH)A second set of cytogenetic metaphase preparations wereartificially aged in 2� saline-sodium citrate /0.1% NonidetP-40 (NP-40) for 30min at 37 1C, successively dehydrated in70, 85 and 100% ethanol, and air- dried. A Vysis LSI p16(9p21) Spectrum Orange/CEP9 Spectrum Green probe(Abbott Molecular, Des Plaines, IL, USA) was hybridizedto metaphase preparations overnight using a Vysis HyBrite(Abbott Diagnostics, Abbott Park, IL, USA) according to theprobe manufacturer’s instructions. Gene copy numbers wereassessed in a blind manner for both the p16INK4A locus and theCEP9 locus for at least 100 interphase nuclei per sample.

Clonogenic survival assaySingle cell suspensions of fibroblasts were electroporated withNTC, topoIIa- or topoIIb-directed siRNA as described. Onethousand cells were plated in triplicate. Colonies were fedevery 2–3 days with fresh medium and stained for14 days afterelectroporation with 0.05% crystal violet/40% methanol.Colonies with 450 cells were counted (Zhou et al., 2006).

Abbreviations

topoII, topoisomerase II; DSB, double-strand break; NHDF,normal human diploid fibroblast; NTC, non-targeting control;MER, mitotic entry rate.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgements

Financial Support: This work was supported by an NIH-NCIRuth L. Kirschstein National Research Service Award Grant# 1F32CA134155-01 (JJ Bower) and PHS grants CA81343 andP30-ES10126 (WK Kaufmann).

References

Akimitsu N, Adachi N, Hirai H, Hossain MS, Hamamoto H,Kobayashi M et al. (2003). Enforced cytokinesis without completenuclear division in embryonic cells depleting the activity of DNAtopoisomerase IIalpha. Genes Cells 8: 393–402.

Austin CA, Marsh KL. (1998). Eukaryotic DNA topoisomerase IIbeta. Bioessays 20: 215–226.

Bandele OJ, Osheroff N. (2007). Bioflavanoids as poisons of humantopoisomerase IIa and IIb. Biochemistry 46: 6097–6108.

TopoIIa and genomic stabilityJJ Bower et al

4797

Oncogene

Bandele OJ, Osheroff N. (2008). (-)-epigallocatechin gallate, a majorconstituent of green tea, poisons human type II topoisomerases.Chem Res Toxicol 21: 936–943.

Barjesteh van Waalwijk van Doorn-Khosrovani S, Janssen J, MaasLM, Godschalk RWL, Nijhuis JG, van Schooten FJ. (2007).Dietary flavonoids induce MLL translocations in primary humanCD34+ cells. Carcinogenesis 28: 1703–1709.

Baxter J, Diffley JFX. (2008). Topoisomerase II inactivation preventsthe completion of DNA replication in budding yeast. Mol Cell 30:790–802.

Beser AR, Tuzlali S, Guzey D, Dolec Guler S, Hacihanefioglu S, DalayN. (2007). Her-2, TOP2A, and chromosome 17 alterations in breastcancer. Pathol Oncol Res 13: 180–185.

Biersack H, Jensen S, Gromova I, Nielsen IS, Westergaard O,Andersen AH. (1996). Active heterodimers are formed from humanDNA topoisomerase IIa and IIb isoforms. Proc Natl Acad Sci USA

93: 8288–8293.Bower JJ, Zhou Y, Zhou T, Simpson DA, Arlander SJ, Paules RS

et al. (2010). Revised genetic requirements for the decatenation G2

checkpoint: the role of ATM. Cell Cycle 9: 1617–1628.Burden DA, Osheroff N. (1998). Mechanism of action of eukaryotic

topoisomerase II and drugs targeted to the enzyme. Biochim Biophys

Acta 1400: 139–154.Capranico G, Tinelli S, Austin CA, Fisher ML, Zunino F. (1992).

Different patterns of gene expression of topoisomerase II isoformsin differentiated tissues during murine development. Biochim

Biophys Acta 1132: 43–48.Carpenter AJ, Porter AC. (2004). Construction, characterization, and

complementation of a conditional-lethal DNA topoisomerase IIalpha mutant human cell line. Mol Biol Cell 15: 5700–5711.

Coelho PA, Queiroz-Machado J, Carmo AM, Moutinho-Pereira S,Maiato H, Sunkel CE. (2008). Dual role of topoisomerase II incentromere resolution and aurora B activity. PLoS Biol 6: 1758–1777.

Cortes F, Pastor N, Mateos S, Dominguez I. (2003). Roles of DNAtopoisomerases in chromosome segregation and mitosis. Mutat Res

543: 59–66.Damelin M, Sun YE, Sodja VB, Bestor TH. (2005). Decatenation

checkpoint deficiency in stem and progenitor cells. Cancer Cell 8:479–484.

Dawlaty MM, Malureanu L, Jeganathan KB, Kao E, Sustmann C,Tahk S et al. (2008). Resolution of sister centromeres requiresRanBP2-mediated SUMOylation of topoisomerase IIa. Cell 133:103–115.

Deming PB, Cistulli CA, Zhao H, Graves PR, Piwnica-Worms H,Paules RS et al. (2001). The human decatenation checkpoint. Proc

Natl Acad Sci USA 98: 12044–12049.Deming PB, Flores KG, Downes CS, Paules RS, Kaufmann WK.

(2002). ATR enforces the topoisomerase II-dependent G2

checkpoint through inhibition of Plk1 kinase. J Biol Chem 277:36832–36838.

Diaz-Martinez LA, Gimenez-Abian JF, Azuma Y, Guacci V,Gimenez-Martin G, Lanier LM et al. (2006). PIASgamma isrequired for faithful chromosome segregation in human cells. PLoS

ONE 1: e53.Doherty SC, McKeown SR, McKelvey-Martin V, Downes CS, Atala

A, Yoo JJ et al. (2003). Cell cycle checkpoint function in bladdercancer. J Natl Cancer Inst 95: 1859–1868.

Downes CS, Clarke DJ, Mullinger AM, Gimenez-Abian JF, CreightonAM, Johnson RT. (1994). A topoisomerase II-dependent G2 cyclecheckpoint in mammalian cells (published erratum appears inNature 1994; 372: 710). Nature 372: 467–470.

Emmons M, Boulware D, Sullivan DM, Hazlehurst LA. (2006).Topoisomerase II beta levels are a determinant of melphalan-induced DNA crosslinks and sensitivity to cell death. BiochemPharmacol 72: 11–18.

Franchitto A, Oshima J, Pichierri P. (2003). The G2-phase decatena-tion checkpoint is defective in Werner syndrome cells. Cancer Res

63: 3289–3295.Goswami PC, Roti Roti JL, Hunt CR. (1996). The cell cycle-coupled

expression of topoisomerase IIa during S phase is regulated by

mRNA stability and is disrupted by heat shock or ionizingradiation. Mol Cell Biol 16: 1500–1508.

Gromova I, Biersack H, Jensen S, Nielsen OF, Westergaard O,Andersen AH. (1998). Characterization of DNA topoisomerase IIa/b heterodimers in HeLa cells. Biochemistry 37: 16645–16652.

Heck MMS, Hittelman WN, Earnshaw WC. (1988). Differentialexpression of DNA topoisomerases I and II during the eukaryoticcell cycle. Proc Natl Acad Sci USA 85: 1086–1090.

Holland AJ, Cleveland DW. (2009). Boveri revisited: chromosomalinstability, aneuploidy, and tumorigenesis. Nat Rev Mol Cell Biol

10: 478–487.Jarvinen TAH, Kononen J, Pelto-Huikko M, Isola J. (1996).

Expression of topoisomerase IIa is associated with rapid cellproliferation, aneuploidy, and c-erbB2 overexpression in breastcancer. Am J Pathol 148: 2073–2082.

Jensen LH, Nitiss KC, Rose A, Dong J, Zhou J, Hu T et al. (2000). Anovel mechanism of cell killing by anti-topoisomerase II bisdiox-opiperazines. J Biol Chem 275: 2137–2146.

Johnson M, Phua HH, Bennett SC, Spence JM, Farr CJ. (2009).Studying vertebrate topoisomerase 2 function using a conditionalknockdown system in DT40 cells. Nuc Acid Res 37: e98.

Ju BG, Lunyak VV, Perissi V, Garcia-Bassets I, Rose DW, Glass CKet al. (2006). A topoisomerase IIb-mediated dsDNA break requiredfor regulated transcription. Science 312: 1798–1802.

Kaufmann WK, Schwartz JL, Hurt JC, Byrd LL, Galloway DA,Levedakou E et al. (1997). Inactivation of G2 checkpoint functionand chromosomal destabilization are linked in human fibroblastsexpressing human papillomavirus type 16 E6. Cell Growth Differ 8:1105–1114.

Larsen AK, Skladanowski A, Bojanowski K. (1996). The roles ofDNA topoisomerase II during the cell cycle. Prog Cell Cycle Res 2:229–239.

Li H, Wang Y, Liu X. (2008). Plk1-dependent phosphorylationregulates functions of DNA topoisomerase II in cell cycleprogression. J Biol Chem 283: 6209–6221.

Lorusso V, Manzione L, Silvestris N. (2007). Role of liposomalanthracyclines in breast cancer. Ann Oncol 18: vi70–vi73.

Lou Z, Minter-Dykhouse K, Chen J. (2005). BRCA1 participates inDNA decatenation. Nat Struct Mol Biol 12: 589–593.

Luo K, Yuan J, Chen J, Lou Z. (2009). Topoisomerase IIa controls thedecatenation checkpoint. Nat Cell Biol 11: 204–210.

Lyu YL, Kerrigan JE, Lin CP, Azarova AM, Tsai YC, Ban Y et al.(2007). Topoisomerase IIb mediated DNA double-strand breaks:implications in Doxorubicin Cardiotoxicity and prevention byDexrazoxane. Cancer Res 67: 8839–8846.

Maeshima K, Eltsov M, Laemmli UK. (2005). Chromosome structure:improved immunolabeling for electron microscopy. Chromosoma

114: 365–375.Mao Y, Desai SD, Ting CY, Hwang J, Liu LF. (2001). 26 S

proteasome-mediated degradation of topoisomerase II cleavablecomplexes. J Biol Chem 276: 40652–40658.

Nakagawa T, Hayashita Y, Maeno K, Masuda A, Sugito N, Osada Het al. (2004). Identification of decatenation G2 checkpoint impair-ment independently of DNA damage G2 checkpoint in human lungcancer cell lines. Cancer Res 64: 4826–4832.

Park I, Avraham HK. (2006). Cell cycle-dependent DNA damagesignaling induced by ICRF-193 involves ATM, ATR, CHK2, andBRCA1. Exp Cell Res 312: 1996–2008.

Robinson HM, Bratlie-Thoresen S, Brown R, Gillespie DA. (2007).Chk1 is required for G2/M checkpoint response induced bythe catalytic topoisomerase II inhibitor ICRF-193. Cell Cycle 6:1265–1267.

Roca J, Ishida R, Berger JM, Andoh T, Wang JC. (1994). Antitumorbisdioxopiperazines inhibit yeast DNA topoisomerase II by trap-ping the enzyme in the form of a closed protein clamp. Proc Natl

Acad Sci USA 91: 1781–1785.Santamaria A, Neef R, Eberspacher U, Eis K, Husemann M,

Mumberg D et al. (2007). Use of the novel Plk1 inhibitor ZK-Thiazolidinone to elucidate functions of Plk1 in early and late stagesof mitosis. Mol Biol Cell 18: 4024–4036.

TopoIIa and genomic stabilityJJ Bower et al

4798

Oncogene

Sehested M, Jensen PB. (1996). Mapping of DNA topoisomerase IIpoisons (etoposide, clerocidin) and catalytic inhibitors (aclarubicin,ICRF-189) to four distinct steps in the topoisomerase II catalyticcycle. Biochem Pharm 51: 879–886.

Simpson DA, Livanos E, Heffernan TP, Kaufmann WK. (2005).Telomerase exression is sufficient for chromosomal integrity in cellslacking p53-dependent G1 checkpoint function. J Carcinog 4: 18.

Wang JC. (2002). Cellular roles of DNA topoisomerases: a molecularperspective. Nat Rev Mol Cell Biol 3: 430–440.

Wang LHC, Schwarzbraun T, Speicher MR, Nigg EA. (2008).Persistence of DNA threads in human anaphase cells suggests

late completion of sister chromatid decatenation. Chromosoma 117:123–135.

Xiao H, Mao Y, Desai SD, Zhou N, Ting CY, Hwang J et al. (2003).The topoisomerase II beta circular clamp arrests transcription andsignals a 26S proteasome pathway. Proc Natl Acad Sci USA 100:3239–3244.

Zhou T, Chou JW, Simpson DA, Zhou Y, Mullen TE, Medeiro Met al. (2006). Profiles of global gene expression in ionizing-radiation-damaged human diploid fibroblasts reveal synchronization behindthe G1 checkpoint in a G0-like state of quiescence. Environ Health

Persp 114: 553–559.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

TopoIIa and genomic stabilityJJ Bower et al

4799

Oncogene