DNA Topoisomerase III Alpha Regulates P53-Mediated Tumor Suppression

Published OnlineFirst February 13, 2014; DOI: 10.1158/1078-0432.CCR-13-1997

Clinical Cancer Human Cancer Biology Research

DNA Topoisomerase III Alpha Regulates p53-Mediated Tumor Suppression

Mei-Yi Hsieh1, Jia-Rong Fan1, Han-Wen Chang1, Hsiang-Chin Chen1, Tang-Long Shen2,3, Shu-Chun Teng1, Yen-Hsiu Yeh1, and Tsai-Kun Li1,3

Abstract Purpose: Human DNA topoisomerase III alpha (hTOP3a) is involved in DNA repair surveillance and cell-cycle checkpoints possibly through formatting complex with tumor suppressors. However, its role in cancer development remained unsolved. Experimental Design: Coimmunoprecipitation, sucrose gradient, chromatin immunoprecipitation (ChIP), real time PCR, and immunoblotting analyses were performed to determine interactions of hTOP3a with p53. Paired cell lines with different hTOP3a levels were generated via ectopic expression and short hairpin RNA (shRNA)-mediated knockdown approaches. Cellular tumorigenic properties were analyzed using cell counting, colony formation, senescence, soft agar assays, and mouse xenograft models. Results: The hTOP3a isozyme binds to p53 and cofractionizes with p53 in gradients differing from fractions containing hTOP3a and BLM. Knockdown of hTOP3a expression (sh-hTOP3a) caused a higher anchorage-independent growth of nontumorigenic RHEK-1 cells. Similarly, sh-hTOP3a and ectopic expression of hTOP3a in cancer cell lines caused increased and reduced tumorigenic abilities, respectively. Genetic and mutation experiments revealed that functional hTOP3a, p53, and p21 are required for this tumor-suppressive activity. Mechanism-wise, ChIP data revealed that hTOP3a binds to the p53 and p21 promoters and positively regulates their expression. Two proteins affect promoter recruitments of each other and collaborate in p21 expression. Moreover, sh-hTOP3a and sh-p53 in AGS cells caused a similar reduction in senescence and hTOP3a mRNA levels were lower in gastric and renal tumor samples. Conclusion: We concluded that hTOP3a interacts with p53, regulates p53 and p21 expression, and contributes to the p53-mediated tumor suppression. Clin Cancer Res; 20(6); 1489–501. 2014 AACR.

Introduction have evolved to prohibit neoplastic transformation (3), and It is generally believed that cancer results when cells thus the function and/or gene expression of the regulatory acquire genetic alterations that cause uncontrolled prolif- factors of senescence and apoptosis safeguards must be eration. Consequently, genetic errors resulting in loss-of- altered in a normal cell to become cancerous. TP53 function of tumor-suppressor genes and/or gain-of-func- , the most frequently mutated gene in cancers tion of oncogenes are critically involved in cancer develop- ( 50%), encodes tumor suppressor p53, which functions ment (1). Supportively, inherited diseases with genomic as a transcription factor that drives the senescent (via p21), instability, such as Bloom syndrome, are prone to cancer apoptotic (via BAX), and DNA repair (via GADD45) path- development (2). In addition, genomic instability is recog- ways (4). In addition to genetic alterations, other p53 nized recently as an enabling characteristic of cancer. Two inactivation mechanisms via BRD7 (bromodomain-con- safeguard mechanisms, namely senescence and apoptosis, taining 7) dysfunction or elevated microRNA expression have been reported in cancers (5–7). Chromatin modifying and remodeling complexes have also shown to regulate Authors' Affiliations: 1Department and Graduate Institute of Microbiology, p53-dependent p21 expression (8, 9). Notably, recent College of Medicine, 2Department of Plan Pathology and Microbiology, reports on cancer genomic analyses revealed the potential College of Bioresources and Agriculture, and 3Center for Biotechnology, National Taiwan University, Taipei, Taiwan contribution of chromatin factors to tumorigenesis (10). Together, these results suggest that cellular factors could Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/). contribute to tumorigenesis via directly modulating p53 expression and/or functionally cooperating with p53-reg- Corresponding Author: Tsai-Kun Li, Department and Graduate Institute of Microbiology, College of Medicine, National Taiwan University, Room 711, ulated gene expression. Section 1, Jen-Ai Road, Taipei 10051, Taiwan. Phone: 886-2231-23456, Two epigenetic mechanisms, chromatin modification and ext. 88287/88294; Fax: 886-2239-15293; E-mail: [email protected] remodeling, are known to influence both chromatin struc- doi: 10.1158/1078-0432.CCR-13-1997 ture and gene expression. Nevertheless, functional roles of 2014 American Association for Cancer Research. the other epigenetic regulators of chromatin DNA topology

www.aacrjournals.org 1489

Hsieh et al.

hTOP3a caused S-phase impairment and rescued some Translational Relevance phenotypes of Ataxia telangiectasia cells (29). It is unclear Transcription factor p53 plays a central role in tumor how TOP3s carries out above specialized essential func- suppression and stress response, thus its expression and tions. Two explanations are offered: (i) TOP3 is present in function must be tightly regulated. Recent studies have multiprotein complexes that perform above functions; or also highlighted the collaborative importance of chro- (ii) both the enzymatic activities and properties of TOP3 are matin and transcription factors in regulated gene expres- greatly modulated by its interacting proteins. Together, sion. Here, we demonstrated that hTOP3a interacts through physical interaction with specific DNA-binding physically and functions genetically with p53 and thus factors, chromatin regulators such as TOP3a participate in contributing to p53/p21-mediated tumor suppression. the specialized biologic functions. Regarding the tumori- Our observations of reduced hTOP3a expression in genic involvement, TOP3s have been shown to exist in tumor samples and its contribution to p53 expression different complexes whose component proteins function and functions further suggest a novel notion that down- in genome stability, checkpoints, and possibly cancer devel- regulation of hTOP3a level is a new way of creating p53 opment (16, 30, 31). Our group also reported that TOP3 dysfunction during tumorigenesis. Because p53 and functions in the alternative lengthening of telomeres via the DNA topoisomerases are clinically useful targets for recombination pathway and hTOP3a knockdown (sh- cancer intervention, our mechanism-based study might hTOP3a) using shRNA technology caused an elevated lead to important therapeutic strategies. telomerase expression (32). Our result resembles telomerase reactivation during cancer progression thus consistent with a potential role for hTOP3a in tumorigenesis. Here, we report a novel role for hTOP3a in tumorigen- in transcription programming remain to be elucidated. esis and describe its mechanism of action via physically Through their enzymatic activity, DNA topoisomerases reg- and genetically interacting with p53 to regulate p53/p21- ulate various biological functions (11–13), and torsional mediated transcription program. Our results demonstrat- stress accompanied with the separation of DNA strands ed that hTOP3a binds to p53 through p530sDNA-binding during transcription can only be relieved by DNA topoi- region. Similar to the phenotypes caused by p53 deficien- somerases. Recent studies revealed that the TOP2b isozyme cy, sh-hTOP3a not only promoted anchorage-indepen- is involved in regulated gene expression through specifically dent growth but also reduced b-galactosidase (b-gal) binding to transcription factors, for example, human TOP2b senescence staining of cells. Through modulating hTOP3a (hTOP2b) binds to the androgen receptor (AR) and subse- expression, we revealed that the cellular levels of hTOP3a quently contributing to AR-directed transcription program- are inversely correlated to tumorigenic growth in vitro ming and to DNA sequence rearrangements during the and in vivo. Molecularly, hTOP3a regulates tumorigenic development of prostate cancer (14). Consistently, our growth in a p53- and p21-dependent manner. Moreover, recent study has demonstrated a specific involvement of the hTOP3a binds to both the p53 and p21 promoter regions nitric oxide-activated TOP2b cleavable complexes in the and positively regulates their expression. Together, our inflammation-associated DNA damage, mutagenesis, and results suggest that hTOP3a acts through and/or together skin melanoma formation (15). Interestingly, hTOP3a plays with p53 as an anticancer block and subsequently pro- roles in maintaining genomic stability by forming DNA vides the first example of how a general chromatin enzyme repair surveillance complexes with BLM, FANCD2, and collaborates with a DNA-binding transcription factor to BRCA1 tumor-suppressor proteins (16–18) and/or acting act on specific transcription programming and subsequent in combination with BLM, RMI1, and RMI2 to promote the cellular functions. dissolution of double Holliday junctions (19, 20). Consid- ering the importance of genomic instability in tumorigenesis Materials and Methods and the specific interactions of hTOP3a with tumor sup- Chemicals, plasmids, and antibodies pressors, above reports suggest that hTOP3a might function All chemicals and pFLAG-CMV2 plasmid were purchased as a tumor suppressor. from Sigma unless otherwise indicated. Plasmid pFLAG- Six topoisomerases have been identified in human cells, hTOP3a was constructed by cloning the hTOP3a PCR namely nuclear hTOP1, 2a,2b,3a,3b, and mitochondrial fragment with pFLAG-CMV2 vector. Antibodies against mTOP1 (12, 13). These enzymes change the topology of p53, p5315p, p21, BLM (cell signaling), GAPDH (Biode- DNA using transesterification reaction involving the cleav- sign), PCNA, HP-1g (Abnova), hTOP3a (Santa Cruz Bio- age and religation of phosphodiester bonds. Although some technology), and histone H3 K9 tri-methylation (Upstate) functions of TOP1 and TOP2 are well studied (11–13), the were obtained commercially. biologic roles of TOP3 remain largely unresolved. Never- theless, embryonic lethality occurs in mice lacking TOP3a Cell culture and cell proliferation assay (21), and TOP3b knockout mice have a reduced lifespan HCT116, RHEK-1, and H1299 cells were incubated in a (22, 23). In addition, unexpected roles in checkpoints and 37 C incubator under 5% CO2. The culture medium con- chromosome segregation were suggested for yeast TOP3s sisted of Dulbecco’s modified Eagle medium (DMEM) (24–28). The expression of dominant-negative or truncated supplemented with 10% FBS, 2 mmol/L L-glutamine, 100

1490 Clin Cancer Res; 20(6) March 15, 2014 Clinical Cancer Research

TOP3a and p53 Cooperate in Tumor Suppression

U/mL penicillin, and 100 mg/mL streptomycin, except for to the transcriptional start site were used. For the p21 the AGS cells (RPMI media). Plasmid transfection was promoter, primers specific for the two 50 p53-binding sites achieved using FuGENE reagent (Promega) as per the man- (2313 to 2212 and 1452 to 1310) and an NS site ufacturer’s instruction. After 2-week selection, resistant col- (control, 4443 to 4199) were used. onies were picked and transferred to 24-well plates. The hTOP3a levels of these clones were confirmed by immu- Tumorigenesis assays: anchorage-independent noblotting analyses. For the cell proliferation assay, 105 growth on soft agar and tumor formation in NOD/SCID cells were seeded in 6 cm dishes for 4 days in triplicate at mice each time point. At each 24-hour interval, the cellular For the soft agar assay, cells (5 or 10 103) were mixed proliferation was determined using an MTT assay. with 2 mL of 0.3% Bacto-agar (Difco)-containing DMEM and then overlaid onto a 2 mL layer of precoated 0.6% Lentivirus-based RNA interference Bacto-agar DMEM in 6-well plates. After incubation for 3 Helper plasmids and lentiviral vectors expressing RNA weeks, the colonies were stained using 0.005% crystal violet interference (RNAi) sequences specifically targeting sh-Luc, and the number of colonies was scored. sh-hTOP3a, sh-p53, and sh-BLM were obtained from the The Animal Use Protocol was designed in accordance National RNAi Core Facility. Nonreplicative viral particles with the guidelines of and approved by the Institutional were prepared as recommended (http://rnai.genmed. Animal Care and Use Committee (IACUC No. 20060072). sinica.edu.tw). The targeted sequences were as follows: For the tumorigenic xenograft model, 4- to 5-week-old nonobese diabetic/severe combined immunodeficient 50-CGAGTTTATTGTTCGCCATTT-30 (sh1-hTOP3a) (NOD/SCID) mice were supplied by the Animal Center 50-GCTTCTCGAAAGTTGAGAATA-30 (sh2-hTOP3a) at the National Taiwan University College of Medicine and 0 0 maintained in the SPF facility (5 mice/cage). A total of 5 5 -CGGCGCACAGAGGAAGAGAAT-3 (sh-p53) 6 0 0 10 cells resuspended in serum-free DMEM were subcuta- 5 -GCTACATATCTGACAGGTGAT-3 (sh-BLM) neously injected into the dorsal regions of the NOD/SCID 50-CCTAAGGTTAAGTCGCCCTCG-30 (sh-luciferase; sh-Luc) mice. The injected mice were examined every 3 to 4 days for tumor appearance. The tumor volumes were estimated Coimmunoprecipitation, immunoblotting, and real- (V ¼ 0.4ab2) from the length (a) and the width (b) time RT-PCR analyses measurements obtained using calipers. Some mouse Coimmunoprecipitation (co-IP) and immunoblotting tumorswereexcisedandmincedinto2-to3-mmpieces experiments were performed as previously described at 37 C in DMEM medium, and the tumor cells were (33). Cellular lysates for immunoblotting were prepared further cultivated. using the radioimmunoprecipitation assay buffer (0.5% C24H39NaO4, 1% NP40, and 0.1% SDS) and 1 Roche Statistical analyses protease inhibitor cocktail. For co-IP, lysates were prepared Statistical analyses were performed using the simple with lysis buffer (20 mmol/L Tris, pH 7.5, 100 mmol/L Student t test. The data were considered to be significant P NaCl, 1% NP40, 100 mmol/L Na3VO4, 50 mmol/L NaF, and if the value was less than 0.01. 30 mmol/L Na4P2O7) and 1 EDTA-free protease inhibitor cocktail. The proteins in the extracts were immunoprecipi- Results tated with specific antibodies, separated by SDS-PAGE, The physical interaction between hTOP3a and p53 transferred onto polyvinylidenedifluoride membranes, Immunoprecipitation experiments were performed to blotted with indicated antibodies, and then detected using analyze the putative interaction of hTOP3a with p53. the enhanced chemiluminescence procedure. In some Using an anti-p53 antibody, we have shown that hTOP3a experiments, cell extracts were treated with DNase I before could be coimmunoprecipitated with p53 and antibodies being subjected to immunoprecipitation using p53 anti- against hTOP3a also precipitated both proteins (Fig. 1A). bodies. Total cellular RNA was extracted using TRIzol Notably, p53 antibodies precipitated both proteins from reagent according to the manufacturer’s protocol (Invitro- the DNase I-treated cell lysates (Fig. 1B). Furthermore, we gen) and cDNA was synthesized using random primers confirmed this interaction using an in vitro pull-down (N6) and SuperScript III reverse transcriptase (Invitrogen). assay with recombinant hTOP3a and p53 proteins (Fig. All PCR reactions were performed with a Bio-Rad real-time 1C). Using a sucrose gradient assay, we found that PCR detection system using DNA-binding SYBR Green dye hTOP3a and p53 proteins coexist in the same fractions for detection of the PCR products. that are different from gradients containing hTOP3a and BLM helicase (Fig. 1D). Immunoprecipitation experi- Chromatin immunoprecipitation assay ments were further performed on cell lysates containing Chromatin immunoprecipitation (ChIP) analysis was various HA-tagged deletion mutants of p53, allowing us performed in HCT116 cells as described in the manufac- to determine the hTOP3a-interacting region within the turer’s instructions (Millipore). For the p53 promoter, the DNA-binding domain of p53 (residues 100-325; Fig. 1E). primers for specific PCR products containing sequences Therefore, our results demonstrated that hTOP3a inter- 1206 to 1082, 320 to 208, and 171 to 90 relative acts with p53 to potentially form a complex.

www.aacrjournals.org Clin Cancer Res; 20(6) March 15, 2014 1491

Hsieh et al.

Figure 1. hTOP3a interacts physically with p53. A, co-IP experiments using p53 or hTOP3a antibodies revealed a physical interaction between hTOP3a and p53. Protein lysates of HCT116 cells expressing either the vector control or hTOP3aOE were immunoprecipitated using the control IgG or the indicated antibodies. B, the binding between hTOP3a and p53 is DNA independent. The co-IP assay with control IgG or anti-p53 antibodies was performed using lysates with (right) or without (left) DNase I pretreatment. C, the GST pull-down assay revealed that recombinant hTOP3a and p53 proteins bind to each other. Bacterial expressed GST-tagged p53 and His-tagged hTOP3a proteins were puriﬁed and a pull-down assay of GST-p53 with glutathione beads was performed as described. The immunoprecipitates and pull-down lysates were immunoblotted (IB) using the indicated antibodies. D, cosedimentation of hTOP3a and p53. Protein lysates from HCT116 hTOP3aOE cells were sedimented using a sucrose gradient. (Continued on the following page.)

1492 Clin Cancer Res; 20(6) March 15, 2014 Clinical Cancer Research

TOP3a and p53 Cooperate in Tumor Suppression

Knockdown of hTOP3a expression promotes tumorigenic growth of HeLa sh-hTOP3a cells in soft agar tumorigenic growth and reduces p53 and p21 (data not shown). Our results thus suggest a novel role for expression hTOP3a in tumor suppression possibly via the specific To investigate the impact(s) of hTOP3a deficiency on interactions with p53. cellular growth and tumorigenic propensities, lentivirus- mediated knockdown (sh1- and sh2-hTOP3a)invarious The collaborative action of and genetic interaction cell lines has been performed. Two targeting sequences could between hTOP3a and p53 on p21 expression and sufficiently, yet differentially, knockdown hTOP3a expres- tumorigenesis sion in RHEK-1 nontumorigenic, HCT116 colorectal, and We investigated the potential mechanism(s) of action AGS gastric cancer cells (Fig. 2A–C) with minimal effects on underlying hTOP3a-mediated tumor suppression. Two cell growth (Fig. 2D–F). We noticed that the expression levels isozymes, TOP3a and TOP3b, were identified in mam- / of p53, p21, and p5315P (phosphorylation at Ser15) were malian cells and top3b mouse embryonic fibroblasts reduced in sh-hTOP3a cells (Fig. 2A–C), suggesting another (MEF) have been reported to be defective in p53 engage- potential interaction between hTOP3a and p53. ment (34). Above results uncovered physical and func- Interestingly, both sh1- and sh2-hTOP3a rendered tional interactions between hTOP3a and p53, prompting RHEK-1, an immortalized human epidermal keratinocyte, us to study the possible genetic link between hTOP3a and þ þ to grow in soft agar (Fig. 2G), an indicative of neoplastic the p53-p21 pathway. First, we generated HCT116 p53 / transformation. Moreover, sh-hTOP3a also promoted and p53 / sh-hTOP3a clones with similar low hTOP3a anchorage-independent growth of HCT116 and AGS cancer levels (Fig. 4A) and found no enhanced tumorigenic cells, in which tumor growth–promoting abilities of growth in the sh-hTOP3a–expressing HCT116 p53 / hTOP3a deficiency were proportional to sh-hTOP3a cells (Fig. 4B–D). knockdown efficiencies (Fig. 2H and I). The above results Second, hTOP3a was ectopically expressed in HCT116 suggest that hTOP3a might play roles in neoplastic trans- wild-type, p53 / or p21 / cells (Fig. 5A, denoted as formation. Notably, when expression levels of hTOP3a and hTOP3aOE) and our data revealed that hTOP3aOE reduced p53 were simultaneously knocked down (Fig. 2J), the tumorigenic growth in HCT116 cells but not in HCT116 þ þ þ þ double knockdown cells conferred the highest tumorigenic p53 / p21 / or p53 / p21 / cells (Fig. 5B), suggesting growth and the lowest p53 activity when compared with the that hTOP3a requires functional p53 and p21 to reduce single knockdown cells (Fig. 2K). Together, our data dem- cellular tumorigenic growth. Nevertheless, ectopic hTOP3b onstrated a functional interaction between hTOP3a and expression did not affect tumorigenic propensity of cells p53. These two proteins function cooperatively to induce (data not shown). Finally, we restored the protein levels of p21 expression and to act in tumor suppression. p53 by expressing p53 or the p53R248W DNA-binding mutant in HCT116 p53 / cells (Fig. 5C). Reintroduction In vitro and in vivo tumorigenic growth abilities of cells of functional p53 not only restored both p5315p and p21 are inversely related to the cellular level of functional expression but also revealed the cooperative action between hTOP3a hTOP3a and p53 to induce p21 expression (compare lane 3 To further support the above notion that hTOP3a to lane 4, Fig. 5C) and the suppression of tumorigenic OE functions as a tumor suppressor, we generated HCT116 growth (Fig. 5D), while p53R248W could not restore hTOP3aOE clones that ectopically express hTOP3a (Fig. p21 expression or tumor-suppressive activity. 3A). We showed that cells with higher hTOP3a expres- Next, we performed cell-cycle analyses on sh-hTOP3a sion always displayed reduced tumorigenic growth abil- and hTOP3aOE cells. Although we found that sh-hTOP3a ities in the soft agar assay and in a mouse xenograft model causes hypersensitivity to cell-cycle–disrupting agents (e.g., (Fig. 3B and C). Moreover, tumor samples were extracted hydroxyurea), both knockdown and overexpression of from mice injected with different cell lines and estab- hTOP3a only minimally affect the cell-cycle progression lished as paired vector control and hTOP3aOE (No. 2 and (data not shown). Moreover, our results also revealed that 3) cell lines. We showed that hTOP3aOE tumor cell lines hTOP3a contributes to the DNA damage-induced activa- exhibit higher hTOP3a levels than paired vector control tion of p53-mediated p21 expression, but only minimally cell lines (Fig. 3D) and sh-hTOP3a knockdown in the No. affect ATM and g-H2AX activation (Supplementary Figs. S1 2 tumor-derived cell line resulted in restoration of its and S2). As shown in the previous reports (35), ATM anchorage-independent growth ability (Fig. 3E and F). inhibitors caffeine and wortmannin effectively reduced Consistently, the complementation with the functional phosphorylation of p5315P, ATM1981P, and g-H2AX and expression of a shRNA-resistant hTOP3a reduced the p21 expression (Supplementary Fig. S2B). Thus, our results

(Continued.) Twenty fractions from the top of the gradient were collected and p53, hTOP3a, and BLM protein level in different fractions (#1–14) were determined by immunoblotting. E, the hTOP3a-interacting motif of p53 was mapped to p530s DNA-binding domain (DBD). Top, schematic diagram of p53 and p53 deletion mutants and the binding of each p53 fragment with hTOP3a (right); TAD, transactivation domain; OD, oligomerization domain; CRD, C-terminal regulatory domain; þ, with interaction; , without interaction. Bottom, HCT116 p53 / cells were cotransfected with hTOP3a- and different HA-tagged p53 deletion mutant-expressing (or an empty vector) plasmids. After transfection, the cellular lysates were then coimmunoprecipitated using anti-hTOP3a or anti-goat IgG antibodies.

www.aacrjournals.org Clin Cancer Res; 20(6) March 15, 2014 1493

Hsieh et al.

Figure 2. Knockdown of hTOP3a expression in vitro promotes tumorigenic growth without affecting cellular proliferation. A to C, expression levels of hTOP3a, p53, p5315P, and p21 in nontransformed RHEK-1 cells (A), HCT116 colorectal (B), and AGS gastric cancer cell lines (C) with knockdown of hTOP3a (sh- hTOP3a) or p53 expression (sh- p53). The lentivirus-mediated shRNA approach was employed to knockdown expression of hTOP3a and p53. Pooled clones from puromycin-resistant cells expressing the indicated shRNAs (e.g., sh-Luc for the luciferase- knockdown control and sh1- hTOP3a for the knockdown by hTOP3a shRNA#1) were subjected to different assays. D to F, neither sh-p53 nor sh-hTOP3a affected cell growth. The MTT assay was used to examine cellular proliferation (n ¼ 3). G to I, similar to sh-p53, sh-hTOP3a expression resulted in enhanced anchorage- independent growth in soft agar. J, both sh-hTOP3a and sh-p53 effectively reduced p21 expression. K, an additional sh-hTOP3a further enhanced tumorigenic growth in HCT116 sh-p53 cells. Pooled clones of sh-Luc, sh-hTOP3a, sh-p53, and the sh-hTOP3 sh-p53 double- knockdown cells were assayed by immunoblotting and the soft agar assay. In the soft agar assay, 1 104 cells were seeded per well and the colonies were counted after 3 weeks. All graphs display the means with SD for 3 to 6 independent experiments (n ¼ 3for G and I; n ¼ 6 for H; n ¼ 4 for K). GAPDH was used as the internal loading control. The asterisk indicates a signiﬁcant difference (, P < 0.001).

suggested that hTOP3a might play a role in p53-mediated The enzymatic activity of hTOP3a is required for tumor suppression and these two proteins might regulate its regulatory roles on tumorigenesis and p53 p21 expression cooperatively. Consistently, we also found expression that hTOP3a could not suppress anchorage-independent DNA topoisomerases required a tyrosine residue in growth of p53-deﬁcient H1299 cells in soft agar (Fig. 5E). their active site for their enzymatic activity (12, 13) and

1494 Clin Cancer Res; 20(6) March 15, 2014 Clinical Cancer Research

TOP3a and p53 Cooperate in Tumor Suppression

Figure 3. Ectopic expression of hTOP3a reduced tumorigenic growth. A, ectopic expression of hTOP3a (hTOP3aOE) in HCT116 cells caused increased protein levels of hTOP3a, p53, p21, and p5315P. B and C, functional expression of hTOP3a reduced the anchorage-independent growth of cells in soft agar (B, n ¼ 4) and tumor growth in mice (C, n ¼ 6). Cells from one vector control (Vec) and two stable clones (HCT116 hTOP3aOE clone 6 and 7) were subjected to immunoblotting analysis (A), soft agar assay (B), and tumor growth assay using a SCID xenograft model (C). D, the protein levels of hTOP3a in the tumor samples from HCT116 Vec and hTOP3aOE cells. E and F, knockdown of hTOP3a expression in tumor-derived HCT116 hTOP3aOE cell line caused reduced levels of hTOP3a, p5315P, p53, and p21 protein expression (E) and restored tumorigenic growth (F, n ¼ 4). GAPDH was used as the internal loading control. The asterisk indicates a signiﬁcant difference (, P < 0.001).

we generated the catalytically dead hTOP3aY337F mutant because sh-hTOP3a expression caused a reduction in accordingly. Differing from HCT116 hTOP3aOE cells, both the p53 and p21mRNA levels (Fig. 6A). Notably, OE OE hTOP3aY337F expression in cells had no effect on the the levels of p53 and p21 mRNA increased in hTOP3a OE levels of p53, p21, and p5315p, nor was there an effect on cells but not in hTOP3aY337F cells(Fig.6B).AsmRNA tumorigenic growth (Fig. 5F and G). These results sug- levels in cells could also be regulated by mRNA stability, gested that the enzymatic activity of hTOP3a is not only RNA decay analyses were performed and transcription is required for tumorigenic propensity but also for the p53/ inhibited by actinomycin D. As shown, TP53 mRNA p21 expression. amounts declined faster in the sh-hTOP3a cells than in Next, we showed that the hTOP3a-mediated p53 and the sh-Luc cells (Supplementary Fig. S3A; quantitative p21 upregulation occurs at the transcriptional level results in the right) and the half-life of p53 protein was

www.aacrjournals.org Clin Cancer Res; 20(6) March 15, 2014 1495

Hsieh et al.

Figure 4. Enhanced tumorigenic growth by knockdown of hTOP3a expression is p53-dependent. A, successful knockdown using sh- hTOP3a in HCT116 cells. B to D, knockdown of hTOP3a in HCT116 þ þ p53 / cells, but not in HCT116 p53 / cells, enhances anchorage- independent growth on soft agar (B, n ¼ 4) and tumor growth in SCID mice (C, n ¼ 5; D, n ¼ 7). The asterisk indicates a signiﬁcant difference (, P < 0.001).

notably shortened when the hTOP3a expression was Both hTOP3a and p53 deficiency reduce b-gal knocked down (Supplementary Fig. S3B), revealing a senescent staining in AGS cells potential role of hTOP3a in regulation of TP53 mRNA Both senescence and apoptosis play important roles in and protein stability. Our results thus suggested that the tumor-suppressor activity of p53 (5–7). The p53-target- hTOP3a regulates p53 expression not only at transcrip- ing gene p21 encoded for a cell-cycle inhibitor that has been tional level but also at the level of mRNA stability. We suggested to play a role in senescence (37). To better further proposed that hTOP3a regulates p53-mediated understand the tumor-suppressive mechanisms of tumor suppression through the modulation of p53 hTOP3a, we determined the effect of sh-hTOP3a on cel- expression and possibly by acting with p53 to upregulate lular senescence. Unlike HCT116 cells (<5% of senescence p21 expression. cells), AGS cells contain approximately 25% of senescent cells during normal growing conditions (38). Therefore, we The hTOP3a isozyme binds to specific regions of p53 have chosen AGS cells for the following studies and we and p21 promoters observed both sh-p53 and sh-hTOP3a reduced levels of To support the mechanisms by which hTOP3a influences b-gal senescent staining in AGS cells (Supplementary Fig. p53 and p21 transcription, ChIP experiments were S4A). Further quantitative analyses of the levels of b-gal performed. Figure 6C showed that hTOP3a was recruited staining and tumorigenic growth exhibited by AGS sh-Luc specifically to the promoter region (171 to 90) of the control, sh-hTOP3a, and sh-p53 cells revealed an inverse TP53 gene, which is a DNA region shown to be important relationship between senescence and tumorigenic propen- for p53 promoter activity (36). In addition, hTOP3a also sity (compare Supplementary Fig. S4B with S4C). These binds to the p53-binding sites of the p21 promoter (1452 results suggest that hTOP3a might, in part, regulate tumor to 1310 and 2313 to 2212) but not to the nonspecific growth in part through its ability to modulate cellular site (NS; 4443 to 4199). Importantly, the recruitment senescence in AGS cells. Nevertheless, two senescence mark- amounts of hTOP3a to the above promoter regions were er proteins, histone H3 K9 tri-methylation and HP-1g, reduced in HCT116 p53 / and HCT116 sh-hTOP3a cells remained unchanged in AGS sh-Luc control, sh-hTOP3a (Fig. 6C and D) and the p53 targeting to the p21 promoter and sh-p53 cells (data not shown). regions was decreased in HCT116 sh-hTOP3a cells (Fig. Cells from patients with cancer-prone Bloom syndrome 6E). Collectively, our results suggested that hTOP3a and undergo cellular senescence and genomic instability. The p53 collaboratively modulate the expression of both p53 BLM helicase forms a complex specifically with hTOP3a to and p21 directly. resolve recombination intermediates (31, 39) and also

1496 Clin Cancer Res; 20(6) March 15, 2014 Clinical Cancer Research

TOP3a and p53 Cooperate in Tumor Suppression

Figure 5. Genetic and functional interactions of hTOP3a and the p53-p21 pathway in the suppression of tumorigenic growth. Ectopic hTOP3a expression in HCT116 cell lines induces p53 and p21 expression (A), but reduces anchorage-independent growth in a p53- and p21-dependent manner (B). C, the p21expression is collaboratively regulated by p53 and hTOP3a. D, the coordinative action of p53 and hTOP3a in tumor suppression. E, successful expression of hTOP3a in H1299 p53-null cells neither changed p53 level nor suppressed anchorage-independent growth. F, expression of hTOP3a and catalytically dead hTOP3aY377F mutant (Tyr to Phe mutation at the active site) in HCT116 cells. G, the suppression of anchorage-independent growth by hTOP3a requires hTOP3a enzymatic activity. Cells were transfected with control (Vec) and/or expressing plasmids (as indicated). Immunoblotting and soft agar assays were performed as mentioned above to analyze the protein expression levels (A, C, E, and F) and anchorage-independent growth abilities of cells (B, n ¼ 8; D, E right and G, n ¼ 4). The asterisk indicates a signiﬁcant difference (, P < 0.001).

www.aacrjournals.org Clin Cancer Res; 20(6) March 15, 2014 1497

Hsieh et al.

Figure 6. Functional hTOP3a is involved in the p53 and p21 expression and in the recruitment of p53 to p21 promoter. A, expression of p53 is regulated by hTOP3a at the transcriptional level. B, the enzymatic activity of hTOP3a is required for the optimal p53 and p21 expression. Expression (n ¼ 3) and knockdown (n ¼ 3) of the indicated genes were performed as described above. Total RNA was isolated from sh-Luc, sh-hTOP3a, sh-p53, Vec, OE OE OE hTOP3a , hTOP3aY337F (active site mutant), and p53 cells and their gene expression profiles were then analyzed using real-time RT-PCR with human b-actin transcript as the normalization control (n ¼ 3). Asterisks indicate a significant difference (, P < 0.05; , P < 0.01; , P < 0.001). C, the ChIP assay revealed that hTOP3a binds to specific promoter regions of p53 and p21 genes. Top, schematic diagram of the p53 and p21 promoter locations in the PCR products; bottom, the targeting levels of hTOP3a to the p53 and p21 promoter regions are reduced in the HCT116 p53 / cells. D and E, knockdown of hTOP3a reduced recruitments of both hTOP3a (D) and p53 to p21 promoter regions (E).

interacts physically with p53 protein (40). We observed that containing 241 pairs of tumor specimens (T) and matched sh-BLM caused an elevation in the level of b-gal senescent normal samples (N) from various anatomic origins, we staining in AGS cells but with only a minimal effect on the examined the relative expression of hTOP3a (T/N ratio). tumorigenic growth (Supplementary Fig. S4D–S4F). As In Supplementary Fig. S5A, the levels of the hTOP3a mRNA reported (31), we found that sh-hTOP3a cells have reduced transcript from tumor samples extracted from stomach and BLM levels and exhibit chromosome instability phenotype kidney cancers were lower compared with the correspond- (data not shown). These results suggest that hTOP3a acts ing normal tissues. Quantitative results revealed that with p53 but not with BLM to mediate senescence and approximately 75%–80% tumor samples from patients tumor suppression. express less hTOP3a mRNA (i.e., T/N ratio < 1; stomach, Supplementary Fig. S5B; kidney, Supplementary Fig. S5C). Reduced hTOP3a expression in human cancer samples Furthermore, we have also obtained 10 pairs of cDNA The expression levels of hTOP1 and hTOP2a in tumor originated from renal carcinoma (RCC) and matched nor- samples are generally higher than in matched normal mal tissues. We found that the hTOP3a mRNA levels were tissues (41). Using the Cancer Proﬁling Array membrane reduced in 80% of the RCC tumor samples (Supplementary

1498 Clin Cancer Res; 20(6) March 15, 2014 Clinical Cancer Research

TOP3a and p53 Cooperate in Tumor Suppression

Fig. S5D). In addition, based on gene expression profiling (43, 44), leading to the discovery of chromatin-modifying data from The Cancer Genome Atlas (TCGA), we found factors, including BRD7, TIP60, and HDAC4 (5, 9, 43). that the hTOP3a mRNA levels are downregulated 53.8% Similar to our findings on hTOP3a, these enzymes either (59.2% altered, 316 cases) and 32.3% (41% altered, 195 serve as cofactors of p53 on tumor suppression or are cases) in ovarian serous cystadenocarcinoma and colon required for its optimal transcriptional activity. In addition, and rectum adenocarcinoma, respectively. These above topoisomerases have been shown to play important roles in results suggest a link of hTOP3a deficiency to tumori- gene expression regulation (12, 13). For example, hTOP2b genesis. Furthermore, hTOP3a deficiency may be a new plays important roles in regulated transcription (14, 45, 46) mechanism that leads to p53 dysfunction in certain and TOP1 is involved in efficient chromatin disassembly at carcinogenic processes. specific promoter regions (11). Collectively, our results suggested a novel regulatory mechanism via collaborative action of chromatin-modifying and/or topology-resolving Discussion enzymes with transcription factors for cellular processes, Tumor suppressor p53 plays a central role in blocking especially for those require regulated gene expression. Our cancer development through its induction of senescence and observations that hTOP3a interacts with p53 and cooperates apoptosis (1, 3, 5). Through the interactions with p53, with p53 to regulate p21 expression and tumor suppression chromatin factors modulate p53 expression and p53-medi- and the recent report on the transcription factor–dependent ated tumor suppression (5, 8, 9). Here, we have identified the recruitment of Isw2 chromatin remodeling enzyme to spe- topology-resolving hTOP3a protein as a candidate antican- cific genomic loci also support the above notion (47). cer block in p53-mediated tumor suppression. The universal Our and other studies (48) have also revealed that CDK inhibitor, p21, contributes to p53-mediated cell-cycle hTOP3a is essential in certain cell lines. Consistently, the arrest and possibly senescence (37). Our results revealed sh-hTOP3a–mediated lethality in HEK293T cells was con- physical, genetic, and functional interactions of hTOP3a firmed by increases in the sub-G1 population of HEK293T with p53. Moreover, hTOP3a binds to p53 and p21 pro- sh-hTOP3a cells at 3 or 4 days after viral infection (data not moters, positively regulates p53 and p21expression, and shown). As for the p53-independent mechanism(s), we helps the recruitment of p53 to the p21 promoter. Knock- found that the level of active form Ras was reduced in down of hTOP3a (sh-hTOP3a) in AGS cells, behaving HCT116 hTOP3aOE cells (data not shown), suggesting an similarly to the sh-p53 phenotypes, caused an elevated tumor involvement of Ras pathway in the hTOP3a-mediated growth but a reduction in senescence. Recently, it has been tumor suppression. In agreement with previous studies in reported that p53 engagement of DNA damage response yeast (24, 25, 27), we observed that sh-hTOP3a results in (DDR) is defective in the top3b / MEF cells (34). Our results hypersensitivity to cell-cycle–disrupting agents and an ele- also showed that p53-mediated DDR was defected in sh- vated genomic instability (data not shown). These results hTOP3a cells and hTOP3aOE expression promoted p53 and suggest that hTOP3a might suppress tumor growth through p21 expression but reduced tumorigenic growth. regulating the cell-cycle checkpoint and preventing geno- The importance of p53 in tumorigenesis is illustrated by mic instability. Therefore, hTOP3a expression level in the fact that p53 is not only highly mutated (50%) in tumors may be a determinant for chemotherapy, such as tumors (42) but it is also rendered inactive by a range of hydroxyl urea-based ones. In the results obtained from indirect mechanisms in other types of cancer (e.g., MDM2 TCGA (http://cancergenome.nih.gov), we found that amplification and loss of ARF). Recent studies have dem- hTOP3a mRNA expression is commonly altered in cancer onstrated microRNA-mediated translational repression as a with different anatomic origins (15 types of cancer in total) novel mechanism for p53 inactivation (6, 7). Our results and 86.67% (13/15) of cancer types are with the altered that p53 and p21 mRNA level are lower in sh-hTOP3a cells gene expression which is dominant by cancers with a lower suggest that hTOP3a positively regulates p53 and p21 hTOP3a level (downregulated). Coupled with our results, expression at the transcriptional level. We have observed we suggested that altered expression of hTOP3a, especially that approximately 54.4% of cancer samples (more than downregulated one, is commonly associated with the 70% in renal and gastric carcinoma) have a reduction in tumorigenic process. It is thus important to study the hTOP3a expression. We therefore suggest that the reduction underlying mechanisms for the reduction in hTOP3a of hTOP3a expression might be an additional way to expression during tumor development. It should be noted inactivate p53 during tumorigenesis. that hTOP3a could be a potential target for the newly Here, we suggest that hTOP3a acts functionally with p53 identified oncogenic miRNA-372 (49). Further characteri- in tumor suppression by the following potential mechan- zation of the potential role of miRNA-372 in the expression isms: (i) promotion of p53 transcriptional expression; (ii) of hTOP3a will help to clarify whether hTOP3a deficiency interaction with p53 to collaboratively regulate p21 expres- contributes to the tumorigenic effect of miRNA-372 and its sion; and/or (iii) p53-mediated senescence program. Nev- potential as a therapeutic target for the intervention of ertheless, our and other studies also suggest an existence of hTOP3a-deficient cancers. p53-independent mechanism(s) in tumor suppression. RNAi technology has allowed the search for novel tumor Disclosure of Potential Conflicts of Interest suppressors using a genetically defined tumorigenic system No potential conflicts of interest were disclosed.

www.aacrjournals.org Clin Cancer Res; 20(6) March 15, 2014 1499

Hsieh et al.

Authors' Contributions providing the expressing plasmids containing the HA-tagged p53 deletion Conception and design: T.-K. Li, M.-Y. Hsieh, J.-R. Fan mutants. Development of methodology: J.-R. Fan, H.-C. Chen, S.-C. Teng Acquisition of data (provided animals, acquired and managed patients, Grant Support provided facilities, etc.): H.-C. Chen, Y.-H. Yeh This work was supported by grants from the National Science Coun- Analysis and interpretation of data (e.g., statistical analysis, biosta- cil, the National Health Research Institutes, and National Taiwan tistics, computational analysis): M.-Y. Hsieh, J.-R. Fan, H.-W. Chang, T.-L. University. Shen The costs of publication of this article were defrayed in part by the Writing, review, and/or revision of the manuscript: T.-K. Li, J.-R. Fan payment of page charges. This article must therefore be hereby marked Study supervision: T.-K. Li advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Acknowledgments The authors thank Dr. M.-C. Huang for technical assistance and for the Received July 21, 2013; revised November 4, 2013; accepted December 17, reagents for the Cancer Proﬁling Array assay, and Dr. P.-C. Yang for 2013; published OnlineFirst February 13, 2014.

References 1. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100: for resolution of meiotic recombination intermediates in Arabidopsis 57–70. thaliana. PLoS Genet 2008;4:e1000285. 2. Hickson ID. RecQ helicases: caretakers of the genome. Nat Rev 20. Wu L, Bachrati CZ, Ou J, Xu C, Yin J, Chang M, et al. BLAP75/RMI1 Cancer 2003;3:169–78. promotes the BLM-dependent dissolution of homologous recombi- 3. Kinzler KW, Vogelstein B. Cancer-susceptibility genes. Gatekeepers nation intermediates. Proc Natl Acad Sci U S A 2006;103:4068–73. and caretakers. Nature 1997;386:761, 763-3. 21. Li W, Wang JC. Mammalian DNA topoisomerase IIIalpha is essential in 4. Hussain SP, Harris CC. p53 biological network: at the crossroads of early embryogenesis. Proc Natl Acad Sci U S A 1998;95:1010–3. the cellular-stress response pathway and molecular carcinogenesis. 22. Kwan KY, Moens PB, Wang JC. Infertility and aneuploidy in mice J Nippon Med Sch 2006;73:54–64. lacking a type IA DNA topoisomerase III beta. Proc Natl Acad Sci U S A 5. Drost J, Mantovani F, Tocco F, Elkon R, Comel A, Holstege H, et al. 2003;100:2526–31. BRD7 is a candidate tumour suppressor gene required for p53 func- 23. Kwan KY, Wang JC. Mice lacking DNA topoisomerase IIIbeta develop tion. Nat Cell Biol 2010;12:380–9. to maturity but show a reduced mean lifespan. Proc Natl Acad Sci U S A 6. Hu W, Chan CS, Wu R, Zhang C, Sun Y, Song JS, et al. Negative 2001;98:5717–21. regulation of tumor suppressor p53 by microRNA miR-504. Mol Cell 24. Chakraverty RK, Kearsey JM, Oakley TJ, Grenon M, de La Torre Ruiz 2010;38:689–99. MA, Lowndes NF, et al. Topoisomerase III acts upstream of Rad53p in 7. Swarbrick A, Woods SL, Shaw A, Balakrishnan A, Phua Y, Nguyen A, the S-phase DNA damage checkpoint. Mol Cell Biol 2001;21:7150–62. et al. miR-380-5p represses p53 to control cellular survival and is 25. Mankouri HW, Hickson ID. Top3 processes recombination intermedi- associated with poor outcome in MYCN-amplified neuroblastoma. Nat ates and modulates checkpoint activity after DNA damage. Mol Biol Med 2010;16:1134–40. Cell 2006;17:4473–83. 8. Hendricks KB, Shanahan F, Lees E. Role for BRG1 in cell cycle control 26. Oakley TJ, Goodwin A, Chakraverty RK, Hickson ID. Inactivation of and tumor suppression. Mol Cell Biol 2004;24:362–76. homologous recombination suppresses defects in topoisomerase III- 9. Xu Y, Zhang J, Chen X. The activity of p53 is differentially regulated by deficient mutants. DNA Repair 2002;1:463–82. Brm- and Brg1-containing SWI/SNF chromatin remodeling com- 27. Oakley TJ, Hickson ID. Defending genome integrity during S-phase: plexes. J Biol Chem 2007;282:37429–35. putative roles for RecQ helicases and topoisomerase III. DNA Repair 10. Stephens PJ, Tarpey PS, Davies H, Van Loo P, Greenman C, Wedge 2002;1:175–207. DC, et al. The landscape of cancer genes and mutational processes in 28. Win TZ, Mankouri HW, Hickson ID, Wang SW. A role for the fission breast cancer. Nature 2012;486:400–4. yeast Rqh1 helicase in chromosome segregation. J Cell Sci 2005;118: 11. Durand-Dubief M, Persson J, Norman U, Hartsuiker E, Ekwall K. 5777–84. Topoisomerase I regulates open chromatin and controls gene expres- 29. Fritz E, Elsea SH, Patel PI, Meyn MS. Overexpression of a truncated sion in vivo. Embo J 2010;29:2126–34. human topoisomerase III partially corrects multiple aspects of the 12. Nitiss JL. DNA topoisomerase II and its growing repertoire of biological ataxia-telangiectasia phenotype. Proc Natl Acad Sci U S A 1997;94: functions. Nat Rev Cancer 2009;9:327–37. 4538–42. 13. Wang JC. Cellular roles of DNA topoisomerases: a molecular perspec- 30. Johnson FB, Lombard DB, Neff NF, Mastrangelo MA, Dewolf W, tive. Nat Rev Mol Cell Biol 2002;3:430–40. Ellis NA, et al. Association of the Bloom syndrome protein with 14. Haffner MC, Aryee MJ, Toubaji A, Esopi DM, Albadine R, Gurel B, et al. topoisomerase IIIalpha in somatic and meiotic cells. Cancer Res Androgen-induced TOP2B-mediated double-strand breaks and pros- 2000;60:1162–7. tate cancer gene rearrangements. Nat Genet 2010;42:668–75. 31. Raynard S, Bussen W, Sung P. A double Holliday junction dissolva- 15. Yang YC, Chou HY, Shen TL, Chang WJ, Tai PH, Li TK. Topoisomerase some comprising BLM, topoisomerase IIIalpha, and BLAP75. J Biol II-mediated DNA cleavage and mutagenesis activated by nitric oxide Chem 2006;281:13861–4. underlie the inflammation-associated tumorigenesis. Antioxid Redox 32. Tsai HJ, Huang WH, Li TK, Tsai YL, Wu KJ, Tseng SF, et al. Involvement Signal 2013;18:1129–40. of topoisomerase III in telomere-telomere recombination. J Biol Chem 16. Meetei AR, Sechi S, Wallisch M, Yang D, Young MK, Joenje H, et al. A 2006;281:13717–23. multiprotein nuclear complex connects Fanconi anemia and Bloom 33. Wang SP, Wang WL, Chang YL, Wu CT, Chao YC, Kao SH, et al. p53 syndrome. Mol Cell Biol 2003;23:3417–26. controls cancer cell invasion by inducing the MDM2-mediated deg- 17. Wang Y, Cortez D, Yazdi P, Neff N, Elledge SJ, Qin J. BASC, a radation of Slug. Nat Cell Biol 2009;11:694–704. super complex of BRCA1-associated proteins involved in the 34. Mohanty S, Town T, Yagi T, Scheidig C, Kwan KY, Allore HG, et al. recognition and repair of aberrant DNA structures. Genes Dev Defective p53 engagement after the induction of DNA damage in cells 2000;14:927–39. deficient in topoisomerase 3beta. Proc Natl Acad Sci U S A 18. Wu L, Davies SL, North PS, Goulaouic H, Riou JF, Turley H, et al. The 2008;105:5063–8. Bloom's syndrome gene product interacts with topoisomerase III. J 35. Blasina A, Price BD, Turenne GA, McGowan CH. Caffeine inhibits the Biol Chem 2000;275:9636–44. checkpoint kinase ATM. Curr Biol 1999;9:1135–8. 19. Hartung F, Suer S, Knoll A, Wurz-Wildersinn R, Puchta H. Topoisom- 36. Tuck SP, Crawford L. Characterization of the human p53 gene pro- erase 3alpha and RMI1 suppress somatic crossovers and are essential moter. Mol Cell Biol 1989;9:2163–72.

1500 Clin Cancer Res; 20(6) March 15, 2014 Clinical Cancer Research

TOP3a and p53 Cooperate in Tumor Suppression

37. Campisi J, d'Adda di Fagagna F. Cellular senescence: when bad things 44. DingJ,HuangS,WuS,ZhaoY,LiangL,YanM,etal.GainofmiR- happen to good cells. Nat Rev Mol Cell Biol 2007;8:729–40. 151 on chromosome 8q24.3 facilitates tumour cell migration and 38. Rohwer N, Dame C, Haugstetter A, Wiedenmann B, Detjen K, Schmitt spreading through downregulating RhoGDIA. Nat Cell Biol 2010;12: CA, et al. Hypoxia-inducible factor 1alpha determines gastric cancer 390–9. chemosensitivity via modulation of p53 and NF-kappaB. PLoS ONE 45. Ju BG, Lunyak VV, Perissi V, Garcia-Bassets I, Rose DW, Glass CK, 2010;5:e12038. et al. A topoisomerase IIbeta-mediated dsDNA break required for 39. Wu L, Hickson ID. DNA helicases required for homologous recombi- regulated transcription. Science 2006;312:1798–802. nation and repair of damaged replication forks. Annu Rev Genet 2006; 46. Tiwari VK, Burger L, Nikoletopoulou V, Deogracias R, Thakurela S, 40:279–306. Wirbelauer C, et al. Target genes of Topoisomerase IIbeta regulate 40. Sengupta S, Linke SP, Pedeux R, Yang Q, Farnsworth J, Garfield SH, neuronal survival and are defined by their chromatin state. Proc Natl et al. BLM helicase-dependent transport of p53 to sites of stalled DNA Acad Sci U S A 2012;109:E934–43. replication forks modulates homologous recombination. EMBO J 47. Yadon AN, Singh BN, Hampsey M, Tsukiyama T. DNA looping facil- 2003;22:1210–22. itates targeting of a chromatin remodeling enzyme. Mol Cell 2013; 41. Li TK, Liu LF. Tumor cell death induced by topoisomerase-targeting 50:93–103. drugs. Annu Rev Pharmacol Toxicol 2001;41:53–77. 48. Temime-Smaali N, Guittat L, Wenner T, Bayart E, Douarre C, Gomez D, 42. Olivier M, Hussain SP, Caron de Fromentel C, Hainaut P, Harris CC. et al. Topoisomerase IIIalpha is required for normal proliferation and TP53 mutation spectra and load: a tool for generating hypotheses on telomere stability in alternative lengthening of telomeres. EMBO J the etiology of cancer. IARC Sci Publ 2004;157:247–70. 2008;27:1513–24. 43. Berns K, Hijmans EM, Mullenders J, Brummelkamp TR, Velds A, 49. Voorhoeve PM, le Sage C, Schrier M, Gillis AJ, Stoop H, Nagel R, et al. Heimerikx M, et al. A large-scale RNAi screen in human cells identifies A genetic screen implicates miRNA-372 and miRNA-373 as onco- new components of the p53 pathway. Nature 2004;428:431–7. genes in testicular germ cell tumors. Cell 2006;124:1169–81.

www.aacrjournals.org Clin Cancer Res; 20(6) March 15, 2014 1501

DNA Topoisomerase III Alpha Regulates p53-Mediated Tumor Suppression

Mei-Yi Hsieh, Jia-Rong Fan, Han-Wen Chang, et al.

Clin Cancer Res 2014;20:1489-1501. Published OnlineFirst February 13, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-13-1997

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2014/02/14/1078-0432.CCR-13-1997.DC1

Cited articles This article cites 49 articles, 23 of which you can access for free at: http://clincancerres.aacrjournals.org/content/20/6/1489.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/20/6/1489. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.