Suppression of RAD21 Gene Expression Decreases Cell Growth and Enhances Cytotoxicity of Etoposide and Bleomycin in Human Breast Cancer Cells

Molecular Cancer Therapeutics 361

Suppression of RAD21 gene expression decreases cell growth and enhances cytotoxicity of etoposide and bleomycin in human breast cancer cells

Josephine M. Atienza,1 Richard B. Roth,1 Introduction 1 1 Caridad Rosette, Kevin J. Smylie, The RAD21 gene codes for a human homologue of Stefan Kammerer,1 Joachim Rehbock,2 1 1 Saccharomyces pombe Rad21 protein. The current knowledge Jonas Ekblom, and Mikhail F. Denissenko about this protein points to a role in modulation of cell 1Sequenom, Inc., San Diego, California and 2Frauena¨rzte growth and in cell defense against DNA damage, both Rosenstrasse, Munich, Germany processes being central to carcinogenesis. Several DNA repair genes including rad21 were initially identified in the fission yeast S. pombe as radiation-sensitive mutants (1, 2). Abstract Specifically, Rad21 has been implicated in homologous A genome-wide case-control association study done in our recombination–mediated double-strand break (DSB) re- laboratory has identified a single nucleotide polymorphism pair, and is unique among the radiation response genes in located in RAD21 as being significantly associated with that it also plays a role in cell cycle regulation (3, 4). Yeast breast cancer susceptibility. RAD21 is believed to function Rad21 and its mammalian homologue were subsequently in sister chromatid alignment as part of the cohesin complex identified as components of a conserved cohesin complex and also in double-strand break (DSB) repair. Following our (5, 6), which is believed to function in aligning sister initial finding, expression studies revealed a 1.25- to 2.5- chromatids during the early stages of cellular division. fold increased expression of this gene in several human Deletion of RAD21/Scc1/Mcd1 in mammalian cells leads breast cancer cell lines as compared with normal breast to abnormal separation of sister chromatids during tissue. To determine whether suppression of RAD21 interphase and improper alignment during metaphase. expression influences cellular proliferation, RNA interfer- These cells also incur increased levels of spontaneous ence technology was used in breast cancer cell lines MCF-7 chromosomal breaks and ionizing radiation–induced chro- and T-47D. Proliferation of cells treated with RAD21- mosomal aberrations, probably due to a reduction in DSB specific small inhibitory RNA (siRNA) was significantly repair efficiency (7). These observations suggest that reduced as compared with mock-transfected cells and cells RAD21 and the cohesin complex not only mediate the transfected with a control siRNA (Lamin A/C). This alignment of chromosomes in preparation for segregation inhibition of proliferation correlated with a significant into daughter cells during mitosis but also facilitate the reduction in the expression of RAD21 mRNA and with an repair of DNA damage incurred during DNA replication increased level of apoptosis. Moreover, MCF-7 cell sensi- by holding sister chromatids together. tivity to two DNA-damaging chemotherapeutic agents, DNA damage is a consequence of exposure to both etoposide and bleomycin, was increased after inhibition of exogenous and endogenous agents, resulting in a diverse RAD21 expression with a dose reduction factor 50 (DRF50) array of DNA modifications. Exogenous agents such as of 1.42 and 3.71, respectively. At the highest concen- ionizing radiation and radiomimetic chemicals and endog- trations of etoposide and bleomycin administered, cells enous agents such as oxygen radicals can all induce DSBs. transfected with a single siRNA duplex targeted against In addition, certain cellular processes including replication, RAD21 showed 57% and 60% survival as compared with meiosis, and V(D)J recombination also give rise to DSBs. control cells, respectively. Based on these findings, we DNA DSBs are among the most deleterious of DNA conclude that RAD21 is a novel target for developing cancer modifications, often resulting in mutagenesis or cytotox- therapeutics that can potentially enhance the antitumor icity (8). DSBs can contribute to tumorigenesis through the activity of chemotherapeutic agents acting via induction of introduction of mutations or chromosomal aberrations DNA damage. [Mol Cancer Ther 2005;4(3):361–8] leading to abnormal regulation of oncogenes or loss of tumor suppressor genes (9). Defects in DSB repair and increased levels of DSBs have both been linked to the development of human tumors (9). Several intricate and Received 9/10/04; revised 12/21/04; accepted 1/5/05. well-regulated DSB repair mechanisms have evolved to The costs of publication of this article were defrayed in part by the counteract these damages. DSBs can be repaired by payment of page charges. This article must therefore be hereby marked homologous recombination or by nonhomologous end advertisement in accordance with 18 U.S.C. Section 1734 solely to joining. indicate this fact. Many common anticancer agents introduce DSBs in Requests for reprints: Mikhail F. Denissenko, Sequenom, Inc., 3595 John Hopkins Court, San Diego, CA 92121. Phone: 858-202-9000; DNA. This therapeutic option takes advantage of the Fax: 858-202-9001. E-mail: [email protected] inefficient DNA-damage checkpoints in cancerous cells Copyright C 2005 American Association for Cancer Research. and, therefore, their compromised ability to efficiently

Mol Cancer Ther 2005;4(3). March 2005

repair the damage caused by the drugs before cell division, competitor oligonucleotide that differs at one base position resulting in increased cytotoxicity. Disabling the cellular from the cDNA target. This competitor is used to calibrate processing and repair of DSBs potentiates cell sensitivity to the assay and quantitate the genes of interest at an absolute traditional chemotherapeutic agents (10–12). Because hu- level by matrix-assisted laser desorption-ionization time-of- man RAD21 is believed to play a dual role in the cell, both flight mass spectrometry (MALDI-TOF) mass spectro- in modulation of sister chromatid alignment during cell metry. The principle of Quantitative Gene Expression by cycle and through repair of DNA DSBs, it presents a good MassARRAY assay was described elsewhere (14). Levels of target candidate for adjuvant gene-specific approaches that gene-specific mRNA were normalized against levels of 18S enhance sensitivity of tumor cells to anticancer agents. rRNA by dividing the observed RAD21 transcript concen- In a large-scale breast cancer case-control study, we tration by the observed concentration of 18S rRNA for each obtained evidence that RAD21 is a susceptibility gene for respective sample. The following oligonucleotides were breast cancer (13). In this article we investigated the role of used: 5V-ACGTTGGATGATATGGATGAGGATGA- RAD21 in cancer cell proliferation by analyzing its TAATGTATC-3V (forward PCR primer), 5V-ACGTTG- expression in a number of human breast tumor cell lines GATGCAGTCATGGTTGGCATTG-3V (reverse PCR and also exploring the utility of silencing its expression in primer), 5V-GTTCAACGGGATCCACTGAAT-3V (MassEX- these cells. Here we present the first report showing that TEND primer), 5V-CAGTCATGGTTGGCATTGGTT- inhibition of RAD21 gene expression in breast cancer cells CAACGGGATCCACTGAATCAGGACTATCAGGCC- results in decreased cellular proliferation, increased apo- CACCCATTGATACATTATCATCCTCATCCATAT-3V ptosis, and increased cell killing after exposure to chemo- (cDNA competitor). Analysis included triplicate experi- therapeutic agents. ments with quadruplicate spotting of reaction products onto MALDI-TOF chips. Data shown represent means of three experimental measurements. Materials and Methods Small InhibitoryRNA Design Cell Lines and Culture Conditions siRNA duplexes were designed according to the guide- MCF-7, T-47D, Hs 578Bst, ZR-75-1, HCC1937, Au-565, Hs lines of Elbashir et al. (15). The following siRNA against 578, BT-474, HCC1395, HCC1428, HCC1500, and SK-BR-3 RAD21 were used: siRad21_272, 5V-AAGCCCAUGU- were acquired from American Type Culture Collection GUUCGAGUGUA-3V; siRad21_1175 5V-AAGAGUUG- (Manassas, VA). Human mammary epithelial cells were GAUAGCAAGACAA-3V, and scrambled siRad21_1175 generously provided by S. Bates (City of Hope Medical siRNA, siRad21_1175s 5V-AAGACAGAUACGAUGAU- Center, Duarte, CA). MCF-7 and T-47D were maintained GAGA-3V. The sequence of Lamin A/C control siRNA in MEME (American Type Culture Collection) and RPMI duplex was 5V-CUGGACUUCCAGAAGAACA-3V. Ready to (American Type Culture Collection) media, respectively, use synthetic siRNA duplexes, including a fluorescent supplemented with 10% fetal bovine serum (Omega transfection control Cy3-modified Luciferase (GL2) siRNA, Scientific, Tarzana, CA) and 10 Ag/mL insulin (Invitrogen, were purchased from Dharmacon (Lafayette, CO) and, Carlsbad, CA). Cells were fed every 5 days and subcultured upon dilution in water, were directly used in transfection when confluent. experiments. RNA, Reverse Transcription ^ PCR, and Quantitative Cell Proliferation and Apoptosis Assays Gene Expression Analysis MCF-7 and T-47D cells were plated in six-well plates Human normal breast tissue total RNA was purchased at 3.5 105 and 2.5 105 cell/mL concentrations, from Ambion (Austin, TX) and used for cDNA synthesis. respectively. Twenty-four hours after plating (day 0), cells To assess cellular expression of RAD21 mRNA, 50 106 were transfected with siRNA (40 nmol/L) using Lipofect- cells were collected and total RNA extracted with a TRIzol amine 2000 (Invitrogen) as suggested by the manufacturer. reagent (Invitrogen, Carlsbad, CA). Reverse transcription On the next day (day 1) cells were trypsinized and plated was done with a SuperScript II kit (Invitrogen) using a in triplicates in 96-well dishes at concentration 1 104 mixture of oligodeoxythymidylate and 18S rRNA–specific cells/well. Cell proliferation was measured using WST-1 primers and aliquots taken for PCR. RAD21 message was assay kit (Roche Diagnostics, Indianapolis, IN) at desig- detected using the following PCR primers: 5V-CAATGC- nated time points, and relative proliferation calculated CAACCATGACTGAT (forward) and 5V-CGGTGTAAGA- by normalizing to day 1 values. Experiments were done at CAGCGTGTAAA (reverse). Amplification (30 cycles) was least thrice. Apoptosis was measured on day 3 using the done at annealing temperature of 60jC. For small inhibi- Vybrant Apoptosis Assay Kit #3 (Molecular Probes, tory RNA (siRNA) experiments, cells were harvested on Eugene, OR) as directed by the manufacturer. day 2 post transfection with siRNA, and total RNA was Clonogenic Survival Assay extracted using standard methods. Cell survival was measured based on colony formation. Levels of transcripts were assessed using competitive MCF-7 cells were transfected with siRad21_272. Two days reverse transcription–PCR and mass spectrometry (Quan- after transfection, cells were exposed to varying concen- titative Gene Expression by MassARRAY assay, Sequenom, trations of etoposide and bleomycin (both from Sigma, St. San Diego, CA). The competitive PCR step of Quantitative Louis, MO) for 2 hours at the concentrations indicated. Gene Expression by MassARRAY includes a synthetic Cells were then trypsinized, washed, and replated at 200

Mol Cancer Ther 2005;4(3). March 2005

cells per 10-cm dishes. Cells were grown for 11 to 12 days (Hs 578Bst and hMEC), and in normal breast tissue, using a (four population doublings), fixed, and stained with crystal Quantitative Gene Expression analysis procedure that violet to detect colonies. Colonies of >50 cells were counted. combines competitive PCR and MALDI-TOF mass spec- Survival of transfected cells was quantitated as a fraction of trometry (14). A comparison of normalized RAD21 levels surviving control (untreated) cells and plotted on a log across all cell lines and tissue tested revealed that RAD21 scale as a function of dose. Increased sensitivity was mRNA expression is lower in normal and immortalized measured as dose reduction factor (DRF), the factor by breast cancer cell lines as compared with 9 out of 11 which the dose of radiation or drug is reduced in the tumorigenic breast cancer lines (Table 1). presence of the sensitizing agent to achieve the same level Null mutations of rad21 in yeast and vertebrate DT40 of cell killing in the absence of the sensitizing agent (10). cells result in a loss of proliferative capacity, underscoring DRF was calculated at the dose that produced 50% cell the importance of Rad21 in mitosis. To determine whether survival (50% clonogenic cell killing). RAD21 plays a role in modulating proliferation of breast Statistical Analysis cancer cells, we used RNA interference technology to Clonogenic survival curves for MCF-7 cells treated silence expression of this gene. Optimal cellular transfec- with etoposide and bleomycin were generated by plotting tion conditions were selected for both cell lines before the average fraction of surviving clones F SD on a log transfection of RAD21-specific siRNAs. Various concen- scale versus the compound concentration. The average trations of Lipofectamine 2000 and Cy3-modified Luciferase fraction of surviving clones was calculated relative to (GL2) siRNA were tested to determine the optimal control (untreated) cells using triplicate samples from combination that resulted in the highest transfection three independent experiments. DRF values were deter- efficiency. Cy3-modified GL2 siRNA was used to monitor mined from the fitted curve generated using XLfit 4.0 transfection efficiency by counting the percentage of Cy3 software package. To determine the significance of fluorescent cells from three to five microscope fields. From differences in sensitivities of transfected and control cells these preliminary experiments, we selected a working to etoposide and bleomycin, Student’s t test was done on transfection concentration of 40 nmol/L siRNA. This condi- values for compound concentrations that gave 50% tion routinely produced a >95% transfection efficiency in survival of clones. P values of <0.05 were considered cells cultured in six-well plates. Other transfection significant. conditions may be optimal for 24-, 48-, or 96-well tissue culture plates. Multiple siRNA were designed according to the guidelines of Elbashir et al. (15) and BLAST verified Results to ensure specificity for RAD21 mRNA. These duplexes Using DNA samples from clinically diagnosed breast were then transiently transfected into breast cancer cancer cases and matched controls, a genome-wide study cells and cellular proliferation (viability) was analyzed on consisting of 25,494 single nucleotide polymorphisms was conducted in our laboratory to discover single nucleotide polymorphisms associated with breast cancer (13). The Table 1. Relative levels of RAD21 expression in normal breast tissue, immortalized breast cell lines, and tumorigenic breast single nucleotide polymorphisms were selected to be gene cancer cell lines based, that is, located within 10 kb of 15,995 LocusLink c annotated genes, and to cover the genome at a median Tissue or cell line* Expression level spacing of about 40 kb. The study identified a C-to-G single nucleotide polymorphism in intron 1 of the RAD21 gene Normal breast 3.5 F 1.1 (rs1374297) as being strongly associated with breast cancer Hs 578Bst 1.0 F 0.3 susceptibility. The frequency of the G allele was increased hMEC 1.3 F 0.5 F in cases (14.8%) compared with controls (7.5%), resulting in ZR-75-1 2.1 0.8 F a P value of 0.0003 and an odds ratio of 2.1. The analysis of HCC1937 3.5 1.2 T-47D 3.8 F 0.9 a total of 27 single nucleotide polymorphisms across the Au-565 4.0 F 0.5 gene in case and control samples pointed toward intron 1 MCF-7 5.1 F 0.7 as the region of highest significance (data not shown). Hs 578 5.2 F 1.1 Therefore, we speculate that the observed disease associ- BT-474 6.9 F 2.3 ation might be due to a variation in a regulatory element in HCC1395 7.9 F 1.3 intron 1. HCC1428 8.7 F 1.5 Gene expression analysis by van’t Veer et al. has HCC1500 8.9 F 1.7 previously shown that RAD21 transcript levels are SK-BR-3 9.8 F 3.7 increased in breast cancer tissue as compared with normal RAD21 *RAD21 cDNA levels were estimated in normal breast tissue, immortalized breast tissue, and that this elevated level of breast cell lines Hs 578Bst and hMEC, and breast cancer cell lines ZR-75-1, expression is also associated with a poor prognosis for HCC1937, T-47D, Au-565, MCF-7, Hs 578, BT-474, HCC1395, HCC1428, breast cancer (16). Prompted by these findings, we HCC1500, and SK-BR-3 using competitive RT-PCR and MALDI-TOF mass spectrometry (MassARRAY). c quantitated RAD21 transcript levels in 11 tumorigenic Expression levels were normalized to the levels of 18S rRNA. Data shown breast cancer cell lines, two immortalized breast cell lines are means F SD.

Mol Cancer Ther 2005;4(3). March 2005

days 1, 2, 4, and 6 post transfection. Two siRNA duplexes, To confirm that gene expression was indeed inhibited siRad21_1175 and siRad21_272, effectively inhibited and that the siRNA-driven inhibition of cell growth was proliferation in both MCF-7 and T-47D cells (Fig. 1A specific to the reduction of RAD21 expression, we did a and B). MCF-7 cells treated with siRad21_272 and quantitation experiment using competitive PCR and siRad21_1175 showed 60% and 34% survival as compared MALDI-TOF mass spectrometry. Although the strongest with control cells treated with Lipofectamine 2000 inhibition of cellular proliferation was observed on day 6, alone, respectively. The effect in T-47D cells was more total RNA for quantitation of gene expression was isolated pronounced: these cells treated with siRad21_272 and from cells on day 2 post transfection. This earlier time point siRad21_1175 showed 32% and 19% survival as compared was used in order to detect the changes in RAD21 message with control, respectively. Because no decrease in prolifer- that were more likely due to siRad21-specific inhibition and ation was observed in cells transfected with siRNA targeted less likely due to general mRNA degradation as a result against Lamin A/C compared with cells mock-transfected of pronounced cell death. We have determined the utility with Lipofectamine 2000 alone, we believe that the effect of early RNA isolation time points on days 1 or 2 in of these siRNA duplexes was likely due to the specific preliminary experiments using a number of unrelated inhibition of RAD21.Furthermore,scrambledsiRNA, siRNA species (data not shown). Consistent with the siRad21_1175S, designed by randomly shuffling the results of the proliferation assay, transfection of both siRad21_1175 nucleotide sequence showed no inhibitory siRad21_1175 and siRad21_272 led to suppression of effect on cell proliferation (data not shown). RAD21 expression in both cell lines (Fig. 1C). Cells

Figure 1. Cell proliferation and RAD21 gene expression levels in MCF-7 and T-47D cells transfected with siRNA specific to RAD21. siRNA specific to RAD21 [siRad21_1175 (x) and siRad21_272 (n)] were transfected in breast cancer cell lines MCF-7 (A) and T-47D (B) and proliferation measured using a WST-1 assay on days 1, 2, 4, and 6 post transfection. Lipofectamine 2000 alone () and siRNA against Lamin A/C (E) were used as controls. Values were normalized to day 1. C, RAD21 mRNA and 18S rRNA levels in cells transfected with RAD21 siRNA. RAD21 expression levels were normalized to the levels of 18s rRNA. Total RNA was isolated on day 2 posttransfection, and quantitative expression analysis by competitive reverse transcription – PCR and MALDI-TOF mass spectrometry was done using primers specific for RAD21 mRNA. Data shown represent means of three experimental measurements, FSD. D, agarose gel electrophoresis of total RNA extracted from siRNA-transfected cells; lane numbering corresponds to that in C.

Mol Cancer Ther 2005;4(3). March 2005

Figure 2. Apoptosis of MCF-7 and T-47D cells transfected with siRNA specific to RAD21. MCF-7 (A) and T-47D (B) cells were transfected with siRad21_1175 and apoptosis assessed on day 3 posttransfection using Annexin-V conjugated to FITC and propidium iodide staining. Several fields were examined and representative fields are shown. transfected with siRad21_1175 showed lower levels of have also done a quantitative assessment of the number of RAD21 mRNA than cells transfected with siRad21_272, Annexin V–stained apoptotic cells at 72 hours post which correlates with the relative decreases in proliferation transfection in MCF-7 and T-47D cells transfected with observed in cells transfected with these siRNAs. This effect siRad21_1175 and siLamin A/C. The respective values is not due to changes in extracted RNA as evidenced in (percentage of Annexin V stained cells) in these two Fig. 1D. samples were 63.6 F 5.1 and 0.02 F 0.01 for MCF-7 cells, To determine whether inhibition of breast cancer cell and 90.8 F 8.9 and 12.3 F 3.3 for T-47D cells. These results proliferation upon treatment with RAD21-specific siRNA suggest that the decrease in proliferation of siRNA- coincided with the induction of programmed cell death, transfected cells could be due to the induction of cells were stained with propidium iodide and Annexin apoptosis. V-FITC 72 hours post transfection and examined by Next we investigated the effect of down-regulating phase contrast and fluorescence microscopy. Both MCF-7 endogenous RAD21 expression on cell killing by known and T-47D cell lines transfected with siRad21_1175 and anticancer drugs that induce DSBs. We reasoned that siRad21_272 contained higher numbers of cells undergo- even a partial down-regulation of RAD21 mRNA level ing apoptosis as compared with cells transfected with would impair DSB repair and thus render cells more control siLamin A/C siRNA and Lipofectamine 2000 sensitive to DSB-generating agents. To assess cell killing alone (Fig. 2 and data not shown). The increased Annexin under partial RAD21 silencing conditions, we did V staining was inversely correlated to the level of clonogenic assays in cells that were initially transfected proliferation of cells transfected with RAD21 siRNA. We with RAD21-specific siRNA and subsequently exposed to

Mol Cancer Ther 2005;4(3). March 2005

varying concentrations of model drugs that induce DNA DSBs, etoposide and bleomycin. In these experiments, we used siRad21_272, which had a less pronounced inhibitory effect on cell proliferation than siRad21_1175. On day 2 posttransfection, MCF-7 cells were treated with etoposide or bleomycin. This time point was chosen based on the above results in which we observed no difference in proliferation between siRad21-transfected and control cells (Fig. 1A) and, at the same time, a marked decrease in RAD21 message in siRad21-transfected but not control cells (Fig. 1). After 11 to 12 days, surviving colonies were counted to determine cellular sensitivity. Cells transfected with siRad21_272 showed increased sensitivity to etoposide (Fig. 3A) compared with control cells transfected with Lamin A/C siRNA or Lipofect- amine 2000 alone. This increased sensitivity corresponded to a DRF50 of 1.42. Interestingly, this modest increase in DRF was accompanied by a substantial decrease in cell survival upon treatment with etoposide. At the highest etoposide concentration, cells transfected with siRad21_272 showed 57% survival as compared with cells transfected with Lamin A/C siRNA or Lipofectamine 2000. In cells treated with bleomycin, a more significant increase in DRF (DRF50 of 3.71) was measured (Fig. 3B). At the highest concentration of bleomycin, cells transfected with siRad21_272 showed 60% survival as compared with control. The increase in cell killing is specific to down- regulation of RAD21 gene expression because the siRad21-transfected cells showed a higher sensitivity compared with control cells over a wide range of concentration of DSB-inducing drugs. These results confirm that RAD21 gene–specific approach augments breast cancer cell killing by DNA-damaging chemotherapeutic agents.

Figure 3. Clonogenic survival of MCF-7 cells transfected with RAD21- Discussion specific siRNA after exposure to etoposide and bleomycin. MCF-7 cells Development of somatic malignancy is characterized by a were transfected with RAD21-specific siRNA followed by exposure to series of changes in processes that regulate cell growth, varying concentrations of etoposide (A) and bleomycin (B). Dose- dependent clonogenic survival was quantitated by colony formation. differentiation, apoptosis, and DNA repair (multistage Curve fitting was done using Xlfit 4.0. Comparison of sensitivity among carcinogenesis). Genetic or epigenetic alterations in one or cells transfected with RAD21 siRNA (siRad21_272; .), with Lipofect- more genes involved in these processes can result in the amine 2000 alone (5), and with siRNA to Lamin A/C (4) at different concentrations of chemotherapeutic agents. P values of 0.014 and loss of coordinated regulation of these cellular events and 0.0002 were calculated for etoposide and bleomycin concentrations that lead to a phenotype of growth advantage. DNA repair gave 50% survival of clones, respectively. Dashed lines, drug concen- mechanisms protect the integrity of the genome and tration that gives 50% survival of clones. provide the main defense against these genetic alterations. Several lines of evidence suggest that cancer cells may genes BRCA1 and BRCA2 function in DNA repair through include and/or acquire deficiencies in genes controlling interactions with Rad51 (20). Furthermore, cytologic exami- DNA repair. It is known that certain genetic syndromes nations of breast cancer cells by several groups have accompanied by predisposition to cancer are due to revealed a very high frequency of chromosomal rearrange- mutations in genes involved in DNA repair (9, 17). For ments (21, 22), a likely consequence of DNA repair example, ataxia telangiectasia is due to a mutation in ATM, dysregulation. a gene involved in S-phase checkpoint response to DSB, In this article, we investigate a possible link between and is characterized by a predisposition to lymphoid the DNA repair gene RAD21 and proliferation of tumor malignancies (18). Li-Fraumeni syndrome, which is accom- cells. Our initial discovery of a single nucleotide panied by a high incidence of breast cancer, is caused by a polymorphism in RAD21 associated with increased mutation in TP53 (19), a gene encoding a key protein in the breast cancer risk stimulated the present study. We transduction of DNA repair signals. Familial breast cancer subsequently showed that RAD21 mRNA levels are

Mol Cancer Ther 2005;4(3). March 2005

higher in multiple breast cancer cell lines compared with numerous DSBs at sites other than the sites of replication. normal tissue and immortalized breast cell lines. Con- The exact molecular basis for eukaryotic cell resistance to sistent with these findings, recent microarray gene bleomycin remains unclear. Our results indicate that expression analyses in human primary breast tumors RAD21 may be an important factor mediating this identified elevated RAD21 expression as associated with resistance. poor prognosis for breast cancer (16). It is therefore In summary, we have shown that down-regulation of plausible that an elevated level of RAD21 expression is RAD21 gene expression in breast cancer cell lines leads to a indicative of a malignant cell growth phenotype. We went marked decrease in cell growth. We also found that this on to show that down regulation of RAD21 expression in treatment enhances the cytotoxicity of two chemothera- breast cancer cell lines using targeted siRNA duplexes peutic drugs, most likely by reducing the efficiency of DSB results in significant loss of cell proliferation and viability. repair. Altogether, these data suggest that RAD21 might be This effect was gene specific and coincided with an a new drug target in breast cancer and that RAD21 gene– induction of programmed cell death. These findings specific therapy may be studied as an adjuvant option establish RAD21 as a potential factor for modulating augmenting the antitumor activity of traditional chemo- growth of breast cancer cells. therapeutic and radiation treatments. Ionizing radiation and surgery are most commonly used to treat various cancers. Exposure of S. pombe rad21 Acknowledgments mutants to g-irradiation results in the accumulation of We thank Dr. John Nestor for helpful discussions and critical reading of the DSBs (4). In vertebrate cells, loss of Rad21 results in an manuscript. increase in spontaneously occurring chromosomal breaks in addition to enhanced ionizing radiation–induced References chromosomal breaks (7). The accumulation of DSBs 1. Groseenbacher-Grunder AM, Thuriaux P. Spontaneous and UV-induced may be explained by the role of Rad21 as a component recombination in radiation-sensitive mutants of Schizosaccharomyces of the cohesin complex (5, 6). This complex facilitates pombe. Mutat Res 1981;81:37 – 48. sister chromatid cohesion during mitosis and homolo- 2. Phipps J, Nasim A, Miller DR. Recovery, repair, and mutagenesis in gous recombination–mediated DSB repair (23). In cells Schizosaccharomyces pombe. Adv Genet 1985;23:1 – 72. 3. Birkenbihl RP, Subramani S. Cloning and characterization of rad21 an lacking functional RAD21, sister chromatids are loosely essential gene of Schizosaccharomyces pombe involved in DNA double- held together, thus reducing the efficiency of homologous strand-break repair. Nucleic Acids Res 1992;20:6605 – 11. recombination–mediated DSB repair. In addition to 4. Birkenbihl RP, Subramani S. The rad21 gene product of Schizosac- ionizing radiation, tumors are frequently treated with charomyces pombe is a nuclear, cell cycle-regulated phosphoprotein. J Biol Chem 1995;270:7703 – 11. chemotherapeutic drugs that induce DSBs. Here we 5. Michaelis C, Ciosk R, Nasmyth K. Cohesins: chromosomal proteins that report that down-regulation of RAD21 in MCF-7 breast prevent premature separation of sister chromatids. Cell 1997;91:35 – 45. tumor cells increases cellular sensitivity to two of such 6. Guacci V, Koshland D, Strunnikov A. A direct link between sister drugs, etoposide and bleomycin. The topoisomerase II chromatid cohesion and chromosome condensation revealed through the inhibitor, etoposide, causes DSBs by stabilizing the DNA- analysis of MCD1 in S. cerevisiae. Cell 1997;91:47 – 57. 7. Sonoda E, Matsusaka T, Morrison C, et al. Scc1/Rad21/Mcd1 is topoisomerase complex, and the radiomimetic antibiotic, required for sister chromatid cohesion and kinetochore function in bleomycin, generates DSBs presumably via binding to vertebrate cells. Dev Cell 2001;1:759 – 70. iron and activating oxygen in the vicinity of DNA. DRF 8. Hoeijmakers JH. Genome maintenance mechanisms for preventing values for etoposide calculated in our experiments are cancer. Nature 2001;411:366 – 74. 9. Pierce AJ, Stark JM, Araujo FD, Moynahan ME, Berwick M, Jasin M. similar to DRF values (1.1–1.4) calculated upon siRNA Double-strand breaks and tumorigenesis. Trends Cell Biol 2001;11:S52 – 9. down-regulation of the DNA repair genes ATM and 10. Collis SJ, Swartz MJ, Nelson WG, DeWeese TL. Enhanced radiation DNA-PK in irradiated prostate cancer cell lines (10). and chemotherapy-mediated cell killing of human cancer cells by small These modest DRF values may be attributed to the fact inhibitory RNA silencing of DNA repair factors. Cancer Res 2003;63: 1550 – 4. that not all surviving cells were transfected and therefore 11. Sak A, Stuschke M, Wurm R, et al. Selective inactivation of DNA- did not have decreased RAD21 transcript levels (10). It is dependent protein kinase with antisense oligodeoxynucleotides: conse- likely that a more detailed dosage/exposure regimen quences for the rejoining of radiation-induced DNA double-strand breaks would determine conditions that achieve more pro- and radiosensitivity of human cancer cell lines. Cancer Res 2002;62: 6621 – 4. nounced cytotoxic effects. The observed sensitivity of 12. Belenkov AI, Paiement JP, Panasci LC, Monia BP, Chow TY. An RAD21 siRNA-transfected MCF-7 cells to etoposide is antisense oligonucleotide targeted to human Ku86 messenger RNA consistent with the role of RAD21 in replication and sensitizes M059K malignant glioma cells to ionizing radiation, bleomycin and etoposide but not DNA cross-linking agents. Cancer Res 2002;62: mitotic chromatid separation, processes critically depen- 5888 – 96. dent on topoisomerase activity (24). The high sensitivity 13. Kammerer S, Roth RB, Reneland R, et al. Large-scale association of transfected cells to bleomycin is an interesting study identifies ICAM gene region as breast and prostate cancer observation that may have implications for anticancer susceptibility locus. Cancer Res 2004;64:8906 – 10. therapy. This effect may be attributed to the role of 14. Ding C, Cantor CR. A high-throughput gene expression analysis technique using competitive PCR and matrix-assisted laser desorption RAD21 as a DNA DSB repair protein active toward metal ionization time-of-flight MS. Proc Natl Acad Sci U S A 2003;100: ion- and oxygen-dependent oxidative damage induced by 3059 – 64. bleomycin. This damage may result in the generation of 15. Elbashir SM, Harborth J, Weber K, Tuschl T. Analysis of gene

Mol Cancer Ther 2005;4(3). March 2005

function in somatic mammalian cells using small interfering RNAs. 20. Scully R, Livingston DM. In search of the tumour-suppressor Methods 2002;26:199 – 213. functions of BRCA1 and BRCA2. Nature 2000;40:429 – 32. 16. van’t Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression profiling 21. Davidson JM, Gorringe KL, Chin SF, et al. Molecular cytogenetic predicts clinical outcome of breast cancer. Nature 2002;415:530 – 6. analysis of breast cancer cell lines. Br J Cancer 2000;83:1309 – 17. 17. van Gent DC, Hoeijmakers JH, Kanaar R. Chromosomal stability 22. Kytola S, Rummukainen J, Nordgren A, et al. Chromosomal and the DNA double-stranded break connection. Nat Rev Genet 2001;2: alterations in 15 breast cancer cell lines by comparative genomic 196 – 206. hybridization and spectral karyotyping. Genes chromosomes Cancer 18. Meyn MS. Ataxia-telangiectasia and cellular responses to DNA 2000;28:308 – 17. damage. Cancer Res 1995;55:5991 – 6001. 23. Koshland DE, Guacci V. Sister chromatid cohesion: the beginning of a 19. Wang XW, Harris CC. TP53 tumour suppressor gene: clues to long and beautiful relationship. Curr Opin Cell Biol 2000;12:297 – 301. molecular carcinogenesis and cancer therapy. Cancer Surv 1996;28: 24. Yanagida M. Frontier questions about sister chromatid separation in 169 – 96. anaphase. Bioessays 1995;17:519 – 26.

Mol Cancer Ther 2005;4(3). March 2005

Downloaded from mct.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. Suppression of RAD21 gene expression decreases cell growth and enhances cytotoxicity of etoposide and bleomycin in human breast cancer cells

Josephine M. Atienza, Richard B. Roth, Caridad Rosette, et al.

Mol Cancer Ther 2005;4:361-368.

Updated version Access the most recent version of this article at: http://mct.aacrjournals.org/content/4/3/361

Cited articles This article cites 24 articles, 7 of which you can access for free at: http://mct.aacrjournals.org/content/4/3/361.full#ref-list-1

Citing articles This article has been cited by 10 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/4/3/361.full#related-urls

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 Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/4/3/361. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.