The Pharmacogenomics Journal (2010) 10, 94–104 & 2010 Nature Publishing Group All rights reserved 1470-269X/10 $32.00 www.nature.com/tpj ORIGINAL ARTICLE

Gene network analysis of oxidative stress-mediated drug sensitivity in resistant ovarian carcinoma cells

AK Maiti Drug resistance in cancer cells involves complex molecular mechanisms and ovarian carcinoma cells become resistant to chlorambucil (Cbl) after A&I, Oklahoma Medical Research Foundation, continuous treatment. This drug- and ionizing radiation-resistant cells have Oklahoma City, OK, USA lower level of endogenous ROS (reactive oxygen species) compared with sensitive cells. Elevation of the cellular ROS level by exogenous ROS Correspondence: generation increases the sensitivity of Cbl to resistant cells. In contrast, Dr AK Maiti, A&I, Oklahoma Medical Research Foundation, 825 NE13th, Oklahoma City, antioxidants prevent the sensitization of resistant cells to Cbl by H2O2,COS À OK 73104, USA. (chronic oxidative stress) or NOO . The molecular mechanism of drug E-mails: [email protected], sensitivity with COS has been investigated by microarray expressions [email protected] followed by gene network analysis and it reveals that a cdc42/rac1 guanine exchange factor, ARHGEF6, with p53 and DNA-Pkc (PRKDC) is central to induce apoptosis in Cblcos (Cbl with COS) cells. mRNA and levels of major gene network pathway differ significantly in Cblcos cells than in Cbl- treated cells. Moreover, DNA-PKc physically interacts with ARHGEF6 and p53 mostly in the nucleus of Cbl-treated cells, whereas in Cblcos-treated cells, its interactions are mostly in the cytoplasm. These results suggest that low doses of Cbl and very low doses of COS together kill Cbl-resistant ovarian carcinoma cells and ARHGEF6 signaling may have an instrumental role in induction of apoptosis in Cblcos cells. The Pharmacogenomics Journal (2010) 10, 94–104; doi:10.1038/tpj.2009.49; published online 17 November 2009

Keywords: ROS; chlorambucil; cancer; microarray

Introduction

To understand the basic mechanism of acquired drug resistance, an ovarian cancer cell line (A2780/100) was developed by exposing the parental cell line (A2780) to escalating doses of an alkylating agent, chlorambucil (Cbl).1 A2780/

100 cells are 10-fold more resistant (on the basis of lethal dose 50 (LD50) values) than the parental cell line to the cytotoxic effect of Cbl, as measured by loss of the colony-forming ability of cells.1,2 Cbl-resistant A2780/100 cells also show cross-resistance to cisplatin (Cpl), bis-chloronitrosourea (BCNU), melphalan and ionizing radiation. Drug resistance is associated with increased expression of antiapoptotic , BCL2L1, Mcl-1 and decreased expression of Bax, a known proapoptotic protein and glutathione S-transferase (GSTp). Modulation of Received 23 March 2009; revised 21 August 2009; accepted 24 September 2009; BCL2L1, Bax and GSTp by transgenic overexpression or gene silencing does published online 17 November 2009 not significantly alter resistance to Cbl, Cpl or etoposide, indicating that the Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 95

primary mechanism of drug resistance in A2780/100 cells is of Cbl for 1 h, followed by addition of growth media not due to enhanced drug detoxification or enhanced containing 10% FBS. Nontoxic doses of H2O2, GO and expression of antiapoptotic proteins.1,2 Cbl first induces NOOÀ are added at 1, 3, 6 and 9 h after Cbl treatment. DNA monoalkyl adducts, followed by alkyl cross-links. Colonies are fixed and stained using cocktail buffer contain- However, Cbl resistance of A2780/100 cells does not ing 2% (w/w) of formaldehyde (Fisher Scientific, Austin, TX, elevate the expressions of N-methylpurine-DNA glycosylase, USA) and 0.01% (w/v) of crystal violet (Sigma-Aldrich). O-6-methylguanine-DNA methyltransferase, ERCC-42,3 or Relative cell survival is calculated from the number of base excision repair proteins such as APE1 and DNA colonies per dish. b-polymerase3 in these cells. Various strategies can be used to sensitize resistant cancer Annexin V apoptotic assay cells to develop successful and improved therapy.4,5 Reactive To identify apoptotic cells, performed flow cytometric oxygen species (ROS), by-products of metabolisms, are toxic analysis on cells stained with Annexin V/PE (phycoerythrin) to cells, but can also exert its effect as signaling molecules and 7-amino actinomycin D (7-AAD).13 leading to the regulation of many cellular processes, Briefly, cells were trypsinized, washed in PBS and resus- including metabolism, cell-cycle control and DNA repair.6 pended in cold Annexin V binding buffer (10 mM 4-(2- ROS indeed has a role in the alteration of the activity hydroxyethyl)-1-piperazineethanesulfonic acid (pH 7.4), of protein kinases (for example, p38 mitogen-activated 0.14 mM NaCl, 0.25 mM CaCl2 Â 2H2O, 0.1% bovine serum kinases, Jun N-terminal kinases, extracellular signal- albumin (BSA; w/v). Annexin-PE (BD Biosciences, San Jose, regulated kinase)7 and p53.8 By modulating endogenous CA, USA) was added to cells for 30 min and then ROS, colon cancer cells have been sensitized9,10 and Mookerjee 7-AAD (2 mgmlÀ1) was added before analysis in a FACScan et al.11,12 sensitized leukemia cell lines with copper (N-2- flow cytometer (Becton Dickinson, Franklin Lake, NJ, USA). hydroxyacetophenone) glycinate,11 which increases ROS. The mean fluorescence for 12 000 cells from three or more Earlier observations indicate that reduced level of ROS independent experiments is expressed as ±s.e.m. in A2780/100 cells is one of the principal reasons for developing and maintaining Cbl resistance. In this study, RNA isolation and hybridization with affymetrix human genomic these A2780/100 Cbl-resistant cells were sensitized with very chips low doses of exogenous COS (chronic oxidative stress). Total RNA is isolated with Trizol (Invitrogen, Carlsbad, CA, Using microarray , I studied the probable USA) and subjected to DNAse1 digestion to remove any molecular mechanisms of sensitizing A2780/100-resistant traces of DNA contamination. RNA is further purified using cells and the involvement of key regulator , DNA-PKc, an RNAeasy column (Qiagen, Valencia, CA, USA). Micro- ARHGEF6, NF-kB2 (p52) and HSP70, in the apoptotic array hybridizations and scanning are carried out in core signaling pathway in Cblcos cells. Unfolding the molecular facility. The quality of RNA is determined by Agilent 2100 mechanism of drug resistance in these cells may help to Bioanalyzer Agilent Technology, Santa Clara, CA, USA. From develop strategies for sensitization of ovarian carcinoma 10 mg of total RNA of each sample, first-strand and double- cells and subsequently for future chemotherapy. strand cDNAs are synthesized using the Superscript First- Strand Synthesis kit (Invitrogen) and Escherichia coli DNA Material and methods polymerase, respectively, and blunt ended with T4 DNA polymerase. Biotinylated cRNA probes are synthesized using Cell culture and drug treatment an in vitro transcription kit (Enzo Laboratories; Applied Human ovarian carcinoma A2780- and A2780/100-resistant Biosystems, Foster City, CA, USA). Each 15 mg of biotinylated cells are maintained in RPMI 1640 (Gibco-Invitrogen Inc., cRNA probe is fragmented to approximately 100 bp using a Carlsbad, CA, USA) with 10% fetal bovine serum (FBS; fragmentation buffer (Applied Biosystems) and used for Sigma-Aldrich, St Louis, MO, USA), glutamine (292 mg lÀ1), hybridization. Human genomic chip U_95AV2 containing penicillin (100 U mlÀ1) and streptomycin (100 mgmlÀ1)at approximately 10 000 full-length genes (Affimetrix Inc.,

37 1C in a humidified 5% CO2 atmosphere. Cells are treated Santa Clara, CA, USA) is hybridized with 15 mg of fragmented for 1 h at 37 1C with Cbl (100 mM) (Sigma-Aldrich; freshly biotinylated cRNA probe. Hybridized chips are scanned using dissolved in methanol/hydrochloric acid (49:1), washed Gene Array Scanner (Hewlett Packard, Palo Alto, CA, USA). with phosphate-buffered saline (PBS) and put in a fresh medium). To generate COS, I added a low dose of glucose Statistical analysis oxidase (GO; 10 ng mlÀ1 (0.0026 U mlÀ1); USB, Cleveland, Hybridized scanned data are analyzed using two softwares, OH, USA) to Cbl-treated cells after 1, 3, 6 and 9 h and, in Spotfire Array Explorer version 7 (Spotfire Inc., Cambridge, some cases, after 12 h. Cell death is observed and counted MA, USA) and GeneSpring (Silicon Genetics, Santa Clara, after 24 h. CA, USA). Genes are analyzed as present and marginal criteria. Normalized genes (those above background and Clonogenic survival assay corrected with the internal chip control) are filtered on the Clonogenic survival assays are carried out as described basis of expression level and 6126 genes those passed the previously.2 Briefly, 200 cells in each 60 mm dish are allowed criteria are further analyzed. These filtered genes are to attach for about 4 h, exposed to increasing concentration subjected to build two-way analysis of variance (ANOVA)

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 96

on the basis of treatment and time, five K-means Cluster are scanned and quantitated with ImageQuant software, Silk Analysis and Gene Tree classification. Statistically significant Scientific, Orem, Utah, USA. The ratio of a given protein (Po0.01) genes are selected for further analysis on the basis of band to the area of its corresponding b-actin band is taken their fold change in Cblcos-treated cells over Cbl-treated cells. into account for protein quantitation and a bar diagram is shown for each protein to visualize the fold change. IPA analysis Genes that are dysregulated in Cblcos cells compared with Co-immunoprecipitation and immunocytochemistry cos Cbl-treated cells are listed in Supplementary Table 1 and Protein lysates (1 mg protein in each) from Cbl -treated cells are subjected to ingenuity pathway analysis (IPA). A gene are incubated at 4 1C overnight with anti-ARHGEF6 antibody network has been constructed to outline the underlying (rabbit), with rabbit immunoglobulin G (IgG) as a control. cos signaling pathways for apoptosis in Cblcos cells. These Cbl- and Cbl -treated cell lysates are incubated with anti- indicate that genes those change expressions are likely to DNA-PKc antibody (mouse), with mouse IgG as a control. be involved in the apoptotic pathway. Both are further incubated overnight with Protein A agarose beads with gentle shaking. Protein-bound beads Validation of microarray gene expression are precipitated and dissolved in sample loading buffer, Expression levels of selected genes are validated by real-time denatured and analyzed in 6% (for DNA-PKc) or 10% PCR (TaqMan method). ARHGEF6, NF-kB2/P52, CYR61, polyacrylamide gel (for all other proteins). For immuno- CACNA2D2, DUSP2/PAC1, RYBP/DEDAF and TIEG probes cytochemistry, Cbl- and Cblcos-treated cells are fixed in a and primers (preassayed) are purchased (Applied Biosystems) mixture of methanol/acetone (50:50), air-dried and incubated and 5-FAM/3-BHQ probes and primers of remaining genes for blocking in TBS-T (TBS with 0.1% BSA and 0.01% Tween are synthesized from Sigma-Genosys, Woodlands, TX, USA. 20) for 1 h at 37 1C. After incubation (1h at 371C) with the Preassayed hypoxanthine phosphorybosyl transferase primary antibody, they are washed in TBS-T and incubated (HPRT) is used as internal control to quantitate total RNA with the corresponding fluorescein isothiocyanate- and within each sample. Quantitative PCR (QPCR) values are rhodamine-labeled secondary antibody. Slides are washed, calculated using the comparative Ct value method and incubated with 46-diamidino-2-phenyl indole (DAP1) to stain represent the fold change of each transcript level. Probes nuclei and air-dried coverslips are viewed using a Zeica and primers of those synthesized are listed in Table 1. immunofluorescence microscope, Tokyo, Japan.

Western blotting Results Pelleted cells are lysed in lysis buffer. After centrifugation, protein concentrations are determined using the Bradford Chronic subtoxic oxidative stress sensitizes A2780/100 to Cbl reagent (Bio-Rad Inc., Hercules, CA, USA). Total protein Chlorambucil-treated A2780 cells continuously generate (50 mg) is loaded onto a 6 or 10% (6% for DNA-PKc) SDS gel ROS, whereas A2780/100-resistant cells normalize ROS and transferred to polyvinylidene difluoride membrane level after the first 2 h of treatment; this observed decrease (Immobilon, Millipore), blocked with 5% dry milk in PBS- in ROS in resistant cells may be attributed to glutathione T (25 mM Tris-HCl (pH 8.0), 125 mM NaCl and 0.025% extrusion from the cell.14 Therefore, the inability of cells Tween 20) for 1 h at room temperature and incubated with to maintain oxidative stress may be the reason for the appropriate primary antibodies (anti-DNA-PKc; BD Phar- observed resistance in A2780/100 cells.15 When A2780/100 mingen, San Diego, CA, USA; anti-NF-kB2/p52, Upstate cells are treated with L-buthionine (SR)-sulfoximine (BSO), Biotechnology; anti-ARHGEF6, Orbigen Biotechnology; a g-glutamylcysteine synthetase (a rate-limiting in anti-HSP70, anti-BCL2L1 and anti-p53, anti-actin and anti- glutathione (GSH) synthesis) inhibitor,16 the percentage of tubulin are from Santa Cruz Biotechnology) for 90 min, cell survival significantly decreased. LD50 of Cbl A2780/100 washed thoroughly, followed by incubation with horse- cells also decreases when cells are exposed to GSH-depleting radish-peroxidase-conjugated secondary antibody (Amer- agent, diethyl maleate.16 In this study, ROS produced in sham Inc., Fairfield, CT, USA; 1:500 to 1:1000 dilution) for GSH-depleted cells is H2O2 because GSH is a substrate for 14 1 h. Secondary antibody is detected by enhanced chemilu- GSH peroxidase to eliminate H2O2. Accordingly, 5 mM 17 minescence (Amersham Inc.). Protein bands of western blots H2O2, a signaling and growth stimulatory dose, is added

Table 1 List of probes and primers sequences synthesized

Gene Probe Forward primer Reverse primer

CDK6 50-cagtggtcgtcacgctgtgg 50-acttcggccttgcccgcatc 50-tctgcaaatatgcagccaac HSP70 50-gagcgcatggtgcaggaggc 50-acaagatcaccatcaccaac 50-tgttgaaggcgtaggactcc TRAF3 50-acctgctgaaggagtggagc 50-caggggacaaaccagcagatc 50-ttgtgcaaactttgtatgctc TSPYL4 50-ccaatccgctggcaccg 50-atgaacgcagatcctctggcc 50-aactagggatagtgttcc CLARP2 50-aaggattacatgggccgaggc 50-atggcagagattggtgagg 50-tctccaactcaaccacaagg

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 97

60 50 40

LD50 30 20 10 0 NO COS H202 NOO- ++H202 COS+NAC NOO-+MCP ++H202+CAT COS+ MnSOD A2780/100 Cbl (100μM)

2 cos Figure 1 Chlonogenic survival assay for A2780/100 cells exposed to chlorambucil (Cbl), H2O2 and Cbl and gene network ingenuity pathway À1 analysis (IPA) lead to apoptosis. Bar 1, 5 mM H2O2; bar 2, 50 mM H2O2; bar 3, 50 mM H2O2 þ 5 mM catalase (Sigma; 5000 U ml ). Addition of H2O2 decreases lethal dose 50 (LD50), which can be reversed with an excess of catalase (CAT) in the medium. H2O2 (5 mM) is likely to be nonlethal to resistant cells. Glucose oxidase treatment (10 ng mlÀ1) (bar 4) and addition of MnSOD (10 ng mlÀ1) (bar 5) with chronic oxidative stress À (COS) further decrease LD50, although COS þ NAC (Bar 6) does not decrease LD50. MnSOD, NAC or MCP is added and NOO is generated after 1 h of Cbl treatment. However, antioxidant NAC partially reverses cell death. Addition of NO has no effect (bar 7) on cell growth, although À À À NOO (1 nM) decreases (bar 8) LD50, similar to COS and H2O2. NOO scavenger MCP reverses (bar 9) the effect of NOO in cell growth. (b) IPA analysis shows the pathways of Cbl–COS treatment in Cblcos cells. Inefficient repair may communicate to nuclear-cytosolic protein and inhibit cell-cycle progression, adhesion and spreading. Simultaneously, cytosolic proteins may signal for apoptosis in Cblcos cells at 1, 3, 6 and 8 h directly into media of Cbl-treated A2780/ extensive cell death of A2780/100 cells is not observed by

100 cells. The effect of H2O2 on Cbl-mediated cell killing is treating with COS multiple times and then once with Cbl determined in clonogenic survival assays. As shown in (100 mM) (75% cell death is observed; data not shown). These

Figure 1a, 50 mM H2O2 (bar 2) mediated an 80% decrease in results suggest that Cbl-resistant A2780/100 cells are the LD50 of Cbl (12.5±0.2 mM)-treated cells, but addition of sensitive to exogenous COS and sublethal doses of COS 5 mM H2O2 (bar 1) alone does not affect cell survival, can kill these cells. suggesting that this concentration of H2O2 is a sublethal dose. Addition of catalase (5 mM, 4000 U mlÀ1) along with Gene expression profiling of A2780/100 cells treated with H2O2 prevents the decrease in LD50. chlorambucil and chronic oxidative stress To further assess the impact of low-dose COS on Cbl- To elucidate the molecular mechanism of COS-mediated mediated cytotoxicity, COS is generated by GO treatment increase in Cbl sensitivity of A2780/100 cells, we performed (10 ng mlÀ1, 0.0026 U mlÀ1). GO, a high molecular weight microarray hybridizations with Affymetrix oligo chips. 18,19 À1 21 protein, generates H2O2 when it reacts with glucose, A2780/100 cells are treated with GO (0.0026 U ml ), Cbl À À1 which in turn produces superoxide anion (O2 ), which is (25 mM,2Â LD50) or Cbl (25 mM) with COS (10 ng ml at converted by dismutases (Cu/ZnSOD to MnSOD and FeSOD) 1, 3 and 6 h after Cbl exposure). Total RNAs are isolated at 20 into H2O2. As determined in preliminary studies, a dose of 0, 4 and 9 h after Cbl exposure and used for microarray GO (10 ng mlÀ1, 0.0026 U mlÀ1) causes an approximately hybridization. As control, RNA is isolated from actively twofold increase in intracellular levels of ROS and gives replicating nontreated A2780 cells. Analysis with two more consistent oxidative insult than does H2O2 alone. software indicates that A2780 and A2780/100 cells have a Moreover, exogenous addition of 1 nM NOOÀ (generated by different expression profile (Supplementary Figure 1A). diethylamine NONOate), but not NO, sensitizes A2780/100 Differential expression in drug-resistant A2780/100 cells cells to Cbl, which can be reversed by a peroxinitrite over Cbl-sensitive cells suggests that a set of genes may scavenger, MCP (3-methyl-1,2-cyclopentanedione; bar 9, contribute to Cbl resistance. Although only COS-treated Figure 1a). In parallel experiments, multiple doses of COS A2780/100 cells have a very different gene expression do not affect survival of control solvent-treated cells. The profile, significant differences in expression profile are cos cells treated with LD50 dose of Cbl under COS conditions observed between Cbl- and Cbl -treated cells. exhibit apoptosis, which is shown by the detection of cell- By K-means (K ¼ 5) cluster analysis, similarly expressed surface expression of phosphatidyl-D-serine on the outer genes are grouped. Global changes of repair, channel, leaflet using Annexin V assay.13 When only Cbl or multiple mitochondrial and apoptotic genes are particularly shown doses (3–5) of COS are added separately, it neither causes (Supplementary Figure 1B). By two-way (treatment and time) Annexin V binding nor has an impact on cell survival ANOVA building and Gene Tree classification, a number of colony formation assays (data not shown). Similarly, genes (at 95% confidence level) are further identified that are

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 98

differentially regulated in Cbl- and Cblcos-treated cells. These treated cells are maintained at both 4 and 9 h, whereas it is genes are involved in various physiological pathways such as reduced at 9 h in Cbl-treated cells, suggesting that the cell division cycle genes, transcription factors, cancerous cell twofold higher level of expression of this gene may also be markers, channel genes, DNA damage-inducible repair- necessary for apoptosis signaling in Cblcos-treated cells. associated genes, apoptotic genes and mitochondrial genes. An apoptotic inducer in tumor cells, Cyr61, is highly Genes that are at least twofold up- or downregulated in expressed in only COS-treated cells and its stress-sensitive Cblcos-treated cells than in Cbl-treated cells are listed in expression has not been reported earlier. However, its Supplementary Table 1. Most of the apoptotic signaling expression level is elevated at 4 h but decreased significantly genes change their expression levels, suggesting that COS (sixfold) in Cbl- and Cblcos-treated cells at 9 h. induces a series of changes in gene expression in the network (Figure 1b) and signals cell death by apoptosis, as earlier evidenced in Cbl-treated A2780-sensitive cells.2,15 Differential expression of apoptotic signaling genes at protein levels Expression profiling of Cbl-sensitive A2780 and -resistant because of Cblcos treatment A2780/100 cells shows the expression difference for a set It would be important to correlate whether transcript and of genes, and those changes in gene expression due to protein levels of ARHGEF6, p53, NF-kB2/p52 and HSP70 are Cbl-induced resistance in A2780/100 cells are not apparent. also differentially regulated and contribute to apoptotic cell The differences in chip value and gene expression values in death of Cblcos-treated cells. A2780 and A2780/100 cells (Supplementary Table 2) suggest Guanine nucleotide depletion triggers cell-cycle arrest and that sensitive and resistant cells significantly differ in their apoptosis in human neuroblastoma cells.26 ARHGEF6, a gene expression profile. However, anticipated changes in the cdc42 guanine exchange factor, mutated in human non- expression of SOD (1.6-fold) or catalase (1.4-fold) are not specific mental retardation, interacts with paxillin, which is extensively observed, and probably minute changes in involved in cell proliferation, spreading and regulating transcript level may have a profound effect in ROS contact inhibition through integrin-mediated cell signal- generation and maintenance. A higher expression of ing.27 Its transcript and protein levels decrease twofold ARHGEF6 and BCL2l1 and a lower expression of CYR61 (Figures 3a and b) in Cblcos-treated cells at 9 h of treatment and TSPYL4 may justify the protection of cells from apop- compared with Cbl treatment alone. This reduced mRNA tosis with Cbl only, although an increased expression expression and protein level in Cblcos cells may be essential of DUSP2 and CACNA2D2 and a decreased expression of for inhibition of tumorigenicity, such as cell proliferation CLARP2 and CDK6 in A2780/100 cells support inclination and spreading, before apoptosis and may induce apoptosis as toward apoptosis. observed in other tumor cells.28 DNA-PKc, a critical DNA repair protein, travels to the cytoplasm and interacts with Validation of microarray gene expression by real-time PCR caspases29 to induce apoptosis. Co-immunoprecipitation Among differentially regulated genes, five genes (CDK6, studies (Figures 3e and f) with the ARHGEF6 antibody CLARP2, TSPYL4, RYBP and CYR61) are selected and their suggest that DNA-PKc physically interacts with ARGEF6 in mRNA expressions are validated by real-time PCR (TaqMan Cblcos-treated cells. Subsequent immunocytochemistry ana- method). A cyclin-dependent kinase, CDK6, helps maintain lyses (Supplementary Figures 2I and II) show that ARHGEF6 a higher ROS level and its reduced expression leads to interacts with DNA-PKc mostly in the cytoplasm of Cblcos cell-cycle arrest and promotes apoptosis.22 Although CDK6 cells, whereas both proteins are predominantly present in the is highly expressed in only Cbl-treated cells, an almost nucleus of Cbl-treated cells. DNA-PKc with impaired repair threefold reduction in the expression level of CDK6 at 4 activity30 may further involve ARHGEF6 for inhibition of and 9 h in Cblcos cells may mediate cell-cycle arrest and tumorigenicity and regulate apoptosis in Cblcos-treated cells. apoptosis (Figure 2a). p53, an apoptotic marker, and its transcript and protein TSPYL4 is a nucleosome assembly protein and may have a levels are twofold elevated in Cblcos-treated cells than in role in gene regulation by histone accessibility.23 However, Cbl-treated cells (Figures 3c and d). Moreover, p53 interacts increased expression of TSPYL4 in Cblcos-treated cells may with DNA-PKc in both Cbl- and Cblcos-treated cells have an unknown significance that remains to be explored. (Figure 3g). Subsequent immunocytochemistry analyses A death-effective domain-containing protein, CLARP2, show that p53 protein is extruded from the nucleus to the interacts with caspase-8 and regulates apoptosis.24 This cytoplasm (Supplemetary Figures 2III and IV) in Cblcos- stress-sensitive gene shows a decreased expression of four- induced cells, whereas in Cbl-treated cells, the p53 protein is fold at 4 h and twofold at 9 h (Figure 2a) in Cblcos-treated mostly localized in the nucleus. Regardless of the mecha- cells than in Cbl-treated cells and implicates its role in nism, the differential expressions of p53-dependent genes apoptotic signaling as observed in other cells.24 (TIEG1, CACNA2D2, TRAF3 and the recently described RYBP, a ryanodine-binding protein, bears a death effector DUSP231) are prominent in Cblcos-treated cells. In fact, at domain (DED) that is crucial for apoptotic signaling. 1 h treatment, a significant increase in DUSP2/PAC1 (20- Increase in RYBP can trigger DEDAF (DED-associated fold), CACNA2D2 (twofold) and TRAF3 (threefold), and a factor)-mediated apoptosis.25 Its expression level is twofold decrease in TIEG1 (1.6-fold) transcripts (Figure 2b) in Cblcos higher at 9 h in Cbl-, COS- and Cblcos-treated cells than in cells but not in Cbl-treated cells indicate early signaling of A2780 and A2780/100 cells. Its higher expressions in Cblcos- apoptosis in Cblcos cells.

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 99

CDK6 TSPYL4 1.4 5 60 DUSP2 50 1.2 4 40 3 0.8 30 2 20 0.4 10 1 0 0 0 10 CLARP2 RYBP 8 6 1.8 TIEG1 1.6 5 6 4 1.2 4 3 0.8 2 2 0.4 0 1 5 0 0 CACNA2D2 4 cos4 cos9 CYR61 A27 3 5 A27 /100 2 4 1 A27 /100 - Cbl4 A27 /100 - Cbl9 3 A27 /100 - cos4 A27 /100 - Cbl A27 /100 - Cbl 0 2 3 1 TRAF3 25 0 2 15 cos4 cos9 A27 chip value 1

A27 /100 qPCR fold change 05 0 A27 /100 - co84 A27 /100 - Cbl4 A27 /100 - Cbl9 -1h -4h -9h -12h A27 /100 - Cbl A27 /100 - Cbl cos cos cos cos A27-NT A27/100-NT A27/100-Cbl-4h A27/100-Cbl-1h A27/100-Cbl-9h A27/100-cos-4h A27/100-cos-9h A27/100-Cbl-12h A27/100-Cbl A27/100-Cbl A27/100-Cbl A27/100-Cbl Figure 2 Validation of gene expressions and expression of chronic oxidative stress (COS)-responsive genes by real-time PCR. (a) Validation of microarray expression by real-time PCR. CDK6, TSPYL4, CLARP2, CYR61 and DEDAF/RYBP transcripts are analyzed from cDNA of A2780, A2780/100 and A2780/100 with untreated, COS-treated, chlorambucil (Cbl)-treated and Cblcos-treated cells. General trend of expression for each gene by QPCR (represented as fold change) is in good agreement with the microarray expression (intensity of hybridization). (b) Expression of these genes changes significantly in COS treatment and shows significant difference in Cblcos-treated cells than in Cbl-treated cells. The expression pattern of these genes is studied at extended time points (1, 4, 9 and 12 h) and their expression pattern favors apoptosis during the early stages of Cblcos treatment. DUSP2/PAC1 expresses 3- to 4-fold higher only after COS treatment in A2780 cells. Compared with Cbl-treated cells, Cblcos cells show a 9- to 10-fold higher expression at 1 h and reduce gradually at later time points. In A2780/100-resistant cells, in excess of COS, TIEG1 expression increases. Strikingly, its expression reduces 1.5- and 3-fold in Cblcos-treated cells at 1 and 12 h, respectively, than in Cbl-treated cells. At a certain level of cellular reactive oxygen species (ROS), this gene expression reduces enough to induce apoptosis. CACNA2D2 expression increases in only COS-treated cells (3- to 5-fold). However, this stress-sensitive gene expresses 2- to 3-fold higher at 1 h in Cblcos-treated cells than in Cbl-treated cells. A higher expression of this gene has been earlier shown to link with CA2 þ changes in mitochondria and subsequently apoptotic signaling. TRAF3 could also generate ROS through nicotinamide adenine dinucleotide phosphate (NADPH) oxidase activity and induce JNK-mediated apoptosis. However, in excess of COS, A2780/100 cells show reduced expression. However, in Cblcos-treated cells at 1 h, its expression increases 2- to 3-fold than that in Cbl-treated cells and may have an important role in apoptosis.

In Cblcos-treated cells, both mRNA and protein (Figures 4a sion of this gene significantly decreases in Cblcos-treated and b) levels of a proapoptotic transcription factor, NF-kB2, cells at 4 h and nominally at 9 h and its protein levels are also increased significantly compared with that in decrease in Cblcos-treated cells than in Cbl-treated cells Cbl-treated cells. Its upregulation in Cblcos-treated cells may (Figures 4c and d) continuously until 9 h. Thus, decrease of exert the synergistic effect of Cbl and COS that favors mRNA and protein levels may facilitate the inhibition of apoptotic signaling. HSP70 transports apoptotic proteins to antiapoptotic activity in Cblcos-treated cells and negatively mitochondria and has antiapoptotic activity.32 The expres- regulate apoptosis.

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 100

ARHGEF6 ARHGEF6 4 4 3 ARHGEF6 3 Tubulin 2 β-Actin 2 1 1 Arbitrary unit

0 - 1h - 4h 0 cos cos A27 - NT - 4h - 9h - 1h - 4h cos cos cos cos A27/100 - NT A27 - Cbl - 1h A27 - Cbl - 4h A27 - Cbl A27 - NT A27 - NT A27/100 - Cbl - 1h A27/100 - Cbl - 4h A27/100 - Cbl A27/100 - NT A27 -Cbl - 1h A27 -Cbl - 4h A27 -Cbl A27/100 - NT A27 - Cbl - 1h A27 - Cbl - 4h A27 - Cbl - 9h A27 - Cbl A27/100 - Cbl A27/100 - Cbl A27/100-Cbl - 1h A27/100-Cbl A27/100 - Cbl - 4h A27/100 - Cbl A27/100 - Cbl - 9h A27/100 - Cbl A27/100 - cos 4h A27/100 - cos 9h A27/100 - Cbl - 4h A27/100 - Cbl A27/100 - Cbl A27/100 - Cbl A27/100 - Cbl A27/100 - Cbl

p53 3 p53 2.5 2 p53 2 1.5 β-Actin 1 1 NT 0.5 Arbitrary unit 0 - 4h - 9h

0 - 12h cos cos NT Cbl - 4h Cbl - 9h Cbl cos - 4h - 9h Cbl - 12h Cbl - 12h NT Cbl Cbl cos cos Cbl - 4h - 9h Cbl - 4h Cbl - 9h Cbl cos - 12h Cbl - 12h Cbl cos cos Cbl Cbl Cbl - 4h Cbl - 9h Cbl cos Cbl Cbl - 12h Cbl Cbl Cbl Cbl

M Cbl Control Cblgo Control

M ARHGEF6 Control M

DNA - PKC Control DNA - PKC 50kd p53 75 kd ARHGEF6 250kd 50kd IgG IgG 75kd AEHGEF6 50 kd 50kd IgG

Figure 3 Protein analysis of p53, DNA-PKc and their interaction with ARHGEF6. (a) A reduced ARHGEF6 expression is observed in A2780-sensitive cells compared with chlorambucil (Cbl)-treated cells from 1 to 4 h. Moreover, in resistant A2780/100 cells, its expression is 2- to 3-fold higher than in A2780-sensitive cells. ARHGEF6 also shows chronic oxidative stress (COS) sensitivity, as its transcript level increases in only COS-treated cells. However, in Cblcos-treated cells, its expression is approximately twofold reduced than in Cbl-treated cells. (b) Protein level is almost similar in A2780 and A2780/100 cells but reduced in Cblcos-treated cells than in Cbl-treated cells. Cbl-sensitive A2780 cells also have less amount of protein than nontreated A2780 cells. (c and d) p53 RNA and protein expressions are shown in A27/100-resistant cells. p53 is increased in Cblcos-treated cells in the first hour of Cbl and COS treatment. (e) ARHGEF6 (73 kDa) protein is co-precipitated with DNA-PKc antibody in Cblcos-treated cells. (f) DNA-PKc ( ¼ 350 kDa) is co-precipitated with ARHGEF6 antibody in Cblcos-treated cells. (g) p53 (53 kDa) is pulled down with DNA-PKc antibody in both Cbl- and Cblcos-treated cells.

Discussion activity of DNA repair pathways or may be synergistic with drugs in promoting cell death through the activation of Understanding the mechanisms by which tumor cells main- proapoptotic p53.8 COS, along with Cbl, is expected to tain resistance is important for improving the therapeutic increase or sustain DNA damage in A2780/100 cells. management of human malignancies. In this study, I report Noncytotoxic doses of H2O2 have the potential to confer that chronic nontoxic doses of ROS can selectively increase its cytotoxic effect in two ways: (1) H2O2 may directly confer drug efficacy and reverse multidrug resistance of ovarian its cytotoxic activity with other stronger ROS (see review, see A2780/100 carcinoma cells, although Cbl or COS alone is Ferriero36) or may increase the penetration and accumula- unable to sensitize Cbl-resistant cells. Sensitization to Cbl tion of Cbl into Cblcos cells; (2) extensive DNA damage may with COS has not been reported earlier, however, moderate be responsible for the inhibition of tumorigenicity and may levels of ROS are shown to be important in redox regulation regulate apoptosis. However, my observations of a differ- and activation of caspases concomitantly in apoptosis.33–35 ential gene expression of DNA damage-inducible and repair Although the role of ROS is unknown, it could affect the genes, cell growth retardation signaling genes and apoptotic

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 101

p52 HSP70

8 4 6 3 4 2 1 2 0 0 - 4h - 9h - 9h - 9h - 4h - 9h

cos cos A27-NT cos cos A27-NT A27/100 - NT A27-Cbl - 1h A27-Cbl A27-Cbl - 4h A27-Cbl - 9h A27-Cbl A27/100 - NT A27/100 - Cbl-4h A27/100 - Cbl-9h A27/100 - cos-4h A27/100 - cos-9h A27/100 - Cbl-4h A27/100 - cos-4h A27/100 - Cbl A27/100 - cos A27/100 - Cbl A27/100 - Cbl A27/100 - Cbl A27/100 - Cbl

p52 HSP70 β-ACTIN β-ACTIN - 4h - 9h - 4h - 9h - 12h - 12h cos cos cos cos Cbl - 4h Cbl - 9h Cbl - 4h Cbl - 9h Cbl cos cos Cbl - 12h Cbl - 12h Cbl Cbl Cbl Cbl Cbl Cbl Cbl A27/100 - NT A27/100 - NT

HSP70 p52 3 4

3 2 2 1

1 Arbitrary Unit Arbitrary Unit 0 0 - 4h - 9h - 4h - 9h - 12h - 12h Cbl-4h Cbl-9h cos cos Cbl-4h Cbl-9h Cbl-12h cos cos cos Cbl-12h cos Cbl Cbl Cbl Cbl Cbl Cbl A27/100 - NT A27/100 - NT Figure 4 RNA and protein analysis of NF-kB2/p52 and Hsp70 in chlorambucil (Cbl)- and Cblcos-treated cells. (a) An increased expression of the p52 transcript is observed in A2780/100-untreated cells than in A2780 cells. However, its expression is always increased in Cblcos-treated cells than in Cbl-treated cells. At 9 h, it specifically increases almost twofold in Cblcos-treated cells. (b) Western blot analysis and bar chart show that a higher amount of active subunit p52 protein is present in Cblcos-treated cells than in Cbl-treated cells. (c and d) HSP70 gene expression significantly differs in Cbl- and Cblcos-treated cells. At 4 h, its expression is less in Cblcos-treated cells than in Cbl-treated cells. However, at 9 h, the expression does not decrease significantly, suggesting an earlier signaling of apoptosis. genes support later explanations. Moreover, three major generally occurs through mitochondria,37 although more cellular events are necessary for cell death in Cblcos cells: (1) evidence is needed to propose a common pathway. excessive DNA damage and breakdown of genomic integrity; Chlorambucil-induced alkylation of DNA results predo- (2) inhibition of oncogenic progress, that is, to stop minantly in N7-guanine adducts and yields apurinic proliferation and cell adhesion, suppression of angiogenic sites, which in turn can cause strand breaks and breaks processes, growth retardation and arrest; (3) signaling to of cross-links.38 Repair of DNA cross-links may involve apoptosis. A diagram depicting the probable mechanism of double-stranded break intermediates through a nonhomo- cellular death by Cblcos is constructed using IPA analysis logous end-joining process, and it require the DNA–PKcs (Figure 1b). It is more likely that one or few master complex, a catalytic subunit of DNA-PKc, the Ku70 and regulatory genes are more activated in CBlcos cells compared Ku86 heterodimers. Earlier it was shown that DNA–PKcs do to Cbl cells and exert its effect on apoptotic and death not change significantly after Cbl treatment until 3 h of pathway genes to signal apoptosis. Hence, the expression treatment, but reduction of kinase activity was observed in differences of many genes in these two cells may be more of Cblcos-treated drug-sensitive cells.15 It is possible that after a secondary effect rather than of a regulatory effect. Several COS treatment in Cbl-treated drug-resistant cells, DNA–PKcs proteins involved in apoptotic signaling pathways (NF-kB2, catalytic subunit is degraded by caspases,29 as previously ARHGEF6, TRAF3 and TIEG1 for JNK, BCL2L1, CACNA2D2; shown in other tumor cells. Figures 2a and b, 3, 4) and cell death-mediated pathways Reduced expression of ARHGEF6, CDK6 and TIEG1 at early (caspase-8, caspase-9, FADD, MORT) are differentially regu- stage and of CYR61 at late stage may also contribute toward lated. However, recently it has been suggested that apoptosis the inhibition of proto-oncogenic activity. Very recently,

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 102

downregulation of cyr61, which is shown to potentiate the depletion of HSP70 induces apoptosis.49 Changes in mito- anticancer activity of zoledronic acid in prostate cancer,39 chondrial membrane permeability with the Ca2 þ pool of justifies its reduction in Cbl and Cblcos cells. ARHGEF6, a G- inner mitochondria and cytochrome c release are a necessary coupled receptor activated by phosphatidylinositol 3-kinase, prerequisite for apoptotic signaling. A subsequent increase in modulates the cellular ROS level through Rac1. It also CACNA2D2, a mitochondrial Ca2 þ pool regulator,50 and a modulates ROS production from nicotinamide adenine decrease in BCL2L1 in transcript and protein levels (data not dinucleotide phosphate (NADPH) oxidase activity in a shown) may explain apoptotic signaling to mitochondria. growth factor-dependent manner,28 and downregulation of More downstream of this cell death pathway, increased ARHGEF6 inhibits cell spreading and adhesion.27 Reduction expression of Caspase-8, FADD (Supplementary Table 1) and of expression in Cblcos may act downstream of DNA repair MORT (1.7-fold) transcripts in Cblcos cells over Cbl-treated signaling and help to signal cells to maintain excess ROS, to cells also favors apoptosis. However, overexpression or stop proliferation and further adhesion of cells. Physical silencing BCL-xl does not improve Cbl sensitivity in interaction of DNA-PKc with ARHGEF6, implicating exten- A2780/100-resistant cells.2 Similarly, BCL2l1 overexpression sive DNA damage, may signal for cell-cycle inhibition, does not decrease etoposide-induced sensitivity in thymic spreading and apoptosis regulation in Cblcos-treated cells. lymphoma cells.51 Nevertheless, no deamidation52 of Interaction of DNA-PKc with p53 has also been earlier shown BCL2L1 or any breakdown of the protein is observed in in fibroblast and glioma cells.40,41 However, interaction of these cell lines with or without Cbl and Cblcos treatment. DNA-PKc with p53 is observed in both the nucleus and the Suppression of deamidation can also occur, although the cytoplasm of Cblcos cells, but only in the nucleus of Cbl- Bcl2l1 protein has a critical role in apoptosis in the hepatoma treated cells. Although both are predominantly nuclear cell line.53 Nevertheless, BCL2L1 may have an important proteins, it is possible that both are extruded from the role preceding apoptosis in Cblcos cells, but by modulating nucleus in response to COS treatment and will move to the downstream of apoptotic genes may not reverse drug cytoplasm or more precisely to the mitochondria. Several sensitivity. Furthermore, dysregulated expression of channel reports suggest that DNA-PKc exists in the cytoplasm in proteins (Supplementary Figure 2B), especially ATP-depen- mitotic phase.42,43 Although p53 is predominantly a nuclear dent membrane proteins, may have a significant role in Cbl protein, it also travels from the nucleus to the mitochondria sensitivity in Cblcos cells. The association of drug resistance in response to death stress.44 Moreover, it seems that, in with P-gp (p-glycoprotein) and MRP (MDR-associated pro- dividing Cblcos cells, DNA-PKc is more in the nucleus than in tein) remains to be tested, as they modulate drug sensitivity the cytoplasm, whereas in dying cells (in which the nucleus in some resistant cancer cells.54,55 In contrast, modulation of appears fragmented, a sign of late-stage apoptosis), it is ROS-sensitive guanine exchange factor (ARHGEF6) may hold present more in the cytoplasm (data not shown). The much the key for increasing sensitivity of Cbl in A2780/100- lesser expression of cyclin-dependant kinase, CDK6, in Cblcos resistant cells. In fact, preliminary results show that siRNA cells also justifies apoptotic signaling in these cells. The knockdown in resistant cells increases the sensitivity of reduced expression of CDK6 inhibits cell-cycle progression resistant cells (data not shown). Nevertheless, low doses and proliferation45 in erythrocyte progenitor cells, support- (100 mM) of Cbl (15 mg daily56 or 0.1 mg kgÀ1 body weight for ing the underlying mechanism that CDK6 action may be 4days57 is mostly used in patients) with very low doses similar for these two types of cells. (10 ng mlÀ1, 0.0026 U mlÀ1; for cytotoxicity, it is generally The DNA damage signaling for apoptosis can also be used in high doses such as 0.1–1 U mlÀ158 and in rat attributed to a higher expression of p52, a TFIIH transcrip- hepatocytes 0.4 U mlÀ159) of COS are useful for increasing tion factor having a death domain that is involved in drug sensitivity in A2780 ovarian carcinoma cells. Successful transcription of proapoptotic proteins and in activation of potential application of COS in conjunction with Cbl may cell death.46,47 The higher level of NF-kB2 transcripts and hold potential for inhibition of tumor progression and can proteins in Cblcos cells than in Cbl-treated cells may have a be tested in a mouse model for subsequent development of significant role in activating apoptosis. chemotherapy of ovarian cancer. Increase in TRAF3, a TNF receptor molecule, can generate ROS through NADPH oxidase and induce JNK-mediated Conflict of interest apoptosis. TRAF3 can bind to nucleoporin in the nuclear membrane, facilitating the transport of the signaling mole- The author declares no conflict of interest. cule either way.48 In A2780/100 cells, TRAF3 transcripts are increased under Cblcos treatment and increase of its expres- sion at 1 h may be significant for the translocation of the Abbreviations signaling molecule across those membranes. Subsequently, HSP70 sequesters antiapoptotic proteins (Bcl2, bcl2l1) in the ERCC-4 excision repair cross-complementing rodent repair deficiency 32 gene, complementation group 4 cytoplasm and transports them to the mitochondria. BCNU bis-chloronitrosourea, a generic anticancer drug named Reduction of its expression in mRNA and protein levels also Carmustine supports its role in the activation of cell death in Cblcos- QPCR quantitative PCR treated cells and suggests that its role in drug resistance GSH glutathione s.e.m. standard error of the mean is similar to that of pancreatic cancer cells, in which

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 103

Acknowledgments 20 Mates JM, Sanchez-Jimenez FM. Role of reactive oxygen species in apoptosis: implications for cancer therapy. Int J Biochem Cell Biol 2000; I thank Dr I Boldogh for gifting A2780 and A2780/100 cells; 32: 157–170. 21 Le Romancer M, Cosulich SC, Jackson SP, Clarke PR. Cleavage and Dr G Das for standardizing COS treatment; Dr M Sinha in the inactivation of DNA-dependent catalytic subunit during bioinformatics division for analyzing microarray data; Silicon apoptosis in Xenopus egg extracts. J Cell Sci 1996; 109: 3121–3127. Genetics and Spotfire Inc. for using GeneSpring and Spotfire 22 Macip S, Igarashi M, Berggren P, Yu J, Lee SW, Aaronson SA. Influence software, respectively. I thank anonymous reviewers for critical of induced reactive oxygen species in p53-mediated cell fate decisions. comments for improving the paper. Mol Cell Biol 2003; 23: 8576–8585. 23 Strausberg RL, Feingold EA, Grouse LH, Derge JG, Klausner RD, Collins FS et al. Generation and initial analysis of more than 15 000 full-length human and mouse cDNA sequences. Proc Natl Acad Sci USA 2002; 99: References 16899–16903. 24 Inohara N, Koseki T, Hu Y, Chen S, Nunez G. CLARP, a death effector 1 Horton JK, Roy G, Piper JT, Van Houten B, Awasthi YC, Mitra S et al. domain-containing protein interacts with caspase-8 and regulates Characterization of a chlorambucil-resistant human ovarian carcinoma apoptosis. Proc Natl Acad Sci USA 1997; 94: 10717–10722. cell line overexpressing glutathione S-transferase mu. Biochem Pharma- 25 Schickling O, Stegh AH, Byrd J, Peter ME. Nuclear localization of DEDD col 1999; 58: 693–702. leads to caspase-6 activation through its death effector domain and 2 Roy G, Horton JK, Roy R, Denning T, Mitra S, Boldogh I. Acquired inhibition of RNA polymerase I dependent transcription. Cell Death alkylating drug resistance of a human ovarian carcinoma cell line is Differ 2001; 8: 1157–1168. unaffected by altered levels of pro- and anti-apoptotic proteins. 26 Messina E, Lupi F, Barile L, Giacomello A. Cyclic nucleotides and Oncogene 2000; 19: 141–150. neuroblastoma differentiation. Nucleosides Nucleotides Nucleic Acids 3 Mitra S, Izumi T, Boldogh I, Bhakat KK, Hill JW, Hazra TK. Choreography 2004; 23: 1551–1554. of oxidative damage repair in mammalian genomes. Free Radic Biol Med 27 Zegers MM, Forget MA, Chernoff J, Mostov KE, ter Beest MB, Hansen 2002; 33: 15–28. SH. Pak1 and PIX regulate contact inhibition during epithelial wound 4 Ruller S, Stahl C, Kohler G, Eickhoff B, Breder J, Schlaak M et al. healing. EMBO J 2003; 22: 4155–4165. Sensitization of tumor cells to ribotoxic stress-induced apoptotic cell 28 Park HS, Lee SH, Park D, Lee JS, Ryu SH, Lee WJ et al. Sequential death: a new therapeutic strategy. Clin Cancer Res 1999; 5: 2714–2725. activation of phosphatidylinositol 3-kinase, beta Pix, Rac1, and Nox1 in 5 Fang J, Seki T, Maeda H. Therapeutic strategies by modulating oxygen growth factor-induced production of H2O2. Mol Cell Biol 2004; 24: stress in cancer and inflammation. Adv Drug Deliv Rev 2009; 61: 4384–4394. 290–302. 29 Srivastava RK, Srivastava AR, Cho-Chung YS, Longo DL. Synergistic 6 Morel Y, Barouki R. Repression of gene expression by oxidative stress. effects of retinoic acid and 8-Cl-cAMP on apoptosis require caspase-3 Biochem J 1999; 342: 481–496. activation in human ovarian cancer cells. Oncogene 1999; 18: 7 Keyse SM. Protein phosphatases and the regulation of mitogen- 1755–1763. activated protein kinase signalling. Curr Opin Cell Biol 2000; 12: 30 Bacsi A, Kannan S, Lee MS, Hazra TK, Boldogh I. Modulation of 186–192. DNA-dependent protein kinase activity in chlorambucil-treated cells. 8 Polyak K, Xia Y, Zweier JL, Kinzler KW, Vogelstein B. A model for p53- Free Radic Biol Med 2005; 39: 1650–1659. induced apoptosis. Nature 1997; 389: 300–305. 31 Yin Y, Liu YX, Jin YJ, Hall EJ, Barrett JC. PAC1 phosphatase is a trans- 9 Laurent A, Nicco C, Chereau C, Goulvestre C, Alexandre J, Alves A et al. cription target of p53 in signalling apoptosis and growth suppression. Controlling tumor growth by modulating endogenous production of Nature 2003; 422: 527–531. reactive oxygen species. Cancer Res 2005; 65: 948–956. 32 Cardile V, Scifo C, Russo A, Falsaperla M, Morgia G, Motta M et al. 10 Kong Q, Beel JA, Lillehei KO. A threshold concept for cancer therapy. Involvement of HSP70 in resveratrol-induced apoptosis of human Med Hypotheses 2000; 55: 29–35. prostate cancer. Anticancer Res 2003; 23: 4921–4926. 11 Mookerjee A, Basu JM, Majumder S, Chatterjee S, Panda GS, Dutta P 33 Vafa O, Wade M, Kern S, Beeche M, Pandita TK, Hampton GM et al. et al. A novel copper complex induces ROS generation in doxorubicin c-Myc can induce DNA damage, increase reactive oxygen species, and resistant Ehrlich ascitis carcinoma cells and increases activity of mitigate p53 function: a mechanism for oncogene-induced genetic antioxidant in vital organs in vivo. BMC Cancer 2006; 6: 267. instability. Mol Cell 2002; 9: 1031–1044. 12 Mookerjee A, Mookerjee Basu J, Dutta P, Majumder S, Bhattacharyya S, 34 Mates JM, Segura JA, Alonso FJ, Marquez J. Intracellular redox status and Biswas J et al. Overcoming drug-resistant cancer by a newly developed copper chelate through host-protective cytokine-mediated apoptosis. oxidative stress: implications for cell proliferation, apoptosis, and Clin Cancer Res 2006; 12: 4339–4349. carcinogenesis. Arch Toxicol 2008; 82: 273–299. 13 Vermes I, Haanen C, Steffens-Nakken H, Reutelingsperger C. A novel 35 Hayden RE, Pratt G, Davies NJ, Khanim FL, Birtwistle J, Delgado J et al. assay for apoptosis. Flow cytometric detection of phosphatidylserine Treatment of primary CLL cells with bezafibrate and medroxyproges- expression on early apoptotic cells using fluorescein labelled Annexin V. terone acetate induces apoptosis and represses the pro-proliferative J Immunol Methods 1995; 184: 39–51. signal of CD40-ligand, in part through increased 15dDelta12,14,PGJ2. 14 Das GC, Bacsi A, Shrivastav M, Hazra TK, Boldogh I. Enhanced gamma- Leukemia 2009; 23: 292–304. glutamylcysteine synthetase activity decreases drug-induced oxidative 36 Ferriero DM. Oxidant mechanisms in neonatal hypoxia–ischemia. stress levels and cytotoxicity. Mol Carcinog 2006; 45: 635–647. Dev Neurosci 2001; 23: 198–202. 15 Boldogh I, Roy G, Lee MS, Bacsi A, Hazra TK, Bhakat KK et al. Reduced 37 Ricci JE, Munoz-Pinedo C, Fitzgerald P, Bailly-Maitre B, Perkins GA, DNA double strand breaks in chlorambucil resistant cells are related to Yadava N et al. Disruption of mitochondrial function during apoptosis is high DNA-PKcs activity and low oxidative stress. Toxicology 2003; 193: mediated by caspase cleavage of the p75 subunit of complex I of the 137–152. electron transport chain. Cell 2004; 117: 773–786. 16 Anderson ME. Glutathione: an overview of biosynthesis and modula- 38 Lawley PD, Phillips DH. DNA adducts from chemotherapeutic agents. tion. Chem Biol Interact 1998; 111–112: 1–14. Mutat Res 1996; 355: 13–40. 17 Davies KJ. The broad spectrum of responses to oxidants in proliferating 39 Marra M, Santini D, Meo G, Vincenzi B, Zappavigna S, Baldi A et al. cells: a new paradigm for oxidative stress. IUBMB Life 1999; 48: 41–47. CYR61 down-modulation potentiates the anti-cancer effects of zole- 18 Panayiotidis M, Tsolas O, Galaris D. Glucose oxidase-produced H2O2 dronic acid in androgen-independent prostate cancer cells. Int J Cancer induces Ca2+-dependent DNA damage in human peripheral blood 2009; 125: 2004–2013. lymphocytes. Free Radic Biol Med 1999; 26: 548–556. 40 Karpinich NO, Tafani M, Rothman RJ, Russo MA, Farber JL. The course of 19 Casentini-Borocz D, Bringman T. Enzyme immunoconjugates utilizing etoposide-induced apoptosis from damage to DNA and p53 activation glucose oxidase and myeloperoxidase are cytotoxic to Candida to mitochondrial release of cytochrome c. J Biol Chem 2002; 277: tropicalis. Antimicrob Agents Chemother 1990; 34: 875–880. 16547–16552.

The Pharmacogenomics Journal Cbl with COS kills Cbl-resistant ovarian carcinoma cells AK Maiti 104

41 Jack MT, Woo RA, Motoyama N, Takai H, Lee PW. DNA-dependent alterations of the intracellular calcium signaling and disruption of protein kinase and checkpoint kinase 2 synergistically activate a mitochondria membrane integrity. Oncogene 2003; 22: 615–626. latent population of p53 upon DNA damage. J Biol Chem 2004; 279: 51 Vilenchik M, Raffo AJ, Benimetskaya L, Shames D, Stein CA. Antisense 15269–15273. RNA down-regulation of bcl-xL expression in prostate cancer cells leads 42 Koike M, Shiomi T, Koike A. Ku70 can translocate to the nucleus to diminished rates of cellular proliferation and resistance to cytotoxic independent of Ku80 translocation and DNA-PK autophosphorylation. chemotherapeutic agents. Cancer Res 2002; 62: 2175–2183. Biochem Biophys Res Commun 2000; 276: 1105–1111. 52 Deverman BE, Cook BL, Manson SR, Niederhoff RA, Langer EM, Rosova I 43 Nilsson A, Sirzen F, Lewensohn R, Wang N, Skog S. Cell cycle- et al. Bcl-xL deamidation is a critical switch in the regulation of the dependent regulation of the DNA-dependent protein kinase. Cell Prolif response to DNA damage. Cell 2002; 111: 51–62. 1999; 32: 239–248. 53 Takehara T, Takahashi H. Suppression of Bcl-xL deamidation in human 44 Mahyar-Roemer M, Katsen A, Mestres P, Roemer K. Resveratrol induces hepatocellular carcinomas. Cancer Res 2003; 63: 3054–3057. colon tumor cell apoptosis independently of p53 and precede by 54 Zhang D, Fan D. Multidrug resistance in gastric cancer: recent research epithelial differentiation, mitochondrial proliferation and membrane advances and ongoing therapeutic challenges. Expert Rev Anticancer potential collapse. Int J Cancer 2001; 94: 615–622. Ther 2007; 7: 1369–1378. 45 Dai C, Chung IJ, Jiang S, Price JO, Krantz SB. Reduction of cell cycle 55 Nooter K, Stoter G. Molecular mechanisms of multidrug resistance in progression in human erythroid progenitor cells treated with tumour cancer chemotherapy. Pathol Res Pract 1996; 192: 768–780. necrosis factor alpha occurs with reduced CDK6 and is partially reversed 56 Jaksic B, Brugiatelli M, Krc I, Losonczi H, Holowiecki J, Planinc-Peraica A by CDK6 transduction. Br J Haematol 2003; 121: 919–927. et al. High dose chlorambucil versus Binet’s modified cyclophos- 46 Hacker H, Karin M. Is NF-kappaB2/p100 a direct activator of phamide, doxorubicin, vincristine, and prednisone regimen in the programmed cell death? Cancer Cell 2002; 2: 431–433. treatment of patients with advanced B-cell chronic lymphocytic 47 Han DC, Lee MY, Shin KD, Jeon SB, Kim JM, Son KH et al. leukemia. Results of an international multicenter randomized trial. 20-Benzoyloxycinnamaldehyde induces apoptosis in human carcinoma International Society for Chemo-Immunotherapy, Vienna. Cancer 1997; via reactive oxygen species. J Biol Chem 2004; 279: 6911–6920. 79: 2107–2114. 48 Gamper C, van Eyndhoven WG, Schweiger E, Mossbacher M, Koo B, 57 Silvennoinen R, Malminiemi K, Malminiemi O, Seppala E, Vilpo J. Lederman S. TRAF-3 interacts with p62 nucleoporin, a component Pharmacokinetics of chlorambucil in patients with chronic lymphocytic of the nuclear pore central plug that binds classical NLS-containing leukaemia: comparison of different days, cycles and doses. Pharmacol import complexes. Mol Immunol 2000; 37: 73–84. Toxicol 2000; 87: 223–228. 49 Aghdassi A, Phillips P, Dudeja V, Dhaulakhandi D, Sharif R, Dawra R 58 Philpott GW, Bower RJ, Parker KL, Shearer WT, Parker CW. Affinity et al. Heat shock protein 70 increases tumorigenicity and inhibits cytotoxicity of tumor cells with antibody-glucose oxidase conjugates, apoptosis in pancreatic adenocarcinoma. Cancer Res 2007; 67: peroxidase, and arsphenamine. Cancer Res 1974; 34: 2159–2164. 616–625. 59 Starke PE, Farber JL. Endogenous defenses against the cytotoxicity of 50 Carboni GL, Gao B, Nishizaki M, Xu K, Minna JD, Roth JA et al. hydrogen peroxide in cultured rat hepatocytes. J Biol Chem 1985; 260: CACNA2D2-mediated apoptosis in NSCLC cells is associated with 86–92.

Supplementary Information accompanies the paper on the The Pharmacogenomics Journal website (http://www.nature.com/tpj)

The Pharmacogenomics Journal