ABL Transcripts at the T315I Point Mutation by Dnazyme Promotes Apoptotic Cell Death in Imatinib-Resistant BCR–ABL Leukemic Cells

ORIGINAL ARTICLE Cleavage of BCR–ABL transcripts at the T315I point mutation by DNAzyme promotes apoptotic cell death in imatinib-resistant BCR–ABL leukemic cells

JE Kim1,7, S Yoon1,7, B-R Choi1, KP Kim2, Y-H Cho3, W Jung4, D-W Kim5,SOh6 and D-E Kim1

The BCR–ABL fusion transcript encodes the BCR–ABL tyrosine kinase (TK), which causes chronic myelogenous leukemia (CML). Although the TK inhibitor imatinib mesylate, which targets the BCR–ABL protein, has been proven to be effective in controlling leukemic growth, imatinib resistance has been observed with disease relapse because of point mutations in the ABL gene that inhibit imatinib efficacy. In this study, we designed oligodeoxyribozymes (DNAzymes) that specifically target and cleave both the junction sequence and the site of the point mutation (T315I), conferring imatinib resistance in BCR–ABL mRNA. DNAzymes significantly induced apoptosis and inhibited proliferation in wild-type and T315I-mutant BCR–ABL-positive cells. Selective cleavage of T315I-mutant ABL mRNA by DNAzyme (T315I Dz) led to cell cycle arrest in G0/G1 phase, with induction of caspase-3/-7 in imatinib-resistant BCR–ABL-positive cells harboring the T315I mutation. Moreover, cotreatment with the DNAzyme targeting the T315I mutation and imatinib resulted in enhanced inhibition of proliferation and induction of apoptosis in T315I leukemic cells as compared with imatinib alone, thereby antagonizing imatinib resistance in CML cells bearing T315I-mutant BCR–ABL. Therefore, cleavage of T315I-mutant ABL mRNA by DNAzyme combined with imatinib treatment may be an alternative approach to overcoming imatinib resistance in leukemic cells.

Leukemia (2013) 27, 1650–1658; doi:10.1038/leu.2013.60 Keywords: BCR–ABL; DNAzyme; imatinib resistance

INTRODUCTION been shown that expression of p210BCR/ABL bestows resistance to Chronic myelogenous leukemia (CML) is a myeloproliferative apoptosis by delaying upstream procaspase-3 activation in 10 disorder that is associated with the presence of the Philadelphia transforming cells. chromosome carrying the chimeric BCR–ABL oncogenes, which is Imatinib mesylate, a competitive inhibitor that binds to the generated by a reciprocal translocation between chromosomes 9 ATP-binding site of BCR–ABL, is an antileukemic drug that was and 22, t(9;22).1,2 This translocation causes the fusion of two developed as a targeted therapy against the chimeric BCR–ABL TK normal genes, the 50-portion of the BCR gene in chromosome and is currently considered a ﬁrst-line agent for newly diagnosed 11–13 22 and the 30-portion of the ABL gene in chromosome 9.3 The ABL CML. However, despite its impressive efﬁcacy, disease relapse 14,15 gene, present in all tissues, encodes for a nonreceptor tyrosine has been observed after initial response to imatinib. Drug kinase (TK) that has an important role in signal transduction and resistance during imatinib treatment is mostly related to point the regulation of cell growth. However, little is known about the mutations occurring within the kinase domain of BCR–ABL at function of the BCR protein translated from the BCR gene.2 more than 40 different amino acid positions.16 These point BCR–ABL mRNA contains two types of possible junctions, b3a2 mutations may either disrupt interactions between imatinib and (composed of BCR exon 3 and ABL exon 2) and b2a2 (composed BCR–ABL or prevent the kinase from converting into the inactive of BCR exon 2 and ABL exon 2).4,5 Both of these mRNAs encode for conformation.17 To overcome this challenge, dasatinib, a SRC a protein of 210 kDa (p210BCR/ABL), which is localized in the (sarcoma)-family TK inhibitor, was developed; dasatinib shows a cytoplasm and which exhibits constitutively activated TK activity, broader spectrum of activity against many mutant forms of leading to uncontrolled cell proliferation with a reduced apoptotic BCR–ABL.14,18 Despite this, the T315I mutant of BCR–ABL has been response to mutagenic stimuli.6 The BCR–ABL TK expressed in cells reported to show resistance to both imatinib and dasatinib.11 The has been shown to mitigate surveillance of the G1–S phase T315I mutant contains a single-base change (C to U) in the ABL transition, thereby enhancing cell proliferation.7 In particular, gene, leading to replacement of the threonine (T) at position 315 BCR–ABL inhibits the activity of the Cdk inhibitor p21 and with an isoleucine (I). Mutations occurring in the ‘gatekeeper’ or decreases p27 expression, both of which are involved in important contact region (for example, T315I) of the BCR–ABL kinase are 8,9 checkpoints for the G1–S phase transition. In addition, it has known to cause resistance to most kinase inhibitors, including

1Department of Bioscience and Biotechnology, Konkuk University, Seoul Republic of Korea; 2WCU Program, Department of Molecular Biotechnology, Konkuk University, Seoul, Republic of Korea; 3Department of Internal Medicine, Konkuk University, Seoul, Republic of Korea; 4Department of Emergency Medicine, Kyunghee University Hospital at Gangdong, Seoul Republic of Korea; 5Division of Hematology, Seoul St. Mary’s Hospital, The Catholic University of Korea, Seoul, Republic of Korea and 6Department of Advanced Fermentation Fusion Science and Technology, Kookmin University, Seoul Republic of Korea. Correspondence: Professor D-E Kim, Department of Bioscience and Biotechnology, Konkuk University, 1 Hyawang-dong, Gwangin-gu, Seoul 143-701, Republic of Korea. E-mail: [email protected] 7These authors contributed equally to this work. Received 18 June 2012; revised 19 February 2013; accepted 20 February 2013; accepted article preview online 25 February 2013; advance online publication, 19 March 2013 A DNAzyme specific to imatinib-resistant BCR-ABL JE Kim et al 1651 19 imatinib, nilotinib, dasatinib and bosutinib. Thus, novel and RNA cleavage activities of DNAzymes were tested in reaction buffer (10 mM more effective therapeutic alternatives are needed to overcome MgCl2 and 20 mM Tris-HCl, pH 7.5) at 37 1C. The reaction was initiated by this ability of mutant kinases to evade the effects of current combining a trace amount of 32P-50-labeled substrate RNA, mixed with BCR–ABL kinase inhibitors used in CML patients. each unlabeled corresponding substrate (to yield a final concentration of RNA-cleaving deoxyribozyme (DNAzyme), which is capable of 100 nM RNA), and DNAzyme (1 mM final concentration). The 20-ml reaction mixture was incubated for 3 h and quenched by adding an equal volume sequence-specific cleavage of target mRNA, was discovered 20 of the gel-loading dye containing 25 mM Na2EDTA and 8 M urea. The through in vitro selection technology. The ‘10–23’ RNA- products were resolved by 12% (w/v) denaturing polyacrylamide gel cleaving DNAzyme can recognize RNA through the Watson– electrophoresis, and the product bands were visualized and quantified on Crick base pairing and can cleave its target at a phosphodiester a Cyclone PhosphorImager (Packard Instruments, Meriden, CT, USA). bond located between an unpaired purine (R) and pyrimidine (Y) (box in Figure 1). DNAzymes that target the BCR–ABL chimeric mRNA junction sequence have been developed, and their efficacy Cell culture, DNAzyme transfection and drug administration toward target RNA cleavage has been tested.21,22 In addition to BaF3 cells, a mouse pro-B cell line that is stably transduced with BCR–ABL DNAzymes, there have been several attempts to suppress the (BaF3/BCR–ABL) or drug-resistant mutant BCR–ABLT315I (BaF3/ 23,24 22 BCR–ABLT315I), were cultured in RPMI 1640 media containing 10% fetal expression of BCR–ABL with antisense DNA, ribozymes, bovine serum, 100 U/ml penicillin and 100 mg/ml streptomycin at 37 1C siRNA,25,26 peptide nucleic acid,27 and locked nucleic acid.28 All of under 5% CO2. Chemically synthesized DNAzymes were transfected into these antisense molecules are directed against BCR–ABL junctions, the cells by electroporation using the Neon Transfection System not the drug-resistant mutants occurring in the ABL gene. (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instruc- In this study, we designed DNAzymes that specifically target tions. T315I DNAzyme, BP DNAzyme and BP-T315I tandem DNAzyme were and cleave the junction sequence and/or the site of point used at concentrations of 4 mg (T315I and BP DNAzyme) and 10 mg mutation conferring T315I drug resistance in BCR–ABL mRNA. (tandem DNAzyme) per 5 Â 105 cells and electroporated at 1100 V with a The designed DNAzymes were introduced into cells harboring double pulse (30 ms pulse width). Following electroporation, cells were the BCR–ABL gene with or without the T315I mutation, and we resuspended in RPMI medium and further incubated at 37 1C for the designated times (usually 48 h). Percentage of cells transfected with the investigated the ability of these DNAzymes to inhibit the oligo DNA was estimated to be about 80% when the electroporation is expression of BCR–ABL and to restore imatinib sensitivity in cells applied to the cells (Supplementary Figure S1). harboring the drug-resistant leukemogenic gene.

Western blotting and quantitative real-time MATERIALS AND METHODS reverse-transcription PCR Oligonucleotides and reagents Cells were lysed in RIPA buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 0.1% The 10–23 oligodeoxyribozymes (shown in Figure 1) and DNA templates SDS, 0.5% deoxycholate, 1.0% NP40, 1 mM phenylmethanesulfonylﬂuoride) for BCR–ABL substrate RNA and for both wild-type and mutant for 1 h on ice. A total of 30 mg of protein was separated by gradient SDS- ABL substrate RNAs were chemically synthesized (Cosmo Gentech, polyacrylamide gel electrophoresis (4–12% polyacrylamide). Proteins were Seoul, Korea). DNA templates harbored the antisense sequence of the T7 then transferred to a nitrocellulose membrane (Whatman) and the promoter and the respective RNA sequence for the in vitro synthesis of membrane was immunoblotted with anti-c-ABL monoclonal (1:500; Santa RNAs using the T7 RNA polymerase.29 In vitro synthesis of RNAs was Cruz Biotechnology, Santa Cruz, CA, USA) and anti-b-actin antibodies performed as described previously.30 Imatinib mesylate STI571 (Gleevec) (1:1500; Santa Cruz Biotechnology). The membrane was subsequently was kindly provided by Novartis, Inc. (Basel, Switzerland). Imatinib stock incubated with horseradish peroxidase-conjugated anti-goat immunoglo- solution (10 mg/ml) was prepared by dissolving the compound in bulin, and signals were detected with an enhanced chemiluminescence system (Intron Biotechnology, Seoul, Korea). RNA isolation for reverse dimethylsulfoxide:H2O (1:1) and was then stored at À 20 1C. transcription of ABL mRNA was carried out as described.31 Primers for ampliﬁcation of T315I ABL transcripts were designed as follows: forward RNA cleavage with DNAzymes primer 50-GCC CCC GTT CTA TAT CAT CAT TGA-30 (the point mutation Substrate RNAs were labeled using T4 polynucleotide kinase and conferring the T315I mutation is underlined); reverse primer 50-TGG ATG [g-32P]ATP (6000 Ci/mmol, GE Healthcare, Piscataway, NJ, USA). Target AAG TTT TTC TTC TCC AG-30. For quantitative real-time PCR, the T315I

Figure 1. Sequences of the DNAzymes targeting BCR–ABL chimeric mRNA. The sequence of BCR–ABL near the junction between the BCR exon (b3) and ABL exon (a2) is shown. The T315I point mutation (C to U) is indicated in the ABL exon. Potential sites of cleavage by the constructed DNAzymes are indicated by arrows.

& 2013 Macmillan Publishers Limited Leukemia (2013) 1650 – 1658 A DNAzyme specific to imatinib-resistant BCR-ABL JE Kim et al 1652 ABL cDNA was amplified using the SYBR Green Premix Ex Taq Kit (TaKaRa, oligo dT(18) spacing nts (Figure 1). This tandem DNAzyme was Shiga, Japan). After 30 cycles at 95 1C for 5 s and 60 1C for 60 s, dissociation expected to cleave both the BCR–ABL junction and the T315I point curve analysis for the DCt was calculated using Thermal Cycler Dice Real mutation in the BCR–ABL mutant mRNA. Time System software (TaKaRa). Specific cleavage of ABL RNA containing the T315I point mutation Cell proliferation assay To examine the specificity of RNA cleavage by the designed Twenty-four hours after transfection of respective DNAzymes into the DNAzyme, we investigated its RNA-cleavage activity and specifi- cells, 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazo- city using short RNA substrates, such as chimeric BCR–ABL RNA, lium (WST-1 reagent; Roche, Mannheim, Germany) was added (10 ml per wild-type ABL RNA and mutant ABL RNA (Figure 2a). Substrate well) to each well of the 96-well plate (with 100 ml medium), and cells were T315I 1 BCR–ABL chimeric RNA consisted of the sequence from 19 nts incubated at 37 C for 20 min. After incubation, absorbance was measured 0 0 at 450 nm in a VICTOR X3 Multilabel Plate Reader (PerkinElmer, Waltham, 5 of the BCR–ABL breakpoint (b3a2 exon junction) to 20 nts 3 of MA, USA). Proliferation of the cells transfected with each DNAzyme was the BCR–ABL junction. The part of the ABL exon (a2) harboring assayed in triplicate, and the mean value was presented with the s.e. for the T315I point mutation (C to U) was chosen as the RNA substrate each sample. Differences in data from the various samples were analyzed (45 nts) for T315I DNAzyme, and the corresponding wild-type ABL by two-tailed unpaired t-tests using SigmaPlot software (SPSS Inc, Chicago, RNA was also prepared as a control RNA substrate. BP DNAzyme IL, USA). exhibited high specificity, cleaving only chimeric BCR–ABL RNA (Figure 2b). When T315I DNAzyme was incubated with mutant Cell cycle analysis and apoptosis detection assay ABL RNA (ABLT315I), RNA cleavage was clearly observed (Figure 2b). Cell cycle distribution was analyzed by measuring DNA content. Cells were In contrast, no RNA-cleavage activity toward the wild-type ABL fixed in ice-cold ethanol (70% v/v) and stained with propidium iodide RNA substrate was observed with both DNAzymes. We further (40 mg/ml; purchased from EMD Chemicals, San Diego, CA, USA) and tested the RNA-cleavage activity of the tandem DNAzyme, which 100 mg/ml RNase in phosphate-buffered saline. After incubation for 30 min was designed to cleave both the BCR–ABL breakpoint and the at room temperature, stained cells were analyzed for cellular DNA content T315I point mutation. This tandem DNAzyme cleaved the RNA by flow cytometry using a FACSCalibur Flow Cytometer (BD Biosciences, substrates of both BCR–ABL and ABLT315I with similar efficiency Franklin Lakes, NJ, USA). Cells with DNA content below 2N were designated (Figure 2c). When the two RNA substrates (BCR–ABL and ABLT315I) as sub-G1, and cells with DNA content between 2N and 4N were were mixed together in the reaction, the tandem DNAzyme designated as G1/G0, S or G2/M phase. The number of cells in each phase of cleaved both RNA substrates. The efficiency of RNA cleavage with the cell cycle was expressed as a percentage of the total number of cells. The progress of apoptosis in cells treated with DNAzyme and/or imatinib the tandem DNAzyme was comparable to that achieved when the was also analyzed using the flow cytometry. After the cells were treated two separate DNAzymes were added to the reaction (Figure 2c). with DNAzyme and/or imatinib for 48 h, cells were washed twice with cold phosphate-buffered saline and resuspended in staining buffer containing Cleavage of ABL transcripts reduced ABL expression and cell fluorescein-5-isothiocyanate, annexin V and propidium iodide, provided in viability the apoptosis detection kit (BD Biosciences), according to the manufacturer’s instructions. Flow cytometric data were obtained using a To demonstrate whether cleavage of mutant ABL RNA with the FACSCalibur Flow Cytometer equipped with BD CellQuest Pro software DNAzyme could reduce expression of the ABL gene (BD Biosciences). Caspase-3/-7 activity in apoptotic cells was measured and subsequently decrease the viability of cells containing the using Caspase-Glo 3/7 reagent (Promega, Madison, WI, USA) according to BCR–ABL gene, we tested the efficacy of the DNAzymes in manufacturer’s instructions, and a VICTOR X3 Multilabel Plate Reader leukemic cells. BaF3 cells that stably expressed BCR–ABL or (PerkinElmer) was used to measure the luminescence emitted from the mutant BCR–ABLT315I were electroporated with DNAzymes target- cleaved substrate. ing either the BCR–ABL breakpoint (BP Dz) or the T315I mutant base (T315I Dz). The tandem DNAzyme composed of the two DNAzymes was also transfected into the cells. A DNAzyme RESULTS targeting a scrambled sequence was also used as a control to Design of DNAzymes targeting BCR–ABL RNA harboring the T315I assess nonspecific effects of oligonucleotide transfection. point mutation Figure 3a shows the amount of remaining ABLT315I RNA transcript In previous studies, several antisense nucleic acids directed upon treatment with T315I DNAzyme in BaF3 cells, in which wild- against the junction of two noncontiguous sequences of BCR type (BaF3/BCR–ABL) cells were mixed with imatinib-resistant and ABL exons were designed and tested for their ability to (BaF3/BCR–ABLT315I) cells in different ratios. When T315I DNAzyme discriminate between the chimeric RNA substrate and the was transfected into BaF3/BCR–ABLT315I cells, about 50% of 21,22,32 nontarget sequences. Among these antisense nucleic ABLT315I mRNA was degraded by 48 h, as demonstrated by acids, we chose a DNAzyme (BP DNAzyme in Figure 1) as a real-time PCR. However, transfection with T315I DNAzyme did not positive control for the cleavage of the RNA substrate; this decrease the amount of ABL mRNA in BaF3/BCR–ABL cells, DNAzyme has been reported to cleave the BCR–ABL junction of indicating that T315I DNAzyme specifically cleaves the T315I 21 the b3a2 splice variant. The cleavage site for BP DNAzyme is point mutation present in ABLT315I RNA in BaF3/BCR–ABLT315I cells. located 5 nucleotides (nts) upstream from the fusion site for b3a2. The specificity of DNAzymes against the target was also Next we set out to design a DNAzyme that could cleave ABL exon demonstrated by reduced expression of ABL proteins. As shown RNA harboring the T315I point mutation. As DNAzymes can in Figure 3b, treatment with each DNAzyme targeting BCR–ABL distinguish single-base mismatches near the cleavage site in RNA led to decreased expression of BCR–ABL protein. Imatinib when complexed with substrate RNA, a target site for DNAzyme treatment did not decrease the expression of BCR–ABL protein, was chosen to specifically cleave only T315I-mutant RNA but treatment with BP DNAzyme significantly reduced the (Figure 1). As shown in Figure 1, the DNAzyme termed T315I expression of BCR–ABL in the cells. As expected, T315I DNAzyme DNAzyme was designed to cleave the phosphodiester bond was more effective at reducing BCR–ABLT315I expression in between A and U, not affecting the A–C bond at the same site in imatinib-resistant cells than in imatinib-sensitive BCR–ABL cells. the wild-type ABL exon. An adenine base was incorporated in the However, BP DNAzyme was effective at inhibiting the expression 50-hybridizing arm of the designed DNAzyme to specifically of BCR–ABL in both wild-type and mutant cells. The decrease recognize the mutant base (U) in the cleavage site (boxed base of BCR–ABL with tandem DNAzyme treatment was similar to that pair in the T315I DNAzyme, Figure 1). We also designed a tandem observed with the T315I DNAzyme, suggesting that the efficacy of DNAzyme that combined BP DNAzyme and T315I DNAzyme with T315I-mutant RNA cleavage was not compromised by the fusion

Leukemia (2013) 1650 – 1658 & 2013 Macmillan Publishers Limited A DNAzyme specific to imatinib-resistant BCR-ABL JE Kim et al 1653 of these two DNAzymes. In contrast, tandem DNAzyme was not as (BP Dz and T315I Dz) or imatinib (Supplementary Figure S2). effective as BP Dz in suppressing the expression of BCR–ABL in Consistent with the result shown in Figure 3c, T315I Dz was the wild-type cells. These results are likely caused by an inequality of most potent in reducing growth rate of the imatinib-resistant BCR– tandem DNAzyme in binding of two target sites on the full length ABL cells. In contrast, the imatinib-sensitive BCR–ABL cells were BCR–ABL mRNA containing the T315I mutation in cytoplasm. more susceptible to cell death with BP Dz than with T315I Dz. BP Dz We next determined how cell proliferation was affected by treatment reduced cell growth rate of both cells that contain BCR– silencing BCR–ABL or the ABLT315I mutant in leukemic cells. ABL junction (BaF3/BCR–ABL and BaF3/BCR–ABLT315I). When the BaF3 cells expressing either BCR–ABL or BCR–ABLT315I were cell viability was measured at 72 h after transfection of DNAzyme treated with each DNAzyme (BP Dz or T315I Dz) twice at intervals (Supplementary Figure S3), BP Dz treatment reduced the cell of 24 h. The cells were also treated with the tandem DNAzyme viability in both cells that contain BCR–ABL junction (BaF3/BCR–ABL targeting both the breakpoint and the T315I mutant base. As and BaF3/BCR–ABLT315I) to a similar extent. In contrast, T315I Dz expected, imatinib treatment decreased the viability of BaF3/BCR– prolonged its efficacy preferably against the imatinib-resistant cells ABL cells, but not BaF3 cells containing the ABLT315I mutation (BaF3/BCR–ABLT315I) for 72 h, resulting in a decrease of cell viability (Figure 3c). Treatment with BP Dz reduced cell viability by about up to 25%. 30% in both cell lines, whereas treatment with T315I Dz significantly reduced the viability of drug-resistant cells by more than 50%, but did not affect BaF3 cells expressing BCR-ABL Imatinib resistance was lost by cleaving mutant ABL transcripts (Figure 3c). Interestingly, the tandem DNAzyme was more with T315I DNAzyme effective at reducing the viability of imatinib-resistant cells than To further investigate the biological consequences of DNAzyme of BaF3 cells lacking the T315I mutation. This result can be treatment on leukemic cells expressing BCR–ABL or BCR–ABLT315I, explained by the inequality of tandem DNAzyme in binding of two the cells were transfected with each DNAzyme and analyzed for target sites in cytoplasm. Treatment with both imatinib and apoptosis. As increased apoptosis may result from DNA fragmen- DNAzyme led to a significant reduction in the viability of BaF3/ tation that can be caused by electroporation, a mock transfection BCR–ABL cells, which was not observed with single treatment of was performed to assess the background level of apoptotic cells. either DNAzyme (Figure 3c). This result suggests that imatinib In both BaF3/BCR–ABL and BaF3/BCR–ABLT315I cells, electropora- treatment was more effective than DNAzyme treatment in tion caused only a small increase in the apoptotic cell population reducing the viability of BCR–ABL-positive cells. In contrast, the (less than 5% of the total cells; see controls in Figure 4). When viability of imatinib-resistant cells was mainly determined by imatinib-resistant BCR–ABL cells were treated with the DNAzyme DNAzyme activity when combined treatment of DNAzyme and targeting the T315I mutant (T315I Dz), a significant increase in the imatinib was applied to the cells. apoptotic cell population, by about 90%, was observed (Figures 4a In addition, we investigated cell growth rate of the leukemic cells and c). In contrast, the DNAzyme targeting the BCR–ABL junction that were incubated up to 24 h after treatment with each DNAzyme (BP Dz) was not effective at enhancing the apoptosis of BCR–ABL

Figure 2. Specificity of the constructed DNAzymes in target RNA cleavage. (a) Three RNA substrates were used to examine the specificity of the DNAzymes toward target RNA cleavage. For efficient RNA synthesis by the T7 RNA polymerase, two G nucleotides were introduced at the 50-end of ABL and BCR–ABL RNAs. P* represents the 50-end labeling of the radioactive phosphate. The T315I point mutation is indicated as a 32 boxed base. (b) Each DNA enzyme (1 mM) and 100 nM 50- P-labeled substrate were incubated at 37 1C for 60 min in a solution that contained 20 mM Tris-HCl (pH 7.5) and 10 mM MgCl2, and reactions were then subjected to electrophoresis on 10% polyacrylamide/8 M urea gels. BP 0 0 DNAzyme generated a 19-nt 5 -fragment of cleaved RNA. Cleavage of mutant ABL RNA (ABLT315I) substrate by T315I DNAzyme generated a 5 - fragment of 23 nts. A DNAzyme targeting a scrambled sequence was also used as a control (lane C). (c) The tandem DNAzyme (1 mM) cleaved both BCR–ABL RNA and mutant ABLT315I RNA (100 nM substrate). The right panel shows that mixed RNA substrates (100 nM each) were present in the RNA cleavage reaction. Similar to the tandem DNAzyme, the mixture of BP DNAzyme and T315I DNAzyme cleaved both BCR–ABL RNA and mutant ABLT315I RNA.

& 2013 Macmillan Publishers Limited Leukemia (2013) 1650 – 1658 A DNAzyme speciﬁc to imatinib-resistant BCR-ABL JE Kim et al 1654 F primer ABL mRNA

R primer BaF3/BCR-ABL BaF3/BCR-ABLT315I 1.1 T315I Dz cleavage site 0.9 BCR-ABL (WT) 0.7 BCR-ABL (MT) T315I ControlImatinibBP DzT315I DzTandem ImatinibDz BP DzT315I DzTandem Dz 0.5 BCR-ABL Relative Quantity Relative

cDNA amplification, 0.3 β-actin 0.1 BaF3 cells WTWT 0.5 : MT 0.5 MT

BaF3/BCR-ABL BaF3/BCR-ABLT315I 100 100 80 **** 80 **** *** *** 60 *** 60 ** * *** ** 40 * 40 * cell viability (%) cell viability (%) 20 20

0 0

BP Dz BP Dz BP Dz BP Dz ControlImatinib ControlImatinib T315I Dz T315I Dz T315I Dz T315I Dz Tandem Dz Tandem Dz Tandem Dz Tandem Dz

+ Imatinib + Imatinib Figure 3. BCR–ABL expression was decreased with transfection of DNAzymes into leukemic cells. (a) Real-time PCR for detection of ABL cDNA was performed in DNAzyme-transfected cells mixed at different ratios, as described in the graph. cDNA ampliﬁed from mutant ABLT315I transcripts was relatively quantiﬁed by normalization to b-actin expression from the same sample. (b) Immunoblotting analysis revealed the expression of BCR–ABL protein from DNAzyme-transfected cells (4 mg for T315I and BP DNAzyme and 10 mg for the tandem DNAzyme per 5 Â 105 cells). For preparation of the mock sample, cells were transfected with oligoDNA containing a scrambled sequence. (c) The viability of each cell line treated with different DNAzymes was measured using the WST (2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H- tetrazolium) assay, 24 h after treatment with DNAzyme (1 mM) and/or imatinib (1 mM). Control cells were untreated. *Po0.001 vs control group; **Po0.005 vs control group; and ***Po0.01 vs control group and ****Po0.05 vs control group.

leukemic cells after 48 h. Moreover, the tandem DNAzyme was less Combined use of imatinib and T315I DNAzyme causes cell cycle effective than the T315I DNAzyme alone at inducing apoptosis in arrest and activation of apoptotic caspases in imatinib-resistant imatinib-resistant BCR–ABL cells. Although the tandem DNAzyme leukemic cells was efficient at cleaving the target RNA at both sites in vitro, this We next investigated the influence of DNAzymes and imatinib DNAzyme may not be able to simultaneously bind to two separate on the cell cycle by analyzing the relative distribution of cells in G1, target sites (BP and T315I) in cells because of the long intervening S, G2 and sub-G1 phases. At 24 h after DNAzyme transfection and stretch of mRNA between the two target sites. imatinib treatment, cells expressing BCR–ABL or BCR–ABLT315I Next, we further tested whether the cleavage of BCR–ABL were collected, and cell cycle distribution was analyzed by transcripts by DNAzyme was sufficient to overcome imatinib fluorescence-activated cell sorting. In BaF3/BCR–ABL cells, the resistance. For analysis of apoptosis resulting from the combined majority of cells treated with imatinib alone (58.7%) or use of DNAzymes and imatinib, we used suboptimal amounts of with imatinib and T315I DNAzyme together (59.6%) were found imatinib (1 mM). Expression of the BCR–ABLT315I kinase in BaF3 cells in the apoptotic sub-G1 phase, whereas single treatment with conferred strong resistance to imatinib, whereas cells with non- T315I DNAzyme or the tandem DNAzyme did not affect cell cycle mutated BCR–ABL were sensitive to imatinib treatment progression (Figure 5). In contrast, in BaF3/BCR–ABLT315I cells, (Figure 4b). Treatment with imatinib and T315I DNAzyme together treatment with T315I DNAzyme alone or with a combination of in BaF3/BCR–ABLT315I cells increased the frequency of apoptosis in T315I DNAzyme and imatinib significantly enhanced apoptotic leukemic cells by 77.3% as compared with single treatment with sub-G1 populations to 46% and 70% of total cells, respectively. imatinib alone (Figures 4b and c). Thus, T315I DNAzyme restored Therefore, T315I DNAzyme was more effective at arresting cell imatinib sensitivity in BaF3/BCR–ABLT315I cells, causing apoptosis cycle progression in imatinib-resistant cells (BaF3/BCR–ABLT315I) in about 80% of the cells. than in BaF3/BCR–ABL cells. We further examined the effects of DNAzymes on BCR– Next, to confirm the dependence of DNAzyme-induced cell ABL-positive human cells (K562 cells) and primary cell lines death in BCR–ABL-positive leukemic cells on caspase activity, we (bone marrow cells) derived from the imatinib-resistant BCR–ABL- examined whether treatment with T315I DNAzyme induced positive leukemia patients. When the imatinib-sensitive BCR– apoptosis through activation of caspase-3 and caspase-7. Leuke- ABL-positive K562 cells were treated with imatinib or BP Dz that mic cells with or without the T315I mutation were transfected with cleaves the BCR–ABL junction, the cells exhibited a significant three DNAzymes, BP Dz, T315I Dz or tandem Dz, for 24 h. Induction increase in apoptotic cell death by about 40B50% as compared of caspase-3/-7 activity was then measured by the appearance of with the control (Supplementary Figure S4a). In contrast, T315I Dz luminescent products generated by the activity of caspase-3/7 (25%) was more effective than BP Dz (8.7%) in enhancing (Figure 6). For a positive control, staurosporine (an inhibitor of apoptosis of the primary cells from imatinib-resistant BCR–ABL- phospholipid/calcium-dependent protein kinase), which has been positive leukemia patients (Supplementary Figure S4b). Combined shown to induce apoptosis by caspase activation in a variety of use of T315I Dz and imatinib further increased apoptotic cell cell types, was used. Staurosporine treatment resulted in induction deaths in the primary cells. These data are consistent with the of caspase-3/-7 activity in both cell types. Treatment with BP results obtained with BaF3 cells, in which T315I Dz was effective in DNAzyme significantly induced the activation of caspases-3/-7 in reducing cell viability of the imatinib-resistant BCR–ABL-positive BaF3/BCR–ABL cells, and this effect was more pronounced with leukemic cells. the addition of imatinib (Figure 6a). In contrast, the relative

Leukemia (2013) 1650 – 1658 & 2013 Macmillan Publishers Limited A DNAzyme speciﬁc to imatinib-resistant BCR-ABL JE Kim et al 1655 Control BP Dz T315I Dz Tandem Dz 4 10 0.8% 0.3% 0.3% 1.4% 103 BaF3/BCR-ABL 102 95.7% 88.7% 93.4% 92.9% 101 3.5%11% 6.3% 5.7% 100 4 PI 10 1.3% 0.8% 10.2% 5.9% 103 BaF3/BCR-ABL T315I 102 97.2% 89% 11.1% 52.3% 101 1.5% 10.2% 78.7% 41.6% 100 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

Annexin V

Control Imatinib Imatinib + T315I Dz 4 10 0.8% 14.7% 9.8% 103

BaF3/BCR-ABL 102 95% 6% 9.7% 101

0 4.2% 79.4% 80.5% 10 PI 104 2% 0.4% 19.5% 103

2 BaF3/BCR-ABL 10 T315I 95.8% 81.5% 4.2% 101 2.2% 18.1% 76.3% 100 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

Annexin V

120 BCR-ABL 110 BCR-ABLT315I 100

30 % of apoptotic cells 20

con BP Dz Imatinib T315I Dz Tandem Dz

T315I Dz +Imatinib Figure 4. DNAzyme and imatinib cooperatively enhanced the induction of apoptosis in imatinib-resistant leukemic cells. Imatinib-sensitive and -resistant cells were treated with DNAzymes (1 mM) and/or imatinib (1 mM) and incubated for 48 h. The cells were analyzed by annexin V/propidium iodide staining to identify the apoptotic population. Percentages of annexin V- and propidium iodide-stained cells are shown in their respective quadrants. The apoptotic populations of cells treated with (a) each DNAzyme and (b) T315I DNAzyme plus imatinib are indicated. Control cells were electroporated, but otherwise untreated. (c) Bar graphs showing the percentage of apoptotic cells analyzed by the flow cytometry. Data from three separate experiments as shown in b and c were collected and averaged. increase in caspase-3/-7 activity observed in imatinib-resistant cells (Supplementary Figure S5). Thus, the synergistic inhibitory cells (BaF3/BCR–ABLT315I) treated with T315I DNAzyme was greater effect by combined treatment of T315I Dz and imatinib has probably than that observed in BaF3/BCR–ABL cells treated with the same resulted from combined efficacies of two different inhibitors, T315I DNAzyme (Figure 6b). DNAzyme and imatinib against BCR–ABLT315I mRNA and phospha- More importantly, induction of caspase-3/-7 activity was greatly tidylinositol-3 kinase/AKT signaling pathway, respectively. enhanced by combined treatment with T315I DNAzyme and imatinib in BaF3/BCR–ABLT315I cells. Several studies suggest a role of imatinib on phosphatidylinositol-3 kinase/AKT signaling DISCUSSION pathway by inhibiting AKT phosphorylation at Ser473 position Development of BCR–ABL kinase inhibitors, such as imatinib and in cancer cells.33–37 We observed that imatinib treatment second-generation agents (dasatinib and nilotinib), brought resulted in a significant dose-responsive reduction in the AKT tremendous advancements to the treatment of BCR–ABL-positive phosphorylation at position Ser473 in the imatinib-resistant leukemia as well as other malignancies. However, despite initial

& 2013 Macmillan Publishers Limited Leukemia (2013) 1650 – 1658 A DNAzyme speciﬁc to imatinib-resistant BCR-ABL JE Kim et al 1656 Control Imatinib T315I Dz BP-T315I T315I Dz +Imatinib Tandem Dz 100 160 160 160 160 80 Sub-G1: 18.4% Sub-G1: 58.7% Sub-G1: 19.4% Sub-G1: 19.1% Sub-G1: 59.6% G1: 44.4% 120 G1: 30% 120 G1: 40.2% 120 G1: 40.2% 120 G1: 27.8% 60 S: 16.6% 80 S: 7.3% 80 S:12.8% 80 S: 16.8% 80 S: 7.7% 40 G2/M: 20.8% G2/M: 4% G2/M: 27.6% G2/M: 18.6% G2/M: 5% 20 40 40 40 40 BaF3/BCR-ABL 0 0 0 0 0

100 160 160 160 Counts Sub-G1: 17.5% 160 Sub-G1: 23.1% Sub-G1: 46% Sub-G1: 23.8% Sub-G1: 70.5% T315I 80 G1: 46.5% 120 G1: 34.2% 120 G1: 28.5% 120 G1: 39.4% 120 G1:18.8% 60 S: 14.6% S: 20.1% S: 14.4% S: 14.4% S: 9% G2/M: 21.4% 80 G2/M: 22.7% 80 G2/M: 11.1% 80 G2/M: 17.1% 80 G2/M: 2% 40 20 40 40 40 40 0 0 0 0 BaF3/BCR-ABL 0 0200 400 6000 200 400 6000 200 400 6000 200 400 6000 200 400 600

DNA content

100 BCR-ABL BCR-ABLT315I 80 **

* 40 Sub-G1 cells (%)

con Imatinib T315I Dz Tandem Dz

T315I Dz +Imatinib Figure 5. Combined use of imatinib and T315I DNAzyme caused cell cycle arrest in leukemic cells. (a) Imatinib-sensitive and -resistant cells were treated with DNAzymes (1 mM) and/or imatinib (1 mM), and incubated for 24 h. Sub-G1 analysis by propidium iodide staining was examined by ﬂow cytometry. The distribution and percentage of cells in sub-G1,G1, S and G2/M phases of the cell cycle are indicated. The sub-G1 population represents apoptotic and dead cells. Control cells were electroporated, but otherwise untreated. (b) Bar graph showing the percentage of sub-G1 cells. Data from three separate experiments as shown in a were collected and averaged; *Po0.001 vs control group, **Po0.05 vs T315I Dz group.

BCR-ABL BCR-ABLT315I 350 350 P<0.005 P<0.005 300 P<0.1 300 P<0.005 200 200 * * 150 * 150 ** ** *** 100 100

50 50 Caspase 3/7 activity (%) Caspase 3/7 activity (%) 0 0 TA TA S S BP Dz BP Dz BP Dz BP Dz Control Imatinib Control Imatinib T315I Dz T315I Dz T315I Dz T315I Dz Tandem Dz Tandem Dz Tandem Dz Tandem Dz + Imatinib + Imatinib Figure 6. DNAzymes induced caspase-dependent apoptosis in leukemic cells. Imatinib-sensitive and -resistant cells were treated with DNAzymes (1 mM) and/or imatinib (1 mM), and incubated for 24 h. The activity of the apoptotic enzyme caspase-3/-7 was evaluated at 24 h after treatment by measuring the luminescence emitted from the cleaved substrate by active caspase-3/-7 in (a) imatinib-sensitive BCR–ABL cells and (b) imatinib-resistant BCR–ABLT315I cells. Control cells were electroporated, but otherwise untreated. *Po0.005 vs control group, **Po0.05 vs control group, ***Po0.1 vs control group.

responses, patients with BCR–ABL-positive leukemia tend to BCR–ABL, conferring imatinib resistance to the affected cells.39 The develop resistance to imatinib and second-generation agents.38 T315I mutation in BCR–ABL is the main cause of imatinib This drug resistance is often caused by mutations in the gate- resistance in CML patients, and the prognosis of CML patients keeper region (for example, T315I and F317L) of the BCR–ABL resistant to imatinib because of this mutation is poor in terms kinase, which disrupt critical contact points between imatinib and of survival.

Leukemia (2013) 1650 – 1658 & 2013 Macmillan Publishers Limited A DNAzyme specific to imatinib-resistant BCR-ABL JE Kim et al 1657 Specific silencing of the chimeric BCR–ABL gene by siRNA has The specific recognition and cleavage of the target sequence by been used to downregulate the expression of leukemogenic DNAzyme may be useful properties for nucleic acid-based drugs, BCR–ABL. Scherr et al.26 demonstrated that siRNA targeting the such as antisense oligonucleotides, siRNA and ribozymes. How- breakpoint in BCR–ABL fusion RNA interfered with the induction of ever, as with other nucleic acid-based applications in vivo, the BCR–ABL expression in BCR–ABL-positive cell lines. Use of siRNA delivery of DNAzyme to target cells remains a considerable directed against the breakpoint in the BCR–ABL fusion site b3a2 challenge for in vivo applications.44 There are several approaches specifically downregulated the BCR–ABL protein, with a significant to introduce DNAzymes into cells, including drug-delivery reduction in cell viability. Using a breakpoint-specific siRNA in systems, in which chemically synthesized DNAzymes are BCR–ABL-positive cells, Wohlbold et al.25 found that siRNA complexed with nanoparticles and delivered to target cells.45,46 targeting the breakpoint of chimeric BCR–ABL RNA partially Another way to introduce DNAzymes into cells is through restored imatinib sensitivity, with a fourfold sensitization in cells electroporation, as used in our study.47 This approach is expressing the imatinib-resistant BCR–ABL kinase domain mutant limited primarily by the necessity for ex vivo manipulations of H396P, which makes leukemic cells fivefold less sensitive to leukemic cells, in which target cells must be removed from the imatinib compared with wild-type cells. However, they failed to body, engineered and returned.48 General advantages and reverse the high imatinib resistance of leukemic cells harboring disadvantages of DNAzymes have been compared with other the T315I mutation, which confers complete resistance to imatinib nucleic acid-based approaches, such as ribozymes, antisense treatment, by siRNA targeting the breakpoint of the BCR–ABL oligonucleotides and RNA interference for silencing of target gene fusion site. expression.44 Compared with these other approaches, DNAzymes Recently, Koldehoff et al.40 reported that coadministration of a are superior for potential in vivo applications, as they do not suffer breakpoint-specific siRNA with imatinib or nilotinib resulted in from mRNA degradation and exhibit improved chemical and enhanced inhibition of cell growth and BCR–ABL gene expression biological stability over RNA and protein agents. Such stability can in BCR–ABL-expressing cells harboring the T315I mutation. be further enhanced through chemical modifications of DNA. In This finding is consistent with our result that transfection addition, unlike RNA interference DNAzymes are catalytically with breakpoint-specific BP DNAzyme increased the activity of independent of cellular machinery and, therefore, may have the apoptotic executioner caspases, caspase-3 and caspase-7, in potential for rational modulation of RNA-cleavage activity. Despite imatinib-resistant BCR–ABL cell lines (Figure 6b). However, our these advantages of DNAzymes, proper folding of the DNAzyme, strategy used in this study to directly cleave ABL RNA at the T315I which may not be guaranteed in vivo due to lower physiological point mutation site provided enhanced efficacy toward sensitiza- concentrations (usually less than 1 mM)49 of Mg2 þ than required tion of imatinib-resistant BCR–ABL cell lines to imatinib treatment. for optimal catalytic activity of DNAzymes in vitro (10–100 mM),20 To the best of our knowledge, cleavage of T315I-mutant ABL RNA still prove to be a challenge. with DNAzyme transfected into BCR–ABL-positive cells has not In summary, we demonstrated that RNA-cleaving DNAzymes previously been reported as a potential method to restore are complementary to the imatinib-resistant point mutation imatinib sensitivity in leukemic cells. Thus, to augment the (T315I) in ABL mRNA and can induce target-specific cleavage efficacy of imatinib treatment in BCR–ABL-positive cells, specific of mutant ABL mRNA. When T315I DNAzyme was transfected targeting of the T315I mutant site would be more effective than into imatinib-resistant BaF3/BCR–ABLT315I cells, expression of the targeting of the breakpoint in the BCR–ABL fusion site when using mutant BCR–ABL kinase was downregulated. DNAzyme treatment antisense oligonucleotides. The use of siRNA or DNAzyme led to inhibition of cell proliferation, accompanied by a strong mixtures targeting both of these sites together may result in induction of apoptotic cell death. A major finding of our study was greater efficacy at restoring imatinib sensitivity. Clinically, that cleavage of T315I mutant RNA sensitized imatinib-resistant both BCR–ABL and T315I silencing approaches may be of great leukemic cells to imatinib treatment, resulting in apoptosis relevance to the conversion of acute or blast crisis phases back to through activation of executioner caspases. These results suggest chronic phase leukemia without concerns of imatinib resistance. that the combined use of the T315I DNAzyme with imatinib may To this end, long-term expression of gene silencing short hairpin be the most effective tool to treat imatinib-resistant BCR–ABL- RNAs targeting these sites could be conducted using lentivirus positive leukemic cells for ex vivo manipulations. If the DNAzyme is vectors in BCR–ABL-positive cells. to be used for ex vivo gene therapy, bone marrow cells derived Previous reports have demonstrated that silencing of BCR–ABL from the patient would be manipulated to be free of mutant expression induces a profound defect in proliferation through BCR–ABL RNA with DNAzymes and/or imatinib, and the cells can decreased expression of cyclin D.41 This result is consistent with be subsequently re-entered into the patient’s bone marrow our findings that mutant RNA cleavage by T315I DNAzyme without need of the bone marrow transplantation. Although triggered cell cycle arrest during the G1 phase of the cell cycle in DNAzymes that have been proved to cleave RNA in vitro are now imatinib-resistant leukemia cells (Figure 5). BCR–ABLT315I-trans- progressing toward clinical tests, they must be combined with a formed cells actively undergo cell proliferation without much safe and appropriate delivery vehicles for clinical use in vivo.If cell death under imatinib treatment. However, combined use of combined with an effective tool to deliver the DNAzymes to cells the T315I DNAzyme and imatinib induced cell cycle arrest at the and tissues, selective interventions in leukemic cell gene regula- sub-G1 phase in up to 70% of total cells through inhibition of tion by DNAzymes may be feasible for the treatment of drug- mutant BCR–ABL expression and restoration of imatinib sensitivity resistant leukemia, such as BCR–ABL-positive CML harboring the (Figure 5). We also observed that selective cleavage of the T315I T315I mutation. mutation in ABL mRNA triggered apoptosis with activation of caspase-3/-7 in imatinib-resistant leukemic cells (Figure 6b). This result indicated that apoptosis was related to decreased BCR–ABL CONFLICT OF INTEREST expression in terms of kinase activity and/or its interacting partners.42 It has been previously reported that ectopic The authors declare no conflict of interest. expression of cyclin D1 inhibits imatinib-induced apoptosis and erythrocyte differentiation in leukemic K562 cells.43 Thus, cell cycle arrest and apoptotic cell death caused by cleavage of mutant ABL ACKNOWLEDGEMENTS mRNA suggests a putative interaction between BCR–ABL and We thank the Korean Leukemia Cell and Gene Bank for providing the CML samples. cyclin D1 to maintain BCR–ABL-positive cells in a state of This work was supported by the National Research Foundation grants funded by the continuous cell cycle and differentiation block. Korean Government MEST (2011-0016385 and 2012M3A9B2028336).

& 2013 Macmillan Publishers Limited Leukemia (2013) 1650 – 1658 A DNAzyme specific to imatinib-resistant BCR-ABL JE Kim et al 1658 REFERENCES 24 Bhatia R, Verfaillie CM. Inhibition of BCR-ABL expression with antisense 1 Daley GQ, Van Etten RA, Baltimore D. Induction of chronic myelogenous leukemia oligodeoxynucleotides restores beta1 integrin-mediated adhesion and prolifera- in mice by the P210bcr/abl gene of the Philadelphia chromosome. Science 1990; tion inhibition in chronic myelogenous leukemia hematopoietic progenitors. 247: 824–830. Blood 1998; 91: 3414–3422. 2 Faderl S, Talpaz M, Estrov Z, O’Brien S, Kurzrock R, Kantarjian HM. The biology of 25 Wohlbold L, van der Kuip H, Miething C, Vornlocher HP, Knabbe C, Duyster J et al. chronic myeloid leukemia. N Engl J Med 1999; 341: 164–172. Inhibition of bcr-abl gene expression by small interfering RNA sensitizes for 3 Rowley JD. Letter: A new consistent chromosomal abnormality in chronic mye- imatinib mesylate (STI571). Blood 2003; 102: 2236–2239. logenous leukaemia identified by quinacrine fluorescence and Giemsa staining. 26 Scherr M, Battmer K, Winkler T, Heidenreich O, Ganser A, Eder M. Specific Nature 1973; 243: 290–293. inhibition of bcr-abl gene expression by small interfering RNA. Blood 2003; 101: 4 Groffen J, Stephenson JR, Heisterkamp N, de Klein A, Bartram CR, Grosveld G. 1566–1569. Philadelphia chromosomal breakpoints are clustered within a limited region, bcr, 27 Rapozzi V, Burm BE, Cogoi S, van der Marel GA, van Boom JH, Quadrifoglio F et al. on chromosome 22. Cell 1984; 36: 93–99. Antiproliferative effect in chronic myeloid leukaemia cells by antisense peptide 5 Shepherd P, Suffolk R, Halsey J, Allan N. Analysis of molecular breakpoint nucleic acids. Nucleic Acids Res 2002; 30: 3712–3721. and m-RNA transcripts in a prospective randomized trial of interferon in 28 Rapozzi V, Cogoi S, Xodo LE. Antisense locked nucleic acids efficiently suppress chronic myeloid leukaemia: no correlation with clinical features, cytogenetic BCR/ABL and induce cell growth decline and apoptosis in leukemic cells. Mol response, duration of chronic phase, or survival. Br J Haematol 1995; 89: Cancer Ther 2006; 5: 1683–1692. 546–554. 29 Jang KJ, Lee NR, Yeo WS, Jeong YJ, Kim DE. Isolation of inhibitory RNA aptamers 6 Goldman JM, Melo JV. Chronic myeloid leukemia--advances in biology and new against severe acute respiratory syndrome (SARS) coronavirus NTPase/Helicase. approaches to treatment. N Engl J Med 2003; 349: 1451–1464. Biochem Biophys Res Commun 2008; 366: 738–744. 7 Cortez D, Reuther G, Pendergast AM. The Bcr-Abl tyrosine kinase activates 30 Kim DE, Joyce GF. Cross-catalytic replication of an RNA ligase ribozyme. Chem Biol mitogenic signaling pathways and stimulates G1-to-S phase transition in 2004; 11: 1505–1512. hematopoietic cells. Oncogene 1997; 15: 2333–2342. 31 Lee B, Kim KB, Oh S, Choi JS, Park JS, Min DH et al. Suppression of hepatitis C virus 8 Gesbert F, Sellers WR, Signoretti S, Loda M, Griffin JD. BCR/ABL regulates genome replication in cells with RNA-cleaving DNA enzymes and short-hairpin expression of the cyclin-dependent kinase inhibitor p27Kip1 through the phos- RNA. Oligonucleotides 2010; 20: 285–296. phatidylinositol 3-Kinase/AKT pathway. J Biol Chem 2000; 275: 39223–39230. 32 Kronenwett R, Haas R, Sczakiel G. Kinetic selectivity of complementary 9 Nimmanapalli R, Fuino L, Bali P, Gasparetto M, Glozak M, Tao J et al. Histone nucleic acids: bcr-abl-directed antisense RNA and ribozymes. J Mol Biol 1996; 259: deacetylase inhibitor LAQ824 both lowers expression and promotes proteasomal 632–644. degradation of Bcr-Abl and induces apoptosis of imatinib mesylate-sensitive 33 Patel BB, He YA, Li XM, Frolov A, Vanderveer L, Slater C et al. Molecular or -refractory chronic myelogenous leukemia-blast crisis cells. Cancer Res 2003; mechanisms of action of imatinib mesylate in human ovarian cancer: a proteomic 63: 5126–5135. analysis. Cancer Genomics Proteomics 2008; 5: 137–149. 10 Dubrez L, Eymin B, Sordet O, Droin N, Turhan AG, Solary E. BCR-ABL delays 34 Chu S, Holtz M, Gupta M, Bhatia R. BCR/ABL kinase inhibition by imatinib mesylate apoptosis upstream of procaspase-3 activation. Blood 1998; 91: 2415–2422. enhances MAP kinase activity in chronic myelogenous leukemia CD34 þ cells. 11 Steinberg M. Dasatinib: a tyrosine kinase inhibitor for the treatment of Blood 2004; 103: 3167–3174. chronic myelogenous leukemia and philadelphia chromosome-positive acute 35 Xu Y, Fang F, Sun St Y, Clair St DK, Clair WH. RelB-dependent differential radio- lymphoblastic leukemia. Clin Ther 2007; 29: 2289–2308. sensitization effect of STI571 on prostate cancer cells. Mol Cancer Ther 2010; 9: 12 Druker BJ, Talpaz M, Resta DJ, Peng B, Buchdunger E, Ford JM et al. Efficacy and 803–812. safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid 36 Matei D, Chang DD, Jeng MH. Imatinib mesylate (Gleevec) inhibits ovarian cancer leukemia. N Engl J Med 2001; 344: 1031–1037. cell growth through a mechanism dependent on platelet-derived growth factor 13 Tsao AS, Kantarijian H, Talpaz M. STI-571 in chronic myelogenous leukaemia. receptor alpha and Akt inactivation. Clin Cancer Res 2004; 10: 681–690. Br J Haematol 2002; 119: 15–24. 37 Klejman A, Rushen L, Morrione A, Slupianek A, Skorski T. Phosphatidylinositol-3 14 Bradeen HA, Eide CA, O’Hare T, Johnson KJ, Willis SG, Lee FY et al. Comparison of kinase inhibitors enhance the anti-leukemia effect of STI571. Oncogene 2002; 21: imatinib mesylate, dasatinib (BMS-354825), and nilotinib (AMN107) in an 5868–5876. N-ethyl-N-nitrosourea (ENU)-based mutagenesis screen: high efficacy of drug 38 Quintas-Cardama A, Kantarjian H, Cortes J. Imatinib and beyond--exploring the combinations. Blood 2006; 108: 2332–2338. full potential of targeted therapy for CML. Nat Rev Clin Oncol 2009; 6: 535–543. 15 Gorre ME, Mohammed M, Ellwood K, Hsu N, Paquette R, Rao PN et al. Clinical 39 La Rosee P, Deininger MW. Resistance to imatinib: mutations and beyond. resistance to STI-571 cancer therapy caused by BCR-ABL gene mutation or Semin Hematol 2010; 47: 335–343. amplification. Science 2001; 293: 876–880. 40 Koldehoff M, Kordelas L, Beelen DW, Elmaagacli AH. Small interfering RNA against 16 Hughes T, Deininger M, Hochhaus A, Branford S, Radich J, Kaeda J et al. BCR-ABL transcripts sensitize mutated T315I cells to nilotinib. Haematologica Monitoring CML patients responding to treatment with tyrosine kinase inhibitors: 2010; 95: 388–397. review and recommendations for harmonizing current methodology for 41 Rangatia J, Bonnet D. Transient or long-term silencing of BCR-ABL alone induces detecting BCR-ABL transcripts and kinase domain mutations and for expressing cell cycle and proliferation arrest, apoptosis and differentiation. Leukemia 2006; results. Blood 2006; 108: 28–37. 20:68–76. 17 von Bubnoff N, Schneller F, Peschel C, Duyster J. BCR-ABL gene mutations in 42 Bueno-da-Silva AE, Brumatti G, Russo FO, Green DR, Amarante-Mendes GP. relation to clinical resistance of Philadelphia-chromosome-positive leukaemia to Bcr-Abl-mediated resistance to apoptosis is independent of constant STI571: a prospective study. Lancet 2002; 359: 487–491. tyrosine-kinase activity. Cell Death Differ 2003; 10: 592–598. 18 Talpaz M, Shah NP, Kantarjian H, Donato N, Nicoll J, Paquette R et al. Dasatinib in 43 Kawano T, Horiguchi-Yamada J, Saito S, Iwase S, Furukawa Y, Kano Y et al. Ectopic imatinib-resistant Philadelphia chromosome-positive leukemias. N Engl J Med cyclin D1 expression blocks STI571-induced erythroid differentiation of K562 cells. 2006; 354: 2531–2541. Leuk Res 2004; 28: 623–629. 19 O’Hare T, Eide CA, Deininger MW. Bcr-Abl kinase domain mutations, drug resis- 44 Baum DA, Silverman SK. Deoxyribozymes: useful DNA catalysts in vitro and tance, and the road to a cure for chronic myeloid leukemia. Blood 2007; 110: in vivo. Cell Mol Life Sci 2008; 65: 2156–2174. 2242–2249. 45 Pun SH, Tack F, Bellocq NC, Cheng J, Grubbs BH, Jensen GS et al. Targeted delivery 20 Santoro SW, Joyce GF. A general purpose RNA-cleaving DNA enzyme. Proc Natl of RNA-cleaving DNA enzyme (DNAzyme) to tumor tissue by transferrin-modified, Acad Sci USA 1997; 94: 4262–4266. cyclodextrin-based particles. Cancer Biol Ther 2004; 3: 641–650. 21 Wu Y, Yu L, McMahon R, Rossi JJ, Forman SJ, Snyder DS. Inhibition of bcr-abl 46 Ryoo SR, Jang H, Kim KS, Lee B, Kim KB, Kim YK et al. Functional delivery of oncogene expression by novel deoxyribozymes (DNAzymes). Hum Gene Ther DNAzyme with iron oxide nanoparticles for hepatitis C virus gene knockdown. 1999; 10: 2847–2857. Biomaterials 2012; 33: 2754–2761. 22 Kuwabara T, Warashina M, Tanabe T, Tani K, Asano S, Taira K. Comparison of the 47 Nunamaker EA, Zhang HY, Shirasawa Y, Benoit JN, Dean DA. Electroporation- specificities and catalytic activities of hammerhead ribozymes and DNA enzymes mediated delivery of catalytic oligodeoxynucleotides for manipulation of vascular with respect to the cleavage of BCR-ABL chimeric L6 (b2a2) mRNA. Nucleic Acids gene expression. Am J Physiol Heart Circ Physiol 2003; 285: H2240–H2247. Res 1997; 25: 3074–3081. 48 Morgan RA, Anderson WF. Human gene therapy. Annu Rev Biochem 1993; 62: 23 Maran A, Waller CF, Paranjape JM, Li G, Xiao W, Zhang K et al. 2’,5’-Oligoadenylate- 191–217. antisense chimeras cause RNase L to selectively degrade bcr/abl mRNA in chronic 49 Grubbs RD. Intracellular magnesium and magnesium buffering. Biometals 2002; myelogenous leukemia cells. Blood 1998; 92: 4336–4343. 15: 251–259.

Supplementary Information accompanies this paper on the Leukemia website (http://www.nature.com/leu)

Leukemia (2013) 1650 – 1658 & 2013 Macmillan Publishers Limited