Combined effects of As₄S₄ and imatinib on chronic myeloid leukemia cells and BCR-ABL oncoprotein

Chronic myeloid leukemia (CML) is a clonal pluripotent hematopoietic stem cell disorder. Leukemic cells from more than 95% of patients with CML carry the Philadelphia (Ph) chromosome, resulting from t(9;22) reciprocal chromosome translocation, and express a BCR-ABL fusion protein with a relative molecular mass of 210 kDa. BCR-ABL protein has more potent tyrosine kinase activity than the wild-type 145-kDa ABL, leading its downstream signal molecules to be phosphorylated and activated, constituting the molecular basis of CML,^1-3  although the critical pathways of CML still remain to be further defined. Targeted treatment against BCR-ABL and/or its downstream effector molecules has become a new hot spot and the introduction of imatinib (STI571, Gleevec, CGP57148B) into CML treatment during the past few years represents the most significant example in this progress.^4-7 

Imatinib competes the adenosine triphosphate (ATP) binding site of the kinase domain of ABL, which allows a selective inhibition of the action of BCR-ABL and thereby blocks the activation of its downstream effector molecules.⁸  Clinical trials have shown that imatinib has very good effects on patients with chronic phase (CP), but its effects on patients with accelerated phase or blast phase (advanced phase, AP) are unsatisfactory since many patients relapse after transient remission.^9-12  In addition, even among patients at CP, imatinib seems unable to eradicate the malignant progenitors and a significant portion of patients develops drug resistance after long-time use.^13-16  Hence, to improve response rates and to circumvent resistance, additional drugs are needed in CML-targeted therapy.

The antileukemic activity of arsenic (Fowler solution) was first reported in the late 1800s.¹⁷  Recently, arsenic compounds have regained attention due to the discovery of their clinical effects and unique mode of action on acute promyelocytic leukemia (APL).^18-20  Moreover, it was reported that As₂O₃ could enhance the selective cytotoxic effects of imatinib against BCR-ABL–positive leukemia cells.^21,22  A drawback of As₂O₃ is that it must be administered by intravenous infusion on a daily basis to achieve a good therapeutic result and to avoid gastrointestinal side effects. More recently, arsenic sulfide (As₄S₄), another arsenic compound, has been shown to be equally effective in the treatment of APL, with fewer side effects and requiring only oral administration.²³  Some preliminary data from our group suggested As₄S₄ could induce apoptosis of CML primary cells and the K562 cell line.²⁴  This orally administered agent would thus contribute to quality of life and also provide easy access to consolidation and maintenance therapy in leukemia therapy.

The aim of this study is to investigate the combined effects of imatinib and As₄S₄ on BCR-ABL and CML cells. The K562 cell line was used since it has been considered as a cellular model of CML for drug screening. On the other hand, in order to address the potential clinical application of the drugs, we used fresh CD34⁺ hematopoietic progenitor cells isolated from patients with CML, which have stronger clonogenic ability and also higher drug sensitivity than their progenies.^25,26  By a variety of assays for cell proliferation and clonogenic potential, cell cycle distribution, programmed cell death, and protein turnover/activity of BCR-ABL tyrosine kinase, we studied the effect of imatinib and/or As₄S₄.

Heparin-treated bone marrow (BM) samples were obtained from 22 patients with Ph⁺ CML and 8 nonleukemic donors. Informed consent was obtained according to institutional guidelines. The 15 men and 7 women with CML ranged in age from 32 years to 63 years and had either CP (14 patients) or AP (8 patients). The diagnosis of CML was established on the basis of morphologic examination, presence of Ph chromosome, and positive reverse transcriptase–polymerase chain reaction (RT-PCR) results for BCR-ABL fusion transcripts. At the time of investigation, 8 patients had received no previous treatment (6 at CP and 2 at AP), whereas 14 had been treated with hydroxyurea, including 4 who also received interferon. Patients were off therapy for a week or on hydroxyurea as the only treatment when the samples were collected.

BM mononuclear cells (BMMNCs) were isolated by means of Ficoll density gradient centrifugation (specific gravity, 1.077; Shanghai Second Reagent Factory, Shanghai, China). CD34⁺ cells were selected from BMMNCs by using positive immunomagnetic column separation (Miltenyi Biotech, Auburn, CA) according to the manufacturer's instructions. The purity of CD34⁺ cells ranged from 88% to 96% as determined by flow cytometry (EPICS XL; Coulter, Hialeah, FL), and the viability of CD34⁺ cells was above 96% by trypan blue exclusion assay.

CD34⁺ cells were cultured in multiwell tissue-culture plates in serumfree medium (StemPro; Gibco BRL, Gaithersburg, MD) supplemented with growth factors (200 ng/L granulocyte macrophage–colony-stimulating factor, 1 μg/L granulocyte colony-stimulating factor, 200 ng/L stem cell factor, 50 ng/L leukemia inhibitory factor, 200 ng/L macrophage inflammatory protein-1 alpha, 1 μg/L interleukin-6 [PeproTech, London, United Kingdom]).¹³ 

The BCR-ABL–positive K562 cell line was derived from a patient with CML in erythroid blast phase. HL60 is a BCR-ABL–negative acute myeloid leukemia cell line. U937 cells are BCR-ABL–negative myeloid precursors blocked at the promonocytic stage, which derived from a patient with histiocytic lymphoma. K562, HL60, and U937 cells were cultured in RPMI 1640, supplemented with 10% fetal bovine serum (PAA, Linz, Austria), in a humidified atmosphere with 5% CO₂ at 37°C. Morphology was determined with Wright staining of cells centrifuged onto slides by cytospin (100 g, 4 minutes; Shandon, Runcorn, United Kingdom).

Imatinib was kindly provided by Novartis Pharma (Basel, Switzerland), and prepared as one mM stock solution in sterile distilled water at -20°C. As₄S₄ (Alfa Aesar, Ward Hill, MA) was dissolved in 0.1 M sodium hydroxide to make a stock solution of one mM. Stock solutions were diluted in RPMI 1640 medium to achieve the final concentration.

K562, HL60, U937, and CD34⁺ cells were plated in triplicate at 1 × 10⁵ cells/mL to 2 × 10⁵ cells/mL. After being incubated with 0 μM to 1 μM imatinib and/or 0 μM to 4 μMAs₄S₄ for one to 3 days at 37°C, viable cells were counted by 0.4% trypan blue exclusion.

The dose-effect curves of single or combined drug treatment were analyzed by the median-effect method of Chou and Talalay²⁷  using the Calcusyn Software (Biosoft, Cambridge, United Kingdom). The combination indexes (CI) less than 1, equal to 1, and greater than 1 indicate synergistic, additive, and antagonistic effects, respectively.²⁸ 

Mixed with 0 μM to 2 μM imatinib and/or 0 μM to 2 μM As₄S₄, 10³ CD34⁺ cells were plated in semisolid methylcellulose progenitor culture (Methocult H4434; Stem Cell Technologies, Vancouver, BC, Canada). After 14 days incubation at 37°C in a fully humidified atmosphere of 5% CO₂, granulocyte macrophage–colony-forming units (CFU-GMs) and erythroid burst-forming units (BFU-Es) were counted. Only clusters with more than 50 cells were counted as a colony. Total CFUs, CFU-GMs, and BFU-Es were expressed as the percentage of inhibition versus control. All clonogenic assays were done in triplicate.

Cells were collected, washed in phosphate-buffered saline (PBS), and fixed in cold 70% ethanol for at least one hour at 4°C. Shortly before flow cytometry analysis, cells were rinsed with PBS, treated by 50 mg/L RNAse A for at least 15 minutes at 37°C, and stained with 50 mg/L propidium iodide (PI). Distribution of cells with different DNA contents was analyzed by flow cytometry. Ten thousand cells were analyzed in each sample.

After drug treatment for 24 to 72 hours, 1 × 10⁵ cells to 5 × 10⁵ cells were washed in PBS and resuspended in 200 μL staining solution containing 5 μL of annexin V–fluorescein isothiocyanate (FITC) and 10 μLof 20 μg/mL PI, according to the protocol of the annexin V staining kit (Clontech, Palo Alto, CA). Flow cytometry was used to analyze 5000 cells. This allowed the discrimination of live cells (unstained with either fluorochrome) from apoptotic cells (stained with annexin V) and necrotic cells (stained with PI). All data were collected, stored, and analyzed by Multigraph software (Coulter, Miami, FL).

After drug treatment for 48 hours, 1 × 10⁵ cells to 5 × 10⁵ cells were washed in PBS and resuspended in 100 μL PBS, containing 5 μL phycoerythrin (PE)–conjugated CD95 antibody (Coulter Immunotech, Margency, France) or PE-conjugated nonspecific immunoglobulin G (IgG). The cells were incubated in the dark at 4°C for 30 minutes and the samples were analyzed by flow cytometry.

After being washed twice with PBS, about 10⁶ cells were incubated with 10 mg/L rhodamine 123 (Rh123) at 37°C for 30 minutes, a cationic lipophilic fluorochrome taken up by mitochondria in proportion to the ΔΨm. Then, 50 mg/L PI, a membrane-impermeable DNA binding dye, was added to cells. Flow cytometry was used to determine 10⁴ cells.

Caspase-3 activity was measured using a kit (Clontech). Briefly, 2 × 10⁶ cells treated by As₄S₄ and/or imatinib were lysed. Assays were performed by incubating 100 μg of cell lysates in 100 μL reaction buffer (1% Nonidet P-40 [NP40], 20 mM Tris-HCl, pH 7.5, 137 mM NaCl, 10% glycerol) containing 5 μL caspase-3 substrate DEVD-pNA at 37°C for 2 hours. Thereafter, the absorbance at 405 nm was measured by a spectrophotometer.

After drug treatment for 48 hours, 5 × 10⁶ cells were washed in PBS, then lysed using 200 μL radioimmunoprecipitation assay (RIPA) buffer (containing 50 mM Tris pH 8.0, 150 mM NaCl, 0.1% SDS, 0.5% deoxycholate, 1% NP-40, 1 mM DTT, 1 μM sodium vanadate, and 0.2 mM phenylmethyl sulfonic fluoride [PMSF]). Protein concentrations were determined with the Bradford method (Dc Protein Assay; BIO-RAD, Hercules, CA). Total protein (100 μg) was run on 8% sodium dodecyl sulfate (SDS)–polyacrylamide gels, then transferred to polyvinylidene difluoride membranes (Amersham, Buckinghamshire, United Kingdom), probed with individual antibodies, and visualized by an enhanced chemiluminescence (ECL) system (Pierce, Rockford, IL). The following antibodies were used: anti-retinoblastoma (anti-Rb), anti–phospho-Rb, anti-cyclin–dependent kinase1 (anti-CDK1), anti–phospho-CDK1, anti–poly adenosine diphosphate (ADP)–ribose polymerase (anti-PARP), anti-FasL, anti–Bcl-2, anti-abl (Santa Cruz Biotechnology, Santa Cruz, CA), and anti-actin (Oncogene, Boston, MA). Secondary antibodies were antirabbit peroxidase–conjugated antibody (New England Biolabs, Beverly, MA), and antimouse or antigoat peroxidase–conjugated antibody (Oncogene).

c-ABL and BCR-ABL in samples containing 100 μg total protein were obtained by immunoprecipitation (Santa Cruz Biotechnology). Immunoprecipitates were washed with RIPA buffer and then resuspended in the assay buffer (Tris-HCl 50 mM pH 7.4, MgCl₂ 40 mM, sodium vanadate 50 mM, dithiothreitol [DTT] 2 mM, MnCl₂ 1 mM). After centrifugation, the precipitated protein was directly assayed for the PTK activity according to the protocol recommended by manufacturer (SGT410; CHEMICON International, Temecula, CA). Each test was triplicated and the results were calibrated with a corresponding phosphopeptide standard curve and control. The absorbance at 450 nm was measured on a spectrophotometer.

Data were reported as the mean plus or minus the standard deviation (SD). Significance levels were determined by SPSS 10.0 for Windows (SPSS, Chicago, IL).

In this study, the K562 cell line was used as a reference since data from several groups, including our own, showed that these cells were sensitive to the imatinib and arsenic trioxide.^21,22,24  Here, we examined the combined effects of As₄S₄ and imatinib on the growth inhibition of K562 cells. Next, we observed the effect of As₄S₄ and imatinib on fresh CML CD34⁺ cells for 48 hours. The median-effect principle was applied not only to analyze dose-response curves for each drug alone and combined but also to quantify synergism or antagonism at various proliferation inhibition levels. Median-effect plot showed the value of linear correlation coefficient of As₄S₄ and/or imatinib was superior to 0.95 (Figure 1A). Fa-CI plot indicated synergistic and additive effects between 2 μM and 4 μMAs₄S₄ and 0.2 μM and 0.4 μM imatinib after 48 hours on K562 cells. However, synergistic effects were mainly observed between 1 μM and 3 μM As₄S₄ and 0.1 μM and 0.3 μM imatinib on CD34⁺ cells among most primary CML samples.

The inhibitory concentration (IC₅₀) values (50% inhibition of proliferation) of As₄S₄ and imatinib in various cells are shown in Figure 1B. Imatinib produced no measurable effect in BCR-ABL–negative HL60 cells, U937 cells, and CD34⁺ cells from nonleukemic donors at therapeutically relevant concentrations (IC₅₀ > 10 μM).

As₄S₄ and imatinib showed different degrees of synergistic inhibitory effects on the growth of progenitors from distinct disease phases of CML (Figure 2). Among samples at CML-CP, additive and synergistic effects were observed between 0.5 μM and 2 μM As₄S₄ and 0.1 μM and 0.5 μM imatinib combination groups, whereas among those at CML-AP, a synergism was observed between all combinations except for the 0.5 μM As₄S₄/0.1 μM imatinib group. With regard to CFU total inhibition, the effects of As₄S₄ at distinct concentrations combined with 0.1 μM imatinib were more powerful in the CML-CP group than in the CML-AP group. However, when 2 μM As₄S₄ was combined with 0.2 μM imatinib, CML-AP cells responded in a similar way as CML-CP cells to the inhibitory effect of the 2 agents. In addition, the effect of the 2 μM As₄S₄ and 0.2 μM imatinib combination was stronger than that of 1 μM imatinib in the CML-CP group (P < .01, n = 13). It is interesting to note that the sensitivity of the colony-forming ability of normal CD34⁺ progenitors to As₄S₄ and/or imatinib was significantly lower than that of CML CD34⁺ cells.

The effects of imatinib and As₄S₄ on the cell cycle distribution of K562 cells were evaluated. As shown in Figure 3A, after 12 hours of exposure to 0.2 μM imatinib, K562 cells began to be arrested at G₁ phase. In contrast, an accumulation of cells arrested at G₂/M phase was observed when K562 cells were exposed to 2 μM As₄S₄ for 12 hours. Interestingly, after combined treatment of imatinib and As₄S₄, K562 cells were arrested at G₂/M phase before 18 hours, while a substantial number of cells were arrested at G₁ phase after 24 hours. To address the mechanism of cell cycle regulation under the 2 drugs, we examined some related protein. As shown in Figure 3B, no significant changes were observed of total CDK1 and Rb proteins upon effect of both compounds. However, imatinib profoundly reduced the phosphorylation of CDK1 and Rb, in contrast to As₄S₄, which exerted no obvious effect on the phosphorylation of the 2 proteins. When cells were exposed to combination treatment, the levels of phosphorylated CDK1 and Rb proteins showed similar changes as those under imatinib. Therefore, the regulatory effect of As₄S₄ on K562 cell cycle may involve a different set of proteins.

The combination of imatinib and As₄S₄ induced more apoptosis than either agent alone (Figure 4). Typical morphologic features of apoptosis, such as nuclear fragmentation and apoptotic bodies, were observed under light microscopic examination (Figure 4A). Since annexin V is the early indicator of apoptosis, the percentages of apoptotic cells were determined using annexin V/PI staining, after K562 cells and CD34⁺ cells from patients with CML were exposed to imatinib and/or As₄S₄ for 24 to 72 hours. Compared with either 0.2 μM to 0.3 μM imatinib or 2 μM to 3 μM As₄S₄ alone for 24 to 72 hours, combined treatment with these agents resulted in significantly more apoptosis of K562 cells in a dose- and time-dependent manner. The CD34⁺ cells from patients with CML with similar treatment exhibited significant apoptotic changes as well, and the results suggested synergistic effects of the drugs (Figure 4B). Parallel results were also obtained when loss of mitochondrial potential was monitored on K562 cells and CML CD34⁺ cells (Figure 4C). As shown in Figure 4D, combination treatment activated caspase-3 in a higher percentage of cells as compared with the result obtained with either drug alone (P < .05, n = 4). More 116 kDa PARPs were cleaved into an 85-kDa fragment by caspase-3 in the combination group as compared with the single treatment group (Figure 4E). Moreover, Western blot showed that Fas-L was up-regulated in the 2 μM As₄S₄ group and the cotreatment group with 0.2 μM imatinib and 2 μM As₄S₄, whereas no obvious change was found in CD95 expression in K562 cells (2.2% ± 0.2%, 2.4% ± 1.0%, 3.3% ± 0.9% versus 3.6% ± 1.5%, for control, 0.2 μM imatinib, 2 μMAs₄S₄ versus 0.2 μM imatinib and 2 μM As₄S₄ combination, respectively; n = 4, P > .05). In addition, Western blot results suggested that the apoptosis induced by 0.2 μM imatinib and 2 μM As₄S₄ was independent of Bcl-2 protein level.

After exposure to As₄S₄ and/or imatinib for 24 hours or 48 hours, the BCR-ABL protein levels and their PTK activities were assessed in K562 cells or primary cells. The 2 drugs exhibited synergistic effects not only on the degradation of BCR-ABL protein but also on the reduction of tyrosine kinase activity. Figure 5A demonstrated that exposure to 2 μM As₄S₄ for 48 hours reduced BCR-ABL protein levels in K562 cells whereas 0.2 μM imatinib could not. However, the cotreatment seemed to induce more degradation of BCR-ABL than did As₄S₄ alone. On the other hand, cotreatment led to more decline of PTK activity than either drug alone (P < .05, n = 3) and the synergistic effects were dose- and time-dependent. To test the effect of combined treatment in primary cells, we used fresh BMMNCs from 3 patients with CML-AP that contained a relatively high proportion of CD34⁺ cells (30%, 65%, and 75%, respectively). Similar results were observed in those samples as compared with K562 cells.

To our knowledge, the present work is the first to address the combined effect of imatinib and As₄S₄ on CML cells. Imatinib inhibits the activity of ABL tyrosine kinase in a highly specific way, whereas As₄S₄ may have a wide range of cellular targets by analogy to the related compound arsenic trioxide (As₂O₃).²⁹  In view of the distinct target profiles of the 2 agents, we selected the CI equation for mutually nonexclusive drugs (with independent modes of action) to study their combined use. Importantly, in both the short-term culture system (within 48 hours) or the relatively long CFU assay (14 days), there was a wide range of synergy between different concentrations of As₄S₄ and imatinib on CD34⁺ cells from 22 patients with CML. Overall, this synergy displayed a dose- and time-dependent feature. In our experiment, a high concentration of imatinib (at 2 μM) could mildly inhibit colony forming of the normal control (35.14% total CFUs, 32.36% CFU-GMs, and 36.6% on average, respectively). Similar results have been reported previously.³⁰  The underlying mechanism could be that high-concentration imatinib inhibits a number of wild-type tyrosine kinases activities (such as C-ABL, C-KIT, and platelet-derived growth factor receptor) that are essential for the normal CD34⁺ cell colony formation. When CFU assays were performed under the 2 μM As₄S₄ and 0.2 μM imatinib combination, the same level of inhibition could be reached in both CML-CP and -AP groups. However, the clonogenic capacity of normal CD34⁺ cells seemed much less sensitive to the same conditions. Thus, 2 μMAs₄S₄ and 0.2 μM imatinib could constitute an ideal combination in achieving therapeutic effect with limited side effects on the normal hematopoiesis. Previous pharmacokinetic studies on patients with APL showed that a standard dose of As₄S₄ achieved a plasma level of 1 μMto2 μM.²³  Under a standard dose of imatinib (300 mg/d to 600 mg/d), on the other hand, the plasma concentrations of the drug fluctuated in the range of 1 μM to 2 μM.⁸  Therefore, the effective in vitro concentrations of both drugs on CML CD34⁺ cells established in this study should also be achievable under an in vivo situation.

The synergistic effect of the imatinib and As₄S₄ combination in terms of antiproliferation might be connected with the distinct but complementary roles of the drugs in interfering with cell cycle progression. Our data showed that As₄S₄ induced G₂/M arrest of K562 cells, whereas imatinib induced G₁ arrest. When these 2 drugs were used simultaneously, the K562 cells were arrested at G₂/M phase at first and then underwent G₁ arrest together with apoptosis. A previous study showed that As₂O₃ disturbs the assembly of spindle microtubules and induces mitotic arrest in a variety of human cancer cell lines.³¹  A similar situation may also exist for As₄S₄. With regard to the cell cycle machinery, it is well known that the phosphorylation of Rb and dephosphorylation of CDK1 play important roles in the transition of G₁/S and G₂/M phases, respectively.^32,33  Rb is a negative regulator of cell proliferation and is inactivated by phosphorylation. Rb could be phosphorylated by RAS/MAPK/ERK–dependent cyclins, CDKs of mitogenic signals, and Fas/p38 kinase of apoptotic signals, leading to dissociation of E2F and increased transcriptional activity.^32,34  It has been reported that phospho-Rb regulation of E2F activity affects cellular potential for G₁/S phase transition in part via its effect on c-myc transcription.³⁵  On the other hand, phospho-CDK1 represents a repressed form of the protein while the dephosphorylation at Thr14/Tyr15 of CDK1 allows its activation and G₂/M progression. Our data demonstrated that imatinib induces significant down-regulation of phosphorylated Rb and CDK1, which is in agreement with the G₁/S but not G₂/M arrest under this drug. However, As₄S₄ shows no obvious effect on these proteins in spite of a visible effect on G₂/M block, suggesting the involvement of other mechanisms such as cytoskeleton regulation.

The cellular mechanisms of As₄S₄ and imatinib have been associated with the induction of apoptosis. It is well known that ΔΨm collapse plays an essential role in mediating apoptosis through the release of apoptotic mediators (eg, cytochrome c/Apaf1/caspase-9 and apoptosis-inducing factor) into cytoplasm, a pathway known as intrinsic apoptotic signaling.³⁶  Ligation of the cell-surface Fas molecule by its ligand, on the other hand, also known as the extrinsic pathway, results in the cleavage and activation of the cysteine protease procaspase-8.^37,38  The activation of procaspase-3 following caspase-8 or caspase-9 activation is considered to play a crucial role in apoptosis.^39,40  It was reported that As₂O₃-induced apoptosis coincided with ΔΨm collapse and activation of caspase-3 in APL cell line NB4 and primary APL cells.³¹  Here we showed that either As₄S₄ or imatinib alone could achieve similar results in CML cell lines and fresh cells. However, the combination of the 2 agents yielded much more significant reduction of ΔΨm and activation of caspase-3, followed by cleavage of PARP and expression of phosphatidyl serine on the cellular surface membrane (detected as annexin V), then displayed morphogenesis features of apoptosis. In addition, the protein level of Fas-L seems to be up-regulated by As₄S₄ and combination treatment, which may cause an autocrine mechanism and lead to an enhancement of the extrinsic pathway of apoptosis.

The molecular basis of CML is the BCR-ABL fusion protein. BCR-ABL induces deregulation of cell proliferation, inhibition of apoptosis, and adhesion abnormality to marrow stroma. Moreover, some important downstream pathways have been significantly altered.²⁵  Therefore, the application of imatinib in CML has been considered as a molecular triumph of cancer research in the genomic era. On the other hand, resistance to imatinib in a number of patients after long-time use has been ascribed to an elevated expression the BCR-ABL gene or to the mutation at key positions of the BCR-ABL coding sequence around the ATP-binding domain.¹⁶  Interestingly, our study showed that one of the targets of As₄S₄ is also BCR-ABL protein. Unlike imatinib, which inhibits the tyrosine kinase activity of BCR-ABL but apparently does not affect the turnover of the protein, As₄S₄ triggers the degradation of BCR-ABL, accompanied by a reduction of the tyrosine kinase activity. Although it is still unknown whether As₄S₄ influences the stability of BCR-ABL in a direct or indirect way, its mode of action seems to be quite different, but highly complementary, to that of imatinib. To this end, combination of imatinib and As₄S₄ represents a new model of synergistic targeting at the molecular level and might be a promising approach to improve response rates and prevent resistance. As a matter of fact, a clinical trial using imatinib and As₄S₄ has recently been launched at the Shanghai Institute of Hematology, to evaluate the potential benefit of the combination of these drugs to patients with CML.

In conclusion, our study suggests As₄S₄ and imatinib have synergistic effects in inhibiting proliferation, inducing apoptosis of cells, and reducing the tyrosine kinase activity of BCR-ABL. Whether this in vitro synergistic activity will translate into better response and survival rates of patients with CML needs to be proven in clinical trials.

We are very grateful to all members of the State Key Laboratory of Medical Genomics and the Shanghai Institute of Hematology for their encouragement and support.