In vitro and in vivo single-agent efficacy of checkpoint kinase inhibition in acute lymphoblastic leukemia

Background Although progress in children, in adults, ALL still carries a dismal outcome. Here, we explored the in vitro and in vivo activity of PF-00477736 (Pfizer), a potent, selective ATP-competitive small-molecule inhibitor of checkpoint kinase 1 (Chk1) and with lower efficacy of checkpoint kinase 2 (Chk2). Methods The effectiveness of PF-00477736 as single agent in B-/T-ALL was evaluated in vitro and in vivo studies as a single agent. The efficacy of the compound in terms of cytotoxicity, induction of apoptosis, and changes in gene and protein expression was assessed using different B-/T-ALL cell lines. Finally, the action of PF-00477736 was assessed in vivo using leukemic mouse generated by a single administration of the tumorigenic agent n-ethyl-n-nitrosourea. Results Chk1 and Chk2 are overexpressed concomitant with the presence of genetic damage as suggested by the nuclear labeling for γ-H2A.X (Ser139) in 68 % of ALL patients. In human B- and T-ALL cell lines, inhibition of Chk1/2 as a single treatment strategy efficiently triggered the Chk1-Cdc25-Cdc2 pathway resulting in a dose- and time-dependent cytotoxicity, induction of apoptosis, and increased DNA damage. Moreover, treatment with PF-00477736 showed efficacy ex vivo in primary leukemic blasts separated from 14 adult ALL patients and in vivo in mice transplanted with T-ALL, arguing in favor of its future clinical evaluation in leukemia. Conclusions In vitro, ex vivo, and in vivo results support the inhibition of Chk1 as a new therapeutic strategy in acute lymphoblastic leukemia, and they provide a strong rationale for its future clinical investigation. Electronic supplementary material The online version of this article (doi:10.1186/s13045-015-0206-5) contains supplementary material, which is available to authorized users.


Background
Acute lymphoblastic leukemia represents a biologically and clinically heterogeneous group of B/T-precursor-stage lymphoid cell malignancies arising from genetic insults that block lymphoid differentiation and drive aberrant cell proliferation and survival. Survival rates for children are approximately 80-85 % with current risk-oriented treatment protocols in contrast to less than 40 % for adults [1][2][3]. Although the outcome of children with relapsed ALL is heterogeneous, many achieve lasting second remissions. In contrast, survival after relapse in adult ALL is short [4,5]. New therapeutic strategies are therefore needed to improve remission rates in adults and to prevent relapse both in adult and pediatric ALL patients.
Recently, in order to enhance DNA-damaging effects inflicted by cytotoxic drugs or radiation, different checkpoint kinase (Chk) inhibitors have been developed and assessed alone or in combination with DNA-damaging agents in preclinical studies and in phase I/II trials for cancer therapy [6][7][8][9][10][11][12][13][14]. Some of them inhibit both Chk1 and Chk2; others have a higher specificity for Chk1. The biological background underlying the development of Chk1 inhibitors and preliminary data deriving from early clinical trials have been recently well reviewed in [15]. Following DNA damage, multiprotein complexes recruit the transducers ataxia telangiectasia mutated (ATM) and ataxia telangiectasia and Rad3-related protein (ATR) which activate Chk2 and Chk1, respectively, despite an existing extensive crosstalk between the ATR-Chk1 and ATM-Chk2 pathways. Once activated by phosphorylation at S317/S345 in response to DNA single strand breaks, Chk1 phosphorylates the phosphatase Cdc25A at the G1/S and S phase checkpoints preventing cells from entering S phase and Cdc25A and Cdc25C at the G2/M checkpoint avoiding the entry in mitosis [16][17][18][19]. If Chk1 is induced by single strand breaks, Chk2 activation is largely restricted to DNA double strand breaks via ATM [20]. Once activated, they mediate cell cycle delay, DNA repair, and apoptosis in response to DNA damage. Moreover, Chk1 is required for proper mitotic spindle assembly and maintenance of chromosomal stability during mitosis [21].
PF-00477736 is a potent, selective, ATP-competitive, small-molecule inhibitor of Chk1, synthesized by Pfizer, which is selective for Chk1 and shows general selectivity over other kinases [6]. Preclinically, PF-00477736 enhanced docetaxel activity in tumor cells and xenografts by abrogating the mitotic spindle checkpoint, as well as the DNA damage checkpoint [10]. The checkpoint abrogating and cytotoxic activities attributed to PF-00477736 in combination with chemotherapy agents (e.g., gemcitabine and carboplatin) showed selectivity for p53-defective cancer cell lines over p53-competent normal cells in vitro [6]. In xenografts, PF-00477736 enhanced the antitumor activity of gemcitabine in a dose-dependent manner. PF-00477736 combinations were well tolerated with no exacerbation of side effects commonly associated with cytotoxic agents [6].
In this study, we investigated the activity of PF-0477736, as a single agent, in B-/T-ALL by the following: assessment of the in vitro efficacy in ALL cell lines and primary blast cells, assessment of the in vivo efficacy in mouse models, and identification and validation of potential biomarkers of functional inhibition. Results demonstrated that in vitro treatment of B-/T-ALL cell lines and primary blast cells with PF-0477736 resulted in inhibition of cell viability, induction of DNA damage, and apoptosis. Moreover, in vivo studies confirmed the efficacy of Chk1 inhibition, suggesting that this therapeutic strategy may be promising in leukemia.
PF-00477736 reduces cell viability in a dose-dependent manner B-/T-ALL cell lines were incubated with increasing concentrations of drug (5-1000 nM) for 24 and 48 h. PF-00477736 inhibition resulted in dose-and time-dependent cytotoxicity with RPMI-8402 (T-ALL) being the most sensitive (IC 50 = 57.4 nM at 24 h) while NALM-6 (B-ALL) , and they are the mean of at least two different replicates. c Immunohistochemical profile of the 60 ALL (36 B-ALL and 24 T-ALL); a red box indicates expression of the marker, whereas a green box signifies negativity (samples were considered positive if 30 % or more of the cells were stained with an antibody); gray box indicates a not evaluable reaction due to a core loss in TMA. The figure allows the assessment of the co-expression of the DNA damage markers in individual samples. d Immunohistochemical expression of Chk1, phosphorylated Chk1 (Ser345), Chk2, phosphorylated Chk2 (Thr68), and phosphorylated H2A.X (Ser139) (γ-H2A.X) in thymus, reactive follicle (RF), T-ALL and B-ALL. The figure highlights the close similarity between expression profiling of B-and T-ALL, characterized by a high percentage of cases Chk1 + , pChk1 + , Chk2 + , Cdc25C, pCdc25C, pH2A.X + and, in a lesser extent, pChk2 + , and its distinction with the expression profiling of the non-tumoral populations as maturing thymocytes, which were substantially negative for these seven proteins and as B lymphocytes of reactive follicles, which resulted Chk1 weakly + , Chk2 + , Cdc25C + but negative for the corresponding phosphorylated forms and pH2A.X.  (Fig. 2a), and with baseline levels of Chk1/2 and ATR/ATM phosphorylation, indicative of intrinsic genetic stress (Fig. 2b).

PF-00477736 induces apoptosis at 24 and 48 h in B-/T-ALL cells
In order to assess whether cytotoxicity was correlated to increased susceptibility to apoptosis, B-/T-ALL cell lines were incubated with increasing concentrations of drug (0.1, 0.5, and 1 μM and 0.05, 0.1, and 0.2 μM only for BV-173 and RPMI-8402, respectively) for 24 and 48 h. Consistent with the viability results, Annexin V/propidium iodide (PI) staining analysis showed a significant increase of apoptosis at 24 and 48 h in B-and T-ALL cells proportional to drug dose and drug exposure times (Additional file 1: Figure S3-A). The induction of apoptosis by PF-00477736 was also assessed by the detection of poly (ADP-ribose) polymerase (PARP) cleavage by Western blot analysis [23]. The PARP-1 cleavage band at 89 kDa was observed in lysates from leukemia cell lines after 24 h of drug exposure, while it was undetectable in cells treated with only DMSO 0.1 % (Fig. 2c).

PF-00477736 perturbs the cell cycle profile in B-/T-ALL cells
The effect of the inhibition of Chk1 pathway on cell cycle progression was evaluated in RPMI-8402, BV-173, SUP-B15, and NALM-6 cell lines. Cells were incubated with increasing concentrations of PF-00477736 for 6 and 24 h and then stained with propidium iodide to quantify the DNA amount. The effect of PF-00477736 on the cell cycle progression was very weak after 6 h of incubation (data not showed) and very heterogeneous among the different cell lines after 24 h. In RPMI-8402, BV-173, and SUP-B15 cell lines, the treatment induced a progressive reduction of the number of cells in S and G2/M phase and a concomitant increment of cell debris. In the less sensitive cell line, NALM-6, the inhibition of Chk1 progressively increased the percentage of cells in G2/M phase and reduced the percentage of cells in G1 phase (Additional file 1: Figure S3-B).

PF-00477736 efficiently targets Chk1 pathway and induces DNA damage
In order to assess whether PF-00477736 efficiently targeted Chk1 pathway, we examined the changes in the downstream phosphorylation of the phosphatase Cdc25C and cyclin-dependent kinase, Cdc2. Moreover, Chk1 itself and γ-H2A.X, which is a marker of stalling replication forks and checkpoint abrogation-induced apoptosis, have been assessed. Functional analyses were initially performed on the most sensitive (BV-173) and the most resistant (NALM-6) B-ALL cell lines, as determined by the viability and apoptosis analyses. Specifically, BV-173 cells were treated with 0.05, 0.1, and 0.2 μM and NALM-6 with 0.1, 0.5, and 1 μM of PF-00477736 for 24 h (Fig. 3a). In the second analysis, BV-173 and NALM-6 cells were treated with the dose of PF-00477736 that goes near to the IC 50 (0.1 μM for BV-173 and 1 μM for NALM-6) for 18, 24, 30, and 48 h (Fig. 3b). Thereafter, the same molecular targets have been assessed on all B-and T-ALL cell lines using the dose that goes near to the IC 50 (Additional file 1: Figure  S4). Active Chk1 phosphorylates Cdc25C at serine 216 throughout the interphase and upon G2 checkpoint activation. This leads to the nuclear export of Cdc25C and its subsequent cytoplasmic sequestration by 14-3-3 protein, which prevents the activation of the downstream target of Cdc25C, the cyclin B/Cdc2 kinase that is responsible for G2/M transition [24]. Upon inhibition of Chk1, we observed decreased levels of phosphorylated Cdc25c (Ser216) and Cdc2 (Tyr 15) as well as reduced levels of their total forms in a dose-and time-dependent manner, suggesting that PF-00477736 efficiently targets Chk1 pathway. In contrast, the phosphorylated forms of Chk1 (Ser317 and Ser345) were dose-dependently increased suggesting that treatment amplifies genomic damage which in a feedback loop increases Chk1 phosphorylation. Although the increment of the phosphorylated forms of Chk1 (Ser317 and Ser345) as a consequence of genomic damage amplification, the complete inhibition of Chk1 functionality was confirmed by the progressive reduction of the autophosphorylated form of Chk1 (Ser296) in an early time point analysis. Indeed, even after 3 h of treatment with PF-00477736 in both BV-173 and NALM-6 cell lines, the auto-phosphorylated form of Chk1 (Ser296) was drastically reduced in comparison with the untreated counterpart (Additional file 1: Figure S4-B). The enhancement of  The phospho-H2A.X accumulates and forms characteristic nuclear foci where the DNA is damaged [25]. Both NALM-6 and BV-173 cell lines show an increased number of γ-H2A.X foci when treated with PF-00477736 compared to untreated controls (Fig. 3c). However, evaluating the mean fluorescence intensity of γ-H2A.X positive cells, BV-173 cells seem to be more damaged than NALM-6. Actually, BV-173 treated cells have five-fold change difference of mean value intensity of γ-H2A.X compared to BV-173 untreated cells, while NALM-6 treated cells have 1.4-fold change difference of mean value intensity of γ-H2A.X compared to their untreated counterparts. In addition to these changes in the overall amount of γ-H2A.X foci among the two different cell types, we observed the presence of hyper-γ-H2A.X-positive cells specifically in BV-173 cells treated with PF-00477736. The so-called hyper-γ-H2A.X-positive cells loose the typical foci signal of γ-H2A.X and emit a much higher and diffuse signal of γ-H2A.X positivity. Almost 6.5 % of the γ-H2A.X-positive cells are hyper-γ-H2A.X-positive in BV-176-treated cells compared to 1.8 % in the untreated controls. These data suggest a higher induction of DNA damage upon PF-00477736 treatment in BV-173 than NALM-6 cells as expected by the fact that BV-173 cells are more responsive to PF-00477736. However, the increase of hyper-γ-H2A.X-positive cells only in BV-176-treated cells would suggest the existence of a specific effect that causes a hyper DNA damage only in a restricted, but significant, population of leukemic cells.
Interestingly, upon treatment, we observed a strong reduction of Chk1. As already hypothesized, this may be the result of cleavage by caspase during apoptosis induced by genotoxic stress [26,27]. Persisting protein levels of Cdc25, pChk1, Ser345, pCdc2 (Tyr 15), and Cdc2 after 48 h of exposure to PF-00477736 differentiated less sensitive leukemia cells from more sensitive ones (Additional file 1: Figure S3c).

Gene expression profiling results
In order to identify gene expression changes specifically correlated with Chk1 inhibition and to better elucidate the mechanism of action of Chk1 inhibitor, gene expression profiling (GEP) analysis was performed by microarray on  Fig. 4). To identify peculiar critical pathways affected by Chk1 inhibition, the differentially expressed genes were analyzed in terms of biological function using the MetaCore pathway mapping software (Gen-eGo Inc.) which categorizes genes in pathway maps, Gene Ontology (GO) cellular processes, and cellular and molecular process networks. Consistent with a Chk1's mechanism of action, the three top scored maps (map with the lowest p value) were "cell cycle: estrogen receptor 1 (ESR1) regulation of G1/S transition" including c-Jun/c-Fos, cyclin A, CARM1, Skp2/TrCP/FBXW, SKP2, c-Fos, CDK4, NCOA3, c-Jun, and CDK2 genes (Additional file 1: Figure S5); "apoptosis and survival: granzyme A signaling" including PHAP1, Ku70/80, Ku80, Ku70, histone H3, NDPK A, SET, histone H2B, and histone H1 genes (Additional file 1: Figure S6); and "DNA damage: ATM/ ATR regulation of G1/S checkpoint" including PCNA, growth arrest and DNA damage 45 alpha (GADD45a), Chk2, cyclin A, NF-kB, cyclin-dependent kinase 4 (CDK4), FANCD2, Claspin, and CDK2 genes (Additional file 1: Figure S7). GO cellular processes highlighted 617 genes involved in cellular metabolic process (Additional file 3). Finally, the three top scored process networks impaired by the treatment with Chk1 inhibitor were as follows: the DNA damage checkpoint, the S phase of cell cycle, and the apoptosis, confirming the specificity of treatment on its target pathways (Additional file 4). Thirty-five genes had a false discovery rate less than 0.05 (Additional file 5: Table S1), and the three most differentially expressed genes were as follows: DNA damage-inducible transcript 3 (DDIT3, fold-change 3.32, p value 6.06 × 10 −5 ), Kruppellike factor 6 (KLF6, fold-change 2.17, p value 8.41 × 10 −5 ), and FBJ murine osteosarcoma viral oncogene homolog (FOS, fold-change 2.40, p value 1.97 × 10 −4 ). Due to its implication in the regulation of the cell cycle and apoptosis, Western blot analysis against c-Jun was performed to better comprehend its biological role in response to PF-00477736. c-Jun is a component of the transcription factor activating protein-1 (AP-1) which is involved in cell cycle progression through the G1 phase by the regulation of the cyclin D1 [28]. As soon after treatment with Chk inhibitor, c-Jun protein levels increased in all B-ALL cell lines but  Figure S8-A). The data found in the gene expression profile analysis were validated using qPCR. Four genes (two upregulated, PLK3 and GADD45a; two downregulated, Chk2 and CDK4), which have been chosen on their biological relevance in the Chk1 pathway, were validated on B-/T-ALL cell lines treated for 24 h with or without PF-00477736 (IC50) (Additional file 1: Figure S8-C).

PF-00477736 impairs survival of leukemic mice
We extended the in vitro and ex-vivo studies by assessing the efficacy of Chk inhibitor in mice transplanted with T-ALL.
Leukemic mice were generated by the usage of the tumorigenic agent N-ethyl-N-nitrosourea (ENU). The obtained leukemia has been immunophenotyped and characterized as a T-ALL. Leukemic blasts coming from the spleen of leukemic animals were transplanted into C57BL/ 6Ly5.1-recipient mice. Three days after transplantation, a time sufficient for leukemic blasts to home the bone marrow of the host, we started to treat the animals. A concentration of 40 mg/kg of PF-00477736 given with a q3dx4 schedule seems not to be toxic for the animals; however, it significantly affects the overall survival of the treated animals (five mice) in comparison to the untreated ones (seven mice) (p value =0.0009). Actually, control animals (intraperitoneally injected with PBS) die all at day 17 posttransplantation while the animals treated with PF-00477736 live longer (up to 59 days post transplantation). The leukemia used for the experiments shown here was very aggressive (untreated animals die with massive infiltration of the spleen and the liver by leukemic blasts at day 17) suggesting a reason why the effect of the Chk inhibitor is significant though quite variable among treated mice (Fig. 5c).

Discussion
Several studies in different tumors have investigated the role of Chk inhibition in combination with conventional chemotherapy demonstrating that this combination enhances tumor cell death [6,10].
Here, we evaluated the in vitro and in vivo effects of the administration of PF-00477736 as a single agent and not in combination with other chemotherapeutic agents in cell lines and primary blast cells from acute lymphoblastic leukemia based on the hypothesis that the intrinsic genomic instability of leukemic clones as demonstrated by the nuclear labeling for γ-H2A.X molecule in 68 % of ALL patients may be itself sufficient to bring the cells to apoptosis. According to this hypothesis, it was possible to highlight how the administration of PF-0477736 as single agent was able to reduce cell viability in all leukemia cell lines treated in this study (BV-173, SUP-B15, REH, NALM-6, NALM-19, MOLT-4, RPMI-8402, and CCRF-CEM) and to induce apoptosis (Fig. 6). Different leukemia cell lines showed different sensitivity to PF-00477736 with RPMI-8402 (T-ALL) being the most sensitive (IC 50 = 57.4 at 24 h) while NALM-6 (B-ALL) the less sensitive (IC 50 = 1423.0 at 24 h) cell line. Interestingly, the sensitivity to Chk1 inhibitor was not related to the mutational status of the tumor suppressor p53.
The results obtained on cell viability and induction of apoptosis have been confirmed by genome-wide studies evaluating global gene expression changes upon treatment and by functional studies performed by Western blot analysis. Interestingly, by GEP analysis, the majority of genes downregulated were involved in the DNA damages response and in particular in DNA repair mechanisms, while the genes upregulated were involved in chromatin assembly, nucleosome organization, DNA packaging, and apoptosis. The efficacy of Chk inhibition has been evaluated and confirmed in terms of reduction of cell viability in primary ALL blast cells but not in normal bone marrow precursor cells. Furthermore, assessing the efficacy of Chk inhibition in mice transplanted with Tlymphoid leukemia, we demonstrated that PF-0477736 increases the survival of treated mice compared with mice treated with vehicle (p = 0.0016).
In conclusion, in vitro, ex-vivo, and in vivo results support the inhibition of Chk1 as a new therapeutic strategy in acute lymphoblastic leukemia, and they provide a strong rationale for its future clinical investigation. The checkpoint kinase inhibitor have been synthetized to increase the effectiveness of conventional chemotherapy, preventing cells to arrest cell cycle, and to repair the DNA damages caused by the exposure to genotoxic agents. Here, we highlight that even the treatment with a checkpoint kinase inhibitor alone associated with the high genetic instability can be enough to kill cancer cells. Indeed, we believe that leukemic cells, thanks to a higher activation of the ATR-Chk1/ATM-Chk2 pathways, can better tolerate high genetic instability. Switching off these mechanisms of survive, we can kill leukemic cell by overcoming the cell cycle checkpoint, by inhibiting the mechanisms of DNA damages repair and by inducing massive damages that cannot be tolerated (Fig. 6).

Conclusions
Inhibition of Chk1/2 represents a novel therapeutic strategy to overcome genetic instability and to promote selective killing of leukemia cells in ALL.

Reagents
The Chk inhibitor PF-00477736 was purchased from Sigma-Aldrich (Sigma-Aldrich Co. St. Louis, Missouri 63103 United States). qPCR analysis of Chk1 and Chk2 mRNA expressions was evaluated in 41 (76 %) newly diagnosed BCR-ABL1-positive (median age 56 years, range 26-81; ratio male to female 20/21) and 13 (23 %) newly diagnosed BCR-ABL1-negative ALL cases (median age 41 years, range 18-69; ratio male to female 5/8). One microgram of RNA was reverse transcribed using the High-Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA, USA). PCR analysis was performed using HS00967506_m1 (Chk1) and HS00200485_m1 (Chk2) assays (Applied Biosystems) and the Fluidigm Dynamic Array 48 × 48 system, a real-time qPCR assay which enables to automatically assemble 48 samples and 48 assays to create individual TaqMan reactions of a final volume of 6.75 nl each (Fluidigm, San Francisco, CA, Fig. 6 Cartoon drawing the rational of the study: in leukemia cells, oncogenes could lead to replication stress, activation of ATR/Chk1 pathway that in conjunction with elevated proliferation promote tolerable level of DNA damage, genomic instability, and tumor progression. Inhibition of Chk1/Chk2 could increase DNA damage leading to apoptosis USA; http://www.fluidigm.com/). RNA integrity was confirmed by PCR amplification of the GAPDH mRNA (Hs99999905_m1), which is expressed ubiquitously in human hematopoietic cells. Results were expressed as 2exp(−ΔΔCt). GraphPad Prism 5 software (GraphPad, Avenida de la Playa La Jolla, CA, USA) was used to plot the data. The basal mRNA expression of Chk1 and of Chk2 was evaluated in all the cell lines using the same assays (Applied Biosystems: HS00967506_m1 Chk1 and HS00200485_m1 Chk2) used for the primary samples, and the statistical validity of the results were confirmed using ANOVA multiple comparisons test (GraphPad Prism 5 software). In addition, reverse transcription and quantitative PCR analysis for GADD45a, PLK3, CDK4, and CHK2 in treated cell lines and on their untreated counterpart was performed as described above using the following assays from Applied Biosystems: Hs00169255_m1 (GADD45a), Hs00177725_m1 (PLK3), Hs00262861_m1 (CDK4), and HS00200485_m1 (Chk2).

TP53 mutation screening
Total cellular RNA was extracted using the RNeasy total RNA isolation kit (Qiagen, Valencia, CA, USA). One microgram of total RNA was reverse transcribed  Table S3) in 25 μl reaction volumes. PCR products were purified using QIAquick PCR Purification Kit (Qiagen) and then directly sequenced using an ABI PRISM 3730 automated DNA sequencer (Applied Biosystems, Foster City, CA, USA) and a Big Dye Terminator DNA sequencing kit (Applied Biosystems). All sequence variations were detected by comparison using the BLAST software tool (www.ncbi.nlm.nih.gov/BLAST/) to reference genome sequence data [GenBank:NM_000546.4].

Immunohistochemistry
Studies evaluating Chk1, phosphorylated Chk1 (Ser345), Chk2, phosphorylated Chk2 (Thr68) Cdc25C, phosphorylated Cdc25C (Ser216), and phosphorylated H2A.X (Ser139) (γ-H2A.X) protein expressions were performed on FFPE samples corresponding to three thymuses, three reactive lymph nodes, and to primary tumors collected from 60 ALL patients at diagnosis, including 36 B-ALL (median age 50 years, range 5-86; ratio male to female 18/ 18) and 24 T-ALL (median age 38 years, range 2-76; ratio male to female 15/9). All the cases were retrieved from the archives of the Haematopathology Unit, Department of Experimental, Diagnostic and Specialty Medicine-DIMES, University of Bologna. The study was conducted according to the principles of the Declaration of Helsinki after approval of the Internal Review Board. Two different tissue microarrays (TMAs) were constructed from these paraffinembedded blocks as previously reported. TMA sections were investigated by antibodies raised against fixation resistant epitopes. The antibody reactivity and sources as well as the antigen retrieval protocols, dilutions, and revelation systems are detailed in Additional file 5: Table S4. A cutoff of staining >30 % of the examined cells was assigned as positive score, according to formerly defined criteria [29]. Immunohistochemical preparations were visualized, and images were captured using Olympus dotSlide microscope digital system equipped with the VS110 image analysis software.
Normal bone marrow progenitors were harvested from bone marrow aspirations performed in lymphoma patients undergoing initial staging procedures, which then resulted negative for lymphoma infiltration. Normal peripheral blood progenitors were harvested from healthy donors.

Cell viability assay
In order to assess the cell viability after treatment with PF-00477736 (Pfizer), ALL cell lines were seeded in 96well plates at 50,000 cell/100 μl/well with increasing concentrations of drug (0.005-2 μM) for 24 and 48 h and incubated at 37°C. Cell viability was assessed by adding WST-1 reagent (Roche Applied Science, Basel, Switzerland) to the culture medium at 1:10 dilution. Cells were incubated at 37°C, and the optical density was measured by microplate ELISA reader at λ = 450 after 3 h. The amount of the formazan formed directly correlates to the number of metabolically active cells. All viability experiments were performed in triplicates and repeated in least two separated experiments. In ex-vivo primary leukemia cells, the effects on cell viability were assessed by counting viable and non-viable cell numbers by the trypan blue dye exclusion method. Cells were seeded in six-well plates at 500,000 cell/1 ml with increasing concentrations of drug (0.1, 0.5, and 1 μM) for 24 h and incubated at 37°C. Cellular viability was calculated as a percentage of the viable cells compared to the untreated controls (DMSO 0.1 %).

Annexin V staining of apoptotic cells
According to the WST-1 results, three different increasing concentrations of PF-00477736 were used to treat leukemia cells lines in order to detect and discriminate apoptotic, necrotic and dead cells. Cell lines were seeded in 12-well plates at 500,000 cell/1 ml with increasing concentrations of drug (0.1, 0.

Immunofluorescence analysis
NALM-6 and BV-173 cells were seeded to poly-D lysinecoated slides, fixed with 4 % PFA (paraformaldehyde), and stained at 37°C with a mouse anti-H2A.X-Phosphorylated Alexa 647 conjugated antibody (BioLegend). Then, they were treated with DAPI (4.6 diamidino-2-phenylindole; Sigma Aldrich), and the slides were mounted with Mowiol (Calbiochem). Images were acquired with wide field fluorescence microscope Olympus BX61 fully motorized driven by Metamorph software, and the analysis was performed using the freeware ImageJ software.

Gene expression profiling
Gene expression profiling on treated and untreated cells (DMSO 0.1 %) after 24 h of exposure to PF-0077736 was performed using Affymetrix GeneChip Human Gene 1.0 ST platform (Affymetrix Inc. Santa Clara, CA, USA) and following manufacturers' instructions. Raw data were normalized by using the RMA algorithm and filtered. Genes differentially expressed were selected by analysis of variance (ANOVA) (p value threshold =0.05) using the Partek Genomics Suite software (Partek Incorporated Saint Louis, MO, USA; http://www.partek.com). The most significantly involved process networks were defined by the Metacore software (GeneGo Inc., www.genego.com).

Cell cycle analysis
In order to evaluate the effect of PF-00477736 on cell cycle progression, NALM-6, SUPB-15, RPMI-8402, and BV-173 cell lines were seeded in a 24-well plate at 500,000 cell/1 ml with increasing concentrations of PF-00477736 (0.1, 0.5, and 1 μM for NALM-6 and SUP-B15; 0.05, 0.1, and 0.2 μM for RPMI-8402 and BV-173). After 6 and 24 h, cells were harvested and fixed overnight using ethanol at 70 %. Then, cells were stained using PI/RNase Staining Buffer according to the manufacturer's instruction (BD Pharmingen), and the cell cycle profile was detected using FACSCanto II instrument. The percentage of the different cell cycle phases was performed using the DNA cell-cycle analysis software for flow cytometry data, ModFit LT (Verity).

In vivo studies
Experiments involving mice were performed in agreement with Italian guidelines and after the approval of the Institutional Review Board of the European Institute of Oncology. Leukemic mice were generated by a single administration of the tumorigenic agent N-ethyl-N-nitrosourea (Sigma, 50 mg/kg intraperitoneally). Spleen cells from leukemic C57BL6/Ly5.2 mice were injected intravenously (2 × 10 6 cells/mouse) into nonirradiated, recipient C57BL6/Ly5.1 mice. PF-0077736 was administered intraperitoneally starting from day 3 after the transplant of leukemic cells. Mice received doses of PF-0077736 (40 mg/kg each dose) every 3 days for four treatments (q3dx4). Kaplan-Meyer rank test was used to compare the survival rate between treated and control mice.