Pontin Is a Critical Regulator for AML1-ETO-Induced Leukemia

ORIGINAL ARTICLE Pontin is a critical regulator for AML1-ETO-induced leukemia

O Breig1, S Bras1, N Martinez Soria2, D Osman1,3, O Heidenreich2, M Haenlin1 and L Waltzer1

The oncogenic fusion protein AML1-ETO, also known as RUNX1-RUNX1T1 is generated by the t(8;21)(q22;q22) translocation, one of the most frequent chromosomal rearrangements in acute myeloid leukemia (AML). Identifying the genes that cooperate with or are required for the oncogenic activity of this chimeric transcription factor remains a major challenge. Our previous studies showed that Drosophila provides a genuine model to study how AML1-ETO promotes leukemia. Here, using an in vivo RNA interference screen for suppressors of AML1-ETO activity, we identiﬁed pontin/RUVBL1 as a gene required for AML1-ETO-induced lethality and blood cell proliferation in Drosophila. We further show that PONTIN inhibition strongly impaired the growth of human t(8;21) þ or AML1-ETO-expressing leukemic blood cells. Interestingly, AML1-ETO promoted the transcription of PONTIN. Moreover, transcriptome analysis in Kasumi-1 cells revealed a strong correlation between PONTIN and AML1-ETO gene signatures and demonstrated that PONTIN chieﬂy regulated the expression of genes implicated in cell cycle progression. Concordantly, PONTIN depletion inhibited leukemic self-renewal and caused cell cycle arrest. All together our data suggest that the upregulation of PONTIN by AML1-ETO participate in the oncogenic growth of t(8;21) cells.

Leukemia (2014) 28, 1271–1279; doi:10.1038/leu.2013.376 Keywords: acute myeloid leukemia; AML1-ETO; PONTIN/RUVBL1; genetic screen; Drosophila

INTRODUCTION Genetic screens in Drosophila have been successfully used to The development of hematopoietic malignancies is frequently identify genes implicated in different human pathologies 16 associated with the presence of recurrent chromosomal rearran- including cancers. Consequently, we have developed a gements that promote blood cell transformation by affecting key Drosophila model to study AML1-ETO. Indeed several aspects regulators of hematopoiesis. One of the most conspicuous is the of blood cell development have been conserved from Drosophila 17 t(8;21)(q22;q22) translocation, which is present in ±12% of all to humans. Notably, reminiscent of AML1 function in cases of acute myeloid leukemia (AML) and brings together AML1 mammals, the RUNX factor Lozenge (LZ) is required for the (also known as RUNX1) and ETO (also known as RUNX1T1).1,2 AML1 differentiation of crystal cells, one of the main Drosophila blood encodes a transcription factor of the RUNX family essential for cell lineages. AML1-ETO expression in this lineage antagonizes several steps of blood cell development and mutations or LZ function, induces a preleukemic-like phenotype characterized translocations affecting AML1 are associated with a variety of by the accumulation of crystal cell progenitors at the larval 18 malignant hemopathies.3 ETO encodes a transcriptional co- stage and causes lethality. Hence, this model provides a repressor that does not seem to participate in normal blood cell genetically tractable system to investigate the conserved basis of 18,19 development. The t(8;21) translocation results in the expression of leukemogenesis. the fusion protein AML1-ETO that comprises AML1 N-terminus, Using an in vivo RNA interference screen, we identified pontin/ including its conserved DNA-binding domain, and the almost RUVBL1 as a suppressor of AML1-ETO in Drosophila. PONTIN and its entire ETO protein. AML1-ETO can compete with AML1 for binding related partner REPTIN/RUVBL2 are conserved proteins of the to DNA and to some of its partners, while it multimerizes and AAA þ (adenosine triphosphatase associated with diverse cellular 20 interacts with other transcriptional regulators via its ETO moiety.4 activities) family. They are found in different complexes involved Therefore, AML1-ETO acts by interfering with the transcriptional in transcriptional regulation, DNA repair or ribonucleoprotein regulation of AML1 target genes but also by deregulating directly assembly. Moreover, several lines of evidence indicate that or indirectly the expression of other genes.5–7 PONTIN participates in cell growth and survival, notably in Cellular and animal models developed to explore how AML1-ETO hepatocellular carcinoma.21 However, its function in AML has promotes leukemia revealed that it acts as a dominant suppressor of not been studied. Our data show that knocking-down PONTIN AML1, impairs myeloid differentiation and promotes hematopoietic inhibits t(8;21) þ cells proliferation and that AML1-ETO expression progenitor self-renewal.8 However, AML1-ETO is not sufficient to sensitizes cells to PONTIN depletion. Moreover, AML1-ETO induce leukemia and secondary mutations are required for disease activates the expression of PONTIN and REPTIN. Genome-wide progression.9–11 Conversely, only a few targets or partners of AML1- expression profiling revealed that PONTIN gene signature ETO have been shown to be important for its oncogenic activity.12–15 correlates with that of AML1-ETO and indicates that PONTIN To gain further insight into t(8;21) leukemia and open new chiefly regulates cell cycle-associated genes. Furthermore, PONTIN avenues for therapeutic intervention, it is critical to delineate the appears required for Kasumi-1 clonogenic potential and cell cycle gene networks implicated in leukemogenesis. progression. We propose that PONTIN is a critical component for

1CNRS, CBD UMR5547, Universite´ de Toulouse, UPS, CBD (Centre de Biologie du De´veloppement), Baˆtiment 4R3, 118 route de Narbonne, Toulouse, France and 2Northern Institute for Cancer Research, University of Newcastle, Newcastle upon Tyne, UK. Correspondence: Dr L Waltzer or Dr M Haenlin, Centre de Biologie du De´veloppement, CNRS, Universite´ de Toulouse, BA˜ btiment 4R3, 118 route de Narbonne, CNRS, CBD UMR5547, Toulouse 31062, France. E-mail: [email protected] or [email protected] 3Present address: Global Health Institute, School of Life Sciences, Station 19, EPFL, 1015 Lausanne, Switzerland. Received 8 November 2013; revised 5 December 2013; accepted 11 December 2013; accepted article preview online 17 December 2013; advance online publication, 14 January 2014 PONTIN function in AML1-ETO leukemia O Breig et al 1272 AML1-ETO-driven leukemic propagation and could be a potential were used. Detection was performed with Lumi-Light PLUS Western therapeutic target for t(8;21) AML. Blotting Substrate (Roche Applied Science) using Hyperﬁlm ECL (Amersham, GE Healthcare, Velizy-Villacoublay, France).

MATERIALS AND METHODS mRNA isolation and real-time RT-PCR analysis Fly crosses and blood cells analyses Total RNA extraction was performed with RNeasy Kit (Qiagen). Reverse To identify suppressors of AML1-ETO-induced lethality, lz-GAL4, UAS-GFP; transcription (RT), and real-time quantitative PCR (qPCR) were performed as previously described using Bio-Rad CFX96 (Bio-Rad, Marnes-la-Coquette, UAS-aml1eto/CyO{tub-GAL80} virgins were crossed to males from individual 24 UAS-dsRNA stocks at 22 1C and were screened for non-CyO adults in their France) detection system. Primers used for qPCR amplification are shown progeny. All UAS-dsRNA stocks allowing the emergence of AML1-ETO- in Supplementary Table 6. expressing adults were re-tested twice and with independent UAS-dsRNA lines when available. The following lines were used to silence pontin: Cell growth, differentiation and clonogenic assays NIG4003R-1 and VDRC105408. LZ þ hemocyte number and differentiation Cells were transfected with siRNA two times (at days 0 and 3). Viable cell status were assessed as previously described.18 numbers were assessed every 3 days using Trypan blue exclusion assay. For differentiation assays, CD11b expression was monitored as marker for Cell culture myeloid differentiation 4 days after siRNA transfection. Cell staining was performed with allophycocyanin-conjugated anti-CD11b antibody (BD Kasumi-1, U937 and U937-AE (kindly provided by Dr M Ruthardt, Frankfurt, Biosciences, Le Pont de Claix, France) according to the manufacturer’s Germany) were cultured in RPMI 1640 (PAA Cell Culture Company, Les instructions. Cells were analyzed by flow cytometry (FACSCalibur, Becton Mureaux, France) plus 10% fetal calf serum (PAA Cell Culture Company). Dickinson, BD Biosciences). For colony formation assays, Kasumi-1 cells SKNO-1 cells were cultivated in RPMI 1640 containing 20% fetal calf serum were transfected with siRNA and 24 h post-transfection 2500 cells were and 10 ng/ml granulocyte macrophage colony-stimulating factor. Droso- plated in 250 ml semisolid medium (containing RPMI 1640, 20% fetal calf phila Kc167 cells were grown in Schneider’s medium (Invitrogen, Cergy serum and 0.56% methylcellulose) in 48-well plates. After 7 days of culture, Pontoise, France) supplemented with 10% fetal calf serum and 50 mgof clusters consisting of 420 cells were counted as colonies. penicillin–streptomycin (Invitrogen). Cell cycle analyses siRNA and dsRNA transfection Cellular DNA content was analyzed using DRAQ5 (Sigma-Aldrich) staining. The following small interfering RNA (siRNA) were purchased from Cells were fixed in 70% ethanol before being stained for 15 min with 2 mM 0 Eurogentec (Seraing, Belgium): AML1-ETO siRNA (5 - CCUCGAAAUCGUACU DRAQ5 and the relative DNA content per cell was measured using flow 0 22 0 GAGAAG-3 ), green fluorescent protein (GFP) siRNA (5 -GCAAGCUGACCC cytometry (FACSCalibur, Becton Dickinson). The percentage of cells in G1, 0 19 0 0 UGAAGUUCAU-3 ), PONTIN siRNA 1 (5 -GGUGAAGUCACAGAGCUAA-3 ) S and G2/M phases was calculated using the ModFit program (Becton and 2 (50-CGGCCAACUUGCUUGCUAA-30). Exponentially growing cells were 7 Dickinson). For phospho-Histone H3 staining, Kasumi-1 cells fixed with concentrated to 10 cells/ml in culture medium, and 500 ml of cell 3.7% formaldehyde were incubated with 1 mg/ml phospho-Histone H3 suspension containing 500 nM siRNAs was transferred into a 4 mm (Ser10) antibody (Cell Signalling Technology, Ozyme) in phosphate- electroporation cuvette (Ozyme/Clonetech, Saint Quentin en Yvelines, buffered saline 0.1% TritonX100, 0.5% bovine serum albumin for 1 h. For France). Electroporation was performed with BTX ECM 830 Square 5-bromo-20deoxyuridine (BrdU) staining, cells were incubated in 10 mM Electroporator (BTX, Harvard Apparatus, Holliston, MA, USA) using a 23 BrdU for 30 min before being fixed and incubated with mouse anti-BrdU rectangle pulse of 300 V for 10 ms as described. Double-stranded RNA antibody (BD Biosciences) for 2 h in phosphate-buffered saline 0.5% Tween (dsRNA) treatments and transfection of Kc167 cells were performed as 24 20, 0.5% bovine serum albumin. Alexa Fluor 488-labeled anti-rabbit or anti- previously described. Briefly, dsRNA targeting different regions of pontin mouse IgG (Invitrogen) were used as secondary antibodies. Fluorescence were produced by in vitro transcription. Kc167 cells were plated on dsRNA was measured by flow cytometry using a FACSCalibur (BD Biosciences). for 24 h before being transfected with pAc-AML1-ETO19 and pAc-Renilla Luciferase expression plasmids using Effectene (Qiagen, Courtaboeuf, France). Cells were harvested and processed for western blot analysis 72 h Expression arrays and microarray gene expression data analysis after transfection. In all, 107 sh1-Pontin Kasumi-1 cells were treated or not with doxycycline for 96 h. Total RNA, prepared from biological quadruplicates was reversed Conditional expression of PONTIN shRNA transcribed and hybridized onto Illumina Human HT-12v4 BeadChip (Illumina, San Diego, CA, USA). Microarray results were analyzed in GenomeStudio For stable expression of PONTIN short hairpin RNA (shRNA) in Kasumi-1 software (Illumina) with background subtraction. The raw gene expression cells, we used the doxycycline-inducible lentiviral expression vector pTRIPZ data were analyzed using R package with quantile normalization. The 5% (Thermo Scientific Open Biosystems, Dutscher, Brumath, France) according threshold (P-value o0.05) was applied to all data. Genes with at least 1.5-fold to the manufacturer’s instructions. Two independent shRNA against changes in expression were selected. Similar results were obtained using 0 0 0 PONTIN were used (sh1: 5 -CGGCCAACTTGCTTGCTAA-3 , and sh2: 5 -T GenomeStudio cubic spline normalization as an alternative method for 0 AGCTCTGTGACTTCACCT-3 ). Lentiviruses were produced as described adjusting microarray data.27 Biological pathways and gene ontology 25 previously. After transduction with each shRNA construct, Kasumi-1 extraction was performed with Genomatix software (Genomatix Software cells harboring an integrated vector were selected with puromycin (2 mg/ml; GmbH, Munchen, Germany). Gene set enrichment analysis (GSEA) analyses Sigma-Aldrich, Lyon, France). Independent clones were isolated for each were performed using either MSigDB gene sets or gene sets associated with construct and PONTIN knockdown following induction with 1 mg/ml AML1-ETO knockdown in Kasumi-1 and SKNO-1 cells.5 Gene sets comprised doxycycline was quantified by real-time PCR and validated by western blot. the top 500 probes with increased or decreased signal intensities upon AML1-ETO knockdown. Before gene set formation, probes with detection Western blot analysis values P40.001 in the lower expressed set were deleted. Proteins were extracted from cultured cells as follows: cells were pelleted and lysed in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 20 mM EDTA, 1% triton X-100, 1% sodium deoxycholate and 0.1% sodium dodecyl RESULTS sulfate) supplemented with Complete (Roche Applied Science, Meylan, Pontin is required for AML1-ETO activity in Drosophila France) protease inhibitor cocktail. Protein samples were resolved by lz-GAL4-driven expression of AML1-ETO induces fly lethality at the sodium dodecyl sulfate–polyacrylamide gel electrophoresis and western 18 26 pupal stage. We thus undertook an in vivo RNA interference blots were performed as indicated in Breig et al. using the following screen in Drosophila using this phenotype as a read-out of AML1- primary antibodies: rabbit anti-ETO (Calbiochem, Merck, Darmstatd, Germany), mouse anti-PONTIN, mouse anti-REPTIN (Sigma-Aldrich), rabbit ETO activity. We first selected a list of 7516 genes detectably expressed in the human t(8;21)-positive Kasumi-1 cells based on anti-Renilla Luciferase (MBL International, Woburn, MA, USA), mouse anti- 5 HA (Covance, Rueil Malmaison, France) and mouse anti-TUBULIN (Sigma- microarray analyses. Using Homologene and reciprocal best-blast Aldrich). As secondary antibody, anti-rabbit IgG and anti-mouse IgG hit, we retrieved putative orthologs for 2558 of them in Drosophila horseradish peroxidase conjugates (Molecular Probes, Saint Aubin, France) melanogaster. UAS-dsRNA transgenic lines without predicted

Leukemia (2014) 1271 – 1279 & 2014 Macmillan Publishers Limited PONTIN function in AML1-ETO leukemia O Breig et al 1273 off-target were available against 591 of these genes at the Similarly, ectopic expression of AML1-ETO in Drosophila Kc167 National Institute of Genetic stock center (Japan) and they were cells was not affected by pontin dsRNA (Figure 1d). These data used in a F1 screen for suppressors of AML1-ETO-induced lethality suggest that Pontin does not control AML1-ETO expression but (see Materials and Methods section). Among all tested lines, we rather acts genetically downstream or in parallel to AML1-ETO. obtained a strong rescue of lethality (±15%) only with a dsRNA targeting pontin/RUVBL1 first exon. A second UAS-dsRNA line PONTIN is required for the growth of human AML1-ETO- directed against pontin exon 3 and devoid of predicted off- expressing leukemic cells target also suppressed AML1-ETO-induced lethality albeit at lower We next tested whether PONTIN participated in AML1-ETO-induced frequency (±5%), strongly suggesting that pontin downregulation leukemia in humans. First, we examined the consequences of is responsible for the observed rescue. However, we could not use PONTIN knockdown in Kasumi-1, a well-studied t(8;21) þ leukemic mutant alleles of pontin to further substantiate this result as cell line that constitutively expresses AML1-ETO and recapitulates pontin-deficient flies die at an early larval stage.28 We then several features of AML1-ETO oncogenic function.5 PONTIN was assessed the effect of pontin knockdown on AML1-ETO-induced readily detected in Kasumi-1 cells and its expression was effectively ‘pre-leukemic’ switch in the LZ þ larval blood cells. lz-driven silenced by two independent siRNAs (Figure 2a). Similar to siRNA- expression of AML1-ETO causes an expansion of this mediated depletion of AML1-ETO23 (Figures 2b and c and hematopoietic lineage (Figure 1a) and blocks its differentiation Supplementary Figure 1), PONTIN silencing blocked the proliferation into crystal cells (Figure 1b).18 Although pontin inhibition did not of Kasumi-1 cells in liquid culture (Figure 2b) and caused significant significantly affect LZ-GFP þ hemocyte number or differentiation cell death by days 6 and 9 (Figure 2c). PONTIN inhibition also in a wild-type larva, it suppressed the increase in LZ-GFP þ blood strongly hindered the growth of another t(8;21) cell lines, SKNO-1 cells caused by AML1-ETO and partially restored their (Supplementary Figure 1). These results suggest that PONTIN is differentiation into crystal cells (Figures 1a and b). Therefore, essential for the proliferation and survival of t(8;21) þ cells. pontin is required for AML1-ETO activity in Drosophila. One hypothesis to explain the suppression of AML1-ETO- induced phenotypes is that pontin might regulate AML1-ETO AML1-ETO-expressing cells are sensitized to PONTIN expression level or subcellular localization. Yet, lz-GAL4-driven downregulation expression of pontin dsRNA did not seem to affect AML1-ETO In Drosophila, pontin inhibition affected LZ þ blood cell number protein accumulation or nuclear localization in vivo (Figure 1c). when they express AML1-ETO but not in wild-type condition.

Figure 1. Pontin is required for AML1-ETO activity in Drosophila.(a, b) The phenotypes induced upon expression of AML1-ETO in the LZ þ blood cell lineage are suppressed when AML1-ETO is coexpressed with a dsRNA against pontin. lz-GAL4, UAS-GFP third instar wandering larvae carrying or not UAS-AML1-ETO and/or UAS-dsPontin transgenes were bled to determine the absolute number of LZ-GFP þ cells (a) as well as the proportion of LZ-GFP þ cells expressing the crystal cell differentiation marker proPO45 (differentiated cells: GFP þ proPO45 þ, progenitors: GFP þ proPO45-)(b). Data points represent the mean of triplicate experiments and error bars the s.d. (c, d) AML1-ETO expression is not affected by pontin dsRNA. (c) Fluorescent immunostaining against AML1-ETO in lz-GAL4, UAS-GFP; UAS-AML1-ETO (control, upper panels) and lz-GAL4, UAS-GFP; UAS-RUNX1-ETO/UAS-dsPontin (lower panels) circulating larval blood cells. Nuclei were stained with DAPI. (d) Western blot analysis of AML1-ETO expression. Kc167 cells treated with dsRNA targeting GFP or pontin were transfected with pAc-AML1-ETO and pAc-Renilla Luciferase expression plasmids. Whole-cell extracts were immunoblotted with anti-AML1. Anti-Renilla Luciferase (R.LUC) and anti-TUBULIN were used as transfection and loading controls, respectively.

& 2014 Macmillan Publishers Limited Leukemia (2014) 1271 – 1279 PONTIN function in AML1-ETO leukemia O Breig et al 1274 AML1-ETO activates PONTIN and REPTIN expression ChIP sequencing has been used to deﬁne AML1-ETO-binding sites at the genome-wide levels in different cellular contexts.5,6 By browsing these results, we found that AML1-ETO bound to the vicinity of PONTIN transcription start site: AML1-ETO ChIP peaks were present in Kasumi-1 and SKNO-1 cells but also in primary AML cells of two (out of three) patients with t(8;21) leukemia.6 Furthermore, AML1- ETO ectopically expressed in MCF7 or K562 cells also bound to this PONTIN cis-element. Strikingly, AML1-ETO ChIP peaks ±300-bp upstream of REPTIN transcription start site were also present in all these cellular contexts.6 These data suggest that PONTIN and REPTIN are direct targets of AML1-ETO. To assess the consequence of AML1- ETO binding, we analyzed PONTIN and REPTIN expression levels in Kasumi-1 cells following siRNA-mediated knockdown of AML1-ETO. As shown in Figure 4, AML1-ETO silencing was paralleled by diminished expression of PONTIN and REPTIN at the transcript and protein levels (Figures 4a and b). Similarly, PONTIN and REPTIN levels were decreased in U937-AE cells following AML1-ETO silencing (Figure 4c). Thus, AML1-ETO is likely to directly activate the transcription of both PONTIN and REPTIN. Of note, in line with previous experiments in epithelial cells showing that PONTIN depletion causes REPTIN degradation,21 we observed that PONTIN knockdown also reduced REPTIN protein but not mRNA levels (Figures 4a and b). Moreover, consistent with our observations in Drosophila, PONTIN silencing did not affect AML1-ETO expression levels (Figures 4a and b).

Correlation between PONTIN and AML1-ETO gene expression signatures To gain further insights into PONTIN function, we examined the consequence of PONTIN depletion on Kasumi-1 cell RNA expression profile. We used lentiviral transduction to establish Kasumi-1 cell lines carrying doxycycline-inducible shRNA directed against PONTIN. In the presence of doxycycline, PONTIN expression Figure 2. PONTIN knockdown inhibits the growth of Kasumi-1 cells. was efficiently decreased (Figures 5a and b) and cell growth was (a) Western blot analysis of PONTIN expression in Kasumi-1 cells strongly impaired (Supplementary Figure 2). Similar results were transfected with control (siGFP) or PONTIN siRNAs. Anti-TUBULIN obtained using a second shRNA (Supplementary Figure 2). was used as a loading control. (b, c) Viable cells numbers (b) and percentage of dead cells (c) were determined at different time We then assessed the changes in transcriptome in sh1-Kasumi-1 points by Trypan blue exclusion assay in Kasumi-1 cells transfected cells after 96-h incubation with or without doxycycline (that is, at a with the indicated siGFP (’), siAML11-ETO ( Â ), siPONTIN1 (m)or time when PONTIN silencing started to affect growth curve, siPONTIN2 (K). Data points represent the means of three Supplementary Figure 2). Whole-genome expression profiling independent experiments and error bars the s.d. using Illumina HT-12 BeadChip revealed that 188 genes were upregulated and 157 were downregulated by 41.5-fold (B.Ho0.05) upon PONTIN depletion (Supplementary Table 1), although most changes remained relatively mild, with only 53 To test whether the situation is similar in humans, we sought to genes deregulated by 42-folds. To validate these results, RT-qPCR analyze the effect of PONTIN inhibition on human leukemic cells analysis was performed on selected genes. The candidates that with or without AML1-ETO expression. Accordingly, we made were either repressed (left panel) or activated (right panel) upon use of a U937 cell line that conditionally expresses AML1-ETO PONTIN knockdown according to the microarray data exhibited a under the control of the copper-inducible mouse methallothio- similar trend in these experiments (Figure 5c), whereas their nein promoter (U937-AE). Yet, U937-AE express low levels of expression was not affected upon doxycycline treatment of the AML1-ETO even in the absence of induction (Figure 3a) parental Kasumi-1 cells (data not shown). requiring the parental U937 cell line as a control. To assess To test for a possible concordance of PONTIN and AML1-ETO the requirement of PONTIN for U937 and U937-AE cell growth, gene expression signatures, we performed GSEA. Indeed, GSEA with these cells were transfected with PONTIN siRNA at two different two gene sets obtained with two distinct t(8;21) cell lines, Kasumi-1 5 concentrations (25 or 500 nM). For a given concentration, the and SKNO-1, demonstrated highly significant correlations between level of PONTIN extinction was similar in the two cell lines as AML1-ETO and PONTIN gene sets (Figure 5d, Supplementary observed by western blot or measured by RT-qPCR (Figures 3b and c). Figure 3, Supplementary Table 2). Genes that were upregulated Importantly, low concentration of PONTIN siRNA slightly dimin- upon AML1-ETO knockdown (AEkd) correlated with PONTIN ished the growth of U937 cells whereas it strongly impaired the knockdown, whereas downregulated genes showed an inverse growth of U937-AE cells (Figures 3d and e). Although high correlation. Restricting the AEkd sets to genes directly bound by concentration of PONTIN siRNA potently suppressed the growth of AML1-ETO5 still showed a similar concurrence (Supplementary both cell lines (Figure 3d), the relative reduction in cell growth was Figure 3). Taken together, these data show that PONTIN and AML1- much higher in U937-AE than in U937 cells (32 and 250-folds ETO establish and maintain a leukemic phenotype by a coordinated vs 8.5 and 21-folds at days 6 and 9, respectively; Figure 3e). modulation of gene expression. Furthermore, direct or indirect All together, these results indicate that AML1-ETO expression target genes shared by both factors form a considerable subset of sensitizes cells to PONTIN inhibition. this transcriptional program.

Leukemia (2014) 1271 – 1279 & 2014 Macmillan Publishers Limited PONTIN function in AML1-ETO leukemia O Breig et al 1275

Figure 3. AML1-ETO-expressing U937 cells are more sensitive to PONTIN depletion. (a) Western blot analysis of AML1-ETO expression in wild- type U937 cells and in U973 cells stably transfected with an expression vector for AML1-ETO (U937-AE). (b) Western blots showing PONTIN expression in U937 and U937-AE cells 4 days after electroporation with siRNA against GFP or PONTIN at the indicated concentrations. (a, b) TUBULIN was used a loading control. (c) Expression of PONTIN mRNA in U937 (light grey) and U937-AE (dark grey) cells 4 days after electroporation with the indicated siRNA. PONTIN expression was measured by RT-qPCR and normalized to GAPDH. Expression levels are shown relative to siGFP control cells. (d) Growth curve showing the number of viable cells at different time points following transfection of U937 (grey, dotted lines) or U937-AE (black, continuous lines) cells with siGFP at 500 nM (D) or siPONTIN at 500 nM (J)or25nM ( Â ). (e) Relative numbers of viable cells in U937 and U937-AE cells at days 3 (light grey), 6 (grey) and 9 (dark grey) following transfection with siRNA PONTIN at 25 or 500 nM. Cells numbers are shown relative to siGFP-treated cells U937 or U937-AE cells. All experiments were performed at least three times. Means and s.d. are indicated.

Figure 4. AML1-ETO activates PONTIN and REPTIN transcription. (a) The expression of AML1-ETO, PONTIN and REPTIN was measured by RT-qPCR in Kasumi-1 cells 4 days after transfection with the indicated siRNA. Expression levels are normalized to GAPDH expression. Means and s.d. of three independent experiments are indicated. (b, c) Western blots showing AML1-ETO, PONTIN, REPTIN and TUBULIN (loading control) in Kasumi-1 cells (b) or U937-AE cells (c) 4 days after electroporation with the indicated siRNA.

GSEA revealed positive correlations of the PONTIN knockdown contributors to these categories (Table 1, Supplementary Table 5). gene signature with gene sets associated with more mature AML This suggests that PONTIN essentially acts by promoting the types or with lymphoid maturation, which may indicate differ- expression of genes involved in cell cycle regulation. For instance, entiation. However, it showed much more signiﬁcant inverse PONTIN inactivation led to the downregulation of several keys correlations with gene sets associated with cell cycle processes regulators of G1/S and G2/M cell cycle progression (for example, including G1-S transition and M-phase progression as well as E2F CDK1/CDC2, CDC25A, CDC45 and E2F2). It also inhibited the and RB signatures (Supplementary Table 3). Indeed, inspection of expression of genes participating in DNA synthesis and replication the differentially expressed (41.5-folds) genes using Genomatix (for example, CDT1, MCM4, MCM5, MCM7 and TK1), chromosome software revealed a striking enrichment for gene ontology terms segregation (for example, CDC20, CENPM, KIF2C and KNTC1) and or signaling pathways related to cell cycle (Table 1, Supplementary genome stability (for example, TIMELESS and FANCG). Network Table 4). Remarkably, repressed genes were by far the main analysis using Genemania database29 highlighted that 40% of the

& 2014 Macmillan Publishers Limited Leukemia (2014) 1271 – 1279 PONTIN function in AML1-ETO leukemia O Breig et al 1276

Figure 5. Genome-wide expression proﬁling indicates that PONTIN expression signature correlates with that of AML1-ETO. (a, b) Analysis of PONTIN expression in sh1-Kasumi-1 cells 4 days after induction of PONTIN shRNA expression ( þ Dox). (a) PONTIN mRNA levels relative to GADPH mRNA were determined by RT-qPCR. (b) PONTIN protein expression was assessed by western blot. (c) The expression of genes inhibited (left panel) or activated (right panel) upon PONTIN depletion in Kasumi-1 according to Illumina HT-12 BeadChips analysis was measured by RT-qPCR and normalized to GAPDH. (d) Gene set enrichment analysis of the PONTIN knockdown gene signature showing positive and negative correlations with sets of genes repressed (left panel) and maintained by AML1/ETO (right panel), respectively.

which we have previously shown to depend on AML1-ETO activity Table 1. Top 15 gene ontology terms (biological processes) over- in Kasumi-1 cells.22,23 Importantly, methylcellulose colony-forming represented in PONTIN-regulated genes data set assays showed that PONTIN depletion severely reduced the GO biological process P-value Genes number of colony formed by these cells in semi-solid medium (Figure 6a). Hence, alike AML1-ETO, PONTIN expression is critical Cell cycle phase 4.55E–15 45 (39) to sustain Kasumi-1 clonogenic potential indicative of leukemic Mitotic cell cycle 1.88E–14 42 (37) self-renewal. However, FACS analysis of CD11b expression cell cycle process 1.67E–14 46 (40) suggested that, in contrast to AML1-ETO depletion, PONTIN cell cycle 2.10E–10 50 (43) downregulation alone was not sufficient to substantially promote M phase of mitotic cell cycle 1.04E–09 24 (22) the myeloid differentiation of Kasumi-1 cells (Figure 6b). Cell cycle DNA metabolic process 1.04E–09 37 (25) analysis with flow cytometry of Draq5-labeled nuclei showed that Nuclear division 1.04E–09 23 (21) Mitosis 1.04E–09 23 (21) PONTIN depletion in sh1-Kasumi-1 cells resulted in a one-third M phase 1.05E–10 28 (25) increase in the relative number of cells in G0/G1 (from 40.7 to Organelle fission 2.25E–09 23 (21) 53.5%) and a concomitant reduction in the proportion of cells in Interphase 3.86E–09 24 (20) either S or G2/M phase (Figure 6c and Supplementary Figure 5). Interphase of mitotic cell cycle 1.27E–08 23 (19) In addition, BrdU incorporation and phospho-Histone H3-Ser10 Cell cycle checkpoint 1.83E–08 18 (16) labeling respectively showed that PONTIN silencing decreased the Cell division 2.25E–08 24 (19) proportion of cells undergoing DNA replication and going through Regulation of cell cycle arrest 6.87E–08 18 (16) mitosis (Figures 6d and e and Supplementary Figure 5). Abbreviation: GO, gene ontology. The number of repressed genes present Thus, PONTIN depletion impairs cell cycle progression across in each category is given between brackets. several checkpoints. These results support the hypothesis that the coordination of cell cycle genes transcription by PONTIN participates in the sustained proliferation of Kasumi-1 cells. proteins in this data set are directly linked to one another, making a densely connected network centered on cell cycle-associated factors (Supplementary Figure 4). Thus, it appears that PONTIN DISCUSSION mainly controls the expression of functionally related components Despite advances in our understanding of the molecular of the cell cycle machinery. mechanisms by which AML1-ETO promotes leukemia, only a few genes are known to be critical for its oncogenic activity. For the PONTIN regulates leukemic self-renewal and cell cycle progression first time, our genetic interaction screen in Drosophila and our In light of these data, we studied the impact of PONTIN on subsequent experiments in human leukemic cells establish a malignant self-renewal, differentiation and cell cycle progression, functional link between AML1-ETO and PONTIN that could have a

Leukemia (2014) 1271 – 1279 & 2014 Macmillan Publishers Limited PONTIN function in AML1-ETO leukemia O Breig et al 1277

Figure 6. PONTIN controls leukemic cells self-renewal and proliferation. (a) Colony numbers obtained in semi-solid medium culture 7 days after transfection of Kasumi-1 cells with the indicated siRNA. (b) FACS histogram plots showing CD11b presentation in Kasumi-1 cells 4 days after transfection with the indicated siRNA. siGFP: red, siAML1-ETO: green, siPONTIN: blue. The percentages of CD11b-expresssing cells (window M1) are indicated. The histogram is representative of three independent experiments. (c–e) sh1-Kasumi-1 cells were treated (Dox þ ) or not (Dox-) with doxycycline for 6 days and their cell cycle status analyzed. (c) The proportion of cells in the different phases of the cell cycle as measured by DRAQ5 nuclear accumulation and FACS analysis. (d) The percentage of cells replicating DNA was measured by FACS following 30 min incorporation of BrdU. (e) The proportion of cells in mitosis was monitored by immunostaining against anti-phosphorylated Histone H3 (Ser10). All experiments were performed in triplicate, means and s.d. are presented. critical role in t(8;21) leukemogenesis by promoting cell reduced hepatocyte growth factor expression. Increased proliferation. hepatocyte growth factor levels and autocrine activation, its Drosophila has emerged as a fruitful model to study AML1- receptor have been found to be critical for the proliferation of ETO18,19,30,31 and we have taken advantage of this model to many AML cells.37 However, addition of this growth factor was not perform an in vivo RNA interference screen aimed at identifying sufficient to rescue the growth of PONTIN-depleted Kasumi-1 cells new genes implicated in AML1-ETO-induced leukemia. AML cells (OB unpublished observations). As the reduction in the expression are characterized by increased proliferation, impaired level of all affected genes was rather modest, we propose that it is differentiation and enhanced self-renewal, all of which are the concomitant downregulation of this gene set rather than affected by AML1-ETO.5 In fly, pontin dsRNA suppressed AML1- individual variations, which explain the profound effect of PONTIN ETO-induced LZ þ blood cell increase and partially restored their silencing on proliferation. Importantly, GSEA showed that PONTIN differentiation. In human, we found that PONTIN is essential in and AML1-ETO regulate the expression of highly correlated sets of Kasumi-1 cells for maintaining proliferation and self-renewal but it genes, further suggesting a functional cooperation between these did not seem to substantially affect their differentiation. This two proteins for the development and maintenance of a leukemic discrepancy may reflect divergence in PONTIN function or phenotype. differences in the way AML1-ETO impairs blood cell The molecular mechanism by which PONTIN activates the differentiation between Drosophila and human. Nonetheless, our expression of these genes remains to be addressed. PONTIN acts findings strengthen the notion that AML1-ETO promotes blood as a coactivator for several transcription factors, including MYC in cell expansion by impinging on conserved gene networks.18,31 human and Drosophila.38,39 MYC is frequently activated in AML PONTIN has been implicated in normal development and and has an important role in the induction of leukemia, notably for cancer in mammals.20 For instance, Pontin is essential for early sustaining cell proliferation.40 Although ‘validated targets of c-Myc development and hematopoietic stem cell survival in mouse,32 transcriptional activation’ was the third over-represented pathway and PONTIN inhibition reduces proliferation and malignant in PONTIN downregulated gene set, ‘E2F transcriptional network’ transformation of different epithelial cell types.21,33–36 Yet, the ranked first (Supplementary Table 4). PONTIN is also a coactivator role of PONTIN in global gene expression had not been analyzed in of E2F134 and PONTIN silencing caused a decrease in the AML. Our expression profiling shows that the main consequence expression of E2F2, a target of MYC required for blood cell of PONTIN depletion in Kasumi-1 cells is a reduced expression of a progenitor proliferation,39,41,42 and E2F7, a target of E2F143 that number of functionally interconnected genes linked to the cell promotes AML cell proliferation.44 Thus, PONTIN could control cycle. Although PONTIN loss eventually caused cell death, we did t(8;21) cell proliferation by promoting MYC-induced transcription not observe apoptotic cell marker expression, suggesting that and the E2F pathway. Alternatively, PONTIN may directly PONTIN is primarily required for cell proliferation rather than cell participate in AML1-ETO-induced transcription as suggested by survival. The strong reduction in proliferation and the cell cycle our GSEA results but we did not observe an interaction between arrest we observed in Kasumi-1 is most likely due to the reduced AML1-ETO and PONTIN (OB unpublished observation). Beside its expression of key regulators of cell cycle progression and role as a coactivator, PONTIN is required for telomerase chromosome replication such as CDC2, CDC25A, CDT1 or MCMs. holoenzyme assembly.45 This could be particularly relevant to Beside intrinsic cell cycle regulators, PONTIN silencing also t(8;21) leukemia considering that AML1-ETO supports Kasumi-1

& 2014 Macmillan Publishers Limited Leukemia (2014) 1271 – 1279 PONTIN function in AML1-ETO leukemia O Breig et al 1278 46 self-renewal by promoting high level of telomerase activity and 6 Martens JH, Mandoli A, Simmer F, Wierenga BJ, Saeed S, Singh AA et al. that PONTIN silencing strongly reduced Kasumi-1 cell ERG and FLI1 binding sites demarcate targets for aberrant epigenetic regulation clonogenicity. by AML1-ETO in acute myeloid leukemia. Blood 2012; 120: 4038–4048. Our data show that AML1-ETO induces PONTIN and REPTIN 7 Saeed S, Logie C, Francoijs KJ, Frige G, Romanenghi M, Nielsen FG et al. Chromatin expression and suggests that t(8;21) AML cells could be selectively accessibility, p300, and histone acetylation define PML-RARalpha and AML1-ETO targeted by partially inhibiting PONTIN. Reminiscent of this binding sites in acute myeloid leukemia. Blood 2012; 120: 3058–3068. situation, REPTIN was shown to be activated by the leukemogenic 8 Hatlen MA, Wang L, Nimer SD. AML1-ETO driven acute leukemia: insights into fusion protein MLL-AF9 and to be required for the proliferation pathogenesis and potential therapeutic approaches. Front Med 2011; 6: 248–262. 9 Schwieger M, Lohler J, Friel J, Scheller M, Horak I, Stocking C. AML1-ETO inhibits and survival of MLL-AF9-expressing cell lines as well as non-MLL 47 maturation of multiple lymphohematopoietic lineages and induces myeloblast rearranged AML cell lines, including Kasumi-1. Moreover, transformation in synergy with ICSBP deficiency. J Exp Med 2002; 196: 1227–1240. another report recently showed that PONTIN is also required for 10 Schessl C, Rawat VP, Cusan M, Deshpande A, Kohl TM, Rosten PM et al. The AML1- 48 the survival of THP1 cells, a MLL-AF9-positive cell line. PONTIN ETO fusion gene and the FLT3 length mutation collaborate in inducing acute and REPTIN are thought to function as complexed together and leukemia in mice. J Clin Invest 2005; 115: 2159–2168. are overexpressed in different cancers.20 Interestingly, REPTIN 11 Wang YY, Zhao LJ, Wu CF, Liu P, Shi L, Liang Y et al. C-KIT mutation cooperates silencing in vivo led to a regression of hepatocellular carcinoma with full-length AML1-ETO to induce acute myeloid leukemia in mice. Proc Natl tumor xenografts49 and recently developed chemical inhibitors Acad Sci USA 2011; 108: 2450–2455. against PONTIN ATPase activity impaired hepatic tumor cell lines 12 Wang L, Gural A, Sun XJ, Zhao X, Perna F, Huang G et al. The leukemogenicity 50 of AML1-ETO is dependent on site-specific lysine acetylation. Science 2012; 333: growth in vitro. Therefore, PONTIN/REPTIN complex is an 765–769. attractive molecular target for cancer therapy. However, more 13 Lo MC, Peterson LF, Yan M, Cong X, Jin F, Shia WJ et al. Combined gene work will be required to characterize PONTIN and REPTIN expression and DNA occupancy profiling identifies potential therapeutic targets functions as they also have distinct and antagonistic of t(8;21) AML. Blood 2012; 120: 1473–1484. activities.28,51,52 14 Zhang Y, Wang J, Wheat J, Chen X, Jin S, Sadrzadeh H et al. AML1-ETO mediates In conclusion, our results suggest a model whereby PONTIN hematopoietic self-renewal and leukemogenesis through a COX/beta-catenin upregulation by AML1-ETO contributes to the expression of genes signaling pathway. Blood 2013; 121: 4906–4916. required for AML cell proliferation and we propose that PONTIN 15 Li Y, Gao L, Luo X, Wang L, Gao X, Wang W et al. Epigenetic silencing of microRNA- might a be relevant therapeutic target in AML. This and other works 193a contributes to leukemogenesis in t(8;21) acute myeloid leukemia by acti- vating the PTEN/PI3K signal pathway. Blood 2013; 121: 499–509. illustrate that lower model organisms such as Drosophila provide 16 Polesello C, Roch F, Gobert V, Haenlin M, Waltzer L. Modeling cancers in powerful systems to perform genetic interaction screens for cancer- 53 Drosophila. Prog Mol Biol Transl Sci 2011; 100: 51–82. relevant genes and to identify new leads for cancer therapy. With 17 Crozatier M, Vincent A. Drosophila: a model for studying genetic and molecular advances in shRNA-based or insertion mutagenesis screens in aspects of haematopoiesis and associated leukaemias. Dis Model Mech 2011; 4: 54,55 mammals, it is anticipated that interaction screens will 439–445. increasingly contribute to the understanding of oncogenic protein 18 Osman D, Gobert V, Ponthan F, Heidenreich O, Haenlin M, Waltzer L. A Drosophila mode of action and help pinpoint critical pathways participating in model identifies calpains as modulators of the human leukemogenic fusion malignant transformation. protein AML1-ETO. Proc Natl Acad Sci USA 2009; 106: 12043–12048. 19 Bras S, Martin-Lanneree S, Gobert V, Auge B, Breig O, Sanial M et al. Myeloid leukemia factor is a conserved regulator of RUNX transcription factor activity involved in hematopoiesis. Proc Natl Acad Sci USA 2012; 109: 4986–4991. CONFLICT OF INTEREST 20 Grigoletto A, Lestienne P, Rosenbaum J. The multifaceted proteins Reptin and Pontin as major players in cancer. Biochim Biophys Acta 2011; 1815: 147–157. The authors declare no conflict of interest. 21 Haurie V, Menard L, Nicou A, Touriol C, Metzler P, Fernandez J et al. Adenosine triphosphatase pontin is overexpressed in hepatocellular carcinoma and coregulated with reptin through a new posttranslational mechanism. Hepatology ACKNOWLEDGEMENTS 2009; 50: 1871–1883. 22 Heidenreich O, Krauter J, Riehle H, Hadwiger P, John M, Heil G et al. AML1/MTG8 We are grateful to members of our teams for comments on the manuscript. We thank oncogene suppression by small interfering RNAs supports myeloid differentiation Toulouse RIO imaging platform and T Jungas for assistance with FACS analysis as well of t(8;21)-positive leukemic cells. Blood 2003; 101: 3157–3163. as L Liaubet and S Bel-Vialar for their help with microarray analysis. We deeply thank 23 Martinez N, Drescher B, Riehle H, Cullmann C, Vornlocher HP, Ganser A et al. ´ B Auge for technical help. We thank the National Institute of Genetics (Japan), the The oncogenic fusion protein RUNX1-CBFA2T1 supports proliferation and inhibits Vienna Drosophila Resource Center (Austria) and Bloomington (USA) for fly stocks. senescence in t(8;21)-positive leukaemic cells. BMC Cancer 2004; 4: 44. This work was supported by grants from the Ligue Re´gionale Midi Pyreńeés, 24 Gobert V, Osman D, Bras S, Auge B, Boube M, Bourbon HM et al. A genome-wide Fondation ARC and Association for International Cancer Research to LW and OB; SB RNA interference screen identifies a differential role of the mediator CDK8 was supported by a fellowship from the Fondation pour la Recherche Me´dicale, and module subunits for GATA/ RUNX-activated transcription in Drosophila. Mol Cell OH and NMS by a research grant from Leukaemia and Lymphoma Research (10033). Biol 2010; 30: 2837–2848. 25 Bomken S, Buechler L, Rehe K, Ponthan F, Elder A, Blair H et al. Lentiviral marking of patient-derived acute lymphoblastic leukaemic cells allows in vivo tracking REFERENCES of disease progression. Leukemia 2013; 27: 718–721. 26 Breig O, Theoleyre O, Douablin A, Baklouti F. Subtle distinct regulations of late 1 Erickson P, Gao J, Chang KS, Look T, Whisenant E, Raimondi S et al. Identification erythroid molecular events by PI3K/AKT-mediated activation of Spi-1/PU.1 of breakpoints in t(8;21) acute myelogenous leukemia and isolation of a fusion oncogene autoregulation loop. Oncogene 2010; 29: 2807–2816. transcript, AML1/ETO, with similarity to Drosophila segmentation gene, runt. 27 Schmid R, Baum P, Ittrich C, Fundel-Clemens K, Huber W, Brors B et al. Comparison Blood 1992; 80: 1825–1831. of normalization methods for Illumina BeadChip HumanHT-12 v3. BMC Genomics 2 Miyoshi H, Kozu T, Shimizu K, Enomoto K, Maseki N, Kaneko Y et al. The t(8;21) 2010; 11: 349. translocation in acute myeloid leukemia results in production of an AML1-MTG8 28 Bauer A, Chauvet S, Huber O, Usseglio F, Rothbacher U, Aragnol D et al. Pontin52 fusion transcript. EMBO J 1993; 12: 2715–2721. and reptin52 function as antagonistic regulators of beta-catenin signalling 3 Lam K, Zhang DE. RUNX1 and RUNX1-ETO: roles in hematopoiesis and leuke- activity. EMBO J 2000; 19: 6121–6130. mogenesis. Front Biosci 2012; 17: 1120–1139. 29 Zuberi K, Franz M, Rodriguez H, Montojo J, Lopes CT, Bader GD et al. GeneMANIA 4 Wichmann C, Becker Y, Chen-Wichmann L, Vogel V, Vojtkova A, Herglotz J et al. prediction server 2013 update. Nucleic Acids Res 2013; 41: W115–W122. Dimer-tetramer transition controls RUNX1/ETO leukemogenic activity. Blood 2010; 30 Wildonger J, Mann RS. The t(8;21) translocation converts AML1 into a constitutive 116: 603–613. transcriptional repressor. Development 2005; 132: 2263–2272. 5 Ptasinska A, Assi SA, Mannari D, James SR, Williamson D, Dunne J et al. Depletion 31 Sinenko SA, Hung T, Moroz T, Tran QM, Sidhu S, Cheney MD et al. Genetic of RUNX1/ETO in t(8;21) AML cells leads to genome-wide changes in chromatin manipulation of AML1-ETO-induced expansion of hematopoietic precursors in structure and transcription factor binding. Leukemia 2012; 26: 1829–1841. a Drosophila model. Blood 2010; 116: 4612–4620.

Leukemia (2014) 1271 – 1279 & 2014 Macmillan Publishers Limited PONTIN function in AML1-ETO leukemia O Breig et al 1279 32 Bereshchenko O, Mancini E, Luciani L, Gambardella A, Riccardi C, Nerlov C. Pontin 45 Venteicher AS, Meng Z, Mason PJ, Veenstra TD, Artandi SE. Identification is essential for murine hematopoietic stem cell survival. Haematologica 2012; 97: of ATPases pontin and reptin as telomerase components essential for holoenzyme 1291–1294. assembly. Cell 2008; 132: 945–957. 33 Feng Y, Lee N, Fearon ER. TIP49 regulates beta-catenin-mediated neoplastic 46 Gessner A, Thomas M, Castro PG, Buchler L, Scholz A, Brummendorf TH et al. transformation and T-cell factor target gene induction via effects on chromatin Leukemic fusion genes MLL/AF4 and AML1/MTG8 support leukemic self-renewal remodeling. Cancer Res 2003; 63: 8726–8734. by controlling expression of the telomerase subunit TERT. Leukemia 2010; 24: 34 Dugan KA, Wood MA, Cole MD. TIP49, but not TRRAP, modulates c-Myc and E2F1 1751–1759. dependent apoptosis. Oncogene 2002; 21: 5835–5843. 47 Osaki H, Walf-Vorderwulbecke V, Mangolini M, Zhao L, Horton SJ, Morrone G et al. 35 Ducat D, Kawaguchi S, Liu H, Yates 3rd JR, Zheng Y. Regulation of microtubule The AAA( þ ) ATPase RUVBL2 is a critical mediator of MLL-AF9 oncogenesis. assembly and organization in mitosis by the AAA þ ATPase Pontin. Mol Biol Cell Leukemia 2013; 27: 1461–1468. 2008; 19: 3097–3110. 48 Huang X, Spencer GJ, Lynch JT, Ciceri F, Somerville TDD, Somervaille TCP. 36 Watkins NJ, Lemm I, Ingelfinger D, Schneider C, Hossbach M, Urlaub H et al. Enhancers of Polycomb EPC1 and EPC2 sustain the oncogenic potential of MLL Assembly and maturation of the U3 snoRNP in the nucleoplasm in a large leukemia stem cells. Leukemia 2013; e-pub ahead of print 12 November 2013; dynamic multiprotein complex. Mol Cell 2004; 16: 789–798. doi:10.1038/leu.2013.316. 37 Kentsis A, Reed C, Rice KL, Sanda T, Rodig SJ, Tholouli E et al. Autocrine activation 49 Menard L, Taras D, Grigoletto A, Haurie V, Nicou A, Dugot-Senant N et al. In vivo of the MET receptor tyrosine kinase in acute myeloid leukemia. Nat Med 2012; 18: silencing of Reptin blocks the progression of human hepatocellular carcinoma 1118–1122. in xenografts and is associated with replicative senescence. J Hepatol 2010; 52: 38 Bellosta P, Hulf T, Balla Diop S, Usseglio F, Pradel J, Aragnol D et al. Myc interacts 681–689. genetically with Tip48/Reptin and Tip49/Pontin to control growth and proliferation 50 Elkaim J, Castroviejo M, Bennani D, Taouji S, Allain N, Laguerre M et al. First during Drosophila development. Proc Natl Acad Sci USA 2005; 102: 11799–11804. identification of small-molecule inhibitors of Pontin by combining virtual 39 Wood MA, McMahon SB, Cole MD. An ATPase/helicase complex is an essential screening and enzymatic assay. Biochem J 2012; 443: 549–559. cofactor for oncogenic transformation by c-Myc. Mol Cell 2000; 5: 321–330. 51 Rottbauer W, Saurin AJ, Lickert H, Shen X, Burns CG, Wo ZG et al. Reptin and 40 Delgado MD, Leon J. Myc roles in hematopoiesis and leukemia. Genes Cancer pontin antagonistically regulate heart growth in zebrafish embryos. Cell 2002; 2010; 1: 605–616. 111: 661–672. 41 Fujiwara K, Yuwanita I, Hollern DP, Andrechek ER. Prediction and genetic 52 Li W, Zeng J, Li Q, Zhao L, Liu T, Bjorkholm M et al. Reptin is required for the demonstration of a role for activator E2Fs in Myc-induced tumors. Cancer Res transcription of telomerase reverse transcriptase and over-expressed in gastric 2011; 71: 1924–1932. cancer. Mol Cancer 2010; 9: 132. 42LiFX,ZhuJW,TessemJS,BeilkeJ,Varella-GarciaM,JensenJet al. The development of 53 Gonzalez C. Drosophila melanogaster: a model and a tool to investigate diabetes in E2f1/E2f2 mutant mice reveals important roles for bone marrow-derived malignancy and identify new therapeutics. Nat Rev Cancer 2013; 13: 172–183. cells in preventing islet cell loss. Proc Natl Acad Sci USA 2003; 100: 12935–12940. 54 Zuber J, Shi J, Wang E, Rappaport AR, Herrmann H, Sison EA et al. RNAi screen 43 Di Stefano L, Jensen MR, Helin K. E2F7 a novel E2F featuring DP-independent identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature 2011; repression of a subset of E2F-regulated genes. EMBO J 2003; 22: 6289–6298. 478: 524–528. 44 Salvatori B, Iosue I, Mangiavacchi A, Loddo G, Padula F, Chiaretti S et al. 55 Bergerson RJ, Collier LS, Sarver AL, Been RA, Lugthart S, Diers MD et al. The microRNA-26a target E2F7 sustains cell proliferation and inhibits monocytic An insertional mutagenesis screen identifies genes that cooperate with Mll-AF9 in differentiation of acute myeloid leukemia cells. Cell Death Dis 2012; 3: e413. a murine leukemogenesis model. Blood 2012; 119: 4512–4523.

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

& 2014 Macmillan Publishers Limited Leukemia (2014) 1271 – 1279