Leukemia (2013) 27, 389–397 & 2013 Macmillan Publishers Limited All rights reserved 0887-6924/13 www.nature.com/leu

ORIGINAL ARTICLE Suppression of the let-7b microRNA pathway by DNA hypermethylation in infant acute lymphoblastic leukemia with MLL rearrangements

M Nishi1,4, M Eguchi-Ishimae2,4,ZWu2, W Gao2, H Iwabuki2, S Kawakami2, H Tauchi2, T Inukai3, K Sugita3, Y Hamasaki1, E Ishii2 and M Eguchi2

MicroRNAs (miRNAs) regulate cell proliferation and differentiation by controlling the expression of involved in many signaling pathways. Recent studies have shown that dysregulation of miRNA expression is associated with increased tumorigenicity and a poor prognosis in several types of . The miRNA let-7b is one of the severely downregulated miRNAs in mixed-lineage leukemia (MLL)-rearranged acute lymphoblastic leukemia (ALL) patients. In vitro transfection of leukemogenic MLL fusion into human embryonic kidney-293 cells suppressed let-7b expression. In leukemic cells with an MLL fusion gene, the regulatory region for let-7b expression was hypermethylated, and its expression was partially recovered after culturing the cells with the demethylating agent 5-azacitidine. These results suggest that loss of let-7b expression may be one of the consequences of oncogenic MLL fusion proteins, and contributes to leukemogenesis possibly through the upregulation of let-7b-regulated target genes with leukemogenic potential in hematopoietic cells. The enforced expression of let-7b in ALL cell lines with an MLL fusion gene inhibited their growth, indicating the possible use of let-7b as a new therapeutic tool for refractory infant ALL with an MLL fusion gene.

Leukemia (2013) 27, 389–397; doi:10.1038/leu.2012.242 Keywords: let-7b; infant; acute lymphoblastic leukemia; MLL; DNA

INTRODUCTION Recently, Schotte et al.5 reported the array-based analysis Acute lymphoblastic leukemia (ALL) in infants is often accom- of miRNA expression in pediatric ALL samples. They identified panied by hyperleukocytosis, hepatosplenomegaly and extrame- let-7b as one of the discriminative miRNAs of MLL-rearranged dullary leukemic infiltration at initial diagnosis.1 Biologically, ALL. leukemic cells are frequently characterized by The miRNA let-7 was originally identified in Caenorhabditis 6 translocations involving 11q23, resulting in the rearrangement of elegans as a regulator of cellular proliferation. There are at least the mixed-lineage leukemia (MLL) gene, which is associated with a nine individual members of the let-7 family in mammals, and the poor outcome in most treatment programs.2 To date, more than sequence and function of this family are highly conserved in 50 genes have been identified at the translocation breakpoint of mammals. Members of the let-7 family functionally inhibit the 7 the MLL gene, which produce leukemogenic fusion proteins with a expression of several such as the Ras family, 8,9 10,11 common structure that always contains the N-terminal portion of HMGA2 and c-Myc, and regulators such as 12 13 MLL. The activation of some oncogenes that drive cellular CDC25A, CDK6 and cyclin D2. These findings indicate that proliferation or perturb differentiation, such as HOXA9, has been the miRNAs of the let-7 family may function as tumor suppressors. identified as one of the effects of MLL fusion genes in In accordance with this hypothesis, several let-7 genes are located 14 hematopoietic precursors/progenitors, and is thought to be in regions that are frequently deleted in human cancers, and involved in the leukemogenic process.3 let-7 expression is downregulated in some malignancies such as 15,16 MicroRNAs (miRNAs), small non-coding of 20–24 nucleo- lung . When ectopically expressed, let-7 miRNA can 8,13,17 tides, have recently been recognized as important molecules that repress cellular proliferation and tumor development. regulate the expression of many proteins by inhibition We show here that let-7b is downregulated in leukemic cells or cleavage of their mRNA transcripts. miRNAs are involved in with an MLL rearrangement via DNA hypermethylation of its many biological pathways, including cell growth and hemato- regulatory region. In vitro experiments showed that downregula- poietic cell differentiation. Several studies have shown that tion of let-7b may be one of the consequences of MLL fusion dysregulation of miRNA expression is a common feature in activity, possibly involved in the first hit pathway of human cancers, including leukemia, indicating that abnormal leukemogenesis that is driven by an MLL fusion oncoprotein, such expression of miRNAs has an important role in oncogenesis.4 as the dysregulation of the family.

1Department of Pediatrics, Saga University, Saga, Japan; 2Department of Pediatrics, Ehime University Graduate School of Medicine, Toon, Japan and 3Department of Pediatrics, University of Yamanashi, Chuo, Japan. Correspondence: Dr M Eguchi, Department of Pediatrics, Ehime University Graduate School of Medicine, Shitsukawa, Toon, Ehime 791-0295, Japan. E-mail: [email protected] 4These authors contributed equally to this work. Received 22 June 2012; revised 13 August 2012; accepted 14 August 2012; accepted article preview online 24 August 2012; advance online publication, 14 September 2012 Suppression of let-7b by DNA hypermethylation M Nishi et al 390 MATERIALS AND METHODS DNA extraction, bisulfite modification and methylation-specific Cell lines real-time PCR Cell lines established from ALL patients were used in this study. These DNA was extracted from the leukemic samples and ALL cell lines using included six cell lines with MLL gene rearrangement established from standard procedures. After DNA extraction, 100 ng to 1 mg of DNA was infant ALL patients: KOCL45, KOCL58 and KOCL69, which carry an MLL–AF4 modified with sodium bisulfite for DNA methylation analysis using the gene fusion, KOPB26, which carry an MLL–AF9 gene fusion, and KOCL44 EpiTect Bisulfite Kit (Qiagen) according to the manufacturer’s instructions. and KOPN1, which carry an MLL–ENL fusion gene. The KOPN79 and REH After bisulfite treatment, the region of a gene located upstream cell lines, which carry a TEL–AML1 fusion gene, and YAMN90, which carry from the let-7b coding region (RP4-695O20_B.10) was amplified using a 0 an E2A–PBX1 fusion gene, were used as MLL control cells. The specific primer pair with an EcoRI recognition site at the 5 -end. PCR details of the cell lines are described elsewhere.18–22 All cell lines were products were electrophoresed and purified using the QIAquick Gel cultured in RPMI1640 medium supplemented with 10% fetal bovine serum Extraction Kit (Qiagen), digested with EcoRI, and cloned into the pBluescript II SK ( ) vector (Stratagene, Santa Clara, CA, USA). Following at 37 1Cin5%CO2 in air. þ bacterial transformation, at least 20 individual colonies were subjected to sequence analysis to detect DNA methylation. The DNA methylation status was also analyzed by methylation-specific, Patient samples real-time PCR of bisulfite-modified DNA. Primers were designed to amplify Peripheral blood or cells were collected from ALL infants at methylated or unmethylated CpG dinucleotides in the promoter region of diagnosis. According to the Declaration of Helsinki, written informed the gene upstream of let-7b. Real-time PCR was performed using the KAPA consent was obtained from the parents, and all research was approved by SYBR FAST qPCR Kit (Kapa Biosystems, Boston, MA, USA), according to the the institutional review board. Mononuclear cells were separated by Ficoll- manufacturer’s instructions, on an Applied Biosystems 7500 Real-time PCR Paque density centrifugation (GE Healthcare, Piscataway, NJ, USA). More System. All the bisulfite–PCR primers were designed using the Methyl than 90% of the isolated cells were leukemic cells in all samples used in Primer Express software (Applied Biosystems). The validity and sensitivity this study. of methylation-specific quantitative PCR was confirmed using standard amplification curves generated from serially diluted vector DNA containing methylated or unmethylated DNA obtained after bisulfite TaqMan miRNA assay and quantitative reverse transcriptase-PCR sequencing analysis of ALL cell lines (Supplementary Figure 1). Details of the primers for bisulfite–PCR are provided in Supplementary Table 1. Total RNA, including miRNA, was extracted from the leukemic samples and cell lines using the miRNeasy Mini Kit (Qiagen, Valencia, CA, USA) with DNase treatment, according to the manufacturer’s instructions. The Demethylating agent TaqMan real-time reverse transcriptase-PCR (RT-PCR) method was used 5-Azacitidine (Sigma-Aldrich, St Louis, MO, USA) was used at a concentra- to quantify the expression of let-7b miRNA. Reverse was tion of 2 mM in cell culture. The ALL cell lines were cultured with performed using the TaqMan miRNA Reverse Transcription Kit (Applied 5-azacitidine for 7 days. The cells were stained with anti-CD10-FITC Biosystems, Carlsbad, CA, USA), and the expression of mature let-7b was (BioLegend, San Diego, CA, USA), and the expression of CD10 after measured using the TaqMan miRNA Assay (Applied Biosystems, Assay ID 5-azacitidine treatment was analyzed using a FACSCalibur Flow Cytometer 002619). RNU44 expression (Applied Biosystems, Assay ID 002619) was also (BD Biosciences). In addition, total RNA (including miRNA) was extracted as measured and used as an internal reference to normalize let-7b expression. mentioned above and subjected to quantitative real-time PCR. For RT-PCR and real-time quantification of genes other than miRNAs, 1 mg of total RNA was reverse transcribed using the PrimeScript RT-PCR Kit (TaKaRa, Otsu, Japan). Real-time PCR was performed using SYBR Premix Ex MTT assay Taq II (TaKaRa) on an Applied Biosystems 7500 Real-Time PCR System with For the introduction of let-7b miRNA into leukemic cell lines, genomic the PCR conditions recommended by the supplier. All measurements were sequences containing the let-7b coding region (691 bp in length) were performed in duplicate, and the difference in the duplicate threshold cycle amplified by PCR and inserted at the 30-end of the EGFP sequence in the was less than 1 cycle in all samples tested. All experiments were repeated pEGFP–C3 expression vector (Clontech, pEGFP-let-7b). In transfected cells, at least three times, and the data were statistically analyzed using the the let-7b coding sequence is transcribed from the 30- t-test. The primers used for real-time PCR are summarized in (UTR) of the EGFP transcript, producing mature let-7b after miRNA Supplementary Table 1. processing. Leukemic cell lines were transfected with this pEGFP-let-7b construct or the pEGFP–C3 empty vector using the Amaxa Nucleofector II Kit V (Lonza, Basel, Switzerland). Twenty-four hours after transfection, the EGFP-positive fraction was sorted using FACSAria and 2 to 6 Â 104 cells Transfection of MLL fusion genes were plated into each well of a 96-well plate in 100 ml culture medium. The neomycin resistance gene of pMSCVneo vector (Clontech, Mountain After 48 h, 10 ml of MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazo- View, CA, USA) was replaced with the EGFP sequence (pMSCV–EGFP lium bromide; Dojindo, Kumamoto, Japan) solution, 5 mg/ml in phosphate- vector), and coding regions of the MLL fusion genes (MLL–AF4, MLL–ENL buffered saline, was added to each well and incubated at 37 1C for another and MLL–AF9) were introduced into the multi-cloning site of the pMSCV– 6 h. The cells were then lysed by adding 100 ml lysis solution (0.04 N HCl in EGFP vector. Briefly, the 50-portion of the MLL coding region ( 2-propanol) per well. The absorbance of each well at 595 nm was 24–3774 in RefSeq NM_001197104) was removed from the MSCV2.1-MLL- measured using a microplate reader (Bio-Rad Laboratories, Hercules, CA, 23 GPHN vector using EcoRI and BamHI. The rest of the MLL coding region USA). The background value (the absorbance of the control well containing (nucleotides 3775–4355) and the fusion partner sequence (AF4: nucleo- medium without cells after the addition of MTT and the lysis solution) was tides 1446–4040 in NM_005935; AF9/MLLT3: nucleotides 1321–1966 in subtracted from all measured values. All experiments were performed in NM_004529; ENL/MLLT1: nucleotides 177–1844 in NM_005934), including quadruplicate and repeated at least three times. the junction sequences, were amplified from the cDNAs of leukemic cell lines and patient samples by PCR using partner gene-specific reverse primers containing XhoI restriction sites. The PCR products were digested Statistical analysis using XhoI and BamHI. These PCR products and the 50-portion of MLL were Statistical significance was determined by calculating two-tailed P-values introduced into the pMSCV–EGFP vector to form the full-length MLL fusion using the t-test. gene in the vector. Primers used for vector construction are listed in Supplementary Table 1. All constructs were checked by direct sequencing to exclude introduced during their construction. Transient RESULTS transfection of human embryonic kidney (HEK)-293 cells was performed In this study, we first examined the expression of let-7b in ALL using Effectene (Qiagen). At 72-96 h after transfection, enhanced green fluorescent protein (EGFP)-positive cells were sorted on a FACSAria Cell patients and cell lines by quantitative RT-PCR using the TaqMan Sorter (BD Biosciences, San Jose, CA, USA), and total RNA, including miRNA, miRNA assay for let-7b. As shown in Figure 1a, let-7b was was extracted from the sorted EGFP-positive cells using the miRNeasy Mini significantly suppressed in MLL-rearranged leukemic samples, Kit (Qiagen) as described above. In all FACS sorting experiments, TO-PRO-3 compared with samples with germline MLL. Considering that the iodide (Molecular Probes, Eugene, OR, USA) was used to exclude dead cells. expression level of let-7b in peripheral blood mononuclear cells

Leukemia (2013) 389 – 397 & 2013 Macmillan Publishers Limited Suppression of let-7b by DNA hypermethylation M Nishi et al 391 let-7b let-7b 1.0 1.6 1.4 0.8 P<0.001 1.2

1 0.6 0.8

0.4 0.6

Relative expression 0.4 0.2 Relative expression 0.2

0 0 REH MLL-R MLL-G KOPN1 KOCL45 KOCL58 KOCL69 KOCL44 KOPB26 KOPN79 YAMN90

MLL-AF4MLL- MLL-ENL TEL-AML1 E2A- AF9 PBX1 Figure 1. Expression of let-7b in infants with MLL-rearranged acute lymphoblastic leukemia (ALL). (a) Expression of let-7b in infants with MLL- rearranged ALL (MLL-R; n ¼ 24) and MLL germline ALL (MLL-G; n ¼ 5) was measured using the TaqMan microRNA assay. let-7b expression is significantly lower in MLL-rearranged ALL than in MLL-germline ALL. Peripheral blood mononuclear cells from healthy volunteers were used as normal controls, and the relative expression of let-7b was calculated by setting the expression level in control peripheral blood as 1. (b) Expression levels of let-7b miRNA in ALL cell lines determined by the TaqMan microRNA Assay. let-7b is downregulated in the MLL- rearranged cell lines, except KOCL44. was set as 1 in Figure 1a, the expression of let-7b was lower in methylation of its promoter.26 To explore the possibility that let-7b infant ALL samples irrespective of the status of the MLL gene. expression was regulated by DNA methylation in MLL-rearranged Cell lines with an MLL rearrangement (except for the MLL–ENL- infant ALL, the methylation status of the let-7b locus was bearing cell line KOCL44) also showed decreased let-7b expression examined by bisulfite–PCR and sequencing. The let-7b coding compared with the cell line with TEL–AML1 or E2A–PBX1 region was highly methylated in all the ALL cell lines studied, (Figure 1b). The expression of pre-let7b (miRBase accession no. irrespective of the MLL gene status (data not shown). The let-7b is MI0000063), the precursor of let-7b with a stem-loop structure located in the 30-UTR of an uncharacterised gene (RP4- before its processing by , was also downregulated in the ALL 695O20_B.10, illustrated in Figure 3a) that has six producing cell lines with an MLL rearrangement, except for the KOCL44 cell several transcripts (Ensembl transcript ID, ENST00000360737, line, exactly corresponding to the expression of mature let-7b ENST00000381051, ENST00000443490 and ENST00000435439). (Supplementary Figure 2). This indicates that the loss of let-7b The expression of RP4-695O20_B.10 in the ALL cell lines was expression may be because of the suppression of examined using quantitative RT-PCR. Although the TEL–AML1- rather than reduced let-7b miRNA processing. positive cell lines KOPN79 and REH showed weak, but definite, The miRNA let-7b is located on chromosome 22q13 in the expression of the gene, it was almost undetectable in the cell lines 30-UTR of a gene with an unknown function (RP4-695O20_B.10, with an MLL fusion gene, KOCL45 and KOCL69 (Figure 3b). As this Ensembl gene ID, ENSG00000197182). We performed PCR and gene had three possible first exons (designated as 1a, 1b and 1c in quantitative PCR to examine whether genomic deletion around Figure 3a), the expression levels of transcripts with each of the first the let-7b coding region was responsible for its low expression. In exons was analyzed by quantitative RT-PCR in the KOPN79 and all cases examined, no deletions were detected in the region, REH cell lines. As shown in Figure 3c, the transcript containing indicating that transcriptional regulation of let-7b may be 1a was the most abundant in the let-7b-expressing KOPN79 responsible for the loss of its expression in MLL-rearranged and REH cell lines. RT-PCR analysis with primers located in exon 1a leukemic cells (data not shown). and exon 6, which is followed by the 30-UTR that contains the To determine whether MLL fusion proteins regulated the let-7b sequence, revealed the expression of a transcript containing expression of let-7b, we measured let-7b expression after the both exons (Supplementary Figure 4), indicating that the introduction of MLL fusion genes into HEK293 cells, which express promoter region of exon 1a may also act as the regulatory region endogenous let-7b. The expression of EGFP and MLL fusion was of let-7b. simultaneously driven so that MLL fusion-transfected cells could The DNA methylation status of the upstream region of exon 1a easily be separated by cell sorting. At 72–96 h after transfection, was analyzed using bisulfite–PCR analysis. Three major CpG the EGFP-positive population in each set of transfected cells was islands were identified by the Methyl Primer Express software separated by cell sorting (Figure 2a), and let-7b expression was (Applied Biosystems), and the proximal CpG island was selected measured by quantitative RT-PCR. As shown in Figure 2b, let-7b for further analysis. The 11 CpG dinucleotides within this region expression was significantly downregulated by the expression of were analyzed by bisulfite–PCR. In the KOCL45 cell line with the all MLL fusion proteins. The expression of MLL fusion genes in the MLL–AF4 fusion gene, all the cytosines in this CpG island were sorted EGFP-positive cells was examined by RT-PCR, and all methylated, presenting a striking contrast to the KOPN79 cell line samples, except the mock-transfected cells, expressed MLL fusion with the TEL–AML1 fusion gene in which almost all the CpG sites transcripts of the predicted size (Supplementary Figure 3). The were unmethylated (Figure 3d). These results were also confirmed results at 72 and 96 h after transfection were comparable. by methylation-specific real-time quantitative PCR analysis using Infant ALL with an MLL gene rearrangement is characterized by bisulfite-treated DNA. In the KOCL45 and KOCL69 cell lines, only the loss of surface CD10 expression, which is a hallmark of MLL methylated CpG-specific amplification was observed, and the gene rearrangements.2,24,25 The suppression of CD10 expression in amplification of unmethylated CpG was virtually negligible MLL-rearranged infant ALL is reported to be caused by the dense (Figure 4a). By contrast, the KOPN79 cell line mainly showed

& 2013 Macmillan Publishers Limited Leukemia (2013) 389 – 397 Suppression of let-7b by DNA hypermethylation M Nishi et al 392 Mock (EGFP) MLL-AF4 100 100

80 80 EGFP(+) EGFP(+) 46.8% 23.7% 60 60

40 40

20 20

0 0 100 101 102 103 104 100 101 102 103 104

MLL-ENL MLL-AF9 100 % of Max 100

80 80 EGFP(+) EGFP(+) 35.6% 34.3% 60 60

40 40

20 20

0 0 100 101 102 103 104 100 101 102 103 104

EGFP

P<0.001

P<0.001

P<0.001 1.2

1

0.8

0.6 let-7b

0.4

0.2

0 Mock MLL-AF4 MLL-AF9 MLL-ENL Figure 2. MLL fusion-induced downregulation of let-7b miRNA. (a) MLL fusion genes (MLL–AF4, MLL–AF9 and MLL–ENL) were introduced into HEK293 cells with EGFP sequence. At 72–96 h after transfection, EGFP-positive MLL fusion-expressing cells were sorted by FACS for further quantification of let-7b expression. FACS analyses of mock- (EGFP only) and MLL fusion-transfected cells 96 h after transfection are shown. The peak with a dark area in each histogram is the fluorescence control from non-transfected cells. (b) Expression of let-7b in the EGFP-positive populations. A significant downregulation of let-7b is observed in MLL–AF4-, MLL–AF9- and MLL–ENL-expressing HEK293 cells 96 h after transfection.

amplification of unmethylated DNA. In the REH cell line and the analyzed ALL cases with an MLL gene rearrangement, 450% of leukemic samples with the TEL–AML1 fusion gene, amplification of the CpG dinucleotides in the promoter were methylated. By both methylated and unmethylated CpG was observed. In most of contrast, unmethylated CpG was dominant in seven of the nine the ALL cases with an MLL gene rearrangement, methylation- cases without an MLL gene rearrangement. As observed in the specific PCR revealed the dominance of methylated CpG in this REH cell line, leukemic cases with a TEL–AML1 fusion gene showed putative promoter region of let-7b (Figure 4b). In seven of the nine almost equal amounts of methylated and unmethylated CpG

Leukemia (2013) 389 – 397 & 2013 Macmillan Publishers Limited Suppression of let-7b by DNA hypermethylation M Nishi et al 393 RP4-695O20_B.10 gene let-7b the mechanisms underlying the aberrant methylation status in 28 1a1b 1c 263 4 5 human cancer. Expression of DNMTs (DNMT1, DNMT3A and DNMT3B) was examined in MLL-rearranged leukemic cell lines by 10 kb quantitative RT-PCR. DNMT expression levels varied between cell lines; however, DNMTs were not specifically overexpressed in MLL-rearranged leukemic cell lines compared with leukemic cells CpG island CTGCCCGGGCTGGGCTCCCAGGCAGCCTGGCACCGTGCCAGGCTCGCCGCCGAGCTCGGG with germline MLL or normal peripheral blood lymphocytes 12345 (Supplementary Figure 5A). Patient samples were not analyzed CAGCAGGGTGGGGCAAACGCGGAGCACGGAGGACGGGGCCGGGCGACCTGGGCACCAGCA because sufficient amounts of samples were not available; 6 7 8 9 10 11 however, data mining of expression array data published by Stam 29 MSP-f MSP-r et al. showed that DNMTs were not overexpressed in MLL-rearranged leukemia compared with MLL germline RP4-695O20_B.10 RP4-695O20_B.10 leukemia. Expression of DNMTs was analyzed after forced 3 3 1a 2.5 2.5 1b expression of MLL fusion genes in HEK293 cells. No changes in 1c expression were observed in the control and MLL fusion gene- 2 2 ) ) -4 -3 transfected samples, indicating that aberrant expression of DNMTs 1.5 1.5 (x10 (x10 is not responsible for downregulation of let-7b in HEK293 cells 1 1 (Supplementary Figure 5B). 0.5 0.5 Relative expression Relative expression We next examined the effect of the demethylating agent 0 0 KOCL KOCL KOPN REH KOCL KOCL KOPN REH 5-azacitidine on let-7b expression in the KOCL58 and KOPN1 cell 45 69 79 45 69 79 lines, which showed heavily repressed expression of let-7b among MLL-AF4 TEL-AML1 MLL-AF4 TEL-AMl1 the MLL fusion-positive ALL cell lines (Figure 1b). CD10,a membrane metallo-endopeptidase, is one of the genes that is KOCL45 CpG site KOPN79 CpG site REH CpG site heavily methylated in leukemic cells with an MLL fusion gene. The 1234567891011 1 234567891011 1 234567891011 ALL cell lines were cultured with 5-azacitidine for 7 days, and the expression of CD10 was measured by FACS analysis. Before treatment with 5-azacitidine, surface expression of CD10 was barely detectable in these cell lines. Seven days after treatment with 5-azacitidine, CD10 expression was upregulated in all the tested cell lines (Figure 5a). As shown in Figure 5b, let-7b expression was increased in MLL fusion-positive cell lines after 5- azacitidine treatment, whereas it did not change significantly in unmethylated methylated TEL–AML1 fusion-positive cells. To determine the effects of let-7b overexpression in MLL- Figure 3. Hypermethylation of the regulatory region of let-7b in MLL- rearranged leukemic cells, the let-7b genomic region was fused to rearranged cell lines. (a) Structure of the gene located upstream of the 30-UTR of EGFP in the pEGFP–C3 expression vector so that let-7b. According to the information available in the Ensembl let-7b could be co-transcribed with EGFP in the ALL cell lines. At database, let-7b is located in the 30-UTR of the RP4-695O20_B.10 gene, which spans B80 kb with three possible transcriptional start 24 h after transfection, EGFP-positive cells were separated by FACS sites (1a, 1b and 1c). There are three major CpG islands (indicated by sorting (Figure 6a), and cell growth was assessed using the MTT the filled circles in the figure) upstream of exon 1a. (b) Expression of assay. The expression of let-7b in the EGFP-positive fraction was RP4-695O20_B.10 in the ALL cell lines. KOCL45 and KOCL69, two cell quantified using the TaqMan miRNA assay. At 24 h after lines with an MLL fusion gene, exhibit no RP4-695O20_B.10 transfection, let-7b expression was elevated in the EGFP-positive expression. By contrast, weak but definite expression of RP4- fraction of all cell lines examined, irrespective of the MLL gene 695O20_B.10 is detectable in the TEL–AML1 fusion-positive cell lines. status (Figure 6b). The expression of let-7b in the EGFP-positive The primers used for the quantitative RT-PCR are indicated as arrows fractions of these cell lines was sustained at 72 h after transfection. with a dotted line in a, and the expression of b-actin was used as an At 72 h after the induction of let-7b expression, no remarkable internal reference. (c) Comparison of the expression of the three possible first exons (1a, 1b and 1c) of RP4-695O20_B.10 in the ALL cell differences in cell viability were observed between MLL-rear- lines. The transcript containing exon 1a is the most abundant in the ranged and MLL germline cell lines by FACS analysis (data not let-7b-expressing cells KOPN79 and REH. The primers used for shown). The MTT assay showed that cell growth of MLL fusion- quantitative RT-PCR analysis are indicated as solid arrows in a, and positive cell lines was significantly suppressed by let-7b expression the expression of b-actin was used as an internal reference. (d) The (Figure 6c). No significant differences were observed in the cell DNA methylation status of the upstream region of exon 1a. A region lines without MLL fusion after the introduction of let-7b.We within the CpG island located most proximal to exon 1a was assessed whether the apoptotic pathway was activated by let-7b analyzed by bisulfite–PCR. In KOCL45 cells, all 11 CpG dinucleotides using annexin V staining; however, there was no difference in the examined in the region (1 to 11 in Figure 3a) are methylated. By annexin V-positive cell populations of the MLL-rearranged and MLL contrast, almost all of the CpG dinucleotides in KOPN79 cells are unmethylated. The first 10 independent clones examined are shown germline cell lines after 6 h of transfection (data not shown). as representative data. DISCUSSION MLL gene rearrangements, leading to the formation of a fusion gene, are frequently observed in infant leukemia cases and in dinucleotides, but the proportion of unmethylated CpG dinucleo- treatment-related leukemia that occurs after chemotherapy for a tides was always sufficient to drive significant residual promoter primary malignancy. The presence of an MLL fusion gene alone is activity for let-7b expression in these cells (Figures 3d and 4a). not thought to be sufficient to promote complete leukemogen- DNA methylation is mediated by DNA esis,30,31 and overt leukemia with an MLL fusion gene is thought to (DNMTs). Mammals have three DNMTs as follows: DNMT1, involve at least one more genetic event (so called ‘second hits’) DNMT3A and DNMT3B.27 DNMT mutations and overexpression other than the formation of the MLL fusion gene, such as the have been reported in cancer, suggesting that it may be one of of FLT3 kinase. The acceleration of leukemogenesis by

& 2013 Macmillan Publishers Limited Leukemia (2013) 389 – 397 Suppression of let-7b by DNA hypermethylation M Nishi et al 394

methylated P=0.35 unmethylated P<0.01 1 1 1

0.8 0.8 0.8

0.6 0.6 0.6

0.4 /total DNA 0.4 0.4 /total DNA DNA/total DNA methylated DNA methylated DNA 0.2 0.2 0.2

methylated or unmethylated 0 0 0 -R -G REH 4571 5071 9225 MLL MLL P13666 P13688 P13696 P13823 P14473 P14749 P14775 P15114 P15259 P12614 P12745 P13535 P14097 P14624 P15295 P15780 P15912 P17352 KOCL45 KOCL69 KOPN79

MLL-R TEL- MLL-R MLL-G TEL- TEL-AML1 AML1 AML1 Infant ALL Figure 4. Methylation-specific real-time PCR analysis of the upstream region of RP4-695O20_B.10.(a) Results of methylation-specific real-time PCR analysis. The CpG dinucleotide at the 11th position shown in Figure 3a was used for methylation-specific PCR analysis. The positions corresponding to the sites of the forward (MSP-f) and reverse primers (MSP-r) are shown as solid arrows in Figure 3a. Total CpG amplification was used as a reference to normalize the methylated and unmethylated CpG amplifications. The ALL cell lines with MLL–AF4 (KOCL45 and KOCL69) contain only methylated sequences, with negligible amplification of unmethylated sequences, in contrast to the dominant amplification of unmethylated sequences in KOPN79 cells. In the REH cell line, methylated CpG is amplified, but a significant amount of unmethylated CpG is also detectable, indicating residual promoter activity for let-7b expression in these cells. These results are comparable to those obtained by bisulfite sequencing (shown in Figure 3d). (b) Results of methylation-specific real-time PCR on leukemic samples. Most of the infant ALL samples with an MLL gene fusion (MLL-R) show dominant amplification of methylated CpG. By contrast, most of the infant ALL samples without an MLL gene fusion (MLL-G) show dominant amplification of unmethylated CpG. All TEL–AML1-positive samples show amplification of both methylated CpG and unmethylated CpG. (c) Average ratio of methylated CpG in leukemeic samples shown in b. Proportion of methylated CpG is significantly higher in the infant ALL samples with an MLL gene fusion (MLL-R) compared with the samples without an MLL gene fusion (MLL-G).

KOCL58 KOPN1 KOPN79 100 100 100 5-Aza (-) 5-Aza (+) 80 CD10(+) 80 CD10(+) 80 8.86% 14.43% 60 60 60

40 40 40 % of Max

20 5-Aza (-) 20 5-Aza (-) 20 5-Aza (+) 5-Aza (+) 0 0 0 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 CD10 let-7b let-7b 0.08 P<0.001 0.6 P=0.29 0.07 5-Aza (-) 5-Aza (+) 0.5 5-Aza (-) 0.06 5-Aza (+) 0.4 0.05

0.04 0.3

0.03 P<0.001 0.2

Relative expression 0.02 0.1 0.01

0 0 KOCL58 KOPN1 KOPN79 Figure 5. 5-azacitidine induces the expression of let-7b in MLL-rearranged ALL cell lines. (a) Expression of CD10 in the ALL cell lines with an MLL fusion gene after 5-azacitidine treatment. CD10-negative ALL cell lines harboring an MLL fusion were cultured in the presence of the demethylating agent 5-azacitidine. After 7 days of culture, surface expression of CD10 is upregulated. The black areas show CD10 fluorescence without 5-azacitidine, and the gray areas correspond to the fluorescence with 5-azacitidine. (b) After 7 days of culture, expression of let-7b was measured by quantitative RT-PCR. Expression of let-7b is upregulated in the ALL cell lines with an MLL fusion gene after 5-azacitidine treatment.

Leukemia (2013) 389 – 397 & 2013 Macmillan Publishers Limited Suppression of let-7b by DNA hypermethylation M Nishi et al 395 a EGFP transfected EGFP-let-7b transfected b (Mock) (let-7b) 24 h after TF 100 100 36.8±10.4 80 80 10 Mock EGFP(+)EGFP(( EGFP(+) let-7b 60 59.4%% 60 41.3% KOCL 9 40 40 58 8 20 20 7 0 0 6 100 101 102 103 104 100 101 102 103 104 let-7b 100 100 5 % of Max 80 80 4 EGFP(+) EGFP(+) 60 60 3 53.4% 39.5% KOPN 40 40 79 2

20 20 1 0 0 0 KOCL KOPN 100 101 102 103 104 100 101 102 103 104 58 79 EGFP

MTT assay Mock c P=0.47 let-7b P<0.001 P=0.81 1.2 P<0.001 P<0.001 P<0.001 P<0.001 P=0.07

1.0

0.8

0.6

absorbance (595-655nm) 0.4

0.2

0 KOCL KOCL KOPB KOCL KOPN KOPN REH YAMN 58 69 26 44 1 79 90

MLL-AF4MLL-AF9 MLL-ENL TEL-AML1 E2A-PBX1 MLL-R Figure 6. Suppression of cell growth by let-7b expression in MLL-rearranged ALL cell lines. (a) FACS analysis of the ALL cell lines at 24 h after the transfection of pEGFP-C3 (mock) and pEGFP-let-7b vectors. The black and gray areas indicate the fluorescence of cells before and after transfection, respectively. The EGFP-positive fractions were separated from these cells by FACS sorting and subjected to the MTT assay. Left- side column, mock-transfected cells; right-side column, pEGFP-let-7b-transfected cells. (b) Expression of let-7b at 24 h after transfection with the pEGFP-C3 (mock) and pEGFP-let-7b vectors. Data are expressed as the average of three independent experiments. The results of the MLL–AF4-positive KOCL58 and TEL–AML1-positive KOPN79 cell lines are shown as representatives in a and b.(c) Representative results of the MTT assay. Values of mock-transfected cells in each cell line were set as 1. Cell growth of MLL fusion-positive cell lines is significantly suppressed by let-7b expression. an MLL fusion protein and a mutated kinase such as FLT332,33 or As shown in the current study and other reports,5 let-7b was RAS34 also supports the importance of second hits in MLL-related significantly suppressed in MLL-rearranged ALL samples, leukemogenesis. compared with germline MLL samples. The let-7 miRNA family Leukemia with an MLL gene rearrangement frequently involves has been reported to exhibit tumor suppressor functions, which is several additional genetic abnormalities such as deletions and consistent with its developmental role in the promotion of cellular translocations.35 The MLL fusion protein itself can also be a potent differentiation.38 Many previous reports have described an inducer of secondary genetic damage in the presence of association between let-7 and human cancer,8,15,16,39–41 genotoxic agents.36 For these reasons, identification of the indicating that let-7 is one of the miRNAs that are primary targets of a leukemogenic MLL fusion protein is downregulated in many cancers. In leukemia, according to a essential, because the MLL fusion protein itself and its primary previous study, let-7b expression was significantly downregulated direct effectors involved in leukemogenesis are preferable targets in ALL patients compared with acute myeloid leukemia (AML) for specific molecular therapy, rather than the effects induced by patients, with let-7b as the most discriminatory miRNA between the secondary genetic abnormalities that vary among leukemic ALL and AML.42 According to this report, it is also true for patients and also among leukemic cells in a single patient.37 MLL-rearranged leukemias, because expression of let-7b in

& 2013 Macmillan Publishers Limited Leukemia (2013) 389 – 397 Suppression of let-7b by DNA hypermethylation M Nishi et al 396 MLL-rearranged AML cells was not suppressed compared with that Daisuke Tomizawa and Professor Shuki Mizutani of Tokyo Medical and Dental in AML cells with t(8;21), inv(16) and t(15;17), or normal control University for providing some of the infant leukemic samples. We also thank Dr Naoto samples.42 These results suggest that a different mechanism may Fujita, Hiroshima Red Cross Hospital and Atomic-bomb Survivors Hospital for the TEL– be involved in the regulation of let-7b in AML cells with MLL gene AML1-positive leukemic samples. We thank all members of the Committee of the rearrangement, which has a better prognosis than ALL with MLL Japan Infant Leukemia Study Group for their contributions to sample collection. The REH cell line was a kind gift from Dr Yoshinobu Matsuo of Fujisaki Cell Center, gene rearrangement. Hayashibara Biochemical Laboratories. This study was supported by the Japan The mechanism by which let-7b is downregulated in ALL has Children’s Cancer Association, a Grant-in-Aid for Scientific Research and a Grant-in- not yet been elucidated. According to our results, MLL fusion Aid for Cancer Research from the Ministry of Health and Labor of Japan. proteins have the capacity to downregulate the expression of let-7b miRNA by suppressing the expression of the upstream RP4-695O20_B.10 gene. This suppression of RP4-695O20_B.10 REFERENCES expression was due to promoter hypermethylation. Here we have 1 Ishii E, Okamura J, Tsuchida M, Kobayashi M, Akiyama Y, Nakahata T et al. Infant identified that DNA hypermethylation is one of the mechanisms leukemia in Japan: clinical and biological analysis of 48 cases. Med Pediatr Oncol underlying the downregulation of let-7b, especially in MLL- 1991; 19: 28–32. rearranged ALL. 2 Pui CH, Kane JR, Crist WM. Biology and treatment of infant leukemias. Leukemia Several studies have revealed global hypermethylation of 1995; 9: 762–769. promoter CpG islands in MLL-rearranged infant leukemia 3 Argiropoulos B, Humphries RK. Hox genes in hematopoiesis and leukemogenesis. 43,44 compared with other types of childhood leukemia. On the 2007; 26: 6766–6776. other hand, MLL and its fusion proteins are 4 Chen J, Odenike O, Rowley JD. Leukaemogenesis: more than mutant genes. Nat methyltransferases, which regulate gene expression by Rev Cancer 2010; 10: 23–36. modulating the methylation status of rather than that 5 Schotte D, De Menezes RX, Akbari Moqadam F, Mohammadi Khankahdani L, of DNA. Till date, there is no direct evidence that MLL fusion Lange-Turenhout E, Chen C et al. MicroRNAs characterize genetic diversity and proteins have DNA methylation capability. Elucidation of the drug resistance in pediatric acute lymphoblastic leukemia. Haematologica 2011; mechanism underlying DNA hypermethylation by MLL fusion 96: 703–711. 6 Reinhart BJ, Slack FJ, Basson M, Pasquinelli AE, Bettinger JC, Rougvie AE et al. The proteins is an important issue, given the future clinical 21- let-7 RNA regulates developmental timing in Caenorhabditis ele- applications of demethylating agents such as 5-azacitidine in gans. Nature 2000; 403: 901–906. MLL fusion-positive leukemic cases. 7 Johnson SM, Grosshans H, Shingara J, Byrom M, Jarvis R, Cheng A et al. RAS is Several promising approaches can be expected for the use of regulated by the let-7 microRNA family. Cell 2005; 120: 635–647. the let-7 family as a therapeutic agent for various kinds of cancer. 8 Lee YS, Dutta A. The tumor suppressor microRNA let-7 represses the HMGA2 In a previous study, expression of let-7a in the lung using an oncogene. Genes Dev 2007; 21: 1025–1030. adenovirus system resulted in a significant reduction in tumor 9 Wang T, Zhang X, Obijuru L, Laser J, Aris V, Lee P et al. A micro-RNA signature formation.45 In another study, mice infected with a let-7g- associated with race, tumor size, and target gene activity in human uterine expressing lentivirus exhibited a significant reduction in tumor leiomyomas. Genes Cancer 2007; 46: 336–347. number and tumor area.17 10 Kumar MS, Lu J, Mercer KL, Golub TR, Jacks T. Impaired microRNA processing Following the introduction of let-7b, the growth of leukemic enhances cellular transformation and tumorigenesis. Nat Genet 2007; 39: cells carrying an MLL fusion gene was suppressed in comparison 673–677. 11 Koscianska E, Baev V, Skreka K, Oikonomaki K, Rusinov V, Tabler M et al. Prediction with leukemic cells without MLL fusion (Figure 6c), indicating that and preliminary validation of oncogene regulation by miRNAs. BMC Mol Biol 2007; let-7b can suppress the growth of MLL fusion-positive leukemic 8: 79. cells. 12 Huang JC, Babak T, Corson TW, Chua G, Khan S, Gallie BL et al. Using expression Several oncogenes, such as HMGA2, CCND1 and BSG, have been profiling data to identify human microRNA targets. Nat Methods 2007; 4: 46,47 identified as targets of let-7b. Loss of let-7b expression in 1045–1049. leukemic cells may result in deregulated expression of some of 13 Johnson CD, Esquela-Kerscher A, Stefani G, Byrom M, Kelnar K, Ovcharenko D these oncogenes, providing greater growth advantage to the et al. The let-7 microRNA represses cell proliferation pathways in human cells. leukemic cells. Further studies will be required to identify the Cancer Res 2007; 67: 7713–7722. targets of let-7b in the context of MLL fusion gene-driven 14 Calin GA, Sevignani C, Dumitru CD, Hyslop T, Noch E, Yendamuri S et al. Human leukemogenesis. A possible mechanism of leukemogenesis by microRNA genes are frequently located at fragile sites and genomic regions MLL fusion proteins via let-7b downregulation is shown in involved in cancers. Proc Natl Acad Sci USA 2004; 101: 2999–3004. Supplementary Figure 6. 15 Takamizawa J, Konishi H, Yanagisawa K, Tomida S, Osada H, Endoh H et al. In conclusion, the complete mechanism underlying the induc- Reduced expression of the let-7 MicroRNAs in human lung cancers in association with shortened postoperative survival. Cancer Res 2004; 64: 3753–3756. tion of leukemogenesis in infant ALL remains obscure; however, 16 Yanaihara N, Caplen N, Bowman E, Seike M, Kumamoto K, Yi M et al. Unique downregulation of the miRNA let-7b by an MLL fusion protein may microRNA molecular profiles in diagnosis and prognosis. Cancer cell have a role in this process. The let-7b downregulation appears to 2006; 9: 189–198. be the direct result of the presence of an MLL fusion protein, 17 Kumar MS, Erkeland SJ, Pester RE, Chen CY, Ebert MS, Sharp PA et al. Suppression which acts as a potent inducer of DNA hypermethylation in of non-small cell lung tumor development by the let-7 microRNA family. Proc Natl 44,48 leukemic cells. In the near future, introduction of let-7b or the Acad Sci USA 2008; 105: 3903–3908. effective induction of let-7b expression in leukemic cells using a 18 Inukai T, Sugita K, Iijima K, Goi K, Tezuka T, Kojika S et al. Leukemic cells with demethylating agent such as 5-azacitidine may be considered a 11q23 translocations express granulocyte colony-stimulating factor (G-CSF) candidate molecular targeting therapy for infant ALL with MLL receptor and their proliferation is stimulated with G-CSF. Leukemia 1998; 12: gene rearrangements. 382–389. 19 Nakamura M, Sugita K, Inukai T, Goi K, Iijima K, Tezuka T et al. p16/MTS1/INK4A gene is frequently inactivated by hypermethylation in childhood acute lympho- CONFLICT OF INTEREST blastic leukemia with 11q23 translocation. Leukemia 1999; 13: 884–890. 20 Iijima K, Sugita K, Inukai T, Goi K, Tezuka T, Uno K et al. Expression of thrombo- The authors declare no conflict of interest. poietin receptor and its functional role in human B-precursor leukemia cells with 11q23 translocation or Philadelphia chromosome. Leukemia 2000; 14: 1598–1605. 21 Inukai T, Zhang X, Goto M, Hirose K, Uno K, Akahane K et al. Resistance of infant ACKNOWLEDGEMENTS leukemia with MLL rearrangement to -related - We thank Ms Risa Sogabe and Ms Chihiro Tanaka for technical assistance, and inducing ligand: a possible mechanism for poor sensitivity to antitumor immunity. Dr Yujiro Sadakane for technical assistance in microarray analysis. We thank Dr Leukemia 2006; 20: 2119–2129.

Leukemia (2013) 389 – 397 & 2013 Macmillan Publishers Limited Suppression of let-7b by DNA hypermethylation M Nishi et al 397 22 Furuichi Y, Goi K, Inukai T, Sato H, Nemoto A, Takahashi K et al. Fms-like tyrosine leukemia in infants with additional chromosomal abnormalities other than 11q23 kinase 3 ligand stimulation induces MLL-rearranged leukemia cells into quies- translocation. Leuk Res 2008; 32: 1523–1529. cence resistant to antileukemic agents. Cancer Res 2007; 67: 9852–9861. 36 Eguchi M, Eguchi-Ishimae M, Knight D, Kearney L, Slany R, Greaves M. MLL chi- 23 Eguchi M, Eguchi-Ishimae M, Greaves M. The small oligomerization domain of meric protein activation renders cells vulnerable to chromosomal damage: an gephyrin converts MLL to an oncogene. Blood 2004; 103: 3876–3882. explanation for the very short latency of infant leukemia. Genes Chromosomes 24 Pui CH, Behm FG, Downing JR, Hancock ML, Shurtleff SA, Ribeiro RC et al. 11q23/ Cancer 2006; 45: 754–760. MLL rearrangement confers a poor prognosis in infants with acute lymphoblastic 37 Anderson K, Lutz C, van Delft FW, Bateman CM, Guo Y, Colman SM et al. Genetic leukemia. J Clin Oncol 1994; 12: 909–915. variegation of clonal architecture and propagating cells in leukaemia. Nature 25 Cimino G, Rapanotti MC, Rivolta A, Lo Coco F, D’Arcangelo E, Rondelli R et al. 2011; 469: 356–361. Prognostic relevance of ALL-1 gene rearrangement in infant acute leukemias. 38 Bussing I, Slack FJ, Grosshans H. let-7 microRNAs in development, stem cells and Leukemia 1995; 9: 391–395. cancer. Trends Mol Med 2008; 14: 400–409. 26 Ikawa Y, Sugimoto N, Koizumi S, Yachie A, Saikawa Y. Dense methylation of types 39 Shell S, Park S-M, Radjabi AR, Schickel R, Kistner EO, Jewell DA et al. Let-7 1 and 2 regulatory regions of the CD10 gene promoter in infant acute lympho- expression defines two differentiation stages of cancer. Proc Natl Acad Sci 2007; blastic leukemia with MLL/AF4 fusion gene. J Pediatr Hematol Oncol 2010; 32: 104: 11400–11405. 4–10. 40 Brueckner B, Stresemann C, Kuner R, Mund C, Musch T, Meister M et al. The 27 Galm O, Herman JG, Baylin SB. The fundamental role of in hemato- human let-7a-3 locus contains an epigenetically regulated microRNA gene with poietic malignancies. Blood Rev 2006; 20: 1–13. oncogenic function. Cancer Res 2007; 67: 1419–1423. 28 Miremadi A, Oestergaard MZ, Pharoah PDP, Caldas C. Cancer of 41 Zhang B, Pan X, Cobb GP, Anderson TA. microRNAs as oncogenes and tumor epigenetic genes. Hum Mol Genet 2007; 16: R28–R49. suppressors. Dev Biol 2007; 302: 1–12. 29 Stam RW, Schneider P, Hagelstein JA, van der Linden MH, Stumpel DJ, 42 Mi S, Lu J, Sun M, Li Z, Zhang H, Neilly MB et al. MicroRNA expression signatures de Menezes RX et al. Gene expression profiling-based dissection of MLL trans- accurately discriminate acute lymphoblastic leukemia from acute myeloid located and MLL germline acute lymphoblastic leukemia in infants. Blood 2010; leukemia. Proc Natl Acad Sci USA 2007; 104: 19971–19976. 115: 2835–2844. 43 Stumpel DJ, Schneider P, van Roon EH, Boer JM, de Lorenzo P, Valsecchi MG et al. 30 Chen W, Li Q, Hudson WA, Kumar A, Kirchhof N, Kersey JH. A murine Mll-AF4 Specific promoter methylation identifies different subgroups of MLL-rearranged knock-in model results in lymphoid and myeloid deregulation and hematologic infant acute lymphoblastic leukemia, influences clinical outcome, and provides malignancy. Blood 2006; 108: 669–677. therapeutic options. Blood 2009; 114: 5490–5498. 31 Montes R, Ayllon V, Gutierrez-Aranda I, Prat I, Hernandez-Lamas MC, 44 Schafer E, Irizarry R, Negi S, McIntyre E, Small D, Figueroa ME et al. Promoter Ponce L et al. Enforced expression of MLL-AF4 fusion in cord blood CD34 þ cells enhances the hematopoietic repopulating cell function and clonogenic hypermethylation in MLL-r infant acute lymphoblastic leukemia: biology and potential but is not sufficient to initiate leukemia. Blood 2011; 117: therapeutic targeting. Blood 2010; 115: 4798–4809. 4746–4758. 45 Esquela-Kerscher A, Trang P, Wiggins JF, Patrawala L, Cheng A, Ford L et al. The 32 Ono R, Nakajima H, Ozaki K, Kumagai H, Kawashima T, Taki T et al. Dimerization of let-7 microRNA reduces tumor growth in mouse models of lung cancer. Cell Cycle MLL fusion proteins and FLT3 activation synergize to induce multiple-lineage 2008; 7: 759–764. leukemogenesis. J Clin Invest 2005; 115: 919–929. 46 Fu TY, Chang CC, Lin CT, Lai CH, Peng SY, Ko YJ et al. Let-7b-mediated suppression 33 Stubbs MC, Kim YM, Krivtsov AV, Wright RD, Feng Z, Agarwal J et al. MLL-AF9 and of basigin expression and in mouse melanoma cells. Exp Cell Res 2010; FLT3 cooperation in acute myelogenous leukemia: development of a model for 317: 445–451. rapid therapeutic assessment. Leukemia 2008; 22: 66–77. 47 Nishino J, Kim I, Chada K, Morrison SJ. Hmga2 promotes neural self- 34 Ono R, Kumagai H, Nakajima H, Hishiya A, Taki T, Horikawa K et al. Mixed-lineage- renewal in young but not old mice by reducing p16Ink4a and p19Arf expression. leukemia (MLL) fusion protein collaborates with Ras to induce acute Cell 2008; 135: 227–239. leukemia through aberrant Hox expression and Raf activation. Leukemia 2009; 23: 48 Stumpel DJ, Schotte D, Lange-Turenhout EA, Schneider P, Seslija L, de Menezes RX 2197–2209. et al. Hypermethylation of specific microRNA genes in MLL-rearranged infant 35 Tauchi H, Tomizawa D, Eguchi M, Eguchi-Ishimae M, Koh K, Hirayama M et al. acute lymphoblastic leukemia: major matters at a micro scale. Leukemia 2011; 25: Clinical features and outcome of MLL gene rearranged acute lymphoblastic 429–439.

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

& 2013 Macmillan Publishers Limited Leukemia (2013) 389 – 397