Skip to main content

MLL rearrangements in pediatric acute lymphoblastic and myeloblastic leukemias: MLL specific and lineage specific signatures

Abstract

Background

The presence of MLL rearrangements in acute leukemia results in a complex number of biological modifications that still remain largely unexplained. Armstrong et al. proposed MLL rearrangement positive ALL as a distinct subgroup, separated from acute lymphoblastic (ALL) and myeloblastic leukemia (AML), with a specific gene expression profile. Here we show that MLL, from both ALL and AML origin, share a signature identified by a small set of genes suggesting a common genetic disregulation that could be at the basis of mixed lineage leukemia in both phenotypes.

Methods

Using Affymetrix® HG-U133 Plus 2.0 platform, gene expression data from 140 (training set) + 78 (test set) ALL and AML patients with (24+13) and without (116+65) MLL rearrangements have been investigated performing class comparison (SAM) and class prediction (PAM) analyses.

Results

We identified a MLL translocation-specific (379 probes) signature and a phenotype-specific (622 probes) signature which have been tested using unsupervised methods. A final subset of 14 genes grants the characterization of acute leukemia patients with and without MLL rearrangements.

Conclusion

Our study demonstrated that a small subset of genes identifies MLL-specific rearrangements and clearly separates acute leukemia samples according to lineage origin. The subset included well-known genes and newly discovered markers that identified ALL and AML subgroups, with and without MLL rearrangements.

Peer Review reports

Background

The MLL gene located on chromosome 11 band q23 normally functions as a transcription regulator of the HOX genes [1] and is essential for normal mammalian development and hematopoiesis [2]. Chromosomal translocations involving MLL gene represent frequent cytogenetic abnormalities found in hematologic malignancies, occurring in 5–6% of patients with acute myeloid leukemia (AML), 7–10% of acute lymphoblastic leukemia (ALL), 60–70% of all acute leukemias in infants, and in most patients with t-AML/t-ALL secondary to therapy that is targeting topoisomerase II [3]. The function of the various MLL fusion genes [4, 5] and proteins is poorly understood but it appears that the fusion proteins disrupt the ability of wild-type MLL to regulate HOX gene expression, leading to leukemogenesis [6]. Recent studies demonstrated that the presence of MLL rearrangements can be associated to specific antigen [7, 8] and gene expression patterns. Pediatric patients with ALL carrying MLL rearrangement have been successfully distinguished from ALL and AML patients without MLL translocation as a distinct subgroup with a specific gene expression profile [9]. Different gene expression signatures for ALL and AML samples, with and without MLL translocation were also identified in adult patients indicating a common method to comprehensively characterize the MLL mutation [10]. However, the involvement of MLL gene in the onset and progression of leukemia event still remains unclear. The present study encompasses the efforts to clarify the relations between MLL translocation and acute leukemias in pediatric patients: we identified common MLL-related markers that are shared between leukemias with different phenotypes (translocation-related signature) and investigated the role of MLL aberration in acute leukemias with different lineage origin (phenotype-related signature). To this end, two independent cohorts of 140 (training set) and 78 (test set) pediatric patients with ALL/MLL+, AML/MLL+, ALL/MLL- and AML/MLL- have been inspected using gene expression profiling. Separated comparisons based on phenotype and translocation information have been applied to find differentially expressed genes using both comparison (SAM) and prediction (PAM) analyses. Each subgroup has been clearly distinguished using a final subset of 15 probes which separated the training cohort samples into phenotype-related and translocation-related signatures. The strength of our predictor was successfully validated on an independent test cohort. The identified markers have been further examined to explain their biological correlations using gene ontology inquiries. We assessed the key role of previously unexplored genes to specifically characterize MLL translocation as well as the impact of well-known genes in separating acute leukemia samples according to phenotype origin.

Methods

Patients and Samples

A total cohort of 140 pediatric patients was enrolled in the training set. Bone marrow samples were collected at diagnosis from 106 and 34 patients with acute lymphoblastic and myeloblastic leukemia, respectively. The presence of MLL rearrangement was detected in 16/116 (13.8%) ALL patients and in 8/34 (23.5%) AML cases (Table 1). Detailed sample information are provided in Additional File 1. A second cohort of 78 patients has been used as test set to validate the results from supervised analyses performed on the training set.

Table 1 Patient cohort

For all patients, ALL and AML diagnoses were performed by morphology, cytochemistry, cytogenetics, immunophenotype and molecular genetics.

Based on the laboratory diagnosis, patients were risk stratified and enrolled in the following AIEOP protocols: LAL2000 (ALL after 2000 year), LAM2002 (AML after 2002 year), LAL95 (ALL before 2000 year), LAM92 (AML before 2002 year) and Interfant99 (ALL and AML less than 1 year old) [11–13]. This study was conducted after obtaining the informed consent from all patients and following the tenets of the Declaration of Helsinki and was approved by the ethics committees of the participating institutions before the initiation.

Morphological classification

The morphological classification was performed by three independent investigators and the conclusive diagnosis for every case was reported according to the FAB criteria. In cases where consensus was not obtained, the three investigators re-analyzed the slides together in order to obtain consensus as to final diagnosis.

Cytogenetic analyses

Samples were processed and cytogenetic studies were performed using the Q-banding technique. About 15–20 metaphases for each sample were acquired (with CASTI System) and analyzed in order to avoid clone loss. Chromosomes were identified and assigned according to the International System for Human Cytogenetic Nomenclature. FISH analysis was performed on interphase nuclei and where possible on metaphases, using an MLL probe (Vysis).

Biomolecular analyses

The AIEOP MLL protocol, which screens for the fusion gene transcripts MLL-AF4 for t(4;11) translocation, MLL-AF10 for t(10;11), MLL-AF9 for t(9;11) and MLL-ENL for t(11;19), was performed as stated in the previously reported method; briefly, total RNA was isolated using the RNAzol-B reagent (Duotech srl Milan, Italy) following manufacturer's instructions. One microgram of total RNA from each specimen was reverse transcribed using the Superscript reverse transcriptase (Life Technologies Milan, Italy) and random hexamers. PCR amplification was performed using AmpliTaq polymerase (Applied Biosystems) according to the BIOMED-1 protocols. An independent PCR reaction was performed with shift primers for confirmation of each positive result. The ABL housekeeping gene expression was assessed to determine the presence of amplifiable RNA and the efficacy of reverse transcriptions. After electrophoresis, the PCR reaction products were stained with ethidium bromide.

RNA isolation for microarray analysis

Microarray analysis was performed on each sample using Affymetrix Human Genome U133 Plus 2.0 GeneChip. Total RNA was extracted from stored frozen cells of leukaemia specimens using TRIzol (Invitrogen) followed by a purification step (RNeasy Mini Kit, Qiagen). RNA quality was assessed on the Agilent Bioanalyzer 2100 using the Agilent RNA 6000 Nano Assay kit (Agilent Technologies, Waldbronn, Germany). RNA concentration was determined using the NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Inc., Wilmington, DE USA). The overall total RNA quality was assessed by A260/A280 ratio.

Study aim and design

We performed comparisons between subgroups carrying the same lineage (Figure 1, ALL/MLL- vs ALL/MLL+ named "L1"; AML/MLL- vs AML/MLL+ named "L3") or the same translocation information (ALL/MLL- vs AML/MLL- named "L2"; ALL/MLL+ vs AML/MLL+ named "L4") using SAM and PAM algorithms. Combining the results from separated lineage-related comparisons (L1 and L3), we identified a specific MLL signature that is shared between ALL and AML subtypes. Matching the gene lists from translocation-specific comparisons (L2 and L4), we described a common lineage signature for samples with and without MLL translocation.

Figure 1
figure 1

Study aim and design. Study aim and design for supervised analyses. MLL signature was obtained by comparing samples with the same phenotype (red arrows) while lineage signature by comparing samples with and without MLL translocation (green arrows).

The results from SAM and PAM analyses have been tested on an independent cohort to separate samples according to their gene expression profiles.

Gene Expression Profiling

Microarrays analyses were performed using Human Genome U133 Plus 2.0 GeneChip Array (Affymetrix, Santa Clara, CA, USA) and specific RNA isolation method as previously described. Microarray data (CEL files) can be found online at GEO repository (Accession Number: GSE14062).

Statistical analyses were performed using open-source BioConductor package [http://www.bioconductor.org, ver.2.1] for R System software [The R Project for Statistical Computing, ver.2.7.0, http://www.r-project.org].

The training and test sets were inspected using the same analysis procedures. Microarray CEL files were analyzed to evaluate potential errors on the arrays using quality control algorithms on PS intensity signals. To allow gene expression comparison, the robust multiple-array average (RMA) normalization has been performed among all arrays. Genes with low expression variation across all samples have been discarded using interquartile range (IQR>1.15) as filtering criteria; the following analyses have been performed starting from 5130 filtered probes.

Significance analysis of microarray (SAM) package has been used in the training set to find differentially expressed genes among ALL/MLL-, AML/MLL-, ALL/MLL+ and AML/MLL+ groups. Less than one false positive-rated gene was found using false discovery rate (FDR) <1% and q-value = 0 cutoffs.

Obviously genes identified by SAM analysis are not necessarily involved in class prediction [14]. Shrunken centroid algorithm was performed on the training set using Tibshirani's prediction analysis of microarrays (PAM) package. All probes but one (204069_at, one of the two MEIS1 probes included in the prediction list) identified by PAM analysis matched the SAM results. The results from supervised analyses were further validated on an independent test set. SAM gene lists were used to perform gene ontology studies using the Database for Annotation, Visualization and Integrated Discovery (DAVID) web-based tool http://david.abcc.ncifcrf.gov/. The phenotype-related signature was analyzed through the use of Ingenuity Pathway Analysis (Ingenuity® System, http://www.ingenuity.com).

Results

SAM results – Training Set

The training cohort of 140 samples was analyzed using Affymetrix HG-U133 Plus 2.0 gene expression chips. As previously described in the study design, four separated comparisons (labeled L1 to L4) were performed using SAM and PAM methods to identify differentially expressed probes for translocation-specific and lineage-specific groups. Table 2 summarizes the results retrieved by SAM analysis using a false discovery rate (FDR) < 1% and q-value = 0 as statistical cut-off values.

Table 2 SAM results

To identify a MLL translocation-specific signature commonly shared by different lineage subtypes, we compared ALL and AML samples separately. For each comparison, upregulated probes in one group are consequently downregulated in the other, and vice versa. The results from ALL/MLL- vs ALL/MLL+ (L1 comparison: 1013 upregulated probes in the former group, 740 upregulated in the latter) and AML/MLL- vs AML/MLL+ (L3 comparison: 155 overexpressed in the first group, 555 overexpressed in the second) were subsequently matched to generate a unique MLL-specific subset. A total of 379 common translocation-specific probe sets were found from L1 and L3 comparison combining 1753 and 710 probes, respectively [see Additional File 2].

The phenotype-associated signature was identified by comparing samples with and without MLL-mutations separately. We matched the results from ALL/MLL- vs AML/MLL- (L2 comparison: 1378 upregulated probes in the former group, 754 upregulated in the latter) and ALL/MLL+ vs AML/MLL+ (L4 comparison: 379 overexpressed in the first group, 601 overexpressed in the second), obtaining 622 shared phenotype-specific probe sets from 2132 (L2) and 980 (L4) differentially expressed markers [see Additional File 2].

SAM results were tested on 140 samples using an unsupervised hierarchical clustering method: patients were clearly distinguished in two groups according to translocation- (379 probes, Figure 2A) and phenotype-related (622 probes, Figure 2B) signatures.

Figure 2
figure 2

MLL-specific and Phenotype-specific signatures. Unsupervised hierarchical clustering performed on 140 samples using (A) translocation-specific (379 probe sets) and (B) phenotype-specific (622 probe sets) signatures. Cluster A clearly separates patients with MLL (orange and blue labels) from patients without MLL translocation (green and red labels). Cluster B distinguishes samples with AML (orange and green labels) from samples with ALL (blue and red labels).

PAM results – Training Set

SAM comparisons were reproduced in the training cohort using a prediction algorithm (PAM). We identified a total of 15 probe sets, corresponding to 14 genes, with predictive value in separating ALL and AML with and without MLL rearrangement (Table 3). Fourteen out of 15 probes extracted by PAM matched SAM analyses results [see Additional File 2].

Table 3 PAM results

To test prediction performances, we applied unsupervised hierarchical clustering on 140 samples using PAM results. In Figure 3, the upper dendrogram clearly separated our samples into 2 branches according to ALL (106 samples) and AML (34 samples) phenotypes. Each branch further divided into MLL positive and MLL negative subgroups showing distinct gene expression profiles. Additionally, the probe sets retrieved by PAM analysis were associated into 2 major clusters as shown in the dendrogram on the left. A translocation-related signature was characterized by overexpression of ZEB2, SRGAP2P1, TMEM30A, AK2, TMED2, HIPK3, FAM62B and MEIS1 genes in MLL positive samples both ALL and AML, while a common upregulation of PAX5, CD72, CSRP2, LOC100130458, TCL1A and EBF1 genes correlated with ALL patients in a phenotype-related signature.

Figure 3
figure 3

Hierarchical clustering for PAM results. Unsupervised hierarchical clustering using the 15 probe sets identified by PAM analyses on the training set. Each column identifies a patient, each row a probe set. The upper dendrogram separates AML (red and blue labels) from ALL (green and orange) samples. Each group further divides into MLL-positive and MLL-negative samples. The dendrogram on the left groups probe sets according to phenotype- and MLL translocation-related signatures.

Validation on Test Set

The predictor (15 probe sets) obtained by SAM and PAM results has been used for validation on a separated cohort of 78 samples. Patients were distributed as follows: ALL/MLL- n = 54, AML/MLL- n = 11, ALL/MLL+ n = 8, AML/MLL+ n = 5. The dendrogram in Additional File 3 validates the strength of our predictor in discriminating ALL and AML with and without MLL rearrangements. Only one sample (PD529) has been misclassified in the independent cohort, probably due to the small sample size of AML/MLL+ subgroup.

MLL Translocation-related signature

ALL patients with and without MLL translocation have been compared (L1). Two probes for MEIS1 gene (204969_at, 15559477_s_at), which encodes a cofactor for HOX proteins that can accelerate Hoxa9-dependent leukemia, were shown to be strongly expressed in MLL positive samples, as previously reported [15]. MEIS1 quantitatively regulates the differentiation arrest, cycling activity, in vivo progression, and self-renewal of MLL leukemia cells, thereby functioning as a critical and rate-limiting determinant of leukemia stem cell potential [16].

ZEB2 (SIP1) gene, coding a zinc finger E-box binding homeobox 2 protein, was already reported as cancer activating factor. In synergy with another transcription factor (Snail), ZEB2 represses transcription of the E-cad gene by binding E-box on E-cad promoter. Loss of E-cadherin (E-cad) triggers invasion, metastasis, and dedifferentiation in various epithelial carcinomas [17].

Interestingly AK2 and HIPK3 genes, here upregulated in MLL rearrangements, share the same target in Fas-mediated apoptosis pathway, an adaptor molecule (FADD) that interacts with various cell surface receptors and mediates cell apoptotic signals. Adenylate kinase 2 (AK2) regulates mitochondrial apoptosis through the formation of an AK2-FADD-caspase-10 (AFAC10) complex. Acting in concert with FADD and caspase-10, AK2 mediates a novel intrinsic apoptotic pathway that may be involved in tumorigenesis [18]. Another Fas/FADD-interacting kinase, HIPK3 (PKY), was first identified as a putative multidrug-resistant protein from studies of cancer cells. Common death receptor target of AK2 and HIPK3 suggests that a principle role of this kinase family is in regulating various aspects of death receptor signaling [19, 20]. The upregulation of AK2, MEIS1 and TMEM30A (transmembrane protein 30A) genes in MLL-rearranged acute leukemias has also been demonstrated by Faber et al. [21] using gene expression profiling.

The genes SRGAP2P1 (SLIT-ROBO Rho GTPase activating protein 2 pseudogene 1), TMED2 (transmembrane emp24 domain trafficking protein 2) and FAM62B (family with sequence similarity 62 (C2 domain containing) member B) shared similar upregulation in both ALL and AML with MLL translocation but their role in leukemogenesis remains to be explored.

Expression of HOX and NG2 genes

HOX gene expression was evaluated in our patient cohort: HOXA10, HOXA9, HOXA7, HOXA5 and HOXA3 were found to be generally upregulated in samples carrying MLL translocation, both in ALL and AML, while HOXA11 and HOXA6 showed common overexpression in AML/MLL+ cases [16]. NG2 gene, an integral membrane chondroitin sulfate proteoglycan expressed by human malignant melanoma [22] and leukemic [7, 23] cells, exhibited higher expression in both ALL and AML with MLL rearrangement similar to HOX genes (data not shown).

Phenotype-related signature

A total of 6 probe sets with predictive value were isolated using PAM analysis to separate ALL from AML samples with and without the presence of MLL aberration. All probes showed a marked downregulation in AML samples as well as a common involvement in enhancing the B-cell signaling pathway.

PAX5 plays a key role in regulating a number of genes identified by PAM and SAM analyses distinguishing ALL from AML phenotypes. PAX5 gene codes for BSAP, a transcription factor expressed in the developing central nervous system, testis and cells of B lymphocyte lineage except terminally differentiated plasma cells [24, 25]. BSAP-binding sites have been identified in several genes, included in SAM list, encoding VpreB1 [26], BLK [27, 28], CD79a [29], CD19 [30], EBF [31], CD72 [32], BLNK [33, 34] and LEF1 [35]. CD72 antigen is a transmembrane glycoprotein that plays a fundamental role in B-cell activation and proliferation. It has been recently shown its function also in preventing the differentiation of naïve B-cells into plasma cells [36]. CD19 antigen is routinely monitored in immunophenotyping diagnosis as the main surface marker for B cell identification. The CD19 tyrosine phosphorylation is induced by CD72 ligation: the activation of B lymphocytes via CD72 resulted in recruitment and activation of PI 3-K, which was mediated by CD19. Moreover, CD19 enhances membrane IgM signaling like CD79a even known as Igα [37]. BLNK is a B-cell linker protein which is fundamental in addressing pro-B cell to pre-B cell transition [38]. Kabak et al. [39] demonstrated the direct recruitment of BLNK to Igα (CD79a) in signaling pathways and Imai et al. [40] showed that BLNK expression is a common leukogenic event in childhood B-lineage ALL similarly to the B lineage-specific polymerase encoded by DNTT gene.

To assess our findings, we compared PAM and SAM results to a minimal self-sustained gene regulatory module for pro-B cell differentiation and proliferation proposed by Medina K. et al. [41]. All module genes (FLT3, PU.1, E2A, BCL11A, IL-7R, EBF, PAX5) except PU.1 were included in SAM lists showing significant over-expression in MLL+/ALL group (Figure 4).

Figure 4
figure 4

Genes involved in pro-B cell differentiation and proliferation. Regulatory role of PAX5 in B cell differentiation and proliferation. Up-regulated and down-regulated genes in ALL/MLL+ from supervised analyses are depicted red and green, respectively. The minimal self-sustaining gene regulatory module was obtained by Medina K. et al. [36] and expanded according to our results. The graph was generated using Ingenuity Pathways Analysis (IPA) software (Ingenuity® System, http://www.ingenuity.com).

CSRP2 is a member of the CSRP (cysteine and glycine-rich protein) family of genes, encoding a group of LIM domain proteins, which may be involved in regulatory processes important for development and cellular differentiation but also in oncogenesis [42].

During lymphocyte differentiation, the expression of T-cell leukemia/lymphoma 1A gene (TCL1A) begins in pre-B cells, is downregulated in germinal centre B cells, and is silenced in memory B and plasma cells [43–45]. Performing comparative Gene-Set Enrichment Analysis (GSEA), Aggarwal et al. [46] demonstrated that the increased expression of TCL1A was significantly associated (P < 0.05) with some of the most important pathways controlling B-cell lymphoma pathogenesis and heterogeneity, including B-cell receptor pathway.

Gene Ontology results

We further inspected the SAM gene lists using a web-based gene ontology (GO) tool (DAVID) to understand the biological meaning behind large list of genes. Additional File 4 summarizes the GO results (Biological Processes) for both SAM signatures ordered by FDR, defined as the median number of false positive genes divided by the number of significant genes.

The analysis of lineage-specific signature (distinguishing ALL from AML) clearly shows the involvement of hematopoietic and immune response-related genes in the progress of disease, in particular leukocyte activation and hematopoietic organ development. MLL-related signature, distinguishing samples with and without MLL translocation, mainly includes regulatory genes involved in developmental process, biological process, apoptosis and programmed cell death.

AML subgroup without MLL translocation and MEIS1 upregulated

In the present study we showed MEIS1 gene to be upregulated both in ALL and AML with MLL translocation. As depicted in Figure 3, a subgroup of AML without MLL mutation displayed clear overexpression for both MEIS1 probes (as well as HOXA9 and HOXA5 genes, see Additional File 5), confirmed by supervised analysis on the 26 patients with AML/MLL-. Patient distribution is described in Table 4: interestingly, all patients with t(8;21)(q22;q22) and t(15;17)(q22;q21) correlated with down-regulation of MEIS1, while 3 out of 4 patients with inv(16)(p13q22) highly expressed MEIS1 gene. Our results confirm the observations by Grubach et al. [47] who determined RQ-PCR expression levels of a series of PcG genes (including MEIS1) and PcG-regulated genes in 126 AML patients and 20 healthy donors.

Table 4 AML without MLL and MEIS1 upregulated

Discussion

The presence of MLL rearrangements in acute leukemias results in a complex number of biological modifications that still remain largely unexplained. Armstrong et al. [9] proposed MLL rearrangement positive ALL as a distinct subgroup with a specific gene expression profile. However, this signature was related only to MLL with ALL phenotype. Here we show that MLL, from ALL and AML origin, shares a signature identified by a small set of genes suggesting a common genetic disregulation that could be at the base of the mixed lineage leukemia in both phenotypes.

Most MLL aberrations have been successfully characterized by cytogenetics, immunophenotyping, molecular biology and, recently, gene expression profiling.

In the present study, a total cohort of 218 patients was analyzed by microarray approach to identify common deregulated MLL targets shared by different leukemia phenotypes and to inspect MLL involvement in acute leukemias with different lineage origin.

We analyzed gene expression data of 140 (training set) and 78 (test set) pediatric samples carrying ALL (N = 106+62) and AML (N = 34+16), with (N = 24+13) and without (N = 116+65) MLL translocations. Supervised analyses on the training set identified two specific signatures according to lineage origin and MLL presence; the results have been subsequently validated on the test set. The genes MEIS1, ZEB2, SRGAP2P1, TMEM30A, AK2, TMED2, HIPK3 and FAM62B showed marked up-regulation in patients with MLL mutation, both in ALL and AML. GO analysis revealed a general regulatory role of these genes in developmental process, biological process, apoptosis and programmed cell death. Our results confirm the primary function of MEIS1 gene in regulating fundamental MLL leukemia-related and HOX genes processes. MEIS1 up-regulation was also observed in a subgroup of AML patients showing inv(16)-like signature in the absence of MLL mutations.

We further identified a phenotype-related signature that distinguishes lymphoblastic and myeloblastic acute leukemias. A set of 6 probes allowed for separation of ALL and AML with and without MLL mutation, including PAX5, CD72, CSRP2, LOC100130458, EBF1 and TCL1A genes. We corroborate the main role of PAX5 gene in orchestrating basic biological processes such as leukocyte and hematopoietic development [10, 48–50]. Mullighan et al. [51] showed that the genes regulating B-cell development and differentiation are mutated in 40% of pediatric ALL and that PAX5 was the most frequent target of somatic mutation being altered in 31.7% of cases. Similarly, we showed that PAX5, EBF1, CD72 and TCL1A are tightly correlated in the capacity to distinguish lymphoblastic and myeloblastic characteristics also in the presence of MLL mutations illustrating the importance of B-cell receptor signaling pathway in this subset of leukemias. We stressed the role of PAX5 in repressing PU.1 (NF-kappaβ) regulated reporter gene as supported by their opposite expression in B lineage cells with MLL rearrangements [52].

Deregulation events of CSRP2, TCL1A, CD72 and EBF genes in acute pediatric leukaemia have been previously reported using gene expression profiling [43].

The role of CSRP2 gene in leukemogenesis still has to be investigated. Bach et al. [53] demonstrated that proteins like CSRP2 with LIM domains play important roles in embryo development and hematopoiesis. Moreover, Bégay-Muller et al. [54] showed that the LIM domain protein Lmo2 binds to AF6, a translocation partner of the MLL oncogene. CRP2 protein, encoded by CSPR2 gene, was shown to transactivate the proximal promoter region of IL-6 [55] whose receptors are expressed in pediatric ALL with the t(4;11)/AF4 translocation [56]. The overexpression of CSRP2 in MLL mutated samples distinguishing ALL and AML supports a novel role for CSRP2 gene in leukemia development.

Conclusion

Our study demonstrates that a small subset of genes identifies leukemias with MLL-specific rearrangements and clearly separates acute leukemia samples according to lineage origin. The subset includes well-known genes and newly discovered markers that allow for characterization of ALL and AML subgroups, with and without MLL rearrangements.

Abbreviations

The following abbreviations have been used throughout the manuscript. ALL :

Acute Lymphoblastic Leukemia

AML :

Acute Myeloblastic Leukemia

MLL :

Mixed Lineage Leukemia

SAM :

Significance Analysis of Microarray

PAM :

Prediction Analysis of Microarray

FAB :

French-American-British

RMA :

Robust Multiple-array Average

FDR :

False Discovery Rate

GO :

Gene Ontology.

References

  1. Lappin TR, Grier DG, Thompson A, Halliday HL: HOX genes: seductive science, mysterious mechanisms. Ulster Med J. 2006, 75 (1): 23-31.

    PubMed  PubMed Central  Google Scholar 

  2. Hilden JM, Smith FO, Frestedt JL, McGlennen R, Howells WB, Sorensen PH, Arthur DC, Woods WG, Buckley J, Bernstein ID, Kersey JH: MLL Gene rearrangement, Cytogenetic 11q23 Abnormalities, and Expression of the NG2 Molecule in Infant Acute Myeloid leukemia. Blood. 1997, 89 (10): 3801-3805.

    CAS  PubMed  Google Scholar 

  3. Rowley JD: Molecular genetics in acute leukaemia. Leukemia. 2000, 14: 513-17. 10.1038/sj.leu.2401600.

    Article  CAS  PubMed  Google Scholar 

  4. Daser A, Rabbitts TH: The versatile mixed lineage leukaemia gene MLL and its many associations in leukaemogenesis. Semin Cancer Biol. 2005, 15 (3): 175-88. 10.1016/j.semcancer.2005.01.007.

    Article  CAS  PubMed  Google Scholar 

  5. Collins EC, Rabbitts TH: The promiscuous MLL gene links chromosomal translocations to cellular differentiation and tumor tropism. Trends Mol Med. 2002, 8 (9): 436-442. 10.1016/S1471-4914(02)02397-3.

    Article  CAS  PubMed  Google Scholar 

  6. Aplan PD: Chromosomal translocations involving the MLL gene: molecular mechanisms. DNA Repair (Amst). 2006, 5 (9–10): 1265-72. 10.1016/j.dnarep.2006.05.034. Review

    Article  CAS  Google Scholar 

  7. De Zen L, Bicciato S, te Kronnie G, Basso G: Computational analysis of flow-cytometry antigen expression profiles in childhood acute lymphoblastic leukemia: an MLL/AF4 identification. Leukemia. 2003, 17 (8): 1557-65. 10.1038/sj.leu.2403013.

    Article  CAS  PubMed  Google Scholar 

  8. Zangrando A, Luchini A, Buldini B, Rondelli R, Pession A, Bicciato S, te Kronnie G, Basso G: Immunophenotype signature as a tool to define prognostic subgroups in childhood acute myeloid leukemia. Leukemia. 2006, 20 (5): 888-91. 10.1038/sj.leu.2404175.

    Article  CAS  PubMed  Google Scholar 

  9. Armstrong SA, Staunton JE, Silverman LB, Pieters R, den Boer ML, Minden MD, Sallan SE, Lander ES, Golub TR, Korsmeyer SJ: MLL translocations specify a distinct gene expression profile that distinguishes a unique leukemia. Nat Genet. 2002, 30 (1): 41-7. 10.1038/ng765.

    Article  CAS  PubMed  Google Scholar 

  10. Kohlmann A, Schoch C, Dugas M, Schnittger S, Hiddemann W, Kern W, Haferlach T: New insights into MLL gene rearranged acute leukemias using gene expression profiling: shared pathways, lineage commitment, and partner genes. Leukemia. 2005, 19 (6): 953-64. 10.1038/sj.leu.2403746.

    Article  CAS  PubMed  Google Scholar 

  11. Flohr T, Schrauder A, Cazzaniga G, Panzer-Grümayer R, Velden van der V, Fischer S, Stanulla M, Basso G, Niggli FK, Schäfer BW, Sutton R, Koehler R, Zimmermann M, Valsecchi MG, Gadner H, Masera G, Schrappe M, van Dongen JJ, Biondi A, Bartram CR, International BFM Study Group (I-BFM-SG): Minimal residual disease-directed risk stratification using real-time quantitative PCR analysis of immunoglobulin and T-cell receptor gene rearrangements in the international multicenter trial AIEOP-BFM ALL 2000 for childhood acute lymphoblastic leukemia. Leukemia. 2008, 22 (4): 771-82. 10.1038/leu.2008.5.

    Article  CAS  PubMed  Google Scholar 

  12. Pession A, Rondelli R, Basso G, Rizzari C, Testi AM, Fagioli F, De Stefano P, Locatelli F, AML Strategy & Study Committee of the Associazione Italiana di Ematologia e Oncologia Pediatrica (AIEOP): Treatment and long-term results in children with acute myeloid leukaemia treated according to the AIEOP AML protocols. Leukemia. 2005, 19 (12): 2043-53. 10.1038/sj.leu.2403869.

    Article  CAS  PubMed  Google Scholar 

  13. Pieters R, Schrappe M, De Lorenzo P, Hann I, De Rossi G, Felice M, Hovi L, LeBlanc T, Szczepanski T, Ferster A, Janka G, Rubnitz J, Silverman L, Stary J, Campbell M, Li CK, Mann G, Suppiah R, Biondi A, Vora A, Valsecchi MG: A treatment protocol for infants younger than 1 year with acute lymphoblastic leukaemia (Interfant-99): an observational study and a multicentre randomised trial. Lancet. 2007, 370 (9583): 240-50. 10.1016/S0140-6736(07)61126-X.

    Article  CAS  PubMed  Google Scholar 

  14. Dudoit S, Shaffer JP, Boldrick JC: Multiple Hypothesis Testing in Microarray Experiments. 2002, U.C. Berkeley Division of Biostatistics Working Paper Series, [http://www.bepress.com/ucbbiostat/paper110]

    Google Scholar 

  15. Rozovskaia T, Feinstein E, Mor O, Foa R, Blechman J, Nakamura T, Croce CM, Cimino G, Canaani E: Upregulation of Meis1 and HoxA9 in acute lymphocytic leukemias with the t(4:11) abnormality. Oncogene. 2001, 20 (7): 874-8. 10.1038/sj.onc.1204174.

    Article  CAS  PubMed  Google Scholar 

  16. Wong P, Iwasaki M, Somervaille TC, So CW, Cleary ML: Meis1 is an essential and rate-limiting regulator of MLL leukemia stem cell potential. Genes Dev. 2007, 21 (21): 2762-74. 10.1101/gad.1602107.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Miyoshi A, Kitajima Y, Sumi K, Sato K, Hagiwara A, Koga Y, Miyazaki K: Snail and SIP1 increase cancer invasion by upregulating MMP family in hepatocellular carcinoma cells. Br J Cancer. 2004, 90 (6): 1265-73. 10.1038/sj.bjc.6601685.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Lee HJ, Pyo JO, Oh Y, Kim HJ, Hong SH, Jeon YJ, Kim H, Cho DH, Woo HN, Song S, Nam JH, Kim HJ, Kim KS, Jung YK: AK2 activates a novel apoptotic pathway through formation of a complex with FADD and caspase-10. Nat Cell Biol. 2007, 9 (11): 1303-10. 10.1038/ncb1650.

    Article  CAS  PubMed  Google Scholar 

  19. Begley DA, Berkenpas MB, Sampson KE, Abraham I: Identification and sequence of human PKY, a putative kinase with increased expression in multidrug-resistant cells, with homology to yeast protein kinase Yak1. Gene. 1997, 200 (1–2): 35-43. 10.1016/S0378-1119(97)00350-8.

    Article  CAS  PubMed  Google Scholar 

  20. Curtin JF, Cotter TG: JNK Regulates HIPK3 Expression and Promotes Resistance to Fas-mediated Apoptosis in DU 145 Prostate Carcinoma Cells. J Biol Chem. 2004, 279 (17): 17090-100. 10.1074/jbc.M307629200.

    Article  CAS  PubMed  Google Scholar 

  21. Faber J, Krivtsov AV, Stubbs MC, Wright R, Davis TN, Heuvel-Eibrink van den M, Zwaan Christian, Kung Andrew, Armstrong Scott: HOXA9 is required for survival in human MLL-rearranged acute leukemias. Blood. 2009, 113 (11): 2375-85. 10.1182/blood-2007-09-113597.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Iida J, Meijne AM, Spiro RC, Roos E, Furcht LT, McCarthy JB: Spreading and focal contact formation of human melanoma cells in response to the stimulation of both melanoma-associated proteoglycan (NG2) and alpha 4 beta 1 integrin. Cancer Res. 1995, 55 (10): 2177-85.

    CAS  PubMed  Google Scholar 

  23. Zangrando A, Intini F, te Kronnie G, Basso G: Validation of NG2 antigen in identifying BP-ALL patients with MLL rearrangements using qualitative and quantitative flow cytometry: a prospective study. Leukemia. 2008, 22 (4): 858-61. 10.1038/sj.leu.2404952.

    Article  CAS  PubMed  Google Scholar 

  24. Barberis A, Widenhorn K, Vitelli L, Busslinger M: A novel B-cell lineage-specific transcription factor present at early but not late stages of differentiation. Genes Dev. 1990, 849-59. 10.1101/gad.4.5.849. 5

  25. Adams B, Dörfler P, Aguzzi A, Kozmik Z, Urbánek P, Maurer-Fogy I, Busslinger M: Pax-5 encodes the transcription factor BSAP and is expressed in B lymphocytes, the developing CNS, and adult testis. Genes Dev. 1992, 6 (9): 1589-607. 10.1101/gad.6.9.1589.

    Article  CAS  PubMed  Google Scholar 

  26. Okabe T, Bauer SR, Kudo A: Pre-B lymphocyte-specific transcriptional control of the mouse VpreB gene. Eur J Immunol. 1992, 22 (1): 31-6. 10.1002/eji.1830220106.

    Article  CAS  PubMed  Google Scholar 

  27. Lin YH, Shin EJ, Campbell MJ, Niederhuber JE: Transcription of the blk gene in human B lymphocytes is controlled by two promoters. J Biol Chem. 1995, 270 (43): 25968-75. 10.1074/jbc.270.43.25968.

    Article  CAS  PubMed  Google Scholar 

  28. Zwollo P, Desiderio S: Specific recognition of the blk promoter by the B-lymphoid transcription factor B-cell-specific activator protein. J Biol Chem. 1994, 269 (21): 15310-7.

    CAS  PubMed  Google Scholar 

  29. Fitzsimmons D, Hodsdon W, Wheat W, Maira SM, Wasylyk B, Hagman J: Pax-5 (BSAP) recruits Ets proto-oncogene family proteins to form functional ternary complexes on a B-cell-specific promoter. Genes Dev. 1996, 10 (17): 2198-211. 10.1101/gad.10.17.2198.

    Article  CAS  PubMed  Google Scholar 

  30. Kozmik Z, Wang S, Dörfler P, Adams B, Busslinger M: The promoter of the CD19 gene is a target for the B-cell-specific transcription factor BSAP. Mol Cell Biol. 1992, 12 (6): 2662-72.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Maier H, Hagman J: Roles of EBF and Pax-5 in B lineage commitment and development. Semin Immunol. 2002, 14 (6): 415-22. 10.1016/S1044532302000763.

    Article  CAS  PubMed  Google Scholar 

  32. Ying H, Healy JI, Goodnow CC, Parnes JR: Regulation of mouse CD72 gene expression during B lymphocyte development. J Immunol. 1998, 161 (9): 4760-7.

    CAS  PubMed  Google Scholar 

  33. Schebesta M, Pfeffer PL, Busslinger M: Control of pre-BCR signaling by Pax5-dependent activation of the BLNK gene. Immunity. 2002, 17 (4): 473-85. 10.1016/S1074-7613(02)00418-1.

    Article  CAS  PubMed  Google Scholar 

  34. Fusaki N, Tomita S, Wu Y, Okamoto N, Goitsuka R, Kitamura D, Hozumi N: BLNK is associated with the CD72/SHP-1/Grb2 complex in the WEHI231 cell line after membrane IgM cross-linking. Eur J Immunol. 2000, 30 (5): 1326-30. 10.1002/(SICI)1521-4141(200005)30:5<1326::AID-IMMU1326>3.0.CO;2-Q.

    Article  CAS  PubMed  Google Scholar 

  35. Nutt SL, Morrison AM, Dörfler P, Rolink A, Busslinger M: Identification of BSAP (Pax-5) target genes in early B-cell development by loss- and gain-of-function experiments. EMBO J. 1998, 17 (8): 2319-33. 10.1093/emboj/17.8.2319.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Yamazaki T, Nagumo H, Hayashi T, Sugane K, Agematsu K: CD72-mediated suppression of human naive B cell differentiation by down-regulating X-box binding protein 1. Eur J Imm. 2005, 35 (8): 2325-34. 10.1002/eji.200425639.

    Article  CAS  Google Scholar 

  37. Carter RH, Doody GM, Bolen JB, Fearon DT: Membrane IgM-induced tyrosine phosphorylation of CD19 requires a CD19 domain that mediates association with components of the B cell antigen receptor complex. J Immunol. 1997, 158 (7): 3062-9.

    CAS  PubMed  Google Scholar 

  38. Minegishi Y, Rohrer J, Coustan-Smith E, Lederman HM, Pappu R, Campana D, Chan AC, Conley ME: An essential role for BLNK in human B cell development. Science. 1999, 286 (5446): 1954-7. 10.1126/science.286.5446.1954.

    Article  CAS  PubMed  Google Scholar 

  39. Kabak S, Skaggs BJ, Gold MR, Affolter M, West KL, Foster MS, Siemasko K, Chan AC, Aebersold R, Clark MR: The direct recruitment of BLNK to immunoglobulin alpha couples the B-cell antigen receptor to distal signaling pathways. Mol Cell Biol. 2002, 22 (8): 2524-35. 10.1128/MCB.22.8.2524-2535.2002.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Imai C, Ross ME, Reid G, Coustan-Smith E, Schultz KR, Pui CH, Downing JR, Campana D: Expression of the adaptor protein BLNK/SLP-65 in childhood acute lymphoblastic leukemia. Leukemia. 2004, 18 (5): 922-5. 10.1038/sj.leu.2403349.

    Article  CAS  PubMed  Google Scholar 

  41. Medina KL, Singh H: Genetic networks that regulate B lymphopoiesis. Curr Opin Hematol. 2005, 12 (3): 203-9. 10.1097/01.moh.0000160735.67596.a0.

    Article  CAS  PubMed  Google Scholar 

  42. Sánchez-García I, Rabbitts TH: LIM domain proteins in leukaemia and development. Semin Cancer Biol. 1993, 4 (6): 349-58.

    PubMed  Google Scholar 

  43. Ross ME, Zhou X, Song G, Shurtleff SA, Girtman K, Williams WK, Liu HC, Mahfouz R, Raimondi SC, Lenny N, Patel A, Downing JR: Classification of pediatric acute lymphoblastic leukemia by gene expression profiling. Blood. 2003, 102 (8): 2951-9. 10.1182/blood-2003-01-0338.

    Article  CAS  PubMed  Google Scholar 

  44. Teitell MA: The TCL1 family of oncoproteins: co-activators of transformation. Nat Rev Cancer. 2005, 5 (9): 750-10.1038/nrc1711.

    Article  CAS  Google Scholar 

  45. Küppers R: Mechanisms of B-cell lymphoma pathogenesis. Nat Rev Cancer. 2005, 5 (4): 251-62. 10.1038/nrc1589.

    Article  PubMed  Google Scholar 

  46. Aggarwal M, Villuendas R, Gomez G, Rodriguez-Pinilla SM, Sanchez-Beato M, Alvarez D, Martinez N, Rodriguez A, Castillo ME, Camacho FI, Montes-Moreno S, Garcia-Marco JA, Kimby E, Pisano DG, Piris M: TCL1A expression delineates biological and clinical variability in B-cell lymphoma. Mod Pathol. 2009, 22: 206-215. 10.1038/modpathol.2008.148.

    Article  CAS  PubMed  Google Scholar 

  47. Grubach L, Juhl-Christensen C, Rethmeier A, Olesen LH, Aggerholm A, Hokland P, Ostergaard M: Gene expression profiling of Polycomb, Hox and Meis genes in patients with acute myeloid leukaemia. Eur J Haematol. 2008, 81 (2): 112-22. 10.1111/j.1600-0609.2008.01083.x.

    Article  CAS  PubMed  Google Scholar 

  48. Tiacci E, Pileri S, Orleth A, Pacini R, Tabarrini A, Frenguelli F, Liso A, Diverio D, Lo-Coco F, Falini B: PAX5 expression in acute leukemias: higher B-lineage specificity than CD79a and selective association with t(8;21)-acute myelogenous leukemia. Cancer Res. 2004, 64 (20): 7399-404. 10.1158/0008-5472.CAN-04-1865.

    Article  CAS  PubMed  Google Scholar 

  49. Li ZY, Liu DP, Liang CC: New insight into the molecular mechanisms of MLL-associated leukemia. Leukemia. 2005, 19 (2): 183-90. 10.1038/sj.leu.2403602.

    Article  PubMed  Google Scholar 

  50. Busslinger M: Transcriptional control of early B cell development. Annu Rev Immunol. 2004, 22: 55-79. 10.1146/annurev.immunol.22.012703.104807.

    Article  CAS  PubMed  Google Scholar 

  51. Mullighan CG, Goorha S, Radtke I, Miller CB, Coustan-Smith E, Dalton JD, Girtman K, Mathew S, Ma J, Pounds SB, Su X, Pui CH, Relling MV, Evans WE, Shurtleff SA, Downing JR: Genome-wide analysis of genetic alterations in acute lymphoblastic leukaemia. Nature. 2007, 446 (7137): 758-64. 10.1038/nature05690.

    Article  CAS  PubMed  Google Scholar 

  52. Linderson Y, French NS, Neurath MF, Pettersson S: Context-dependent Pax-5 repression of a PU.1/NF-kappaB regulated reporter gene in B lineage cells. Gene. 2001, 262 (1–2): 107-14. 10.1016/S0378-1119(00)00546-1.

    Article  CAS  PubMed  Google Scholar 

  53. Bach I: The LIM domain: regulation by association. Mec Dev. 2000, 91: 5-17. 10.1016/S0925-4773(99)00314-7.

    Article  CAS  Google Scholar 

  54. Bégay-Müller V, Ansieau S, Leutz A: The LIM domain protein Lmo2 binds to AF6, a translocation partner of the MLL oncogene. FEBS Lett. 2002, 521 (1–3): 36-8. 10.1016/S0014-5793(02)02814-4.

    Article  PubMed  Google Scholar 

  55. Bretz JD, Williams SC, Baer M, Johnson PF, Schwartz RC: C/EBP-related protein 2 confers lipopolysaccharide-inducible expression of interleukin 6 and monocyte chemoattractant protein 1 to a lymphoblastic cell line. Proc Natl Acad Sci USA. 1994, 91 (15): 7306-10. 10.1073/pnas.91.15.7306.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  56. Gu L, Zhou M, Jurickova I, Yeager AM, Kreitman RJ, Phillips CN, Findley HW: Expression of interleukin-6 receptors by pediatric acute lymphoblastic leukemia cells with the t(4;11) translocation: a possible target for therapy with recombinant IL6-Pseudomonas exotoxin. Leukemia. 1997, 11 (10): 1779-86. 10.1038/sj.leu.2400757.

    Article  CAS  PubMed  Google Scholar 

Pre-publication history

Download references

Acknowledgements

This work was supported by Fondazione Città della Speranza, P.R.I.N. 40%, MIUR and AIRC. We gratefully acknowledge Dr. Manuela Frascella and Dr. Martina Pigazzi for biomolecular analyses, Dr. Laura Sainati and Dr. Monica Spinelli for morphological analyses, Dr. Anna Leszl for cytogenetic analyses, Dr. Colette Case for manuscript revision.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Andrea Zangrando.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors' contributions

AZ established study aim and design, performed gene expression analyses and wrote the manuscript, MCDO performed microarray experiments, GTK and GB supervised the study and writing of the manuscript. All authors read and approved the final manuscript.

Geertruy te Kronnie and Giuseppe Basso contributed equally to this work.

Electronic supplementary material

12920_2009_100_MOESM1_ESM.xls

Additional file 1: Patient information. Supplementary information about the 140 patient cohort enrolled in the present study. (XLS 30 KB)

12920_2009_100_MOESM2_ESM.xls

Additional file 2: SAM and PAM results. SAM results for MLL translocation-specific (379 probes) and phenotype-specific (622 probes) signatures. In addition, PAM results (15 probes) are listed in a separated column. Probe set number and gene symbol have been provided for each analysis. (XLS 74 KB)

12920_2009_100_MOESM3_ESM.png

Additional file 3: Validation on Test Set. Unsupervised hierarchical clustering of 78 pediatric leukaemia patients using the 15-probes predictor identified by SAM and PAM analyses. The dendrogram separates AML (red and blue labels) from ALL (green and orange) samples. Each group further divides into MLL-positive and MLL-negative samples. (PNG 19 KB)

12920_2009_100_MOESM4_ESM.xls

Additional file 4: Gene Ontology results. Gene ontology results for phenotype-specific (379 probes) and MLL translocation-specific (622 probes) signatures using DAVID web-based tool. Biological processes are ordered by FDR relevance; highlighted rows depict GO terms with FDR <10%. (XLS 198 KB)

12920_2009_100_MOESM5_ESM.xls

Additional file 5: AML patients without MLL mutation and MEIS1 upregulated. SAM results from the analysis of 26 AML patients without MLL translocation. Class comparison analysis was performed between AML/MLL- patients with MEIS1-upregulated and MEIS1-downregulated (see hierarchical clustering results for PAM analyses, Figure 3). Probes up-regulated in one group are considered down-regulated in the other, and vice versa. (XLS 16 KB)

Authors’ original submitted files for images

Rights and permissions

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and permissions

About this article

Cite this article

Zangrando, A., Dell'Orto, M.C., te Kronnie, G. et al. MLL rearrangements in pediatric acute lymphoblastic and myeloblastic leukemias: MLL specific and lineage specific signatures. BMC Med Genomics 2, 36 (2009). https://doi.org/10.1186/1755-8794-2-36

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/1755-8794-2-36

Keywords