ARTICLE Acute Lymphoblastic Leukemia Ferrata Storti Foundation Unraveling the cellular origin and clinical prognostic markers of infant B-cell acute lymphoblastic leukemia using genome-wide analysis Antonio Agraz-Doblas,1,2 Clara Bueno,2# Rachael Bashford-Rogers,3# Anindita Roy,4,# Pauline Schneider,5 Michela Bardini,6 Paola Ballerini,7 Gianni Cazzaniga,6 Thaidy Moreno,1 Carlos Revilla,1 Marta Gut,8,9 Maria G. Valsecchi,10 Haematologica 2019 Irene Roberts,4,11 Rob Pieters,5 Paola De Lorenzo,10 Ignacio Varela,1,$,* Pablo Volume 104(6):1176-1188 Menendez2,12,13,$,* and Ronald W. Stam5 1 Instituto de Biomedicina y Biotecnología de Cantabria (IBBTEC), Universidad de Cantabria- CSIC, Santander, Spain; 2Josep Carreras Leukemia Research Institute-Campus Clinic, Department of Biomedicine, School of Medicine, University of Barcelona, Spain; 3 Department of Medicine, University of Cambridge, Cambridge Biomedical Campus, UK; 4 Department of Paediatrics, University of Oxford, UK; 5Princess Maxima Center for Pediatric Oncology, Utrecht, the Netherlands; 6Centro Ricerca Tettamanti, Department of Pediatrics, University of Milano Bicocca, Fondazione MBBM, Monza, Italy; 7Pediatric Hematology, A. Trousseau Hospital, Paris, France; 8CNAG-CRG, Center for Genomic Regulation, Barcelona, Spain; 9Universitat Pompeu Fabra, Barcelona, Spain; 10Interfant Trial Data Center, University of Milano-Bicocca, Monza, Italy; 11MRC Molecular Haematology Unit, MRC Weatherall Institute of Molecular Medicine, University of Oxford, UK; 12Instituciò Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain and 13Centro de Investigación Biomédica en Red de Cáncer (CIBERONC), ISCIII, Barcelona, Spain These authors contributed equally to this work. # These senior authors contributed equally to this work. $ ABSTRACT B -cell acute lymphoblastic leukemia is the commonest childhood cancer. In infants, B-cell acute lymphoblastic leukemia remains fatal, especially Correspondence: in patients with t(4;11), present in ~80% of cases. The pathogenesis of PABLO MENÉNDEZ t(4;11)/KMT2A-AFF1+ (MLL-AF4+) infant B-cell acute lymphoblastic leukemia pmenendez@carrerasresearch.org remains difficult to model, and the pathogenic contribution in cancer of the IGNACIO VARELA reciprocal fusions resulting from derivative translocated-chromosomes Ignacio.varela@unican.es remains obscure. Here, “multi-layered” genome-wide analyses and validation were performed on a total of 124 de novo cases of infant B-cell acute lym- Received: September 7, 2018. phoblastic leukemia uniformly diagnosed and treated according to the Accepted: December 20, 2018. Interfant 99/06 protocol. These patients showed the most silent mutational landscape reported so far for any sequenced pediatric cancer. Recurrent muta- Pre-published: January 24, 2019. tions were exclusively found in K-RAS and N-RAS, were subclonal and were doi:10.3324/haematol.2018.206375 frequently lost at relapse, despite a larger number of non-recurrent/non-silent mutations. Unlike non-MLL-rearranged B-cell acute lymphoblastic Check the online version for the most updated leukemias, B-cell receptor repertoire analysis revealed minor, non-expanded information on this article, online supplements, B-cell clones in t(4;11)+ infant B-cell acute lymphoblastic leukemia, and RNA- and information on authorship & disclosures: sequencing showed transcriptomic similarities between t(4;11)+ infant B-cell www.haematologica.org/content/104/6/1176 acute lymphoblastic leukemias and the most immature human fetal liver ©2019 Ferrata Storti Foundation hematopoietic stem and progenitor cells, confirming a “pre-VDJ” fetal cellular Material published in Haematologica is covered by copyright. origin for both t(4;11) and RASmut. The reciprocal fusion AF4-MLL was All rights are reserved to the Ferrata Storti Foundation. Use of expressed in only 45% (19/43) of the t(4;11)+ patients, and HOXA cluster published material is allowed under the following terms and genes are exclusively expressed in AF4-MLL-expressing patients. Importantly, conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode. AF4-MLL/HOXA-expressing patients had a significantly better 4-year event- Copies of published material are allowed for personal or inter- free survival (62.4% vs. 11.7%, P=0.001), and overall survival (73.7 vs. 25.2%, nal use. Sharing published material for non-commercial pur- P=0.016). AF4-MLL expression retained its prognostic significance when ana- poses is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, lyzed in a Cox model adjusting for risk stratification according to the sect. 3. Reproducing and sharing published material for com- Interfant-06 protocol based on age at diagnosis, white blood cell count and mercial purposes is not allowed without permission in writing response to prednisone. This study has clinical implications for disease out- from the publisher. come and diagnostic risk-stratification of t(4;11)+ infant B-cell acute lym- phoblastic leukemia. 1176 haematologica | 2019; 104(6) Cellular origin and clinical prognostic markers of infant MLLr B-ALL Introduction Methods B-cell precursor acute lymphoblastic leukemia (BCP- Patients ALL) is the most frequent cancer in children.1 Current 5- Bone marrow or peripheral blood samples from 124 infants year survival rates in pediatric BCP-ALL approach 90%. (<12 months old) diagnosed with either pro-B or pre-B-cell ALL However, BCP-ALL in infants (iBCP-ALL; <1 year of age) were used in this study. The discovery cohort of patients was remains clinically challenging with an aggressive early composed of 42 de novo cases: 27 with the t(4;11) encoding for clinical presentation in uniquely vulnerable hosts.2 MLL-AF4, five with the t(9;11) encoding for KMT2A-MLLT3 Approximately 80% of iBCP-ALL are diagnosed with (MLL-AF9) and ten without MLLr (non-MLL B-other BCP-ALL chromosomal rearrangements involving the mixed-lin- without numerical or structural chromosomal abnormalities eage leukemia (KMTA2, also called MLL) gene, located reported at diagnosis). Additionally, for eight MLL-AF4+ iBCP-ALL on 11q23,3–5 which confers a dismal prognosis especially patients matched diagnostic-relapse samples were available allow- in patients carrying the t(4;11)/KMT2A-AFF1+ (MLL- ing for longitudinal studies. MLL rearrangements were confirmed AF4+).6–8 by fluorescence in-situ hybridization.31,32 For validation, an addi- MLL is a H3K4 histone methyltransferase required for tional cohort of patients, comprising 43 MLL-AF4+, 11 MLL-AF9+, normal hematopoiesis and HOX gene expression.9,10 and 28 non-MLL iBCP-ALL cases, was used. All patients were Leukemia transformation by MLL fusions requires the enrolled in the Interfant99 treatment study. Bone marrow samples recruitment of the H3K79 histone methyltransferase were collected at Erasmus MC-Sophia Children’s Hospital Dot1L to the MLL transcriptional complex.11,12 Indeed, an (Rotterdam, the Netherlands), Armand Trousseau Hospital (Paris, H3K79 methylation profile defines both mouse and France), and San Gerardo Pediatric Hospital (Monza, Italy). human t(4;11)/MLL-AF4+ BCP-ALL.13 Importantly, MLL Complete remission bone marrow samples were available for all rearrangements (MLLr) occur prenatally during embryon- patients. The clinical and genetic features of the patients are pre- ic/fetal hematopoiesis, and the concordance rate for sented in Online Supplementary Table S1. As a control for the RNA- iBCP-ALL in identical twins with a monochorionic pla- sequencing studies, CD34+CD19+ healthy B-cell progenitors were centa is close to 100%.14–17 This, coupled to the extremely purified by fluorescence-activated cell sorting (FACS) from 22- short latency, suggests that MLL fusions might be suffi- week old human fetal livers (FL) as previously described.32 FL cient for leukemogenesis.4 Accordingly, genome-wide hematopoietic stem and progenitor cells (HSPC) were processed studies using both single nucleotide polymorphism and FACS-purified from second trimester human FL as previously arrays and whole-genome sequencing revealed that described.33 Briefly, cells were processed and stained for flow MLLr iBCP-ALL has a very low frequency of somatic cytometry with up to ten fluorophore-conjugated monoclonal mutations with the predominant clone carrying ~1.3 antibodies [antibodies (clone): CD34PECy7 (8G12), CD45RA non-silent mutations and one copy number alteration.18– FITC (HI100), CD19APC (HIB19), CD123PE (9F5), CD90 PECy5 20 Although these studies were performed at low cover- (5E10), CD38 Pacific blue (HIT2), lineage cocktail APC (CD2 age sequencing they reinforce the concept that MLLr (RPA-2.10)/CD3 (OKT3)/CD14 (61D3)/ CD16(CB16)/ CD19 iBCP-ALL requires few additional mutations to induce (HIB19)/CD56 (TULY56)/CD235a (HIR2)]. FACS was performed full transformation. In contrast, MLL-AF4-induced leuke- using a BD FACSAria II (Becton Dickinson). Gates were set with mogenesis has proven difficult to model.4,9 With the unstained and fluorescence minus one controls, on viable cells. exception of a recent work by Lin et al.21,22 who fused Data were analyzed using FlowJo software (Tree Star). Gating human MLL to murine Af4, creating an artificial leuke- strategies are as described in the results section. The study was mogenic human-mouse chimeric fusion, current murine approved by the Barcelona Clinic Hospital (2013/8529) and and humanized models of MLL-AF4+ BCP-ALL do not Hammersmith and Queen Charlotte’s Hospital (04/Q0406/145) faithfully recapitulate the disease pathogenesis/pheno- research ethics committees. type, suggesting that MLL-AF4 per se is insufficient to ini- tiate leukemogenesis.23–28 Statistical analysis The few mutations and copy number alterations pres- For quantitative variables, a one-tailed t-test was used to identi- ent in MLLr iBCP-ALL seem subclonal and not always fy significant differences between groups. For qualitative vari- retained at relapse.20 Intratumor heterogeneity drives ables, a Fisher exact test was used in order to identify significant clonal evolution in response to microenvironmental cues differences between groups of patients. Software for analysis of and cytotoxic treatment and therefore recurrent muta- mutations and gene expression have their own statistical models tions at diagnosis and relapse may be found in minor but explained in detailed in the references. Where multiple tests were clinically relevant subclones.29 Here we aimed to address performed the significance is shown corrected for multiple testing. the clinical relevance of subclonal mutations and gene Mutation allele frequency evolution was plotted with the R pack- expression signatures in a large cohort of iBCP-ALL. To age distribution Fishplot. Patterns Fisher exact test was used to do this, we performed deeper exome sequencing along assess the association between clinical characteristics and pres- with whole-genome DNA- and RNA-sequencing on a ence of RAS mutations or AF4-MLL expression. Event-free sur- large cohort of 50 MLLr and non-MLL iBCP-ALL patients vival was defined as time from diagnosis to first event, i.e. resist- uniformly treated and followed up according to an ance, relapse, death from any cause, or second malignant neo- Interfant treatment protocol.30 Similarly to Anderson et plasm. Observation periods were censored at the time of last con- al.,20 we report a silent mutational landscape in iBCP-ALL tact when no events were reported. Event-free survival curves irrespective of the MLL rearrangement/status. However, were estimated with the Kaplan-Meier method and standard strikingly, our genome-wide DNA and RNA analyses errors (SE) were calculated according to Greenwood. Differences revealed new, clinically relevant information about dis- in event-free survival and overall survival between groups were ease outcome and cell-of-origin for t(4;11) and RAS compared with the log-rank test. Analysis of the prognostic rele- mutations. vance of AF4-MLL/HOXA expression in combination with risk haematologica | 2019; 104(6) 1177 A. Agraz-Doblas et al. stratification according to the Interfant-06 protocol (based on age Andersson et al.,20 likely reflecting the 3-fold larger at diagnosis, white blood cell count and response to prednisone) sequencing coverage (Figure 1A, Online Supplementary was performed with the Cox model and the Wald test. All tests Figure S1 and Online Supplementary Table S3). All mutations were two-sided. Analyses were performed using SAS 9.2. found at diagnosis were validated using orthogonal meth- ods. This mutational frequency is the lowest described for DNA, RNA and B-cell receptor (VDJ) repertoire any other pediatric tumor type according to recent genome-wide analyses and data analysis reports34 (Online Supplementary Figure S2). Intriguingly, one Preparation and analysis of all DNA and RNA genome-wide third of the mutations validated showed a mutant allele high-throughput sequencing is detailed in the Online frequency (MAF) <20% indicating that iBCP-ALL contains Supplementary Methods, Online Supplementary Figure S1 and Online genetically different intratumoral subclones despite its Supplementary Table S2. genomic stability, likely explaining the higher mutational load than that reported by Andersson et al.20 (Figure 1A and Online Supplementary Table S3). Despite the paucity of Results mutations, ~80% of the validated protein-coding muta- tions (90/116) are predicted to produce deleterious effects At diagnosis infant B-cell precursor acute on the protein (Online Supplementary Figure S3A) which lymphoblastic leukemia shows a silent mutational might support a strong selective pressure in iBCP-ALL. To landscape irrespective of MLL gene status gain insights into the molecular mechanisms underlying Whole-exome sequencing and whole-genome sequenc- the accumulation of mutations, we analyzed the enrich- ing analyses showed a silent mutational landscape in the ment of specific mutational signatures as described by three iBCP-ALL subtypes studied here: MLL-AF4+, MLL- Alexandrov et al.35 In the MLL-AF4+ iBCP-ALL subgroup AF9+ and non-MLL (n=42 patients, Online Supplementary we identified a significant enrichment of signature 1 char- Table S1). Our study revealed an average of one genomic acterized by the accumulation of C>T/G>A transitions, rearrangement and 2.5 non-silent single nucleotide vari- linked to a spontaneous deamination of 5-methylcytosine ants, a 2-fold higher number than that reported by (Online Supplementary Figure S3B,C).35 This mutational sig- A B Figure 1. Somatic mutations detected by whole-exome sequencing in the discovery cohort of infant B-cell precursor acute lymphoblastic leukemia. (A) Total num- ber of mutations identified in each individual patient. The total number of non-synonymous mutations (yellow area, right Y axis) and mutant allele frequency (MAF) for each mutation (individual dots, left Y axis) are represented. (B) Oncodrive software identified the PI3K-RAS pathway as the only recurrently mutated pathway in infant B-cell precursor acute lymphoblastic leukemia. The distribution of mutations in genes of the PI3K-RAS pathway is shown for all patients within the three iBCP- ALL subgroups: [total 42 patients: 27 t(4;11)+, 5 t(9;11)+ and 10 MLLwt]. 1178 haematologica | 2019; 104(6) Cellular origin and clinical prognostic markers of infant MLLr B-ALL nature has also been described in other pediatric tumors, (32% vs. 6%, Fisher exact test P=0.01) (Figure 2A,B). suggesting that iBCP-ALL is not subjected to a specific Surprisingly, we observed that many iBCP-ALL patients mutational signature. had mutations in both KRAS and NRAS, or more than one We also determined the molecular breakpoint of all (different) mutation in the same gene (Figure 2A,C). To MLLr at the base-pair level. In t(4;11)/MLL-AF4+ iBCP- further analyze the biological contribution of KRAS and ALL, the AF4 breakpoints were almost invariably localized NRAS mutations, we calculated the MAF of individual within intron 3 whereas MLL breakpoints were found mutations and observed that the majority of patients who between introns 9 and 11 (Online Supplementary Table S4).3 had a single RAS mutation (either KRAS or NRAS) had We found whole-genome sequencing reads compatible MAF scores between 0.20 and 0.45, suggesting that the with an AF4-MLL reciprocal rearrangement in all samples mutation is present in a major leukemic subclone (Online Supplementary Figure S4). AF4-MLL genomic break- (P=0.0025). By contrast, those patients harboring two or points were validated by polymerase chain reaction capil- more RAS mutations displayed MAF scores between 1% lary sequencing and they were located nearby MLL-AF4 and 20%, compatible with these RAS mutations being in breakpoints, confirming a reciprocal chromosomal distinct and smaller leukemic subclones. We then ana- translocation. lyzed the impact of RAS mutations on disease outcome and found no clinical correlation of RAS mutations with RAS-PI3K is the only recurrently mutated pathway either clinical outcome (overall survival, event-free sur- in infant B-cell precursor acute lymphoblastic vival, central nervous system infiltration) or diagnostic leukemia with NRAS mutations being significantly parameters (gender, age, percentage of blasts and white more frequent in t(4;11)+ patients blood cells) (Online Supplementary Figure S5). Despite the low number of mutations found per sample, 38% of the sequenced iBCP-ALL patients displayed acti- Evidence of clone selection and genomic instability vating/gain-of-function mutations in either KRAS or at relapse NRAS. Additional mutations in other genes members of Paired diagnostic-relapse samples were available for the RAS-PI3K pathway such as FGFR4, JAK2, PTPN11, eight MLL-AF4+ iBCP-ALL patients, permitting longitudi- SETD2, or FLT3 were also identified (Figure 1B). To further nal studies. Whole-exome sequencing revealed an 8-fold validate the unique recurrence of KRAS and NRAS muta- increase in the number of somatic non-synonymous tions, we performed targeted sequencing of these muta- mutations at relapse (19.5 mutations/patient, range:1-434, tions in a large, additional validation cohort of infant paired t-test P=0.03) (Figure 3A,B and Online Supplementary patients (n=82) and confirmed that 34% of the iBCP-ALL Table S3). We performed orthogonal validation for 160 cases carry mutations in either KRAS or NRAS36 (Figure random mutations, and 90% and 75% of mutations with 2A). Interestingly, the overall frequency of RAS mutations MAF >15% and <15%, respectively, were confirmed (data differed slightly between the different cytogenetic sub- not shown). Similarly to diagnosis, the majority of the groups of iBCP-ALL, with the MLL-AF4+ subgroup show- somatic mutations found at relapse had MAF commonly ing the highest frequency (42%) and the MLL-AF9+ sub- <30%, suggesting the existence of multiple leukemic sub- group the lowest (19%). This difference was basically clones (Figure 3A). Importantly, none of the new de novo attributed to the frequency of NRAS mutations, which somatic mutations found at relapse was found in more was 6-fold more common in the MLL-AF4+ subgroup than one patient, likely reflecting an intrinsic genomic A B C Figure 2. Frequent somatic mutations in RAS genes in both the discovery and validation cohorts. (A) Specific KRAS and NRAS mutations recurrently found in each patient by high coverage targeted sequencing. (B) Proportions of patients with mutations in KRAS (brown), NRAS (yellow) or both (gray) within the three infant B-cell precursor acute lymphoblastic leukemia subgroups. (C) Mutant allele frequency of KRAS (brown squares) or NRAS (yellow circles) mutations in each individual patient. Discovery cohort, n=42; validation cohort, n=82. haematologica | 2019; 104(6) 1179 A. Agraz-Doblas et al. instability of leukemic clones surviving induction/consoli- nosis to relapse, we performed high-coverage targeted dation chemotherapy. This is further reflected by a signif- sequencing on the identified mutations in paired diagnos- icant enrichment of signature 6 associated with defective tic-remission-relapse samples.37 Importantly, the main DNA mismatch repair, including a higher number of small leukemic clone at relapse was always present at diagnosis indels, observed in MLL-AF4+ patients at relapse (Online although in some cases with a very low MAF, suggesting Supplementary Figure S6A). a chemotherapy-induced clonal pressure selecting for To delineate the evolutionary clonal structure from diag- resistant/adapted leukemic subclones (Figure 3C). A B C Figure 3. Clonal evolution and genomic instability at relapse. (A) Total number of mutations identified for each patient in paired diagnostic- relapse samples. Total number of non-synonymous mutations (yellow area, right Y axis) and mutant allele frequency (MAF) for each mutation (individual dots, left Y axis) are repre- sented for paired diagnostic and relapsed (R) samples. (B) Circos plot representation of the total number of mutations identified at diagnosis and relapse for a representative patient (MA4_17). Genomic rearrangements are represented by lines connecting both breakpoints. Copy number alter- ations (blue=gains, red=losses) are represented by the outer gray circle. Somatic mutations (both single D nucleotide variants and indels) are depicted in the center of the circle and the affected gene is indicated. (C) Graphic representation of clonal evo- lution in paired diagnostic (DX)- relapsed (RL) samples. The number of unique somatic mutations called at diagnosis (orange), relapse (yellow) or shared between DX and REL (red) are indicated. Bigger gene names indicate higher MAF for the mutations shared at DX and REL. (D) Dynamics of RAS- mutated clones identified as MAF in matched DX-Remission-REL trios (n=8). 1180 haematologica | 2019; 104(6) Cellular origin and clinical prognostic markers of infant MLLr B-ALL Interestingly, we found a correlation between the number of the RAS mutations varied among patients: one-third of of mutations and time to relapse in MLL-AF4+ patients, the iBCP-ALL patients had RAS-mutated clones at relapse with a trend towards a higher mutational load in patients (MA4_20 and MA4_22 increased the size of the RAS- with late relapses (Online Supplementary Figure S6B). We mutated initial clone and in MA4_14 a de novo RAS muta- next analyzed the clonal evolution of RAS-mutated tion emerged), whereas it was lost in two-thirds of the leukemic clones at relapse. We found that the contribution patients (MA4_17, MA4_18, MA4_23, MA4_24) (Figure A B C D Figure 4. Transcriptional signature of infant B-cell precursor acute lym- E phoblastic leukemia samples. (A) Heatmap representing FLT3, PROM1, MEIS1 and HOXA gene expression according to the infant B-cell precur- sor acute lymphoblastic leukemia (iBCP-ALL) cytogenetic group and RAS mutations. (B) Top panel: heatmap showing HOXA cluster gene expres- sion according to the expression of the reciprocal fusion AF4-MLL. Bottom panel: quantitative poly- merase chain reaction validating high expression of HOXA cluster genes in t(4;11) iBCP-ALL patients expressing AF4-MLL. (C,D) Four-year event-free survival (C) and overall survival (D) Kaplan-Meier curves for t(4;11) iBCP- ALL patients according to AF4-MLL expression, n=43 t(4;11)+ patients. (E) Heatmap representation of select- ed genes for the signaling pathways most significantly deregulated. Right panels represent positive pathway enrichment called by gene set enrich- ment analysis software. Total 42 patients: 27 t(4;11)+, 5 t(9;11)+ and 10 MLLwt. haematologica | 2019; 104(6) 1181 A. Agraz-Doblas et al. 3D). This indicates that infants with MLL-AF4+ BCP-ALL the pathogenesis of iBCP-ALL, by performing an unbiased relapse irrespective of the status of KRAS and NRAS. transcriptional analysis of the RNA-sequencing data from Thus, subclones carrying KRAS mutations do not exert an the iBCP-ALL patients. We found deregulated expression advantage over non-mutated clones, despite representing of a total of 3,905 genes, of which 2,575 (66%) were a recurrent genetic insult at diagnosis. Hence, this would upregulated and 1,330 (34%) downregulated as compared argue against a leukemia-initiating role for RAS with those of healthy FL-derived B-cell progenitors, illus- mutations.38 Alternatively, RAS mutations might indeed trating the global transcriptional activation nature of MLL be leukemogenic drivers, but the treatment-induced fusions (Online Supplementary Figure S8).25,41 Furthermore, a genetic instability observed at relapse may compensate de significant upregulation of genes involved in the control of novo RAS mutations, acting as new leukemia drivers coop- cell growth, including the CDK inhibitors P21, P16, P19, erating with MLL-AF4 during relapse. P27 and components of the transforming growth factor-b pathway such as TGFB1, SMAD and ACVR1B, was HOXA cluster genes are only expressed in t(4;11)+ observed in iBCP-ALL (Figure 4D and Online patients expressing the reciprocal fusion AF4-MLL Supplementary Figure S9). By contrast, iBCP-ALL showed a which determines clinical outcome robust downregulation of genes involved in DNA integri- To gain insights into the mechanisms underlying leuke- ty checkpoints such as CHEK1, CHEK2, ATM, ATR and mogenesis in these mutationally silent MLLr and MLL RAD17, and in double-strand break repair genes including germline iBCP-ALL patients, we performed RNA- ERCC4, BRCA1, POLA1 and RAD51 (Figure 4E and Online sequencing in the discovery cohort of patients (n=42) Supplementary Figure S9). These transcriptional changes using FL-derived CD34+CD19+ healthy B-cell progenitors were validated by quantitative reverse transcriptase poly- as controls, as these cells most likely represent the healthy merase chain reaction in ten patients per group (Online counterparts of the leukemic blast stalled at the pro/pre-B- Supplementary Figure S10). Deregulation of DNA integrity cell differentiation stage. We first surveyed the expression checkpoints and double-strand break repair genes may of the genes previously reported to be specific to either well contribute to the genomic instability observed at MLLr iBCP-ALL or specifically to MLL-AF4+ iBCP-ALL39. relapse, and might explain the enrichment in C>T/G>A RNA-sequencing profiling confirmed that these genes seg- transitions, associated with the spontaneous deamination regate patients according to the molecular subtype, MLL- of 5-methylcytosine (Online Supplementary Figures S3 and AF4+, MLL-AF9+ and MLL germline (Online Supplementary S6). Figure S7). We also observed, at diagnosis, a strong upreg- By using FL-derived normal B-cell progenitors as con- ulation of the MLL target genes FLT3,40 MEIS1, PROM1 trols, differences between leukemic blasts and their nor- and HOXA genes in many of our MLLr iBCP-ALL samples mal counterparts could be identified but this does not but not in MLL germline samples (t-test, P<0.05) (Figure allow the definition of transcriptomic differences within 4A), thus validating our RNA-sequencing approach. the iBCP-ALL cytogenetic groups. We, therefore, analyzed Strikingly, the reciprocal AF4-MLL fusion gene was dis- the RNA-sequencing data comparing the genes differen- cernibly expressed in 19/43 (45%) of the t(4;11)+ iBCP- tially expressed in MLL-AF4+ versus MLL-AF9+ and MLL- ALL samples, and its expression was always maintained at wildtype iBCP-ALL patients, without considering normal relapse (data not shown). We then compared the genes dif- B-cell progenitors as controls. A gene ontology analysis ferentially expressed between AF4-MLL-expressing and (gene set enrichment analysis, GSEA) performed with the non-expressing t(4;11)+ patients and found a striking posi- genes differentially expressed revealed that MLL-AF4+ tive correlation between the expression of the HOXA gene patients show, as compared to both MLL-AF9+ and MLL- cluster and overexpression of the reciprocal AF4-MLL wildtype patients, a significant upregulation of genes asso- fusion (t-test, P=0.002) (Figure 4B). These AF4- ciated with cellular catabolism, coupled to a significant MLL/HOXA-expressing patients (n=19) had a significantly downregulation of negative regulators of the PI3-MAPK better prognosis than those lacking AF4-MLL/HOXA pathway, as well as of genes involved in lymphoid differ- expression (n=24). Four-year event-free and overall sur- entiation and RNApol II transcriptional regulation (Figure vival rates were 62.4% (SE, 11.3%) versus 11.7% (SE, 5). This suggests, respectively, a metabolic change in MLL- 10.2%) (P=0.001) (Figure 4C), and 73.7% (SE, 10.1%) ver- AF4+ cells towards rapid energy generation while reinforc- sus 25.2% (SE, 10.3%) (P=0.016) (Figure 4D), respectively. ing the basal hyperactivation of the PI3-MAPK pathway When “AF4-MLL expression” was analyzed in a Cox by RAS mutations (Figures 1 and 2), a poorly differentiated model adjusting for risk stratification (medium risk or high cellular origin of t(4;11), and an impairment of the normal risk according to the Interfant-06 protocol based on age at function of AF4, a key component of the RNApol II tran- diagnosis, white blood cell count and response to pred- scription complex. nisone), it retained its prognostic significance with a haz- ard ratio for patients lacking AF4-MLL expression of 3.42 Deep-sequencing analysis of B-cell receptor [95% confidence interval (95% CI): 1.35-8.63; P=0.01) repertoires suggests a hematopoietic stem cell/early compared to those expressing AF4-MLL/HOXA, while pre-VDJ progenitor as the cell-of-origin for t(4;11) risk group was not significant (HR for high risk vs. medi- and RAS mutations um risk, 1.34; 95% CI: 0.59-3.03; P=0.49). This is the first We next analyzed BCR repertoires to gain insights into study showing that AF4-MLL overexpression correlates the immunoglobulin heavy chain (IgH) rearrangement very well with transcriptional deregulation of the HOXA clonal composition of paired diagnostic-relapse samples gene cluster in iBCP-ALL and that the co-expression of from t(4;11)+ iBCP-ALL (4 pairs). BCR are generated AF4-MLL and HOXA gene cluster identifies a subgroup of through DNA recombination during B-cell differentiation t(4;11)+ iBCP-ALL with a very more favorable clinical out- and represent unique markers for each B-cell clone. come. Because the BCR sequence provides a molecular tag for We next explored new molecular pathways involved in each B-cell clone, high-throughput sequencing of BCR 1182 haematologica | 2019; 104(6) Cellular origin and clinical prognostic markers of infant MLLr B-ALL provides a detailed analysis of B-cell population dynamics (Figure 6B). This is in contrast to non-MLL BCP-ALL and clone tracking.42,43 BCR sequencing was therefore per- patients (n=5) including three patients with t(1;19)/TCF3- formed to address whether t(4;11)+ iBCP-ALL cells PBX1 (EF2-PBX1), one with t(12;21)/ETV6-RUNX1 (TEL- expressed fully rearranged BCR from which increased lev- AML1) and one with t(9;22)/BCR-ABL1, which were all els of B-cell clonal expansion may be observed and to found to be significantly clonal, with large B-cell clones determine whether there are detectable levels of B-cell comprising ~3-40% of total BCR (Figure 6B,C).37 Given clonal persistence over time indicative of B-cell clonal sur- the persistence of both t(4;11) and RAS mutations in MLL- vival. BCR sequencing was performed on t(4;11)/MLL- AF4+ iBCP-ALL, the lack of B-cell clonal expansion or per- AF4+ iBCP-ALL peripheral blood samples (blasts >98%) sistence supports the model that t(4;11)/MLL-AF4+ iBCP- using a polymerase chain reaction-based method37 with ALL malignant cells are developmentally stalled at the pro- additional incorporation of unique molecular barcodes, B stage, and that the cellular origin of such genomic driv- allowing for accurate quantitation of relative B-cell clone ers has to be pre-VDJ stem/progenitor cells. frequency. After BCR sequence filtering, each sample Finally, in order to understand whether the fetal cell-of- yielded between 1,583-46,863 BCR (1,213-38,426 unique origin in iBCP-ALL lies upstream of committed B progen- BCR) (Online Supplementary Table S5). itors, we compared the transcriptome of iBCP-ALL blasts We first delineated the relative clonality in these (n=42) with that of highly purified human FL HSPC pop- patients, and found that the BCR repertoires from t(4;11)+ ulations (3-7 for each population) (Figure 7A,B and Online patients did not exhibit significantly expanded VDJ- Supplementary Table S6) by RNA-sequencing. In keeping rearranged B-cell clones (Figure 6A) either at diagnosis or with the results of the BCR analysis, our principal compo- relapse compared to healthy peripheral blood samples nent analysis revealed a gene expression signature for Figure 5. Specific transcriptional differences between MLL-AF4+ and MLL-AF9+ or MLLwt infant B- cell precursor acute lymphoblas- tic leukemia patients. Here, FL- derived CD34+CD19+ progenitors were not included as normalizers in the analysis in order to avoid potential bias. Gene set enrich- ment analysis (GSEA) was per- formed with the genes differen- tially expressed between MLL- AF4+ patients and MLL-AF9+ or MLLwt patients. MLL-AF4+ infant B-cell precursor acute lym- phoblastic leukemia (iBCP-ALL) patients showed a significant overexpression of genes associ- ated with cellular catabolism, coupled to a significant downreg- ulation of negative regulators of the PI3-MAPK pathway, as well as of genes involved in lymphoid differentiation and RNApol II transcriptional regulation as compared to both MLL-AF9+ and MLLwt iBCP-ALL patients. The bottom panels represent positive pathway enrichment called by GSEA software. Total 42 patients: 27 t(4;11)+, 5 t(9;11)+ and 10 MLLwt. haematologica | 2019; 104(6) 1183 A. Agraz-Doblas et al. primitive Lin-CD34+CD38-CD19- FL HSPC (hematopoietic significantly higher in t(4;11)+ patients. In line with our stem cells, multipotent progenitors, and lymphoid-primed previous work we found no recurrent mutations in the multipotent progenitors, which lie upstream of B progen- FLT3 gene.40 itors) very similar to t(4;11)+ iBCP-ALL, while FL-commit- Analysis of clonal evolution of RAS-mutated clones ted B progenitors clustered as a transcriptionally different from diagnosis to relapse revealed that one-third of the entity (Figure 7C). patients still carry RAS mutations at relapse, whereas the other two-thirds of patients who relapse have lost the diagnostic RAS mutation. This is in accordance with Discussion recently published data by Trentin et al.,36 and suggests that the therapy is able to eliminate the RAS-mutated We set out to perform multi-layered sequencing on a clone in some patients, while in other patients the RAS large cohort iBCP-ALL patients, all enrolled in the interna- mutation seems to confer chemoresistance, allowing these tional, collaborative Interfant treatment protocol. The fact clones to evade treatment.44 Intriguingly, ~25% of the that all patients were identically treated provides legitima- patients carry more than one RAS-mutated clone at diag- cy and confidence in potential correlations of clinical nosis, indicating a selection bias towards mutations in the value. Our study revealed an average of 2.5 non-silent sin- RAS genes, or activated RAS pathways during leukemic gle nucleotide variants, a 2-fold higher number than that transformation. From this perspective, the occurrence of reported by Andersson et al.,20 likely reflecting the 3-fold patients carrying multiple distinct clones with activated larger sequencing coverage. This silent mutational land- RAS pathways may point to convergent evolution of scape, even in non-MLL iBCP-ALL, likely reflects the very clones capable of controlling the proliferation rate. young age of these patients, reinforcing the notion that However, arguing against this is the substantial represen- infant cancer is a developmental disease with not enough tation of patients not carrying RAS mutations at all. time to develop somatic mutations. We also found the Hence, the role of RAS mutations in t(4;11)+ iBCP-ALL only recurrent, but subclonal, mutations occur in the remains obscure, and the available data suggest that RAS KRAS and NRAS genes (gain-of-function mutations), pathway mutations are unlikely leukemia-initiating although the frequency of subclonal NRAS mutations is lesions. Indeed, Tamai et al.45 showed that leukemogenesis A B Figure 6. Analysis of B-cell receptor repertoires suggest a hematopoietic stem cell/early pre-VDJ progenitor as the cell- of-origin for t(4;11)/MLL-AF4+ infant B-cell precursor acute lymphoblastic leukemia. (A) Cloud-plots of B-cell receptor (BCR) repertoires from two rep- resentative t(4;11)+ infant B-cell precursor acute lymphoblastic leukemia (iBCP-ALL) patients depicting the existence of many minor non-expanded B-cell clones either at diagnosis or relapse. Each vertex represents a unique BCR sequence, and the relative vertex size is propor- tional to the number of identical reads. (B) Largest BCR clone size in t(4;11)+ iBCP-ALL, healthy individuals and non- C t(4;11)+ pediatric BCP-ALL. (C) Cloud-plots of BCR repertoires of representative t(1;19)/E2A- PBX1+, t(12;21)/TEL-AML1+ and t(9;22)/BCR-ABL+ patients showing high clonality of B-cell clones. The samples from the iBCP-ALL patients who were BCR-sequenced were four MLL- AF4+ diagnostic-relapse pairs, three E2A-PBX1+ samples, one TEL-AML1+ sample and one BCR-ABL+ sample. 1184 haematologica | 2019; 104(6) Cellular origin and clinical prognostic markers of infant MLLr B-ALL of transgenic mice expressing human MLL-AF4 could be status and parameters associated with diagnosis or disease significantly accelerated by KRAS mutations. However, outcome such as overall survival, event-free survival, cen- although activated KRAS did cooperate with MLL-AF4 in tral nervous system infiltration, gender, percentage of human cord blood-derived CD34+ HSPC to promote blasts and white blood cells and age, further supporting extramedullary infiltration and central nervous system the concept that RAS mutations are not leukemia-initiat- infiltration it failed to initiate leukemia in engrafted mice.27 ing/propagating lesions. Importantly, we report a lack of correlation between RAS Clearly, this brings us back to the central question of A B C Figure 7. Comparison of the transcriptome of human fetal CD34+ hematopoietic stem and progenitor cell populations to infant B-cell precursor acute lymphoblastic leukemia. (A) Schematic representation of B-cell development in human fetal liver (FL) showing immunophenotypic definitions for hematopoi- etic stem cells (HSC), multipotent progenitors (MPP), lymphoid- primed multipotent progenitors (LMPP), committed B progeni- tors (CBP) and B cells. The onset and expected patterns of IgH rearrangements59,60 are depicted as red arrows. (B) Sorting strat- egy for FL hematopoietic stem and progenitor (HSPC) popula- tions by fluorescence-activated cell sorting. The sorting gates for each population are shown in representative flow plots on the left. The purity of the sorted populations is depicted on the right demonstrating >95% purity. (Lin, Lineage cocktail). (C) Principal component analysis of gene expression of infant B-cell precursor acute lymphoblastic leukemia (iBCP-ALL) samples (n=42) and FL HSPC populations (n=3-7) using the top 1,000 variably expressed genes, as determined by RNA-sequencing. FL HSPC as in (A); MAF4, MLL-AF4+ iBCP-ALL; MA9, MLL-AF9+ iBCP-ALL; MLLwt, MLL wildtype iBCP-ALL. haematologica | 2019; 104(6) 1185 A. Agraz-Doblas et al. whether or not MLL-AF4 by itself is sufficient to initiate MLL genomic fusion was previously detected in 80-85% BCP-ALL in humans. The silent mutational landscape of t(4;11)+ patients.5,52 Our “multi-layered omics” observed in this study and by others20 certainly votes in approach allowed for the exact characterization of the favor of MLL-AF4+ iBCP-ALL being initiated by a single t(4;11) molecular DNA/RNA break points and the identi- “big-bang” transformation hit, probably in a short-lived fication of those patients expressing the reciprocal AF4- but highly proliferative prenatal B-cell progenitor.4 This MLL fusion. We now report that the AF4-MLL reciprocal hypothesis is supported by recent work by Lin et al., who fusion is expressed in only 50% of t(4;11)+ iBCP-ALL indeed demonstrated that enforced expression of a fusion patients. Strikingly, there was a previously unrecognized transcript consisting of human MLL and murine Af4 in and very significant positive correlation between the cord blood-derived CD34+ HPSC is sufficient to induce upregulation of the HOXA gene cluster and the expres- pro-B ALL in xenografted immunodeficient mice.21,22 Yet, sion of AF4-MLL. Of note, a recent study showed that similar results using a human MLL-AF4 transcript remain approximately half of t(4;11)+ patients do not have an to be established. activated HOXA signature.44,53,54 Furthermore, in the Although MLL-AF4 by itself may be sufficient to induce recent MLL-Af4-induced B-ALL xenograft model MLL- BCP-ALL without significant contributions from coopera- Af4 failed to bind to HOXA genes and therefore HOXA tive genetic lesions, the contribution of the MLL-AF4 and gene expression was not upregulated.21 This is experi- RAS mutations to leukemogenesis should take into mentally supported by chromatin immunoprecipitation- account the nature of both the fetal target cell for transfor- sequencing analysis performed in human embryonic stem mation and the leukemia-initiating cell, according to the cells transduced with MLL-AF4, AF4-MLL or both show- increasingly accepted stochastic stem cell model of B- ing a significant enrichment of H3K79 methylated regions ALL.46,47 Here, we employed high-throughput BCR- specifically associated with HOX-A cluster genes in dou- sequencing of the IgH locus to delineate the dynamics of ble fusion-expressing hematopoietic derivatives, estab- clonality of B-cell populations in paired diagnosis-relapse lishing a functional and molecular cooperation between samples of t(4;11)/MLL-AF4+ iBCP-ALL. While pediatric MLL-AF4 and AF4-MLL fusions during human patients with E2A-PBX1+, TEL-AML1+ and BCR-ABL1+ B- hematopoietic development (data not shown). Strikingly, ALL all had significantly clonal disease, with a major VDJ AF4-MLL-expressing patients had a 5-fold longer event- rearranged B-cell IgH clone accounting for up to 40% of all free survival and a 3-fold longer overall survival compared BCR, infants with MLL-AF4+ BCP-ALL exhibited a BCR to t(4;11)+ iBCP-ALL patients lacking AF4-MLL expres- repertoire composed of thousands of minor, non-expanded sion, which is in line with previous reports suggesting VDJ rearranged IgH B-cell clones. Because MLL fusions are that high HOXA gene expression is associated with clonal and RAS mutations are found in clones of relative improved survival and lower risk of relapse.22,39 Because big size, this suggests that MLL fusions with or without the expression of AF4-MLL is not analyzed in routine RAS mutations are likely to originate in primitive fetal pro- molecular diagnosis, our “multi-layered omics” approach genitors that have a germline or an incompletely was critical to unraveling the association between AF4- rearranged (DJ) IgH locus.48 Indeed, an unsupervised com- MLL and HOXA expression, thus identifying a novel sub- parison of the transcriptome of FL HSPC populations and group of t(4;11)+ iBCP-ALL with better clinical outcome. iBCP-ALL blasts suggests that while the gene expression of It is very important for routine diagnostic and clinical primitive FL HSPC (Lin-CD38-CD34+CD19- populations) is practice that when the expression of AF4-MLL was eval- similar to that of iBCP-ALL, FL B progenitors uated in a Cox model adjusting for risk stratification (CD34+CD19+) are transcriptionally distinct. Our data ele- (medium risk or high risk according to the Interfant-06 gantly reinforces previous fluorescence in-situ hybridiza- protocol), it retained its prognostic significance. tion findings suggesting that a primitive “pre-VDJ” Mechanistically, AF4-MLL contains the SET domain dis- stem/progenitor cell (perhaps CD34+CD19-) may represent rupted from its "specification domain", the N-terminal the cell in which both t(4;11) and RAS mutations arise.14,31,49 portion of MLL, which binds to MEN1 and LEDGF thus Cooperative leukemogenic events in iBCP-ALL may shaping the gene-targeting module of the MLL gene. need to be sought beyond genetic insults; for instance, epi- When AF4-MLL is expressed, the N-terminal portion is genetic and transcriptomic deregulation. MLL-AF4 might substituted by the AF4 N-terminus (AF4N) which is the only induce BCP-ALL in cells that meet certain epigenetic crucial domain for binding to and strongly activating RNA and transcriptomic make-up criteria, either influenced by polymerase II (RNAP II) for transcriptional elongation. microenvironmental cues, or characteristic of the cell-of- Thus, expression of A4M-MLL may induce robust RNAP origin.31 Indeed, lesions such as RAS mutations may con- II-dependent gene transcription by overwriting the elon- tribute to disease pathogenesis only against certain intrin- gation control process in a dominant fashion.55–58 We sic epigenetic or transcriptomic backgrounds present in hypothesize that a likely function of AF4-MLL could be to the cell in which the MLL translocations occurred50,51. This prepare the ground for MLL-AF4 or other transcription is supported by the limited impact of RAS mutations in factors to skew normal and leukemic hematopoietic cell transcriptomic signatures associated with leukemia origin, fate decisions. This also explains why MLL-AF4, but not development and pathogenesis, although this is likely due AF4-MLL, seems to be necessary in 100% of patients. to the subclonal nature of RAS mutations.38 However, in Despite being a developmental cancer, iBCP-ALL line with the reported contribution of RAS mutations to patients did not show reactivation of pluripotent or extramedullary infiltration of MLLr BCP-ALL blasts,27 embryonic-like gene expression signatures as revealed by RAS-mutated patients displayed a transcriptomic signa- RNA-sequencing. Additional research is required to deci- ture associated with migration. pher the nature of the insults initiating MLLr iBCP-ALL, as The functional and molecular contribution of the recip- so far we can only speculate on the data currently avail- rocal fusion genes resulting from the derivative translo- able. Whole-genome pyrosequencing will likely provide cated chromosomes remains obscure in cancer. The AF4- unique insights into the DNA methylome landscape of 1186 haematologica | 2019; 104(6) Cellular origin and clinical prognostic markers of infant MLLr B-ALL this mutationally silent iBCP-ALL. This study has clinical Española Contra el Cáncer (AECC-CI-2015), FERO implications in the diagnostic risk-stratification of t(4;11)+ Foundation, and the ISCIII (PI14-01191) to CB. PM/IV also iBCP-ALL. acknowledge financial support from The Obra Social La Caixa- Fundaciò Josep Carreras, The Inocente Inocente Foundation, Acknowledgments Fundación Ramón Areces and The Generalitat de Catalunya We would like to thank the Santander Supercomputing Service (SGR330). IR was supported by a Programme Grant from for IT support. The human fetal material was provided by the Bloodwise (LLR 13001) and by the Oxford NIHR Biomedical Joint MRC/Wellcome Trust (grant# MR/R006237/1) Human Centre based at Oxford University Hospitals NHS Trust and Developmental Biology Resource (http://hdbr.org). This work University of Oxford. PM is an investigator of the Spanish Cell was supported by the European Research Council (CoG-2014- Therapy cooperative network (TERCEL). AR was supported by 646903 to PM; and StG-2014-637904 to IV), the Spanish a Clinician Scientist Fellowship from Bloodwise (14041). This Ministry of Economy and Competitiveness (SAF-SAF2013- work was motivated by our patients and it honors the vital exam- 43065 to PM and SAF2016-76758-R to IV), the Asociación ple given to us by the family of AMC. References oncogene in infant leukaemias. Nature. al. Activated KRAS cooperates with MLL- 1993;363(6427):358–360. AF4 to promote extramedullary engraftment 16. Gale KB, Ford AM, Repp R, et al. and migration of cord blood CD34+ HSPC 1. Pui C-H, Evans WE. A 50-year journey to Backtracking leukemia to birth: identifica- but is insufficient to initiate leukemia. cure childhood acute lymphoblastic tion of clonotypic gene fusion sequences in Cancer Res. 2016;76(8):2478–2489. leukemia. Semin Hematol. 2013;50(3):185– neonatal blood spots. Proc Natl Acad Sci U S 28. Krivtsov A V., Armstrong SA. MLL translo- 196. A. 1997;94(25):13950–13954. cations, histone modifications and 2. Pui CH, Mullighan CG, Evans WE, Relling 17. Bueno C, Montes R, Catalina P, Rodríguez leukaemia stem-cell development. Nat Rev MV. Pediatric acute lymphoblastic leukemia: R, Menendez P. Insights into the cellular ori- Cancer. 2007;7(11):823–833. where are we going and how do we get gin and etiology of the infant pro-B acute 29. Varela I, Menendez P, Sanjuan-Pla A. there? Blood. 2012;120(6):1165–1174. lymphoblastic leukemia with MLL-AF4 Intratumoral heterogeneity and clonal evolu- 3. Meyer C, Burmeister T, Gröger D, et al. The rearrangement. Leukemia. 2011;25(3):400– tion in blood malignancies and solid tumors. MLL recombinome of acute leukemias in 410. Oncotarget. 2017;8(39):66742–66746. 2017. Leukemia. 2018;32(2):273–284. 18. Bardini M, Galbiati M, Lettieri A, et al. 30. Driessen E, Lorenzo P, M C, et al. Outcome 4. Sanjuan-Pla A, Bueno C, Prieto C, et al. Implementation of array based whole- of congenital acute lymphoblastic leukemia Revisiting the biology of infant t(4;11)/MLL- genome high-resolution technologies con- treated on the Interfant-99 protocol. Blood. AF4+ B-cell acute lymphoblastic leukemia. firms the absence of secondary copy-number 2009;114(18):3764–3768. Blood. 2015;126(25):2676–2685. alterations in MLL-AF4-positive infant ALL 31. Menendez P, Catalina P, Rodríguez R, et al. 5. Marschalek R. Mechanisms of leukemogen- patients. Leukemia. 2011;25(1):175–178. Bone marrow mesenchymal stem cells from esis by MLL fusion proteins. Br J Haematol. 19. Dobbins SE, Sherborne AL, Ma YP, et al. The infants with MLL-AF4+ acute leukemia har- 2011;152(2):141–154. silent mutational landscape of infant MLL- bor and express the MLL-AF4 fusion gene. J 6. Ribeiro RC, Pui CH. Prognostic factors in AF4 pro-B acute lymphoblastic leukemia. Exp Med. 2009;206(13):3131–3141. childhood acute lymphoblastic leukemia. Genes Chromosom Cancer. 2013;52(10): 32. Muñoz-López A, Romero-Moya D, Prieto Hematol Pathol. 1993;7(3):121–142. 954–960. C, et al. Development refractoriness of MLL- 7. Pieters R, Schrappe M, De Lorenzo P, et al. A 20. Andersson AK, Ma J, Wang J, et al. The land- rearranged human B cell acute leukemias to treatment protocol for infants younger than scape of somatic mutations in infant MLL- reprogramming into pluripotency. Stem Cell 1 year with acute lymphoblastic leukaemia rearranged acute lymphoblastic leukemias. Reports. 2016;7 (4):602–618. (Interfant-99): an observational study and a Nat Genet. 2015;47(4):330–337. 33. Roy A, Cowan G, Mead AJ, et al. multicentre randomised trial. Lancet. 21. Lin S, Luo RT, Ptasinska A, et al. Instructive Perturbation of fetal liver hematopoietic 2007;370(9583):240–250. role of MLL-fusion proteins revealed by a stem and progenitor cell development by tri- 8. Biondi A, Cimino G, Pieters R, Pui C-H. model of t(4;11) pro-B acute lymphoblastic somy 21. Proc Natl Acad Sci U S A. Biological and therapeutic aspects of infant leukemia. Cancer Cell. 2016;30(5):737–749. 2012;109(43):17579–17584. leukemia. Blood. 2000;96(1):24–33. 22. Lin S, Luo RT, Shrestha M, Thirman MJ, 34. Gröbner SN, Worst BC, Weischenfeldt J, et 9. Milne TA. Mouse models of MLL leukemia: Mulloy JC. The full transforming capacity of al. The landscape of genomic alterations recapitulating the human disease. Blood. MLL-Af4 is interlinked with lymphoid line- across childhood cancers. Nature. 2018;555 2017;129(16):2217–2223. age commitment. Blood. 2017;130(7):903– (7696):321-327. 10. Nakamura T, Mori T, Tada S, et al. ALL-1 is 907. 35. Alexandrov LB, Nik-Zainal S, Wedge DC, et a histone methyltransferase that assembles a 23. Yamamoto H. Successful sustained engraft- al. Signatures of mutational processes in supercomplex of proteins involved in tran- ment after reduced-intensity umbilical cord human cancer. Nature. 2013;500(7463):415– scriptional regulation. Mol Cell. 2002;10(5): blood transplantation for adult patients with 421. 1119–1128. severe aplastic anemia. Blood. 2011;116(26): 36. Trentin L, Bresolin S, Giarin E, et al. 11. Chen CW, Armstrong SA. Targeting DOT1L 6123–6132. Deciphering KRAS and NRAS mutated and HOX gene expression in MLL- 24. Montes R, Ayllón V, Prieto C, et al. Ligand- clone dynamics in MLL-AF4 paediatric rearranged leukemia and beyond. Exp independent FLT3 activation does not coop- leukaemia by ultra deep sequencing analy- Hematol. 2015;43(8):673–684. erate with MLL-AF4 to immortalize/trans- sis. Sci Rep. 2016;6:34449. 12. McLean CM, Karemaker ID, van Leeuwen F. form cord blood CD34+ cells. Leukemia. 37. Bashford-Rogers RJM, Nicolaou KA, The emerging roles of DOT1L in leukemia 2014;28(3):666–674. Bartram J, et al. Eye on the B-ALL: B-cell and normal development. Leukemia. 25. Bueno C, Montes R, Melen GJ, et al. A receptor repertoires reveal persistence of 2014;28(11):2131–2138. human ESC model for MLL-AF4 leukemic numerous B-lymphoblastic leukemia sub- 13. Krivtsov AV, Feng Z, Lemieux ME, et al. fusion gene reveals an impaired early clones from diagnosis to relapse. Leukemia. H3K79 methylation profiles define murine hematopoietic-endothelial specification. 2016;30(12):2312–2321. and human MLL-AF4 leukemias. Cancer. Cell Res. 2012;22(6):986–1002. 38. Prelle C, Bursen A, Dingermann T, 2009;14(5):355–368. 26. Bueno C, Ayllón V, Montes R, et al. FLT3 Marschalek R. Secondary mutations in 14. Greaves MF, Maia AT, Wiemels JL, Ford AM. activation cooperates with MLL-AF4 fusion t(4;11) leukemia patients. Leukemia. Leukemia in twins: lessons in natural histo- protein to abrogate the hematopoietic spec- 2013;27(6):1425–1427. ry. Blood. 2003;102(7):2321–2333. ification of human ESCs. Blood. 39. Stam RW, Schneider P, Hagelstein JAP, et al. 15. Ford AM, Ridge SA, Cabrera ME, et al. In 2013;121(19):3867–3878. Gene expression profiling-based dissection utero rearrangements in the trithorax-related 27. Prieto C, Stam RWRW, Agraz-Doblas A, et of MLL translocated and MLL germline haematologica | 2019; 104(6) 1187 A. Agraz-Doblas et al. acute lymphoblastic leukemia in infants. and equipotent founder cells establish and pendent gene signatures in pediatric t(4;11) Blood. 2010;115(14):2835–2844. maintain acute lymphoblastic leukaemia. acute lymphoblastic leukemia patients. Eur J 40. Chillón MC, Gómez-Casares MT, López- Leukemia. 2017;31(12):2577–2586. Haematol. 2009;83(5):406–419. Jorge CE, et al. Prognostic significance of 47. Prieto C, López-Millán B, Roca-Ho H, et al. 54. Kühn A, Löscher D, Marschalek R. The FLT3 mutational status and expression levels NG2 antigen is involved in leukemia inva- IRX1/HOXA connection: insights into a in MLL-AF4 and MLL-germline acute lym- siveness and central nervous system infiltra- novel t(4;11)- specific cancer mechanism. phoblastic leukemia. Leukemia. 2012;26(11): tion in MLL-rearranged infant B-ALL. Oncotarget. 2016;7(23):35341–35352. 2360–2366. Leukemia. 2018;32(3):633–644. 55. Wilkinson AC, Ballabio E, Geng H, et al. 41. Boelens JJ, Aldenhoven M, Purtill D, et al. 48. Jansen MWJC, Corral L, van der Velden VHJ, RUNX1 is a key target in t(4;11) leukemias Outcomes of transplantation using various et al. Immunobiological diversity in infant that contributes to gene activation through hematopoietic cell sources in children with acute lymphoblastic leukemia is related to an AF4-MLL complex interaction. Cell Rep. Hurler syndrome after myeloablative condi- the occurrence and type of MLL gene 2013;3(1):116–127. tioning Key Points. Blood. 2013;121(10): rearrangement. Leukemia. 2007;21(4):633– 56. Benedikt A, Baltruschat S, Scholz B, et al. 3981–3987. 641. The leukemogenic AF4-MLL fusion protein 42. Bashford-Rogers RJM, Palser AL, Huntly BJ, 49. Hotfilder M, Röttgers S, Rosemann A, et al. causes P-TEFb kinase activation and altered et al. Network properties derived from deep Leukemic stem cells in childhood high-risk epigenetic signatures. Leukemia. 2011;25(1): sequencing of human B-cell receptor reper- ALL/t(9;22) and t(4;11) are present in primi- 135–144. toires delineate B-cell populations. Genome tive lymphoid-restricted CD34+CD19-cells. 57. Scholz B, Kowarz E, Rössler T, Ahmad K, Res. 2013;23(11):1874–1884. Cancer Res. 2005;65(4):1442–1449. Steinhilber D, Marschalek R. AF4 and AF4N 43. Bruggemann M, Schrauder A, Raff T, et al. 50. Bergmann AK, Castellano G, Alten J, et al. protein complexes: recruitment of P-TEFb Standardized MRD quantification in DNA methylation profiling of pediatric B- kinase, their interactome and potential func- European all trials: Proceedings of the cell lymphoblastic leukemia with KMT2A tions. Am J Blood Res. 2015;5(1):10–24. Second International Symposium on MRD rearrangement identifies hypomethylation 58. Mück F, Bracharz S, Marschalek R. DDX6 assessment in Kiel, Germany, 18-20 at enhancer sites. Pediatr Blood Cancer. transfers P-TEFb kinase to the AF4/AF4N September 2008. Leukemia. 2010;24(3):521– 2017;64(3):e26251. (AFF1) super elongation complex. Am J 535. 51. Malouf C, Ottersbach K. The fetal liver lym- Blood Res. 2016;6(3):28–45. 44. Driessen EMC, van Roon EHJ, Spijkers- phoid-primed multipotent progenitor pro- 59. van Zelm MC, van der Burg M, de Ridder D, Hagelstein JAP, et al. Frequencies and prog- vides the prerequisites for the initiation of et al. Ig gene rearrangement steps are initiat- nostic impact of RAS mutations in MLL- t(4;11) MLL-AF4 infant leukemia. ed in early human precursor B cell subsets rearranged acute lymphoblastic leukemia Haematologica. 2018 Jun 14. [Epub ahead of and correlate with specific transcription fac- in infants. Haematologica. 2013;98(6):937– print] tor expression. J Immunol. 2005;175(9): 944. 52. Kowarz E, Burmeister T, Lo Nigro L, et al. 5912–5922. 45. Tamai H, Miyake K, Takatori M, et al. Complex MLL rearrangements in t(4;11) 60. Böiers C, Richardson SE, Laycock E, et al. A Activated K-Ras protein accelerates human leukemia patients with absent AF4MLL human IPS model implicates embryonic B- MLL/AF4-induced leukemo-lympho- fusion allele. Leukemia. 2007;21(6):1232– myeloid fate restriction as developmental mogenicity in a transgenic mouse model. 1238. susceptibility to B acute lymphoblastic Leukemia. 2011;25(5):888–891. 53. Trentin L, Giordan M, Dingermann T, Basso leukemia-associated ETV6-RUNX1. Dev 46. Elder A, Bomken S, Wilson I, et al. Abundant G, Te Kronnie G, Marschalek R. Two inde- Cell. 2018;44(3):362–377.e7. 1188 haematologica | 2019; 104(6)