Mouse Models of NPM1-Mutated Acute Myeloid Leukemia: Biological and Clinical Implications

CONCISE REVIEW Mouse models of NPM1-mutated acute myeloid leukemia: biological and clinical implications

P Sportoletti1, E Varasano1, R Rossi1, A Mupo2, E Tiacci1, G Vassiliou2, MP Martelli1 and B Falini1

Acute myeloid leukemia (AML) carrying nucleophosmin (NPM1) mutations displays distinct biological and clinical features that led to its inclusion as a provisional disease entity in the 2008 World Health Organization (WHO) classification of myeloid neoplasms. Studies of the molecular mechanisms underlying the pathogenesis of NPM1-mutated AML have benefited greatly from several mouse models of this leukemia developed over the past few years. Immunocompromised mice xenografted with NPM1-mutated AML served as the first valuable tool for defining the biology of the disease in vivo. Subsequently, genetically engineered mouse models of the NPM1 mutation, including transgenic and knock-in alleles, allowed the generation of mice with a constant genotype and a reproducible phenotype. These models have been critical for investigating the nature of the molecular effects of these mutations, defining the function of leukemic stem cells in NPM1-mutated AML, identifying chemoresistant preleukemic hemopoietic stem cells and unraveling the key molecular events that cooperate with NPM1 mutations to induce AML in vivo. Moreover, they can serve as a platform for the discovery and validation of new antileukemic drugs in vivo. Advances derived from the analysis of these mouse models promise to greatly accelerate the development of new molecularly targeted therapies for patients with NPM1-mutated AML.

Leukemia (2015) 29, 269–278; doi:10.1038/leu.2014.257

INTRODUCTION detected in myelodysplasia and T-cell lymphoblastic leukemia (for 9,10 Recently, the sequencing of 200 acute myeloid leukemia (AML) example, DNMT3A) or solid tumors (for example, IDH1/2 and 11 exomes or genomes1 revealed that they carry hundreds of gene NRAS). Fourth, NPM1 mutations are mutually exclusive with other mutations. The majority of these are likely to represent ‘passenger’ recurrent genetic abnormalities that define distinct AML entities in mutations and only about 20 are regarded as ‘drivers’ on the basis the World Health Organization (WHO) classification of lympho- 12 that they occur in at least 2% of AML patients.1 Interestingly, the hemopoietic tumors. Finally, NPM1-mutated AMLs show 13,14 AML genome appears less complex than that of other adult cancer distinctive mRNA and miRNA expression profiles, regardless types. Specifically, the median mutation frequency in AML is 0.28 of the presence or absence of karyotypic abnormalities.15 per megabase (Mb), whereas other tumors show an average of Deregulated genes in these signatures include several homeobox over 1 mutation per Mb.2 (HOX), CD34, miR-10a, miR-10b and let-7 family members. Overall, The most commonly mutated genes in AML include nucleo- these molecular effects of NPM1 mutations deeply influence phosmin (NPM1), FLT3 (each in about 30% of cases) and DNMT3A hematopoietic development and maintenance of stem/progenitor (in ~ 20% of cases).1 Mutations of other genes (for example, cell properties. IDH1/2, NRAS) occur at a frequency ⩽ 10%.1 Mutated genes in AML NPM1 mutations also associate with distinctive clinicopatholo- are organized into functional categories including transcription- gical features,4,5 including female sex, higher white blood cell factor fusions, tumor suppressors, DNA-methylation-related genes, count with increased blast percentage (especially in association activated signaling genes, chromatin modifiers, myeloid transcrip- with FLT3-ITD), frequent M4/M5 morphology, absent or low CD34 tion factors, cohesion complex genes, spliceosome complex genes expression, strong positivity for CD33, good response to induction and NPM1, which occupies its own category.1 therapy and favorable prognosis (mostly in the absence of FLT3- We discovered NPM1 mutations in AML in 20053 and ITD).16 For these reasons, NPM1-mutated AML has been included subsequently found that they associate with unique biological as a new provisional entity in the 2008 WHO classification.17 and clinical features.4 Several lines of evidence point to NPM1 In spite of these advances, the pathogenic role of NPM1 mutations as a driving event defining a distinct AML entity.5 First, mutations in AML remains incompletely understood. In vitro they are highly recurrent in AML (about one-third of cases) and analysis of cells transfected with mutant NPM1, as well as studies represent the most common genetic alteration underlying AML of NPM1-mutated human AML cell lines18 and primary blasts from with normal cytogenetics, accounting for 50–60% of patients.3,4 patients, contributed to the clarification of mechanisms respon- Second, NPM1 mutations remain stable over the course of sible for the aberrant accumulation of nucleophosmin in the disease4,6 and are usually detected at relapse, even many years cytoplasm of leukemic cells19,20 and AML development.21 How- after the initial diagnosis of AML.7 Third, NPM1 mutations and ever, to gain insight into the early transformation events leading aberrant cytoplasmic expression of nucleophosmin are specific for to the generation of leukemic stem cells, there is a need for AML.2,3,8 In contrast, other common AML-associated mutations are prospective in vivo models.

1Institute of Hematology, University of Perugia, Perugia, Italy and 2The Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK. Correspondence: Dr P Sportoletti or Professor B Falini, Institute of Hematology, University of Perugia, Ospedale S. Maria della Misericordia, S. Andrea delle Fratte, 06132 Perugia, Italy. E-mail: [email protected] or [email protected] Received 24 July 2014; revised 25 August 2014; accepted 26 August 2014; accepted article preview online 2 September 2014; advance online publication, 23 September 2014 Mouse models of NPM1-mutated AML P Sportoletti et al 270 In this review, we provide a synopsis of three types of mouse Altogether, these findings strongly suggest that the DNMT3A models (xenotransplant, transgenic and knock-in) that are mutations preceded NPM1 mutations during leukemogenesis. currently available to understand NPM1-driven leukemogenesis. Notably, the DNMT3A-mutated preleukemic populations in NPM1-mutated AML appear to be functionally competent. In fact, in xenograft repopulation assays, DNMT3A-mutated preleukemic XENOTRANSPLANTATION MODELS OF NPM1-MUTATED AML cells gave rise mostly to multilineage (lymphoid and myeloid) Xenografting of immunocompromised mice with human primary engraftments devoid of NPM1 mutations.26 Moreover, the leukemic cells represents a valuable tool for studying the biology of xenograft models highlighted a competitive growth advantage AML in vivo.22 Severe combined immunodeficiency (SCID) mice that of the preleukemic DNMT3A-mutated HSCs over non-mutated were subcutaneously injected with NPM1-mutated AML blasts from normal HSCs.26 These findings support a model of leukemogenesis apatient6 served as the first in vivo model for the disease. The wherein an ancestral DNMT3A-mutated HSC generates an engraftment represented the expansion of true leukemia-initiating expanded pool of HSCs and downstream multilineage progeni- cells, as the engraftment potential and genetic characteristics were tors, but not leukemia. Overt AML develops when NPM1 mutations retained in secondary and tertiary recipients through many passages are subsequently acquired, probably within a DNMT3A mutant for 48years.6 Moreover, the characteristic aberrant expression of HSC or possibly a granulocyte monocyte or multilymphoid nucleophosmin in the cytoplasm of leukemic cells3 was maintained progenitor, which is transformed by the mutation.26 This over time. These findings fully mimic those observed in patients observation is of potential clinical relevance because the experiencing late relapses of NPM1-mutated AML.7 persistence of chemoresistant preleukemic cells at remission Xenotransplantation studies have shown that AML is a stem-cell represents a reservoir from which a relapse may potentially arise disease, with individual leukemia cases containing a variable through acquisition of de novo mutations.26 population of CD34-positive leukemia-initiating cells (LICs).22 NPM1-mutated AML is frequently CD34-negative3 and in only GENETICALLY ENGINEERED MOUSE MODELS OF MUTANT NPM1 about 10% of cases it expresses CD34 at low intensity.23 Two studies in NPM1-mutated AML demonstrated the presence of After the description of xenotransplant models, efforts focused on NPM1 mutations in the rare fraction of CD34-positive cells, the development of transgenic and knock-in models of NPM1- providing evidence that they belong to the leukemic clone.23,24 mutated AML using several strategies (Figures 2 and 3). This NPM1-mutated gene and/or protein was also confirmed in a approach allows the generation of large cohorts of inbred mice subpopulation of CD34-positive cells with the immunophenotype with a constant genotype and a reproducible phenotype, thus of leukemic stem cells (that is, CD38-/CD123+/CD33+/CD90 − ).23 facilitating detailed functional studies. It is noteworthy that NPM1- Interestingly, NPM1-mutated/CD34-positive blasts could be mutated alleles were sometimes referred to as NPMc+ or NPM1c to successfully transplanted into immunocompromised NOD/SCID indicate the aberrant cytoplasmic localization of the mutant fi 3 IL2 receptor gamma chain knockout (NSG) mice to generate an protein, as rst reported by Falini et al. (resulting from the AML that recapitulated the features of the original patient's disruption of a C-terminal nucleolar localization signal and disease, including the loss of CD34 expression in the leukemic generation of a de novo nuclear export signal). Because NPM1 fi 2,8 population bulk (Figure 1). In contrast, engraftment in NSG mice mutations are AML speci c, some mouse models have been fi transplanted with NPM1-mutated/CD34-negative leukemic cells engineered to express NPMc+ speci cally in either the myeloid fi was less consistent and LIC activity was not always present, progenitor or HSC compartments using speci c Cre transgenic suggesting that this fraction was relatively depleted in LICs. lines (Figure 3). Examples of such models are discussed below. The existence of an ancestral population of preleukemic hemopoietic stem cells (HSCs) capable of differentiation has been Transgenic model of NPM1-mutated gene (NPMc+) postulated in AML.25 The analysis of nonleukemic populations of Cheng et al.28 developed a transgenic mouse expressing the HSCs and progenitors, as well as of mature B and T cells sorted from NPM1-mutated gene under the control of the human MRP8 NPM1-mutated/DNMT3A-mutated AML patients at diagnosis and promoter (Figures 2 and 3a). This promoter drives the expression remission, revealed that, unlike AML blasts, they usually contain the of NPMc+ in common myeloid progenitors, as well as in mature DNMT3A but not the NPM1 mutations.26 Similarly, Corces- granulocytes and monocytes.28 The transgenic NPMc+ mice Zimmerman et al.27 using a genomic and functional analysis of de developed a late-onset myeloproliferative disease with splenome- novo AML and patient-matched HSCs demonstrated that mutations galy and increased number of Gr-1/Mac-1+ve mature myeloid in NPM1 or in genes involved in activated signaling were significantly cells in the bone marrow and spleen.28 None of these transgenic absent in preleukemic cells, whereas mutations in genes involved in mice developed acute leukemia. processes such as DNA methylation, histone modification and A similar expansion of myeloid cells was obtained in zebrafish chromatin looping were significantly enriched in preleukemic cells. embryos overexpressing mutant NPMc+ ubiquitously.29 In this

Figure 1. A xenograft model of human NPM1-mutated AML: CD34+ cells from NPM1-mutated AML generate CD34-negative NPMc+ AML in immunocompromised NOD/SCID IL2 receptor gamma chain knockout (NSG) mice. (a) Human NPM1-mutated AML xenotransplant model. (b) Flow cytometric analysis showing CD34 antigen expression in the CD34+ cell fraction (carrying NPM1 mutation) isolated from a patient with NPM1-mutated AML and injected in NSG mice (left). Flow cytometric analysis of mouse bone marrow (8 weeks after the injection) shows engraftment of NPM1-mutated human myeloid (CD33+/CD117+) cells that are mainly CD34 − (3% CD34+ cells). (c) Femoral head paraffin section showing extensive bone marrow infiltration by human leukemic cells with aberrant cytoplasmic expression of nucleophosmin (NPM1), characteristic of NPM1-mutated AML. In comparison, a nonleukemic cell showing nuclear NPM1 staining is indicated (arrow). Immunostaining was performed with an anti-NPM1 mouse monoclonal antibody (clone 376) (Alkaline Phosphatase/Anti-alkaline Phosphatase (APAAP) method; hematoxylin counterstain). (d) Mouse bone marrow cells express mutant NPM on western blot analysis with a specific anti-mutant NPM1 rabbit polyclonal antibody (anti-NPMmut), as is also seen with the unprocessed original patient sample (total) and the CD34+ cell fraction (CD34+). Lysate from the OCI/AML3 human cell line harboring NPM1 mutation is used as positive control. (e and f) Tibial paraffin sections showing preferential leukemic infiltration of bone marrow endosteal regions at low (e) and high (f) magnification. Immunostaining was performed with a specific anti-human CD45 monoclonal antibody (APAAP; hematoxylin counterstaining). Images were collected using an Olympus B61 microscope (Olympus Italia, Segrate, Italy) a UPlanApo 40 × /0.85 (c and f) and UPlanApo 10 × /0.40 (e); Camedia 4040, Dp_soft Version 3.2 (Olympus Italia); and Adobe Photoshop 7.0 (Adobe Systems, Mountain View, CA, USA).

Transgenic

Myeloproliferation NPMc+ hMRP8 Some mice Cheng et al. 2010

“Non conventional” conditional Knock-in Block of Mk differentiation Sportoletti et al. 2013

NPMc+ Mx1-Cre AML late-onset Some mice Mallardo et al. 2013

Conditional Knock-in MPD Some mice Chou et al. 2012

NPMc+ Mx1-Cre AML late-onset Some mice Vassiliou et al. 2011

Compound Mutants

AML rapid-onset All mice NPMc+ Flt3-ITD Mupo et al. 2013 Mallardo et al. 2013 Rau et al. 2013

Figure 2. Summary of NPM1 mutant mouse models and their phenotypic characteristics. The ﬁrst genetically engineered mouse model of NPM1 mutation was a transgenic mouse expressing the mutated gene under the control of the MRP8 hematopoietic promoter. These mice developed myeloproliferation with low frequency and late onset. Two different ‘nonconventional’ conditional knock-in mouse models targeted the human NPM1 mutation A into a highly recombinogenic mouse genomic locus in embryonic stem cells. Conditional expression of the mutant into the hematopoietic compartment was obtained using Mx1-Cre mice. Mice developed hematopoietic defects including megakaryocytic expansion, myeloid proliferation and a late-onset leukemia. A conventional knock-in strategy has been used in two different models in order to induce the expression of either a human or a mouse NPMc+ in BM cells. The human NPM1 mutant led to late leukemia development, whereas the mouse mutant induced myeloproliferation with low penetrance. Three different groups have generated compound mutant mice carrying the NPM1 mutation and Flt3-ITD expression in the hematopoietic compartment. All these models developed a fully penetrant leukemic phenotype rapidly, although with different latencies. AML, Acute Myeloid Leukemia; LICs, Leukemia-Initiating Cells; Mk, Megakaryocytes; MPD, Myeloproliferative Disease; NPMc+, NPM1 mutation; NSG, NOD/SCID IL2 receptor gamma chain knockout.

setting, embryos injected with the human NPM1-mutated mRNA characterized by the upregulation of HOX genes,13 pointing to demonstrated abnormal cytoplasmic localization of the mutated the possible need to speciﬁcally target the HSC compartment. protein, similar to human NPM1-mutated AML cells. The expres- Second, NPM1 mutations in AML patients are consistently sion of the NPM1 mutant perturbed primitive myelopoiesis heterozygous4 and result in a mutant to wild-type NPM1 expression inducing a marked proliferation of early pu.1-expressing myeloid ratio that is instrumental to induce cytoplasmic delocalization of cells. In addition, embryos showed increased numbers of both NPM1 forms.20 In contrast, the degree of cytoplasmic hematopoietic stem cells (c-myb+/cd41+), but follow-up for AML mislocalization of the NPM1 forms and any effects of NPM1 development was not possible owing to the transient expression heterozygosity was lower in the NPMc+ transgenic model. This is of the NPM1 mutant.29 likely owing to the fact that the expression levels of the NPM1 These initial in vivo observations indicated that the NPM1 mutant mutant were too low compared with endogenous expression, as was sufﬁcient to induce a myeloproliferative disorder, but was both Npm1 wild-type alleles were preserved. This view is supported unable to generate AML recapitulating the phenotypic features by in vitro transfection experiments showing that an excess of the observed in humans. The absence of a leukemic phenotype may wild-type NPM1 protein relocated the NPM1 mutant from the have been due to a number of reasons. First, NPM1-mutated AML cytoplasm to the nucleoli dampening the mutant's gain-of-function has a characteristic stem cell-like gene expression signature activity and impairing its transforming abilities.30 Besides NPM1

Random * Activated In progenitor transgene hMRP8 promoter FLAG hNPMc+ insertion and mature myeloid cells

Targeted * Rosa26/Hprt pA hNPMc+ STOP Neo CAG Pr. locus

Mx1Cre + pIpC

Cre-mediated * excised Induced in adults pA hNPMc+ CAG Pr. Rosa26/Hprt hematopoietic cells locus

* Humanized mutant exon11

2 6 9 10 11 11* Targeted Npm1 puΔTK locus

Mx1Cre + pIpC

2 6 9 10 11* Cre-mediated Induced in adults excised Npm1 locus hematopoietic cells

* Mouse mutant exon 11

Targeted 2 6 9 10 11* Npm1 locus Neo (before injection into foster mothers)

loxP-neo-loxP cassette In vitro removal

2 6 9 10 11* Targeted Activated Npm1 locus constitutively Figure 3. Schematic representation of the targeting strategies for generating NPM1-mutated mouse models. (a) Transgenic mouse model. The transgene construct was generated by cloning the NPM1 mutation A cDNA (hNPMc+) under the control of the human MRP8 promoter (hMRP8 promoter). The construct was flag-tagged at the N terminus (FLAG) of the NPM1 mutation. The NPMc+ protein was expressed in common myeloid progenitors and mature granulocytes and monocytes. (b)Sportolettiet al.33 and Mallardo et al.34 models: the human cDNA of the Type A mutation of NPM1 (hNPMc+) was inserted into the Rosa26 and Hprt genetic loci, respectively. A stop cassette (STOP) flanked by loxP sites (triangles) was inserted between a strong CAG promoter (CAG pr.) and the hNPMc+ cassette, thus allowing the conditional expression of the NPM mutant protein upon Cre-mediated recombination. NPM1 mutant mice were crossed with Mx1-Cre mice, and excision of the STOP-Neo cassette was induced by polyinosinic-polycytidylic acid (pIpC) treatment in vivo.(c) Vassiliou et al.38 model: a constitutive knock-in allele was engineered to have NPM1 TCTG duplicationinexon11(asterisk)ofthemouseNpm1 gene. The humanized mutant exon 11 was directly used to replace the wild-type mouse sequence (mouse exon 11 is homologous to human exon 12). Two loxP sites (triangle) were flanking the wild-type mouse exon 11 and a Puro- delta-TK cassette (puΔTK). The NPM1 mutant was expressed under the control of the endogenous mouse Npm1 regulatory elements. The conditional allele allowed the expression of a mouse NPM1 protein that was converted into a humanized mutant NPM1 upon Cre recombination. NPM1 conditional knock-in mice were crossed with Mx1-Cre mice, and the expression of the mutant in the hematopoietic compartment was induced by pIpC treatment in vivo.(d)Chouet al.39 model: the conditional knock-in allele was generated by introducing a mutated NPM1 exon 11 with a floxed Neo cassette (Neo) in intron 10 of the mouse NPM1 gene. This mouse-mutated exon 11 was obtained by the insertion of a TCTG after nucleotide c.857 of murine Npm1 coding sequence. This pattern mimics human NPM1 mutation without any ‘humanized’ sequence. The loxP-neo- loxP cassette was excised in mouse embryonic stem cells before their implantation into foster mothers. The mouse Npm1 mutant was constitutively expressed under the control of the endogenous mouse Npm1 regulatory elements. cytoplasmic dislocation effects, reduced levels of wild-type NPM1 Knock-in models expressing NPMc cDNA from permissive loci owing to the loss of one NPM1 functional allele may also contribute A number of permissive genomic loci are routinely used for to leukemogenesis. Indeed, NPM1 haploinsufficiency led to knocking-in gene cDNAs by homologous recombination, so that myeloid-specificdefectsex vivo31 and Npm1 heterozygous mice mRNAs are then expressed under the control of the endogenous developed both myeloid and lymphoid leukemias.32 regulatory elements of these loci. Mice harboring the NPM1 Third, the association between NPM1 and other frequent mutation A (representing 80% of all NPM1 mutations)4 mutations, such as FLT3-ITD, DNMT3A and IDH11,5 in human AML have been generated using this approach to circumvent some strongly suggests that additional genetic lesions are required to limitations of conventional transgenics, such as low expression induce a leukemic phenotype in vivo. and susceptibility to epigenetic silencing after random genomic

© 2015 Macmillan Publishers Limited Leukemia (2015) 269 – 278 Mouse models of NPM1-mutated AML P Sportoletti et al 274 integration into susceptible sites. Two mouse models have been perturbation in the niche function with a decline of cobblestone developed through insertion of the human NPM1 mutation area formation and a defective CXCR4/CXCL12 pathway that was A cDNA into the Rosa26 and Hprt loci, respectively33,34 similar to NPM1-mutated AML patients.39 (Figures 2 and 3b). In addition, the targeting vector was designed to contain a strong and ubiquitous pCAG promoter along with a stop cassette flanked by loxP sites between the promoter and the COMPOUND MUTANT MICE RELEVANT TO NPM1-MUTATED cDNA (Figure 3b), thus allowing the conditional expression of AML the NPM1 mutant protein upon Cre‐mediated recombination. Although data gained from knock-in mouse models have added These mouse models expressed the mutant NPM1 gene in new insights into the pathogenesis of NPM1-mutated AML, they addition to the two normal copies of the endogenous mouse have also demonstrated that sustained NPM1 mutant expression Npm1 gene. Nevertheless, the mutant NPM1 expression levels in the HSC compartment is not sufficient for leukemia onset. were comparable to the levels of NPM1 wild-type protein encoded Additional stochastic mutations that some of the animals acquire by the endogenous loci at least in homozygous transgenic mice.33 during their life span are likely needed to drive a full leukemia Upon Cre induction, the NPM1 mutant expression significantly phenotype. This view is also supported by in vitro co-transfection perturbed adult hematopoiesis in both conditional models. As studies showing that cooperative genetic events are required for described by Sportoletti et al.,33 NPMc+ induced myeloproliferation NPM1 mutation leukemogenesis.21 Moreover, it is consistent with in a fraction of mice and a fully penetrant block of megakaryocytic the experimental observation that, as with NPM1, other AML development. Although none of these mice developed leukemia oncogenes (for example, RUNX1-RUNX1T1 or CBF-MYH11) are not after a long follow-up, the model mirrored some features of the sufficient to cause leukemia alone in mice, but require cooperating 40,41 human NPM1-mutated AML, such as megakaryocytic expansion35,36 events to do so. Indeed, some animal studies support the 42 (Figure 4) and deregulation of specificmiRNAsinbonemarrow.14 classic two-hit model of leukemogenesis in which the coopera- In particular, expanded CD41+ megakaryocytes of NPMc+ mice tion between two classes of genetic alterations is necessary for overexpressed miR-10a, miR-10b and miR-20a that are also leukemic transformation: for example, a class I mutation, which deregulated in NPM1-mutated AML patients.14 Interestingly, these activates signal transduction pathways (for example, FLT3-ITD) and miRNAs are known to control megakaryocytic lineage development confers a proliferative advantage, and a class II mutation, which 40 and platelet function, as their downregulation allows expression of affects transcription (for example, RUNX1/RUNX1T1) and causes a target genes involved in megakaryocytic differentiation.37 differentiation block. Findings from whole-genome sequencing In the model developed by Mallardo et al.,34 the mutant NPM1 studies point to an even more complex configuration in many induced AML after a very long latency and in a minority of mice. In AMLs that might involve several mutations, serving complemen- particular, 30% of animals developed leukemia and about 50% of tary roles including those affecting chromatin landscaping and epigenetic pathways. On the basis of such observations, an leukemias expressed both myeloid and B-lymphoid markers. 43 In contrast, human acute lymphoblastic leukemias never carry alternative ‘slot machine’ model has been recently proposed, in NPM1 mutations.3 which the late steps would be, to some extent, constrained by the initial ones (clonal dominance, cooperations/exclusions). A number of clinical and experimental lines of evidence point Conventional knock-in models of NPM1 mutation to a particular complementarity in AML development between This strategy has been used to introduce the mutation into the NPM1 and FLT3-ITD mutations, which are present in 30–40% endogenous NPM1 locus, thus keeping its expression under the of NPM1-mutated cases.44 Moreover, in conditional Npm1c+/ control of the endogenous promoter and genocopying what Sleeping Beauty transposon compound mutant mice, recurrent happens in patients' leukemic cells. This replacement reduces the activating integrations were identified in Flt3 and the related gene normal Npm1 gene to heterozygosity such that the relative Csf2 (encoding GM-CSF).38 expression levels of the normal and mutated genes are, in principle, To better understand whether FLT3-ITD and NPM1 mutations similar and comparable with the situation in human AML (NPM1 can significantly cooperate to induce leukemia in vivo, three mutations are consistently heterozygous in AML patients). Two different groups have generated compound mutant mice carrying NPM1-mutated knock-in mice have been developed including a both these mutations in the hematopoietic compartment34,45,46 conditional model that allows the induction of NPMc+ expression in (Figure 2). Notably, the combination of Flt3/ITD and NPMc+ always HSCs at controlled time points after birth and a constitutive model resulted in the development of AML with features that closely expressing the NPM1 mutant in all tissues (Figure 2; Figures 3c and d). recapitulated those observed in patients, including monocytic 38 Vassiliou et al. (Figure 3c), directly used the human NPM1 type differentiation and very high white blood cell counts. Moreover, A mutation nucleotide sequence to replace the wild-type mouse spontaneous loss of heterozygosity (LOH) of the wild-type Flt3 sequence. The goal was to have the same consequences at the allele occurred with a high frequency, and the extent of LOH protein level as seen in human AML. Activation of the humanized closely correlated with the level of leukocytosis. Notably, FLT3-LOH Npm1 knock-in allele in mouse HSCs caused Hox gene over- is a well-described feature of NPM1-mutated/FLT3-ITD+ AML expression (a characteristic feature of NPM1-mutated AML13,14), patients and is associated with a poor outcome.47 enhanced self-renewal and expanded myelopoiesis. One-third of The rapidity of leukemia onset depended on the NPM1 the mice developed late-onset leukemia, demonstrating that the mutant mouse line that had been crossed with Flt3-ITD mice NPM1 mutation is an AML-driving lesion. Furthermore, to accelerate and reflected different NPM1 mutant expression levels. In leukemogenesis, these mice were subjected to hemopoietic- transgenic NPMc+/Flt3-ITD mice described by Rau et al.,46 specific insertional mutagenesis with the Sleeping Beauty transpo- leukemia occurred with a rather long latency probably because son, and in this context 480% of mice developed AML. of the low expression of NPMc+ and the high levels of wild-type More recently, Chou et al.39 developed a constitutive knock-in NPM1 (two endogenous alleles). By comparison, in the model mouse model by introducing the same TCTG duplication seen in developed by Mallardo et al.,34 NPM1 mutant expression was human type A mutations, but without ‘humanizing’ the surround- significantly increased and the NPMc+/Flt3-ITD mice died of ing sequence leading to a mutant protein sequence predicted to leukemia between 35 and 161 days (median 72 days). Mupo have a weaker nuclear export signal and not seen in human AML et al.45 described an even shorter latency to disease onset using (Figure 3d). As described in other NPM1-mutated models, a the humanized NPM1-mutated model, in which all double-mutant percentage of mice developed a myeloproliferative disease with mice developed leukemia in o2 months without the need for extramedullary hematopoiesis. In addition, mice showed a formal Cre induction (Figure 5a), but driven instead by low-level

Figure 4. NPM1 conditional mutant mice show a megakaryocytic compartment expansion mimicking that observed in human NPM1-mutated AML patients. (a) The NPM1 mutation induces an expansion of the immature megakaryocytes in mice. Representative colonies ( ×10 magniﬁcation) from (ii) NPM1-mutated compared with (i) wild-type control mice showing an increase in the CFU-MK potential of bone marrow (BM) cells from NPM1-mutated mice. Flow cytometric analysis of BM cells from representative conditional NPM1-mutated (lower dot plot) and wild-type control (upper dot plot) mice demonstrates an increase in the percentage of Lin–Kit1Sca-1–CD150+CD41+ megakaryocytic progenitor populations. (b) Megakaryocyte expansion in BM trephines of NPM1-mutated AML patients. Representative BM sections from a nonleukemic control were stained with (i) hematoxylin and eosin and (ii) a mouse monoclonal antibody against the linker for activation of T-cell (LAT) protein, which is an excellent marker for megakaryocytes in BM biopsies. (iii-iv) BM trephine sections from an NPM1-mutated AML patient with an increased number of megakaryocytes (40 × magniﬁcation), as assessed by (iii) hematoxylin and eosin and (iv) immunostaining for human LAT.

‘leaky’ expression of Cre from the Mx1-Cre allele and by sub- An in vivo insertional mutagenesis study conducted by sequent acquisition of LOH for Flt3-ITD (Figure 5b). This is not Vassiliou et al.38 showed that NPM1 mutations could also surprising because in this setting both tumor suppressor effects of cooperate with activated Ras signaling to cause AML in NPM1 NPM1 haploinsufﬁcency and oncogenic properties of the mutated knock-in mice. Indeed, compound Npmc+/Nras-G12D mutated NPMc+ protein operate to drive leukemogenesis. The rapid onset mice develop AML with 80% penetrance and a 3-month of leukemia in all compound mutant mice suggests that median survival.48 Interestingly, exome analysis of these coexpression of NPM1 and FLT3‐ITD mutations may be enough leukemias revealed the accumulation of additional mutations, to initiate and promote leukemogenesis. However, it cannot be including an Idh1R132Q (synonymous to human R132H) substitu- excluded that additional mutations are acquired very rapidly in a tion that is known to co-occur with NPM1 and NRAS-G12D population of murine cells prone to leukemic transformation. in human AML.

Wild type (pIpC)

** * * Npmc+ (pIpC)

Flt3-ITD (No pIpC) % alive Npmc+ / Flt3-ITD (No pIpC)

Survival (days)

~30-70 days

Mx1-Cre Expansion of Expansion of Npmc+ “leak” Flt3-ITD/Npmc+ Flt3-ITD-LOH/Npmc+

& acquisition of cells LOH for Flt3-ITD in one or more cells

Flt3-ITD cell Flt3-ITD/Npmc+cell Flt3-ITD-LOH/Npmc+ cell Figure 5. Molecular synergy and clonal evolution in Npmc+ /Flt3-ITD-driven murine AML. (a) Npmc+ and Flt3-ITD single-mutant mice have a signiﬁcantly reduced survival compared with wild-type mice owing to an excess of myeloid malignancies (*Po0.01). All double-mutant mice develop very early-onset universal AML, in keeping with a strong synergy between the two mutations. (b) Clonal evolution of AML in double- mutant mice: the process starts in prenatal or early postnatal life with activation by ‘leaky’ Mx1-Cre expression of the Npmc+ conditional allele to generate a small number of double-mutant HSCs. Double-mutant cells grow rapidly to dominate hemopoiesis with an evidence of AML by the time of weaning (3–4 weeks), while a small number of these cells acquire LOH for Flt3-ITD. LOH gives the latter cell an additional advantage and they outgrow double-mutant cells without LOH by the time mice develop full-blown AML.

NPM1-TARGETED THERAPY AND MOUSE MODELS mutant to the nucleoplasm but unfortunately not in the nucleolus NPM1 mutations are a particularly attractive therapeutic target as (because of the lack of one or both C-terminal tryptophans at 19 they are common, behave as founder genetic lesions and lead to positions 288 and 290). AML development through mechanisms that appear to differ from An alternative therapeutic strategy derives from the assumption 5 that the nucleolus in NPM1-mutated AML cells is more vulnerable those of other mutations. Clinically, NPM1 mutations are 51 associated with a favorable prognosis. However, about 40% of than in other cells, because it contains a lower amount of wild- fi NPM1-mutated/FLT3-ITD-negative and the majority of NPM1- type NPM1 (because of both haploinsuf ciency and cytoplasmic dislocation).19 This establishes the rationale for using compounds mutated/FLT3-ITD-positive AML patients succumb to their 51,53 disease.16 Thus, there is clearly a need for new therapeutic that are capable of disrupting nucleolar integrity by dislocat- strategies against NPM1-mutated AML. ing the residual nucleolar wild-type NPM1 to the nucleoplasm. Because NPM1 is a hub protein that in its oligomeric form is Wild-type NPM1 is a nucleolar protein that shuttles between the 54 20 49 essential for maintaining nucleolus homeostasis, its dislocation nucleus and the cytoplasm and exerts multiple functions. (and possibly that of other nucleolar components) to the Characteristically, all NPM1 mutations cause similar changes at the nucleoplasm55 can trigger apoptotic signals.56 We predict that in C-terminal portion of the wild-type protein, disrupting a nucleolar 19 NPM1-mutated AML a number of compounds affecting ribosome localization and creating a new nuclear export motif in its place. 57 53 fi biogenesis or inhibiting the oligomerization of NPM1 may These alterations perturb the nucleo-cytoplasmic traf cof exert antileukemic activity. Recent studies indicate that wild-type nucleophosmin, leading to its aberrant accumulation in the NPM1 binds to the nucleolus by interacting with G-quadruplexes cytoplasm of AML cells, in what appears to be a critical 19,20 of ribosomal DNA, suggesting also the potential of using selective leukemogenic event. However, we found that a small amount ligands of the G-quadruplexes that compete with NPM1 binding,58 of residual wild-type NPM1 is always detectable in the nucleolus of to disrupt the structure of the nucleolus. NPM1-mutated leukemic cells, strongly suggesting that it may be 19 Preclinical testing of compounds against NPM1-mutated AML required for their survival. This is in keeping with the observation cells, both in vitro and in animal models, will be critical for their 4,20 that NPM1 mutations in AML are consistently heterozygous and rapid translation into clinical use. The OCI–AML318 and the IMS- that the complete knockdown of Npm1 alleles is embryonic lethal M259 human cell lines, which recapitulate the features of NPM1- 50 in mice. mutated AML, are useful tools for in vitro studies. Genetically These data suggest that AML cells harboring NPM1 mutations engineered mouse models of NPM1 mutation and compound could be potentially targeted using at least two different mutants may further contribute to high-throughput drug screen- strategies that make use of the aberrant localization of the ing in vivo. For example, they could be used for the generation of mutant protein (reviewed by Falini et al.51). One approach is to immortalized cell lines harboring selected genetic lesions and be focus on relocation of the cytoplasmic NPM1 mutant to the used in drug sensitivity screens. Mouse models could be also nucleus using Leptomycin-B19 or newer Crm1 inhibitors with useful in designing clinical trials. As an example, Schlenk et al.60 improved therapeutic windows.52 These compounds redirect the found that all-trans retinoic acid (ATRA) significantly improved

Leukemia (2015) 269 – 278 © 2015 Macmillan Publishers Limited Mouse models of NPM1-mutated AML P Sportoletti et al 277 survival only in AML patients harboring the NPM1 mutation in the GV is funded by a Welcome Trust Senior Fellowship in Clinical Science. AM is funded absence of FLT3-ITD. However, these findings were not confirmed by a Kay Kendall Leukaemia Fund project grant. in another study from the UK MRC AML12 trial.61 Different ATRA administration schedules and/or inclusion of etoposide in the REFERENCES regimen have been claimed to account for these conflicting results. The availability of NPM mutant mouse models could help 1 Cancer Genome Atlas Research Network. Cancer Genome Atlas Research Network. clarify this controversial issue. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. A strong association has been found between NPM1 mutations N Engl J Med 2013; 368: 2059–2074. and those affecting the FLT3, DNMT3A, IDH1/2, RAS and cohesin- 2 Kandoth C, McLellan MD, Vandin F, Ye K, Niu B, Lu C et al. Mutational 1 landscape and significance across 12 major cancer types. Nature 2013; 502 complex genes in AML with normal cytogenetics. This suggests 333–339. that, in order to eradicate the leukemic clone, drugs targeting 3 Falini B, Mecucci C, Tiacci E, Alcalay M, Rosati R, Pasqualucci L et al. Cytoplasmic more than one mutation could be combined, hence the nucleophosmin in acute myelogenous leukemia with a normal karyotype. N Engl J importance of having compound mutant mouse models as a Med 2005; 352254–266. platform for testing drug combinations. As an example, therapy of 4 Falini B, Nicoletti I, Martelli MF, Mecucci C. Acute myeloid leukemia carrying NPM1-mutated AML with a high burden of FLT3-ITD is a medical cytoplasmic/mutated nucleophosmin (NPMc+ AML): biologic and clinical features. need because this genotype is usually associated with a poor Blood 2007; 109874–885. prognosis. The double-mutated (NPMc+/FLT3-ITD) mouse 5 Falini B, Martelli MP, Bolli N, Sportoletti P, Liso A, Tiacci E et al. Acute myeloid 34,45,46 leukemia with mutated nucleophosmin (NPM1): is it a distinct entity? Blood 2011; models represent a valuable preclinical tool for testing 117 – drugs targeting the residual nucleolar wild-type NPM1 (see above) : 1109 1120. 6 Falini B, Martelli MP, Mecucci C, Liso A, Bolli N, Bigerna B et al. Cytoplasmic in combination with second-generation FLT3-ITD inhibitors (for 62 mutated nucleophosmin is stable in primary leukemic cells and in a example, quizartinib) or with bromodomain and extra-terminal xenotransplant model of NPMc+ acute myeloid leukemia in SCID mice. 63 (BET) inhibitors. Haematologica 2008; 93:775–779. Generation of mouse models mimicking clinically relevant 7 Meloni G, Mancini M, Gianfelici V, Martelli MP, Foa R, Falini B. Late relapse of acute examples of cooperative mutations other than NPMc+/FLT3-ITD, myeloid leukemia with mutated NPM1 after eight years: evidence of NPM1 for example, NPM1, DNMT3A and IDH1/2 mutations in various mutation stability. Haematologica 2009; 94:298–300. combinations, would be very useful for evaluating multitargeted 8 Liso A, Bogliolo A, Freschi V, Martelli MP, Pileri SA, Santodirocco M et al. In human therapeutic approaches—for instance, the use of compounds genome, generation of a nuclear export signal through duplication appears targeting NPM1 together with epigenetic drugs such as unique to nucleophosmin (NPM1) mutations and is restricted to AML. Leukemia 2008; 22: 1285–1289. 5-azacytidine (that have already shown activity in the pre- 64 9 Traina F, Visconte V, Elson P, Tabarroki A, Jankowska AM, Hasrouni E et al. emptive therapy of NPM1-mutated AML ) or the recently Impact of molecular mutations on treatment response to DNMT inhibitors in 65 66 developed inhibitors of IDH1 and IDH2 mutants. myelodysplasia and related neoplasms. Leukemia 2014; 28:78–87. Finally, xenotransplant models may serve as a tool for assessing 10 Grossmann V, Haferlach C, Weissmann S, Roller A, Schindela S, Poetzinger F et al. the capability of drugs to kill LSCs. Recent microarray analyses of The molecular profile of adult T-cell acute lymphoblastic leukemia: mutations in enriched CD34+ LSCs from NPM1-mutated AML patients revealed RUNX1 and DNMT3A are associated with poor prognosis in T-ALL. Genes overexpression of genes involved in T-cell immunogenicity, such Chromosomes Cancer 2013; 52:410–422. as CD96 and IL12RB1 in NPM1-mutated as compared with NPM1 11 Yan H, Parsons DW, Jin G, McLendon R, Rasheed BA, Yuan W et al. IDH1 and IDH2 360 – wild-type cases.67 Because CD96 is not significantly expressed in mutations in gliomas. N Engl J Med 2009; :765 773. normal HSCs, specific anti-CD96 antibodies may have the potential 12 Falini B, Mecucci C, Saglio G, Lo Coco F, Diverio D, Brown P et al. NPM1 mutations and cytoplasmic nucleophosmin are mutually exclusive of recurrent genetic to eradicate NPM1-mutated CD34+ leukemic stem cells. abnormalities: a comparative analysis of 2562 patients with acute myeloid leukemia. Haematologica 2008; 93:439–442. CONCLUSIONS 13 Alcalay M, Tiacci E, Bergomas R, Bigerna B, Venturini E, Minardi SP et al. Acute myeloid leukemia bearing cytoplasmic nucleophosmin (NPMc+ AML) shows a Current mouse models of NPM1-mutated AML have been distinct gene expression profile characterized by up-regulation of genes involved important for (i) demonstrating that the NPM1 mutation alone in stem-cell maintenance. Blood 2005; 106: 899–902. displays a low leukemogenic activity in vivo but can lead to 14 Garzon R, Garofalo M, Martelli MP, Briesewitz R, Wang L, Fernandez-Cymering C leukemia after a long latency required to acquire collaborating et al. Distinctive microRNA signature of acute myeloid leukemia bearing cyto- 105 – mutations; (ii) showing that the combination of mutant NPM1 with plasmic mutated nucleophosmin. Proc Natl Acad Sci USA 2008; :3945 3950. Flt3-ITD can lead to the rapid development of AML in mice, often 15 Haferlach C, Mecucci C, Schnittger S, Kohlmann A, Mancini M, Cuneo A et al. AML with mutated NPM1 carrying a normal or aberrant karyotype show overlapping displaying LOH for Flt3-ITD and mimicking the situation in AML biologic, pathologic, immunophenotypic, and prognostic features. Blood 2009; patients; (iii) identifying a functionally competent population of 114: 3024–3032. preleukemic hemopoietic stem cells; and for (iv) improving our 16 Schlenk RF, Dohner K, Krauter J, Frohling S, Corbacioglu A, Bullinger L et al. understanding of the molecular mechanisms underlying deregu- Mutations and treatment outcome in cytogenetically normal acute myeloid lated growth induced by NPM1 mutants. They also represent leukemia. N Engl J Med 2008; 358: 1909–1918. robust in vivo platforms for the evaluation of novel therapies alone 17 Arber DA, Brunning R.D, Le Beau M.M, Falini B, Vardiman J.W, Porwit A, Thiele J, or in combination with established drugs. Information gained Bloomfield C.D. Acute myeloid leukaemia with recurrent genetic abnormalities. from these models can help the development of new rational WHO Classification of tumours of haematopoietic and lymphoid tissues. I.A.R.C. – treatments for NPM1-mutated AML patients. International Agency for Research on Cancer: Lyon, 2008. 110 123. 18 Quentmeier H, Martelli MP, Dirks WG, Bolli N, Liso A, Macleod RA et al. Cell line OCI/AML3 bears exon-12 NPM gene mutation-A and cytoplasmic expression of CONFLICT OF INTEREST nucleophosmin. Leukemia 2005; 19: 1760–1767. 19 Falini B, Bolli N, Shan J, Martelli MP, Liso A, Pucciarini A et al. Both B. Falini applied for a patent on the clinical use of NPM1 mutants. The remaining carboxy-terminus NES motif and mutated tryptophan(s) are crucial for aberrant fl authors declare no con ict of interest. nuclear export of nucleophosmin leukemic mutants in NPMc+ AML. Blood 2006; 107: 4514–4523. 20 Falini B, Bolli N, Liso A, Martelli MP, Mannucci R, Pileri S et al. Altered nucleo- ACKNOWLEDGEMENTS phosmin transport in acute myeloid leukaemia with mutated NPM1: molecular This work was supported by the Associazione Italiana Ricerca Cancro (AIRC) (IG 2013 basis and clinical implications. Leukemia 2009; 23: 1731–1743. n.14595), the Associazione Umbra contro le Leucemie e i Linfomi (AULL) and Italian 21 Cheng K, Grisendi S, Clohessy JG, Majid S, Bernardi R, Sportoletti P et al. Minister of Health, Project ‘Ricerca Finalizzata 2008’ (Grant n. RF-UMB-2008-1198396). The leukemia-associated cytoplasmic nucleophosmin mutant is an oncogene

© 2015 Macmillan Publishers Limited Leukemia (2015) 269 – 278 Mouse models of NPM1-mutated AML P Sportoletti et al 278 with paradoxical functions: Arf inactivation and induction of cellular senescence. 44 Martelli MP, Sportoletti P, Tiacci E, Martelli MF, Falini B. Mutational landscape of Oncogene 2007; 26: 7391–7400. AML with normal cytogenetics: biological and clinical implications. Blood reviews 22 Lapidot T, Sirard C, Vormoor J, Murdoch B, Hoang T, Caceres-Cortes J et al. A cell 2013; 27:13–22. initiating human acute myeloid leukaemia after transplantation into SCID mice. 45 Mupo A, Celani L, Dovey O, Cooper JL, Grove C, Rad R et al. A powerful molecular Nature 1994; 367:645–648. synergy between mutant Nucleophosmin and Flt3-ITD drives acute myeloid 23 Martelli MP, Pettirossi V, Thiede C, Bonifacio E, Mezzasoma F, Cecchini D et al. leukemia in mice. Leukemia 2013; 27: 1917–1920. CD34+ cells from AML with mutated NPM1 harbor cytoplasmic mutated 46 Rau R, Magoon D, Greenblatt S, Li L, Annesley C, Duffield AS et al. NPMc+ nucleophosmin and generate leukemia in immunocompromised mice. Blood cooperates with Flt3/ITD mutations to cause acute leukemia recapitulating 2010; 116: 3907–3922. human disease. Exp Hematol. 2013; 42: 101–113. 24 Taussig DC, Vargaftig J, Miraki-Moud F, Griessinger E, Sharrock K, Luke T et al. 47 Whitman SP, Archer KJ, Feng L, Baldus C, Becknell B, Carlson BD et al. Absence of Leukemia-initiating cells from some acute myeloid leukemia patients with the wild-type allele predicts poor prognosis in adult de novo acute myeloid mutated nucleophosmin reside in the CD34(− ) fraction. Blood 2010; 115: leukemia with normal cytogenetics and the internal tandem duplication of FLT3: a 1976–1984. cancer and leukemia group B study. Cancer Res 2001; 61: 7233–7239. 25 Jan M, Snyder TM, Corces-Zimmerman MR, Vyas P, Weissman IL, Quake SR et al. 48 Dovey O, Grove C, Cooper J, Mupo A, Wright P, Bradley A et al. A study of the Clonal evolution of preleukemic hematopoietic stem cells precedes human acute leukemogenic interaction between mutant NPM1 and NRASG12D in conditional 98 myeloid leukemia. Sci Tansl Med 2012; 4: 149ra18. knock-in mice. Haematologica 2013; (s1): 464, abstract S1126. fi 26 Shlush LI, Zandi S, Mitchell A, Chen WC, Brandwein JM, Gupta V et al. Identifi- 49 Grisendi S, Mecucci C, Falini B, Pandol PP. Nucleophosmin and cancer. Nat Rev 6 – cation of pre-leukaemic haematopoietic stem cells in acute leukaemia. Nature Cancer 2006; :493 505. 2014; 506: 328–333. 50 Grisendi S, Bernardi R, Rossi M, Cheng K, Khandker L, Manova K et al. Role of 437 27 Corces-Zimmerman MR, Hong WJ, Weissman IL, Medeiros BC, Majeti R. nucleophosmin in embryonic development and tumorigenesis. Nature 2005; : – Preleukemic mutations in human acute myeloid leukemia affect epigenetic 147 153. regulators and persist in remission. Proc Natl Acad Sci USA 2014; 111: 2548–2553. 51 Falini B, Gionfriddo I, Cecchetti F, Ballanti S, Pettirossi V, Martelli MP. Acute 28 Cheng K, Sportoletti P, Ito K, Clohessy JG, Teruya-Feldstein J, Kutok JL et al. The myeloid leukemia with mutated nucleophosmin (NPM1): any hope for a targeted 25 – cytoplasmic NPM mutant induces myeloproliferation in a transgenic therapy?. Blood Rev. 2011; :247 254. mouse model. Blood 2010; 115: 3341–3345. 52 Mutka SC, Yang WQ, Dong SD, Ward SL, Craig DA, Timmermans PB et al. Identi- fi 29 Bolli N, Payne EM, Grabher C, Lee JS, Johnston AB, Falini B et al. Expression of the cation of nuclear export inhibitors with potent anticancer activity in vivo. Cancer 69 – cytoplasmic NPM1 mutant (NPMc+) causes the expansion of hematopoietic cells Res. 2009; :510 517. 53 Falini B, Martelli MP. NPM1-mutated AML: targeting by disassembling. Blood 2011; in zebrafish. Blood 2010; 115: 3329–3340. 118:2936–2938. 30 Bolli N, De Marco MF, Martelli MP, Bigerna B, Pucciarini A, Rossi R et al. A dose- 54 Emmott E, Hiscox JA. Nucleolar targeting: the hub of the matter. EMBO Rep. 2009; dependent tug of war involving the NPM1 leukaemic mutant, nucleophosmin, 10:231–238. and ARF. Leukemia 2009; 23: 501–509. 55 Chan PK, Chan FY. A study of correlation between NPM-translocation and 31 Khanna-Gupta A, Abayasekara N, Levine M, Sun H, Virgilio M, Nia N et al. apoptosis in cells induced by daunomycin. Biochem Pharmacol 1999; 57: Up-regulation of translation eukaryotic initiation factor 4E in nucleophosmin 1 1265–1273. haploinsufficient cells results in changes in CCAAT enhancer-binding protein 56 Rubbi CP, Milner J. Disruption of the nucleolus mediates stabilization of p53 in alpha activity: implications in myelodysplastic syndrome and acute myeloid response to DNA damage and other stresses. EMBO J 2003; 22: 6068–6077. leukemia. J Biol Chem 2012; 287: 32728–32737. 57 Burger K, Muhl B, Harasim T, Rohrmoser M, Malamoussi A, Orban M et al. Che- 32 Sportoletti P, Grisendi S, Majid SM, Cheng K, Clohessy JG, Viale A et al. Npm1 is a motherapeutic drugs inhibit ribosome biogenesis at various levels. J Biol Chem haploinsufficient suppressor of myeloid and lymphoid malignancies in 2010; 285: 12416–12425. the mouse. Blood 2008; 111: 3859–3862. 58 Chiarella S, De Cola A, Scaglione GL, Carletti E, Graziano V, Barcaroli D et al. 33 Sportoletti P, Varasano E, Rossi R, Bereshchenko O, Cecchini D, Gionfriddo I et al. Nucleophosmin mutations alter its nucleolar localization by impairing G-quad- The human NPM1 mutation A perturbs megakaryopoiesis in a conditional ruplex binding at ribosomal DNA. Nucleic Acids Res 2013; 41: 3228–3239. 121 – mouse model. Blood 2013; : 3447 3458. 59 Chi HT, Vu HA, Iwasaki R, Nagamura F, Tojo A, Watanabe T et al. Detection of 34 Mallardo M, Caronno A, Pruneri G, Raviele PR, Viale A, Pelicci PG et al. NPMc+ and exon 12 type A mutation of NPM1 gene in IMS-M2 cell line. Leuk Res 2010; 34: FLT3_ITD mutations cooperate in inducing acute leukaemia in a novel 261–262. 27 – mouse model. Leukemia 2013; : 2248 2251. 60 Schlenk RF, Dohner K, Kneba M, Gotze K, Hartmann F, Del Valle F et al. Gene 35 Pasqualucci L, Liso A, Martelli MP, Bolli N, Pacini R, Tabarrini A et al. Mutated mutations and response to treatment with all-trans retinoic acid in elderly nucleophosmin detects clonal multilineage involvement in acute myeloid patients with acute myeloid leukemia. Results from the AMLSG Trial AML HD98B. fi 108 – leukemia: Impact on WHO classi cation. Blood 2006; : 4146 4155. Haematologica 2009; 94:54–60. 36 Falini B, Macijewski K, Weiss T, Bacher U, Schnittger S, Kern W et al. Multilineage 61 Burnett AK, Hills RK, Green C, Jenkinson S, Koo K, Patel Y et al. The impact on dysplasia has no impact on biologic, clinicopathologic, and prognostic features of outcome of the addition of all-trans retinoic acid to intensive chemotherapy in 115 – AML with mutated nucleophosmin (NPM1). Blood 2010; : 3776 3786. younger patients with nonacute promyelocytic acute myeloid leukemia: overall 37 Garzon R, Pichiorri F, Palumbo T, Iuliano R, Cimmino A, Aqeilan R et al. MicroRNA results and results in genotypic subgroups defined by mutations in NPM1, FLT3, fi ngerprints during human megakaryocytopoiesis. Proc Natl Acad Sci USA 2006; and CEBPA. Blood 2010; 115:948–956. 103: 5078–5083. 62 Pratz KW, Luger SM. Will FLT3 inhibitors fulfill their promise in acute meyloid 38 Vassiliou GS, Cooper JL, Rad R, Li J, Rice S, Uren A et al. Mutant nucleophosmin leukemia?. Curr Opin Hematol 2014; 21:72–78. and cooperating pathways drive leukemia initiation and progression in mice. 63 Dawson MA, Gudgin EJ, Horton SJ, Giotopoulos G, Meduri E, Robson S et al. Nat Genet 2011; 43: 470–475. Recurrent mutations, including NPM1c, activate a BRD4-dependent core tran- 39 Chou SH, Ko BS, Chiou JS, Hsu YC, Tsai MH, Chiu YC et al. A knock-in Npm1 scriptional program in acute myeloid leukemia. Leukemia 2014; 28:311–320. mutation in mice results in myeloproliferation and implies a perturbation in 64 Wermke M, Thiede C, Kiani A, Ehninger G, Bornhauser M, Platzbecker U. Successful hematopoietic microenvironment. PloS one 2012; 7: e49769. treatment of molecular relapse in NPM1-positive AML using 5-azacytidine. 40 Schessl C, Rawat VP, Cusan M, Deshpande A, Kohl TM, Rosten PM et al. The Leukemia 2010; 24:236–237. AML1-ETO fusion gene and the FLT3 length mutation collaborate in inducing 65 Chaturvedi A, Araujo Cruz MM, Jyotsana N, Sharma A, Yun H, Gorlich K et al. acute leukemia in mice. J Clin Invest 2005; 115: 2159–2168. Mutant IDH1 promotes leukemogenesis in vivo and can be specifically targeted in 41 Zhao L, Melenhorst JJ, Alemu L, Kirby M, Anderson S, Kench M et al. KIT with D816 human AML. Blood 2013; 122:2877–2887. mutations cooperates with CBFB-MYH11 for leukemogenesis in mice. Blood 2012; 66 Wang F, Travins J, DeLaBarre B, Penard-Lacronique V, Schalm S, Hansen E et al. 119: 1511–1521. Targeted inhibition of mutant IDH2 in leukemia cells induces cellular 42 Gilliland DG. Hematologic malignancies. Curr Opin Hematol 2001; 8: 189–191. differentiation. Science 2013; 340: 622–626. 43 Murati A, Brecqueville M, Devillier R, Mozziconacci MJ, Gelsi-Boyer V, Birnbaum D. 67 Schneider V, Zhang L, Bullinger L, Rojewski M, Hofmann S, Wiesneth M et al. Myeloid malignancies: mutations, models and management. BMC cancer 2012; Leukemic stem cells of acute myeloid leukemia patients carrying NPM1 mutation 12: 304. are candidates for targeted immunotherapy. Leukemia 2014; 28:1759–1762.