Erythroid and Megakaryocytic Transformation

A Wickrema1 and JD Crispino2

1Section of Hematology/Oncology, University of Chicago, Chicago, IL, USA and 2Division of Hematology/Oncology, Northwestern University, Chicago, IL, USA

Red blood cells and megakaryocytes arise from a common Accumulated evidence mostly from studies with precursor, the megakaryocyte-erythroid progenitor and mouse models and human primary cells suggests that share many regulators including the transcription factors cellular expansion and differentiation occur concur- GATA-1 and GFI-1B and signaling molecules such as JAK2 rently until the late stages of erythroid differentiation and STAT5. These lineages also share the distinction (polychromatic/orthochromatic) at which point the cells of being associated with rare, but aggressive malignancies exit the cell cycle and undergo terminal maturation that have very poor prognoses. In this review, we (Wickrema et al., 1992; Ney and D’Andrea, 2000; will brieﬂy summarize features of normal development of Koury et al., 2002). A disruption of the balance between red blood cells and megakaryocytes and also highlight erythroid cell expansion and differentiation results in events that lead to their leukemic transformation. It is either myeloproliferative disorders (MPDs) such as clear that much more work needs to be done to improve our polycythemia vera, myelodysplastic syndrome, or rarely understanding of the unique biology of these leukemias in erythroleukemia. Furthermore, some patients initially and to pave the way for novel targeted therapeutics. diagnosed with MPDs ultimately progress to erythro- Oncogene (2007) 26, 6803–6815; doi:10.1038/sj.onc.1210763 leukemia. In this ﬁrst part of the review, we will highlight key factors that regulate the erythroid Keywords: AMKL; erythroleukemia; myeloproliferative differentiation program and their contribution to diseases; GATA1; RUNX1; Fli-1; Friend virus erythroleukemia.

Clinical characteristics of erythroleukemia Erythroleukemia is a relatively rare disease in humans Normal and malignant erythropoiesis accounting for approximately 5% of all acute myeloid leukemia (AML). The disease is more prevalent in Development of erythroid cells males, and 50% of all cases are considered therapy- Commitment of hematopoietic stem cells/early progeni- related leukemias due to prior exposure to chemother- tors to the erythroid lineage takes place within the bone apy or immunosuppressive agents. In addition, some marrow microenvironment under the influence of multi- cases occur as end-stage transformation of myelodys- ple regulatory factors. Pleotrophic and lineage-specific plastic syndrome. The disease has a bimodal distribution cytokines play a key role in the commitment and with a small peak in patients under the age of 20 and an differentiation of erythroid cells (Figure 1). In addition, increasing incidence in patients over the age of 70. transcription factors, the bone marrow microenviron- Patients typically present with fatigue and anemia and ment and numerous signaling proteins also play a vital occasional hepato- or splenomegaly. The large majority function in guiding and promoting expansion and of patients have cytogenetic abnormalities in the bone differentiation of erythroid progenitors. During the past marrow, many of them associated with an adverse two decades, especially since the discovery of erythro- prognosis (such as deletions of chromosome 5 or 7) poietin in mid-1980s, it has been possible to system- (Michiels et al., 1997; Park et al., 2002). Unlike chronic atically study the contribution of various transcription myelogenous leukemia, there is no single characteristic factors and signaling proteins in the expansion and cytogenetic abnormality associated with erythroleuke- differentiation of erythroid progenitors into mature mia. The diagnosis is established by examination of the circulating erythroid cells (Cantor and Orkin, 2002; peripheral blood and bone marrow, which is hypercel- Koury et al., 2002). These studies have greatly lular and commonly displays trilineage dysplasia. Two contributed to our understanding of regulatory factors subtypes of acute erythroleukemia are recognized in the responsible for both erythroid cell expansion as well as current World Health Organization classification: ery- differentiation. throleukemia (erythroid/myeloid), also commonly named M6a, is defined by the presence in the bone marrow of >50% erythroid precursors in the entire Correspondence: Professor JD Crispino, Division of Hematology/ Oncology, Northwestern University, 303 E. Superior Street, Lurie nucleated cell population and >20% myeloblasts in the 5-113, Chicago, IL 60611, USA. non-erythroid population, (that is, the myeloblasts are E-mail: [email protected] calculated as a percentage of the non-erythroid bone Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6804

Figure 1 A model for human erythroid cell development and potential factors inﬂuencing the transformation into erythroleukemia. The scheme outlines the stages of maturation and cytokine requirements at each stage of erythroid differentiation in addition to temporal expression pattern of transcription factors during differentiation. Expression pattern of cytokine receptors (in parentheses) and stage-speciﬁc activation of signal transduction proteins are also depicted in the scheme. Morphological changes in size, shape and color (due to hemoglobin) are indicated as part of the differentiation program. Aberrant expression of transcription factors and activation of signaling proteins in human and/or murine erythroleukemia are shown in bold italics (red).

marrow cells). The second subtype, pure erythroleuke- Transformation of human erythroid cells mia, also known as M6b, represents a neoplastic Transformation of normal erythroid progenitors to proliferation of immature cells committed exclusively leukemia cells is a multistep process that occurs over a to the erythroid lineage (>80% of marrow cells) with no long period of time. It is characterized by expansion of evidence of a signiﬁcant myeloid component (Jaffe, erythroid progenitors at the expense of terminal differ- 2001). Cytochemical stains sometimes show aberrant entiation and can occur at any of the distinct stages of periodic acid Schiff (PAS) positivity in erythroid erythroid maturation. For example, minimally differenti- precursors, while Prussian blue stains demonstrate ated erythroleukemia includes cells that are mostly at increased iron stores and sometimes ringed sideroblasts. early stages after commitment to erythroid lineage Immunophenotyping gives variable results, however, the (burst-forming unit erythroid, BFU-E). Such immature classic erythroid markers glycophorin A and CD71 erythroleukemia cells exhibit an HLADRÀ/CD36 þ (transferrin receptor) are useful in classiﬁcation of phenotype. On the other hand, erythroleukemia ambiguous cases. affecting the late stages of maturation (colony-forming Treatment for patients with erythroleukemia is similar unit erythroid, CFU-E), exhibit more mature morpho- to that of other AML patients, but their outcome tends logy and are generally positive for glycophorin A (Park to be poor, with the majority of patients relapsing after et al., 2002). Although the molecular mechanisms an initial response. Allogeneic bone marrow transplan- underlying the transformation of erythroid cells into tation is considered a curative treatment, although its erythroleukemia in humans are not well understood, cell use is limited by comorbidities in elderly patients and by lines generated from patient samples have been quite lack of suitable donors. Erythroleukemia is a rare useful in characterizing some of the key biochemical and disease, and as a result relatively few clinical samples cellular characteristics of this malignancy. Approxi- have been available for research. The ability to study mately 21 human erythroleukemia cell lines have been primary erythroleukemia in vitro has been further utilized in studies to understand the proliferation and hampered by the fact that most erythroleukemic differentiation program of the erythroid lineage (Drexler patients have their myeloid compartment also affected et al., 2004). The best-known one is the K562 cell line, by leukemia. Therefore, much of the research has been which has been extensively used over the last 25 years to performed using mouse models or human cell lines. study the ‘normal’ erythroid differentiation program,

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6805 but which originated from a patient with chronic The differences in the manifestation of the disease in myelogenous leukemia in blasts crisis (and thus ex- these two strains are due to differences in a few amino presses the BCR/ABL fusion protein). These available acids within the gp55 protein (Chung et al., 1989; Fang erythroleukemia cell lines exhibit a wide range of et al., 1998). Mice infected with SFFV anemia strain functional characteristics such as variable cytokine exhibit clonal expansion with terminal differentiation in requirements and differing abilities to differentiate in the presence of Epo. During the initial stage of infection, culture with or without differentiation agents. There- erythroid progenitors are found mostly at the colony- fore, despite their common use for this purpose, these forming unit erythroid stage of differentiation, and the cell lines are not particularly accurate models for the spleens of these mice are enlarged due to accumulation study of the normal differentiation program of erythroid of colony-forming unit erythroids. However, as the cells. With the discovery of c-kit-ligand (stem cell factor) Friend disease progresses, the leukemic clones emerge and improved methodologies to isolate CD34 þ early and yield a preponderance of mouse erythroleuke- progenitors from healthy individuals, it is now possible mia cells. The molecular mechanisms involved in this to expand and differentiate large numbers of primary transformation have been studied extensively. Of note, erythroid progenitors. This approach allows us to study erythroid progenitors obtained during the initial phase the differentiation program of human erythroid cells of infection with SFFV-A have been used to study under more physiological conditions (Freyssinier et al., the molecular program of terminal erythroid differen- 1999; Uddin et al., 2004; Giarratana et al., 2005). The tiation, especially because they respond to Epo during knowledge of cytokine requirements and the regulation in vitro culture and terminally differentiate to reticulo- of transcription factor expression learned by studying cytes (Koury et al., 1984; Wickrema et al., 1991). Prior the normal differentiation program in primary cells can to development of model systems to study human then be applied to gain insights into the molecular erythroid differentiation in vitro this was one of the program of erythroleukemia. Nevertheless, our under- few primary in vitro cell culture systems that were standing of multistep transformation process in human available to study the molecular program of late stage erythroid progenitors remains incomplete. At present erythropoiesis. this process has been most elegantly studied in mice and One of the proteins involved in the leukemic chicken using retroviral-infected primary erythroid transformation in mice is the gene product of spi-1, progenitors. PU.1. PU.1 belongs to the E26 transformation specific (ETS) family of transcription factors, which are also known to play a role in avian erythroleukemia (Blair Transformation of murine and avian erythroid and Athanasiou, 2000). On the basis of a large number progenitors of studies on the role of PU.1, it is well established that Transformation of normal avian and murine erythroid in order for erythroid progenitors to continue their cells to erythroleukemia occurs in a multistep process as normal differentiation program PU.1 expression must a result of infection by retroviruses (Ney and D’Andrea, be switched off very early at the time of commitment of the 2000; Rietveld et al., 2001). Even though erythroleuke- erythroid lineage. Chromosomal rearrangements that mia in humans has not been associated with a specific occur due to infection by retroviruses such as SSFV viral infection, studies performed using avian and affect Spi-1/PU.1 expression in mice, and contribute to murine model systems have provided a great deal of leukemogenesis (Kosmider et al., 2005). Although insight into the molecular events involved in progression previous studies have clearly shown that PU.1 blocks of erythroblastosis to erythroleukemia. In the case of differentiation, these recent studies in spi-1 transgenic avian erythroleukemia, E26 retrovirus is responsible for mice indicate that PU.1 also impacts Epo-mediated cell severe erythroblastosis that rapidly progresses to ery- survival through AKT and STAT5 signaling pathways. throleukemia (Sotirov, 1981; Samarut and Gazzolo, In addition to PU.1, several other transcription regula- 1982). This avian virus contains the human homologs of tors have been implicated in promoting erythroleukemia myb and ets-1 genes, which are responsible for in mice as a result of chromosomal rearrangements. One transformation of avian erythroblasts. In the case of such transcriptional regulator is FLI-1, another member murine erythroleukemia, Friend virus is responsible for of the ETS family of transcription factors. FLI-1 progression of erythroblastosis to murine erythroleuke- induces erythroleukemia in newborn mice (as a result mia (Friend, 1957). The Friend virus has two variants, of infection by a component of the Friend virus) as the spleen focus-forming virus (SFFV) polycythemia opposed to in adult mice in the case of PU.1 (Blair and strain (SFFV-P) and SFFV anemia strain (SFFV-A). Athanasiou, 2000). Nevertheless, expression of these The pathogenicity of SFFV is due to the envelope transcriptional regulators alone is not sufficient to protein gp55, which binds the erythropoietin (Epo) produce erythroleukemia in mice. Subsequent genetic receptor and activates it in the absence of Epo (Blair and alterations and/or loss of genes such as p53 is necessary Athanasiou, 2000; Ney and D’Andrea, 2000). Both for erythroleukemia to develop, reaffirming the notion SFFV polycythemia and SFFV anemia strains cause that leukemogenesis is a multistep process that occurs disease in the erythroid lineage although they act quite over time. PU.1 and FLI-1 are the best-studied trans- differently. In the case of infection by SFFV poly- cription factors in the context of viral-induced erythro- cythemia strain, the mice become polycythemic, whereas leukemia, but other transcription factors, especially mice infected with SFFV anemia strain become anemic. those that promote expansion and/or differentiation,

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6806 are likely to be important as well in human erythroleu- GATA1/2, SCL/TAL, NF-E2 and EKLF have been kemia. extensively studied using transgenic mice and cell lines. Differentiation blocks due to deletion of GATA1, SCL/ TAL, NF-E2 or EKLF in mice do not result in Regulatory pathways and molecules influencing erythroid uncontrolled proliferation or leukemia in vivo, at least transformation based on mouse models where most of the studies have Rearrangement, deletions and/or specific mutations in been performed (Aplan et al., 1992; Andrews, 1998; human and mouse chromosomes can result in clonal Cantor and Orkin, 2002; Xu et al., 2003; Drissen et al., expansion and subsequent transformation to erythro- 2005). This lack of transformation likely may stem from leukemia. In order for erythroleukemia to emerge, one an additional requirement for these transcription factors or more molecules regulating erythroid differentiation in cell survival (for example, Weiss and Orkin, 1995). must be affected. However, recent data also indicate Also, many of these transcription factors, such as leukemogenesis of a particular lineage may arise as a NF-E2, participate in various aspect of globin gene result of altered balance or expression of a transcription regulation. Interestingly, integration of the Friend virus factor that regulates a lineage besides the one that is within the sequences encoding the p45 subunit of the affected by leukemia (Cantor and Orkin, 2002). NF-E2 gene results in a block of erythroid maturation in mice (Lu et al., 1994). This phenomenon is limited to the Transcription factors influencing erythroleukemia.A Friend disease in mice but has not been implicated in large number of transcription factors guide the expan- erythroleukemia in humans. In addition to the transcrip- sion and differentiation of committed erythroid pro- tion factors mentioned above, nuclear factor-kB and genitors during normal erythropoiesis. Although lineage- forkhead transcription factors (FOXO) are expressed specific transcription factors are pivotal in induction during the erythroid differentiation program (Zhang of the commitment and subsequent progression of the et al., 1998; Mahmud et al., 2002). The FOXO family of differentiation process, transcription factors that are transcription factors has also been implicated in several important for guiding other hematopoietic lineages also forms of solid tumors in pediatric patients: in some play a role in manifestation of erythroleukemia, such cases, they form fusion proteins with PAX and inacti- as in mice infected with the Friend virus. There is vate the normal function of FOXO1a (Keller et al., now direct evidence for involvement of transcription 2004). Up to now, neither nuclear factor-kB nor FOXO factors that promote erythroid proliferation in the has been directly implicated in erythroleukemia. development of erythroleukemia. In a recent study, STAT1 and STAT3 transcription factors were found to be constitutively activated in primary erythroleukemia Signaling proteins influencing erythroleukemia. Signal- cells (Kirito et al., 2002). The same study also showed ing proteins play a central role in leukemogenesis. that mutation of STAT-binding sites within the c-myc Involvement of signaling proteins in erythroleukemia gene promoter was able to block promoter activity, occurs as a result of mutations in cytokine receptors, suggesting that c-myc activation may contribute to the which will result in constitutive activation downstream proliferative potential of the malignant clone in ery- signaling pathways independent of cytokines. An excel- throleukemia. Another transcription factor that has lent example is that of JAK2 mutations prevalent in been implicated in erythroid leukemia is the growth polycythemia vera (Baxter et al., 2005; James et al., 2005; factor-independent 1B protein (GFI-1B). In a recent Kralovics et al., 2005; Levine et al., 2005; Zhao et al., study, it was demonstrated that patients with AML M6 2005). In SFFV-infected mice, acquisition of specific and AML M7 (acute megakaryocytic leukemia (AMKL, activating mutations in c-kit also results in clonal see below) express high levels of GFI-1B mRNA in bone expansion of erythroid progenitors (Kosmider et al., marrow cells (Elmaagacli et al., 2007). Furthermore, this 2005). These activating mutations in receptor tyrosine report also demonstrated a decrease in the proliferative kinases allow constitutive activation of multiple down- capacity of K562 and HEL after silencing GFI-1B stream signaling cascades important in cell proliferation expression using siRNA. However, other hematologic and/or cell survival. In addition to these mechanisms of malignancies such as chronic myelogenous leukemia activation, it has been demonstrated that autocrine or and other AMLs did not exhibit an increased level of paracrine production of proinflammatory cytokines, GFI-1B expression. In normal erythropoiesis, GFI-1B such as tumor necrosis factor-a due to oncogenic has been shown to play a vital role in both promoting changes in cells, may provide a proliferative advantage differentiation as well as expansion of early committed by activating signaling pathways (Jacobs-Helber et al., erythroid progenitors (Osawa et al., 2002; Saleque et al., 2003). Other studies have shown that Lyn kinase and 2002; Garcon et al., 2005). the signaling intermediate Gab2 also play a role in Figure 1 summarizes the expression pattern of supporting the growth of erythroleukemia cells. Lyn transcription factors important in the erythroid maturation kinase, which belongs to the Src family of kinases, is and highlights those that have been implicated in human involved in normal signaling as well as in the develop- erythroleukemia. In addition to transcription factors ment of erythroleukemia in mice (Tilbrook et al., 2001), promoting proliferation, a number of transcription while Gab2, an adaptor protein involved in Epo- factors have been identified that only promote erythroid mediated signaling events in normal cells is recruited commitment and terminal differentiation. Among them by the truncated form of Stk tyrosine receptor in Friend

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6807 virus-infected erythroid cells to promote the expansion additional regulation provided by signaling molecules of mouse erythroid cells (Wickrema et al., 1999; Teal such as JAK and Src family kinases. Many of these et al., 2006). Finally, besides signal transduction proteins same factors are shared between erythroid cells and that are actively involved in supporting proliferation, it is megakaryocytes, but have distinct targets in the two conceivable that signaling proteins such as phosphatidy- different settings. Below, we will highlight the roles of linositol 3-kinase and AKT also could support the these factors in abnormal megakaryopoiesis, with an survival of erythroleukemia cells through constitutive emphasis on recently published ﬁndings and on the activation of the cell survival pathways. relationship between gene expression and leukemic transformation.

Normal and malignant megakaryopoiesis Clinical characteristics and classification of acute megakaryocytic leukemia Development of megakaryocytes Megakaryocytic leukemias comprise a subgroup of rare, Similar to erythroid cells, megakaryocytes also arise but aggressive, AMLs. There are three main groups of from the megakaryocyte-erythroid progenitor (MEP) and patients who develop distinct forms of AMKL: children progress through discrete maturation stages (Figure 2). with Down syndrome (DS) (DS-AMKL); infants with- Committed megakaryocyte progenitors, including the out DS (non-DS pediatric AMKL); and adults. AMKL colony-forming unit megakaryocyte, proliferate to a comprises approximately 5–7% of AML in children limited extent, giving rise to promegakaryoblasts. without DS (Barnard et al., 2006) and 1% of adult AML Individual megakaryocytes then undergo terminal dif- (Tallman et al., 2000; Pagano et al., 2002). In the context ferentiation and eventually shed platelets. In concert of DS, however, megakaryocytic disorders are relatively with cytoplasmic maturation that leads to platelet common, occurring in as many as 10% of newborns. production, megakaryocyte nuclei undergo a matura- Here, we will address the clinical and genetic features of tion process that involves repeated rounds of DNA these intriguing megakaryocytic malignancies. synthesis without cell division, a variant cell cycle termed endomitosis (Ravid et al., 2002). This phenom- AMKL in children with DS. Children with DS are enon allows megakaryocytes to accumulate DNA predisposed to two related megakaryocytic disorders: content typically up to 64N and greatly increases their transient MPD (TMD) and AMKL. TMD, which is size and protein production (Hancock et al., 1993; characterized by the expansion of abnormal megakar- Raslova et al., 2003). These increases in cell size, DNA yoblasts in the peripheral blood and liver, is estimated to content and protein levels are associated with the occur in nearly 10% of DS newborns, although the development of long cytoplasmic extensions, termed majority of cases are subclinical (Zipursky, 2003). proplatelet forms, that eventually shed platelets (Cramer Interestingly, this disease often undergoes spontaneous et al., 1997). Although it remains unclear whether remission and does not require treatment except in rare polyploidization is essential for platelet production, a cases in which TMD blasts promote severe hepatic recent gene profiling study has shown that polyploidiza- fibrosis or cardiopulmonary failure (Al-Kasim et al., tion does not directly regulate gene expression per se in 2002). Although TMD disappears in most cases, it megakaryocytes (Raslova et al., 2006). reappears as AMKL in 20–30% of infants diagnosed Megakaryocyte maturation is governed by the activity with more severe cases of TMD. The incidence of of a group of transcription factors, including GATA-1, AMKL in children with DS is 1 of 500, with a median Friend of GATA-1 (FOG-1) and RUNX1, with age of presentation of 2 years, which corresponds to a

Figure 2 A model for megakaryocyte development. Megakaryocyte maturation proceeds in an ordered manner from the MEP through proplatelet-producing cells. Important transcriptional and signaling regulators that are required for the initial stages of megakaryopoiesis are indicated above and below the cells, respectively. Genes labeled with an asterisk are either mutated or misexpressed in human megakaryocytic disorders. One of the more interesting areas of megakaryocyte research is the investigation of factors that regulate the choice of a 4N megakaryoblast to undergo cell division and generate two 2N daughter cells, or to commit to polyploidization and proplatelet formation. It is possible that deregulation of this decision may contribute to AMKL. AMKL, acute megakaryocytic leukemia; MEP, megakaryocyte-erythroid progenitor; CFU-Mk, colony forming unit-megakaryocyte; Mk, megakaryocyte.

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6808 nearly 500-fold increased risk of AMKL in comparison malignancy that comprises nearly 1% of adult AML to children without DS (Lange, 2000). In contrast to cases (Tallman et al., 2000; Pagano et al., 2002). Similar TMD, AMKL is a life-threatening acute leukemia, to the other subtypes of AMKL, this leukemia is characterized by the presence of megakaryoblasts in the characterized by the excessive production of primitive peripheral blood and bone marrow, and is frequently megakaryoblasts within the bone marrow and extensive accompanied by myelofibrosis and thrombocytopenia myelofibrosis. Peripheral blood features frequently (Lange, 2000). DS-AMKL blasts express the myeloid include anemia and thrombocytopenia, in addition to markers CD33 and/or CD13 in addition to at least one circulating megakaryoblasts, which express platelet- platelet-associated antigen (CD36, CD41a, CD41b or specific antigens, such as CD41 and CD61, but lack CD61). A proportion of these blasts also express myeloperoxidase or lymphoid antigens. Adult AMKL erythroid-specific mRNAs, such as g-globin and ery- frequently arises in individuals who had an antecedent throid d-aminolevulinate synthase, suggesting that they blood disorder or myelodysplastic syndrome (Oki et al., have both erythroid and megakaryocytic properties and 2006). Although some patients achieve complete remis- perhaps arise from an MEP-like progenitor cell (Ito sion, the long-term outcome is significantly worse for et al., 1995). AMKL than other forms of adult AML, with a median Mutations in the essential hematopoietic transcription survival of 40 weeks or less (Tallman et al., 2000; factor GATA1 are associated with nearly all cases of Pagano et al., 2002; Oki et al., 2006). TMD and DS-AMKL, but are not found in any other To date, no specific chromosomal rearrangements or form of acute leukemia (for a review, see Muntean et al., genetic mutations have been found in adult AMKL. In 2006). The role of GATA1 mutations and trisomy 21 in one study, it was noted that nearly 50% of patients had the transformation of megakaryocytes is an area of one or more cytogenetic abnormalities, including À5, intense research (discussed below). Interestingly, the À7, þ 8 or 11q involvement (Oki et al., 2006). These GATA1 mutation in these leukemic cells may be findings partially explain the poor outcome of this responsible for their favorable response to chemother- subtype of AMKL, although the poor prognosis is not apy: Taub and colleagues (Ge et al., 2004) have shown fully dependent on cytogenetic abnormalities. With that DS-AMKL cells express reduced levels of the respect to genetic mutations, multiple studies have GATA-1 target gene cytidine deaminase, and as a now reported that the tyrosine kinases JAK2 or JAK3 consequence, the cells are hypersensitive to Ara-C are mutated in a subset of AMKL patients, although the chemotherapy. Indeed, the current treatment regimen mechanisms by which these mutations contribute to has resulted in favorable outcomes for this group of acute leukemia have not yet been elucidated (discussed AMKL patients, with an overall survival of 70–80% at 5 below). years (Gamis, 2005; Al-Ahmari et al., 2006). Transcriptional regulation in normal and transformed Pediatric non-DS AMKLs. Childhood non-DS AMKL megakaryocytes is characterized by an expansion of megakaryoblasts Leukemias often arise as a consequence of deregulation within the bone marrow and is frequently accompanied of transcription factors that are essential for normal by myelofibrosis, hepatosplenomegaly and pancytopenia, development (Rosenbauer and Tenen, 2007), such as usually including thrombocytopenia (Carroll et al., 1991; mutations in RUNX1/AML1 and C/EBPA in AML, Bernstein et al., 2000; Barnard et al., 2006; Rubnitz et al., translocation of SCL in T-ALL, and mutations in 2007). The majority of infant cases of AMKL are GATA1 in DS-AMKL. associated with the (1;22) translocation, which was initially discovered in 1991 (Carroll et al., 1991) and GATA-1. GATA-1 and its partner FOG-1 are abso- recently found to result in fusion of the RNA-binding lutely required for normal development of megakaryo- motif protein 15 (RBM15; aka OTT) and MKL1 (aka cytes and platelets. Mice that lack GATA-1 expression MAL) genes (Mercher et al., 2001; Ma et al., 2001b). In specifically in the megakaryocyte lineage (GATA-1low or contrast, many different cytogenetic abnormalities are GATA-1 knockdown mice) develop megakaryocytes, observed in childhood non-DS AMKL, including but they are abnormal in many respects (Shivdasani t(10;11), which leads to the CALM–AF10 fusion et al., 1997; Vyas et al., 1999; Vannucchi et al., 2002). (Abdelhaleem et al., 2007), t(9;11), þ 8or þ 21 (Barnard For example, they show reduced polyploidization, a et al., 2006). These latter cases share many features with failure to produce platelets, and a marked expansion t(1;22) leukemias, but present slightly later in life, at a both in vivo and when cultured in vitro. Of note, rescuing median age of 2.2 years (Barnard et al., 2006). Of note, all their polyploidization defect by ectopic expression of groups of children with non-DS AMKL show signifi- cyclin D1/cdk4 failed to restore platelet biosynthesis, cantly inferior overall survival and event-free survival confirming that polyploidization and terminal matura- compared to children diagnosed with other myeloid tion can be uncoupled (Muntean et al., 2007). The defect leukemias (FAB M0-M5) or with DS-AMKL (Lange in GATA-1low mice also extends to the progenitor stage, et al., 1998; Barnard et al., 2006; Rubnitz et al., 2007). as MEPs isolated from GATA-1low mice generate not only red blood cells and megakaryocytes, but also mast Adult AMKL. Another class of aggressive megakar- cells (Ghinassi et al., 2007). This surprising observation yocytic leukemias is adult AMKL. This is a rare shows that the level of GATA-1 is not only important in

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6809 terminally differentiating cells, but also hematopoietic amount of RUNX1 is also essential for normal platelet progenitors. homeostasis, as Runx1-heterozygous animals display a The interacting partner of GATA-1, FOG-1, is also mild thrombocytopenia, which is accompanied by an essential for megakaryopoiesis, although it plays a more inversion of the CD4( þ ) to CD8( þ ) T-cell ratio and a crucial role than GATA-1 in early megakaryocyte decrease in long-term repopulating hematopoietic stem progenitors, as FOG-1-deficient mice do not generate cells (Sun and Downing, 2004). RUNX1 gene rearran- any cells of the megakaryocyte lineage (Tsang et al., gements are involved in some of the recurring transloca- 1998). Mice that express a GATA-1 mutant that fails to tions associated with leukemia, while mutations are interact with FOG-1 show a phenotype that resembles associated with familial platelet disorder with predis- that of GATA-1 knockdown mice, suggesting that position to AML and sporadic cases of AML (Nucifora FOG-1 has GATA-1-independent functions, likely and Rowley, 1995; McLean et al., 1996; Song et al., mediated through GATA-2 (Chang et al., 2002). 1999; Osato et al., 1999; Preudhomme et al., 2000; Furthermore, mutations in the N-terminal zinc finger Michaud et al., 2002; Peterson and Zhang, 2004). that led to the dissociation of the GATA-1/FOG-1 Although mutations in RUNX1 were not found in interaction, or to a decreased ability of GATA-1 to bind patients with DS-AMKL (Wechsler et al., 2002), altered DNA, have been associated with a spectrum of benign gene dosage of RUNX1 may be involved in the hematologic diseases in humans, including dyserythro- development of AMKL. Indeed, several groups have poietic anemia and thrombocytopenia (for a review, see speculated that increased dose of RUNX1 contributes to Crispino, 2005), gray platelet syndrome (Tubman et al., leukemia in DS (Osato and Ito, 2005; Yanagida et al., 2007) and congenital erythropoietic porphyria (Phillips 2005; Langebrake et al., 2006). In one study, transgenic et al., 2007). Of note, there have been no reports of overexpression of RUNX1, driven by the GATA-1 mutations in FOG-1 in human hematopoietic diseases. promoter, did not cause spontaneous leukemia, but did This may reflect the essential early role for FOG-1 in shorten the latency of leukemia development in BXH-2 megakaryopoiesis or the redundancy of the multiple zinc mice and increase the frequency of tumors with a fingers within FOG-1 that may compensate for any megakaryoblastic phenotype (Yanagida et al., 2005). deficiency in one finger. Conversely, reduced expression of RUNX1 also facili- In addition to mutations in the N-finger of GATA1 in tated leukemia in animals, as Runx1-heterozygous benign blood disorders, distinct mutations in the BXH-2 mice developed myeloid leukemias with a sequences encoding the N-terminal activation domain shortened latency (Yamashita et al., 2005). Further- are found in patients with TMD and DS-AMKL more, reduced expression of RUNX1 may contribute to (Muntean et al., 2006). In all cases, the GATA1 the pathogenesis of human TMD and DS-AMKL mutations block expression of the full-length protein, (Bourquin et al., 2006). Taken together, these studies but allow for expression of a shorter isoform named suggest that dosage of RUNX1 must be tightly GATA-1s (Wechsler et al., 2002). This short isoform coordinated in vivo. retains the two zinc fingers and can bind DNA and interact with cofactors including FOG-1, but lacks the N-terminal transcriptional activation domain and thus GFI-1B, NF-E2 and c-myb. Other transcription factors is likely deficient in its ability to maintain the normal that contribute to megakaryopoiesis, as well as erythro- transcriptional program in megakaryocytes. Indeed, poiesis, include GFI-1B, NF-E2 and c-myb. Gfi-1b is several recent studies have shown that although essential for terminal megakaryocyte maturation, as GATA-1s can promote terminal differentiation of Gfi-1b-deficient fetal liver progenitors give rise to small, megakaryocytes, it is unable to properly regulate the poorly formed megakaryocyte colonies comprised of proliferation of fetal liver megakaryocyte progenitors cells that express significantly reduced levels of von- (Kuhl et al., 2005; Li et al., 2005; Muntean and Willebrand factor, gpIIb, c-mpl and NF-E2 compared Crispino, 2005). Thus, it appears that mutagenesis of to wild-type cells (Saleque et al., 2002). Interestingly, GATA1 cooperates with trisomy 21 to promote the GFI-1B appears to be overexpressed in both erythroid abnormal proliferation of megakaryocytes in TMD and and megakaryocytic human malignancies where it may DS-AMKL. For additional insights into this leukemia, contribute to the increased proliferation of leukemic the reader is directed to recent reviews of this topic cells (Elmaagacli et al., 2007). Another transcriptional (Izraeli, 2006; Vyas and Roberts, 2006; Vyas and regulator that is essential for platelet biosynthesis is Crispino, 2007). NF-E2, which promotes expression of mid-late stage megakaryocyte genes including b1-tubulin (Shivdasani et al., 1995; Chen et al., 2007). Overexpression of NF-E2 RUNX1/AML1. RUNX1/AML1 has established func- may play a role in the progression of MPDs, including tions in hematopoietic stem cell development, megakar- polycythemia vera (Goerttler et al., 2005), although a yopoiesis and leukemia. In particular, the absence of strong association with megakaryocytic leukemias has RUNX1 has profound effects on the polyploidization not been found. Finally, c-myb also plays an essential and terminal maturation of megakaryocytes, leading to role in megakaryocyte development. Suboptimal levels a significant reduction in polyploidization and platelet of c-myb favor the differentiation of megakaryocytes formation in vivo (Ichikawa et al., 2004; Growney et al., and macrophages (Emambokus et al., 2003), and loss- 2005; Putz et al., 2006). Expression of the proper of-function mutations are also associated with excess

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6810 megakaryopoiesis and increased platelet counts in differentiation of these cells. Furthermore, analysis of animals (Carpinelli et al., 2004). Point mutations in Rbm15/Ott conditionally deleted mice has revealed that c-myb, however, are unlikely to be a common cause Rbm15/Ott plays inhibitory roles in development of the of human myelofibrosis (Steensma et al., 2006b) or myeloid, megakaryocytic and progenitor compartments megakaryocytic leukemia. and suggests that dysregulation of Ott-dependent pathways may contribute to infant AMKL (Raffel et al., ETS family members. Several ETS proteins, including 2007). However, the specific role of RBM15 in mega- GABPa, ETS1, ETS2, ERG and Fli-1, are expressed in karyocyte progenitors remains undefined and awaits megakaryocytes and play key roles in megakaryocyte further studies. development. Although a role for Fli-1 in regulation of The notable features of the MKL1/MAL fusion megakaryocyte genes has been well established (Szalai partner include the presence of a SAP domain, which et al., 2006), recent work has shown that GABPa is a has been postulated to connect DNA to the nuclear regulator of early megakaryocyte-specific genes, includ- scaffold, a C-terminal proline-rich region, and a ing aIIb and c-mpl (Pang et al., 2006). In addition, this glutamine-rich segment reminiscent of MLL family study showed that the ratio of GABPa/Fli-1 decreases members. MKL1/MAL is weakly similar to myocardin with megakaryocyte maturation, consistent with a throughout its length and shows high similarity in the dependence of late megakaryocyte-specific genes on SAP domain. Interestingly, MKL1/MAL has been Fli-1 for their expression. Similarly, another recent found to physically interact with myocardin via et al study has shown that ETS1 is involved in megakar- conserved leucine zipper domains (Du ., 2004). yocyte development: ETS1 was found to increase during MKL1/MAL functions as a potent transcriptional et al megakaryocyte maturation, and its overexpression was cofactor for serum response factor (SRF) (Cen ., associated with enhanced differentiation and upregula- 2003) and also regulates SRF-dependent smooth muscle et al tion of GATA-2 and megakaryocyte-specific genes cell differentiation (Du ., 2004). These observations (Lulli et al., 2006). ETS2 and ERG, which are localized suggest that perhaps aberrant expression of SRF- to human chromosome 21, are also expressed within dependent genes, resulting from the translocation, megakaryocytes. Aberrant expression of these genes contributes to AMKL. Of note, the RBM15–MKL1 may be involved in megakaryocytic leukemia in children fusion protein retains all the functional domains of both with DS (Rainis et al., 2005), but future studies are proteins and is predicted to aberrantly modulate necessary to define the contribution of these factors to chromatin organization, Hox differentiation pathways et al AMKL. or extracellular signaling (Mercher ., 2001). Indeed, the RBM15–MKL1 fusion protein was able to activate a subset of SRF-dependent reporters to a much greater Transformation by translocation: OTT and MKL1 in extent than MKL1 alone (Cen et al., 2003). These pediatric AMKL findings suggest that the SRF co-activator activity of The majority of infant cases of AMKL are associated MKL1 is enhanced by its fusion to RBM15. However, with the t(1;22), which results in fusion of RBM15, also there have not been any reports to date of the known as one-twenty-two (OTT), to megakaryocyte consequences of overexpression of the fusion protein leukemia (MKL1), also known as megakaryocytic acute in hematopoietic cells in vivo. Thus, future studies to leukemia (MAL) (Mercher et al., 2001; Ma et al., uncover the mechanisms by which this fusion protein 2001a). The RBM15/OTT partner contains three RNA leads to aberrant expansion and differentiation of recognition motifs (RRMs) and its C terminus is megakaryocytes are needed. Interestingly, a homolog comprised of a ‘spen paralogs and orthologs C-terminal’ of RBM15, RBM6, has been found as a fusion partner domain. This latter domain is derived from the of CSF1R in the MKL1 cell line, although it has yet to Drosophila spen gene, which encodes an RRM-contain- be found in human megakaryocytic leukemia samples ing protein that can modulate both Hox function and (Gu et al., 2007). The presence of a second fusion Ras signaling (Wiellette et al., 1999; Chen and Rebay, containing the RBM motif suggests that perhaps 2000). A related member of the Drosophila Spen family, aberrant regulation of Hox function and/or Ras SHARP, has been found to interact with RBP-Jk to signaling is behind these megakaryocytic diseases. repress transcription of Notch target genes, including HES-1, and thus inhibit signaling in the absence of activated Notch (Oswald et al., 2002). A recent study Signaling proteins influencing megakaryocyte has demonstrated that RBM15 itself can modulate development Notch signaling by interacting with RBP-Jk within JAK/STAT signaling. The main physiologic mediator hematopoietic cells (Ma et al., 2007). Interestingly, the of megakaryocyte growth is thrombopoietin, whose association of RBM15 with RBP-Jk enhanced Notch- receptor c-MPL is expressed on megakaryocytes (re- induced HES promoter activity in hematopoietic cells, viewed by (Kirito and Kaushansky, 2006). Engagement but inhibited promoter activity in non-hematopoietic of c-MPL by thrombopoietin then leads to activation of cells. Consistent with a role in leukemogenesis, over- JAK kinases and their downstream effectors, including expression of RBM15 inhibited differentiation of the STAT3 and STAT5. Recently, activating mutations in 32D myeloid cell line, whereas reducing its expression c-MPL have been found associated with two human by shRNA-mediated knockdown-enhanced terminal MPDs, myelofibrosis with myeloid metaplasia and

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6811 essential thrombocythemia, at a frequency of about 5 with megakaryocytic malignancies, but perhaps their and 1% of patients, respectively (Pardanani et al., 2006; dysregulation will be found associated with a subset of Pikman et al., 2006). Mutations in JAK2 are much more AMKL cases. common in human MPDs, having been found associated with polycythemia vera, essential thrombocythemia and chronic idiopathic myelofibrosis as well as in rare cases of myelodysplastic syndrome, chronic myelo- Conclusions and future directions monocytic leukemia and atypical myeloid disorder (Baxter et al., 2005; Frohling et al., 2005; James et al., Erythroid cells and megakaryocytes share many com- 2005; Kralovics et al., 2005; Levine et al., 2005; Zhao mon features, but in general the events that led to et al., 2005; Nelson and Steensma, 2006). Of note, malignant transformation appear to be distinct between several studies have shown that JAK family members the two lineages. For example, mutagenesis of GATA1 is can be mutated in AMKL. For example, Jelinek et al. exclusively associated with megakaryocytic leukemia, (2005) found the V617F JAK2 mutation in 2/11 patients although maturation of the erythroid lineage is certainly with megakaryocytic AML, while Steensma et al. affected by non-leukemic, somatic GATA1 mutations. (2006a) identified the V617F JAK2 mutation in 3/27 Likewise, integration of the Friend virus or misexpres- AMKL samples. Recently, Walters et al. (2006) sion of PU.1 leads to transformation of erythroid, but identified JAK3-activating mutations in the CMK cell not megakaryocyte, progenitors in animals. line, which harbors a GATA1 mutation and was derived Going forward, the challenges for researchers in the from a patient with DS-AMKL, as well as in 1 of 3 field of erythroid and megakaryocytic leukemia include DS-AMKL samples and 1 of 16 non-DS AMKL deciphering the specific mechanisms by which known sample. Interestingly, transplantation of bone marrow genetic alterations, including mutagenesis of GATA1 or cells expressing the JAK3 A572V mutant induced several JAK2, and fusion of the RBM15 and MKL1 proteins, features of megakaryocytic leukemia in recipient contribute to transformation. The dual approach of animals, in addition to a T-cell lymphoproliferative mouse modeling and analysis of patient samples should disorder (Walters et al., 2006). These findings raise the provide new insights to these diseases, although patient intriguing possibility that mutations in JAKs may studies are hampered by the rarity of these diseases. facilitate the development of leukemia, although it is Furthermore, since the prognoses for patients with likely that additional genetic events are required to drive erythroleukemia and AMKL (apart from DS-AMKL) AMKL. are very poor, another great challenge for researchers is to develop novel therapeutics that specifically target Src family kinases. Another class of signaling mole- these leukemias. Perhaps by first improving our under- cules that govern proliferation and growth is the Src standing of the molecular nature of these diseases, we kinase family. Within megakaryocytes, Lyn and Fyn are will be better positioned to design improved therapies the most highly expressed family members and appear for patients. to act to negatively regulate megakaryopoiesis downstream of thrombopoietin signaling (Lannutti et al., 2003). Evidence that supports this negative role for Src Acknowledgements family kinases (SFKs) include the observation that Lyn- We thank Dr Koen van Besien for helpful discussions and null mice show increased formation of megakaryocytes apologize to those whose work could not be discussed due to in vivo, and these megakaryocytes reached a higher space limitations. Dr Wickrema is supported by grants from ploidy state when compared to those of wild-type the NCI and the Leukemia and Lymphoma Society. Dr littermates (Lannutti et al., 2006). To date, misexpres- Crispino is a scholar of the Leukemia and Lymphoma Society sion or mutations in these genes has not been associated and acknowledges support from the NCI and NIDDK.

References

Abdelhaleem M, Beimnet K, Kirby-Allen M, Naqvi A, Hitzler J, Aplan PD, Nakahara K, Orkin SH, Kirsch IR. (1992). The SCL gene Shago M. (2007). High incidence of CALM-AF10 fusion and the product: a positive regulator of erythroid differentiation. EMBO J identiﬁcation of a novel fusion transcript in acute megakaryo- 11: 4073–4081. blastic leukemia in children without Down’s syndrome. Leukemia Barnard DR, Alonzo TA, Gerbing RB, Lange B, Woods WG. (2006). 21: 352–353. Comparison of childhood myelodysplastic syndrome, AML FAB Al-Ahmari A, Shah N, Sung L, Zipursky A, Hitzler J. (2006). Long- M6 or M7, CCG 2891: report from the children’s oncology group. term results of an ultra low-dose cytarabine-based regimen for the Pediatr Blood Cancer 49: 17–22. treatment of acute megakaryoblastic leukaemia in children with Baxter EJ, Scott LM, Campbell PJ, East C, Fourouclas N, Down syndrome. Br J Haematol 133: 646–648. Swanton S et al. (2005). Acquired mutation of the tyrosine kinase Al-Kasim F, Doyle JJ, Massey GV, Weinstein HJ, Zipursky A. (2002). JAK2 in human myeloproliferative disorders. Lancet 365: Incidence and treatment of potentially lethal diseases in transient 1054–1061. leukemia of Down syndrome: pediatric oncology group study. Bernstein J, Dastugue N, Haas OA, Harbott J, Heerema NA, Huret JL J Pediatr Hematol Oncol 24: 9–13. et al. (2000). Nineteen cases of the t(1;22)(p13;q13) acute megakary- Andrews NC. (1998). The NF-E2 transcription factor. Int J Biochem blastic leukaemia of infants/children and a review of 39 cases: report Cell Biol 30: 429–432. from a t(1;22) study group. Leukemia 14: 216–218.

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6812 Blair DG, Athanasiou M. (2000). Ets and retroviruses—transduction Friend C. (1957). Cell free transmission in adult Swiss mice of a disease and activation of members of the Ets oncogene family in viral having the character of leukemia. J Exp Med 105: 307–318. oncogenesis. Oncogene 19: 6472–6481. Frohling S, Scholl C, Gilliland DG, Levine RL. (2005). Genetics of Bourquin JP, Subramanian A, Langebrake C, Reinhardt D, Bernard myeloid malignancies: pathogenetic and clinical implications. J Clin O, Ballerini P et al. (2006). Identification of distinct molecular Oncol 23: 6285–6295. phenotypes in acute megakaryoblastic leukemia by gene expression Gamis AS. (2005). Acute myeloid leukemia and Down syndrome profiling. Proc Natl Acad Sci USA 103: 3339–3344. evolution of modern therapy—state of the art review. Pediatr Blood Cantor AB, Orkin SH. (2002). Transcriptional regulation of ery- Cancer 44: 13–20. thropoiesis: an affair involving multiple partners. Oncogene 21: Garcon L, Lacout C, Svinartchouk F, Le Couedic JP, Villeval JL, 3368–3376. Vainchenker W et al. (2005). Gfi-1B plays a critical role in terminal Carpinelli MR, Hilton DJ, Metcalf D, Antonchuk JL, Hyland CD, differentiation of normal and transformed erythroid progenitor Mifsud SL et al. (2004). Suppressor screen in MplÀ/À mice: c-Myb cells. Blood 105: 1448–1455. mutation causes supraphysiological production of platelets in the Ge Y, Jensen TL, Stout ML, Flatley RM, Grohar PJ, Ravindranath Y absence of thrombopoietin signaling. Proc Natl Acad Sci USA 101: et al. (2004). The role of cytidine deaminase and GATA1 6553–6558. mutations in the increased cytosine arabinoside sensitivity of Carroll A, Civin C, Schneider N, Dahl G, Pappo A, Bowman P et al. Down syndrome myeloblasts and leukemia cell lines. Cancer Res (1991). The t(1;22) (p13;q13) is nonrandom and restricted to infants 64: 728–735. with acute megakaryoblastic leukemia: a pediatric oncology group Ghinassi B, Sanchez M, Martelli F, Amabile G, Vannucchi AM, study. Blood 78: 748–752. Migliaccio G et al. (2007). The hypomorphic Gata1low mutation Cen B, Selvaraj A, Burgess RC, Hitzler JK, Ma Z, Morris SW et al. alters the proliferation/differentiation potential of the common (2003). Megakaryoblastic leukemia 1, a potent transcriptional megakaryocytic-erythroid progenitor. Blood 109: 1460–1471. coactivator for serum response factor (SRF), is required for serum Giarratana MC, Kobari L, Lapillonne H, Chalmers D, Kiger L, induction of SRF target genes. Mol Cell Biol 23: 6597–6608. Cynober T et al. (2005). Ex vivo generation of fully mature human Chang AN, Cantor AB, Fujiwara Y, Lodish MB, Droho S, Crispino red blood cells from hematopoietic stem cells. Nat Biotechnol JD et al. (2002). GATA-factor dependence of the multitype zinc- 23: 69–74. finger protein FOG-1 for its essential role in megakaryopoiesis. Proc Goerttler PS, Kreutz C, Donauer J, Faller D, Maiwald T, Marz E Natl Acad Sci USA 99: 9237–9242. et al. (2005). Gene expression profiling in polycythaemia vera: Chen F, Rebay I. (2000). Split ends, a new component of the overexpression of transcription factor NF-E2. Br J Haematol Drosophila EGF receptor pathway, regulates development of 129: 138–150. midline glial cells. Curr Biol 10: 943–946. Growney JD, Shigematsu H, Li Z, Lee BH, Adelsperger J, Rowan R Chen Z, Hu M, Shivdasani RA. (2007). Expression analysis of primary et al. (2005). Loss of Runx1 perturbs adult hematopoiesis mouse megakaryocyte differentiation and its application in identify- and is associated with a myeloproliferative phenotype. Blood 106: ing stage-specific molecular markers and a novel transcriptional 494–504. target of NF-E2. Blood 109: 1451–1459. Gu TL, Mercher T, Tyner JW, Goss VL, Walters DK, Cornejo MG Chung SW, Wolff L, Ruscetti SK. (1989). Transmembrane domain of et al. (2007). A novel fusion of RBM6 to CSF1R in acute the envelope gene of a polycythemia-inducing retrovirus determines megakaryoblastic leukemia. Blood 110: 323–333. erythropoietin-independent growth. Proc Natl Acad Sci USA 86: Hancock V, Martin JF, Lelchuk R. (1993). The relationship between 7957–7960. human megakaryocyte nuclear DNA content and gene expression. Cramer EM, Norol F, Guichard J, Breton-Gorius J, Vainchenker W, Br J Haematol 85: 692–697. Masse JM et al. (1997). Ultrastructure of platelet formation by Ichikawa M, Asai T, Saito T, Seo S, Yamazaki I, Yamagata T et al. human megakaryocytes cultured with the Mpl ligand. Blood 89: (2004). AML-1 is required for megakaryocytic maturation and 2336–2346. lymphocytic differentiation, but not for maintenance of hemato- Crispino JD. (2005). GATA1 in normal and malignant hematopoiesis. poietic stem cells in adult hematopoiesis. Nat Med 10: 299–304. Semin Cell Dev Biol 16: 137–147. Ito E, Kasai M, Hayashi Y, Toki T, Arai K, Yokoyama S et al. (1995). Drexler HG, Matsuo Y, MacLeod RA. (2004). Malignant hemato- Expression of erythroid-specific genes in acute megakaryoblastic poietic cell lines: in vitro models for the study of erythroleukemia. leukaemia and transient myeloproliferative disorder in Down’s Leuk Res 28: 1243–1251. syndrome. Br J Haematol 90: 607–614. Drissen R, von Lindern M, Kolbus A, Driegen S, Steinlein P, Beug H Izraeli S. (2006). Perspective: chromosomal aneuploidy in leukemia— et al. (2005). The erythroid phenotype of EKLF-null mice: defects in lessons from Down syndrome. Hematol Oncol 24: 3–6. hemoglobin metabolism and membrane stability. Mol Cell Biol 25: Jacobs-Helber SM, Roh KH, Bailey D, Dessypris EN, Ryan JJ, Chen J 5205–5214. et al. (2003). Tumor necrosis factor-a expressed constitutively in Du KL, Chen M, Li J, Lepore JJ, Mericko P, Parmacek MS. (2004). erythroid cells or induced by erythropoietin has negative and Megakaryoblastic leukemia factor-1 transduces cytoskeletal signals stimulatory roles in normal erythropoiesis and erythroleukemia. and induces smooth muscle cell differentiation from undifferen- Blood 101: 524–531. tiated embryonic stem cells. J Biol Chem 279: 17578–17586. Jaffe ES. (2001). Pathology and Genetics: Tumors of Haematopoietic Elmaagacli AH, Koldehoff M, Zakrzewski JL, Steckel NK, Ottinger H, and Lymphoid Tissues. IARC Press: Lyon. Beelen DW. (2007). Growth factor-independent 1B gene (GFI1B) is James C, Ugo V, Le Couedic JP, Staerk J, Delhommeau F, Lacout C overexpressed in erythropoietic and megakaryocytic malignancies and et al. (2005). A unique clonal JAK2 mutation leading to constitutive increases their proliferation rate. Br J Haematol 136: 212–219. signalling causes polycythaemia vera. Nature 434: 1144–1148. Emambokus N, Vegiopoulos A, Harman B, Jenkinson E, Anderson G, Jelinek J, Oki Y, Gharibyan V, Bueso-Ramos C, Prchal JT, Verstovsek Frampton J. (2003). Progression through key stages of haemopoiesis S et al. (2005). JAK2 mutation 1849G>T is rare in acute leukemias is dependent on distinct threshold levels of c-Myb. EMBO J 22: but can be found in CMML, Philadelphia chromosome-negative 4478–4488. CML, and megakaryocytic leukemia. Blood 106: 3370–3373. Fang C, Choi E, Nie L, Li JP. (1998). Role of the transmembrane Keller C, Hansen MS, Coffin CM, Capecchi MR. (2004). Pax3:Fkhr sequence of spleen focus-forming virus gp55 in erythroleukemogen- interferes with embryonic Pax3 and Pax7 function: implications esis. Virology 252: 46–53. for alveolar rhabdomyosarcoma cell of origin. Genes Dev 18: 2608– Freyssinier JM, Lecoq-Lafon C, Amsellem S, Picard F, Ducrocq R, 2613. Mayeux P et al. (1999). Purification, amplification and characteriza- Kirito K, Kaushansky K. (2006). Transcriptional regulation of tion of a population of human erythroid progenitors. Br J Haematol megakaryopoiesis: thrombopoietin signaling and nuclear factors. 106: 912–922. Curr Opin Hematol 13: 151–156.

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6813 Kirito K, Nagashima T, Ozawa K, Komatsu N. (2002). Constitutive McLean TW, Ringold S, Neuberg D, Stegmaier K, Tantravahi R, activation of Stat1 and Stat3 in primary erythroleukemia cells. Int J Ritz J et al. (1996). TEL/AML-1 dimerizes and is associated with Hematol 75: 51–54. a favorable outcome in childhood acute lymphoblastic leukemia. Kosmider O, Denis N, Lacout C, Vainchenker W, Dubreuil P, Blood 88: 4252–4258. Moreau-Gachelin F. (2005). Kit-activating mutations cooperate Mercher T, Coniat MB, Monni R, Mauchauffe M, Nguyen Khac F, with Spi-1/PU.1 overexpression to promote tumorigenic progression Gressin L et al. (2001). Involvement of a human gene related to during erythroleukemia in mice. Cancer Cell 8: 467–478. the Drosophila spen gene in the recurrent t(1;22) translocation of Koury MJ, Sawyer ST, Bondurant MC. (1984). Splenic erythroblasts acute megakaryocytic leukemia. Proc Natl Acad Sci USA 98: in anemia-inducing Friend disease: a source of cells for studies 5776–5779. of erythropoietin-mediated differentiation. J Cell Physiol 121: Michaud J, Wu F, Osato M, Cottles GM, Yanagida M, Asou N et al. 526–532. (2002). In vitro analyses of known and novel RUNX1/AML1 Koury MJ, Sawyer ST, Brandt SJ. (2002). New insights into mutations in dominant familial platelet disorder with predisposition erythropoiesis. Curr Opin Hematol 9: 93–100. to acute myelogenous leukemia: implications for mechanisms of Kralovics R, Passamonti F, Buser AS, Teo SS, Tiedt R, Passweg JR pathogenesis. Blood 99: 1364–1372. et al. (2005). A gain-of-function mutation of JAK2 in myeloproli- Michiels JJ, van der Meulen J, Brederoo P. (1997). The natural history ferative disorders. N Engl J Med 352: 1779–1790. of trilinear myelodysplastic syndrome and erythroleukemia. Hae- Kuhl C, Atzberger A, Iborra F, Nieswandt B, Porcher C, Vyas P. matologica 82: 452–454. (2005). GATA1-mediated megakaryocyte differentiation and Muntean AG, Crispino JD. (2005). Differential requirements for growth control can be uncoupled and mapped to different domains the activation domain and FOG-interaction surface of GATA-1 in GATA1. Mol Cell Biol 25: 8592–8606. in megakaryocyte gene expression and development. Blood 106: Lange B. (2000). The management of neoplastic disorders of 1223–1231. haematopoiesis in children with Down’s syndrome. Br J Haematol Muntean AG, Ge Y, Taub JW, Crispino JD. (2006). Transcription 110: 512–524. factor GATA-1 and Down syndrome leukemogenesis. Leuk Lange BJ, Kobrinsky N, Barnard DR, Arthur DC, Buckley JD, Lymphoma 47: 986–997. Howells WB et al. (1998). Distinctive demography, biology, and Muntean AG, Pang L, Poncz M, Dowdy SF, Blobel GA, Crispino JD. outcome of acute myeloid leukemia and myelodysplastic syndrome (2007). Cyclin D-Cdk4 is regulated by GATA-1 and required in children with Down syndrome: children’s cancer group studies for megakaryocyte growth and polyploidization. Blood 109: 2861 and 2891. Blood 91: 608–615. 5199–5207. Langebrake C, Klusmann JH, Wortmann K, Kolar M, Puhlmann U, Nelson ME, Steensma DP. (2006). JAK2 V617F in myeloid disorders: Reinhardt D. (2006). Concomitant aberrant overexpression of what do we know now, and where are we headed? Leuk Lymphoma RUNX1 and NCAM in regenerating bone marrow of myeloid 47: 177–194. leukemia of Down’s syndrome. Haematologica 91: 1473–1480. Ney PA, D’Andrea AD. (2000). Friend erythroleukemia revisited. Lannutti BJ, Minear J, Blake N, Drachman JG. (2006). Increased mega- Blood 96: 3675–3680. karyocytopoiesis in Lyn-deficient mice. Oncogene 25: 3316–3324. Nucifora G, Rowley JD. (1995). AML1 and the 8;21 and 3;21 Lannutti BJ, Shim MH, Blake N, Reems JA, Drachman JG. (2003). translocations in acute and chronic myeloid leukemia. Blood 86: Identification and activation of Src family kinases in primary 1–14. megakaryocytes. Exp Hematol 31: 1268–1274. Oki Y, Kantarjian HM, Zhou X, Cortes J, Faderl S, Verstovsek S Levine RL, Wadleigh M, Cools J, Ebert BL, Wernig G, Huntly BJ et al. (2006). Adult acute megakaryocytic leukemia: an analysis et al. (2005). Activating mutation in the tyrosine kinase JAK2 in of 37 patients treated at MD Anderson cancer center. Blood 107: polycythemia vera, essential thrombocythemia, and myeloid meta- 880–884. plasia with myelofibrosis. Cancer Cell 7: 387–397. Osato M, Asou N, Abdalla E, Hoshino K, Yamasaki H, Okubo T Li Z, Godinho FJ, Klusmann JH, Garriga-Canut M, Yu C, Orkin SH. et al. (1999). Biallelic and heterozygous point mutations in the runt (2005). Developmental stage-selective effect of somatically muta- domain of the AML1 gene associated with myeloblastic leukemias. ted leukemogenic transcription factor GATA1. Nat Genet 37: Blood 93: 1817–1824. 613–619. Osato M, Ito Y. (2005). Increased dosage of the RUNX1/AML1 gene: Lu SJ, Rowan S, Bani MR, Ben-David Y. (1994). Retroviral a third mode of RUNX leukemia. Crit Rev Eukaryot Gene Expr 15: integration within the Fli-2 locus results in inactivation of the 217–228. erythroid transcription factor NF-E2 in Friend erythroleukemias: Osawa M, Yamaguchi T, Nakamura Y, Kaneko S, Onodera M, evidence that NF-E2 is essential for globin expression. Proc Natl Sawada K et al. (2002). Erythroid expansion mediated by the Acad Sci USA 91: 8398–8402. Gfi-1B zinc finger protein: role in normal hematopoiesis. Blood 100: Lulli V, Romania P, Morsilli O, Gabbianelli M, Pagliuca A, Mazzeo S 2769–2777. et al. (2006). Overexpression of Ets-1 in human hematopoietic Oswald F, Kostezka U, Astrahantseff K, Bourteele S, Dillinger K, progenitor cells blocks erythroid and promotes megakaryocytic Zechner U et al. (2002). SHARP is a novel component of the Notch/ differentiation. Cell Death Differ 13: 1064–1074. RBP-Jk signalling pathway. EMBO J 21: 5417–5426. Ma X, Renda MJ, Wang L, Cheng EC, Niu C, Morris SW et al. Pagano L, Pulsoni A, Vignetti M, Mele L, Fianchi L, Petti MC et al. (2007). Rbm15 modulates notch-induced transcriptional activation (2002). Acute megakaryoblastic leukemia: experience of GIMEMA and affects myeloid differentiation. Mol Cell Biol 27: 3056–3064. trials. Leukemia 16: 1622–1626. Ma Z, Morris SW, Valentine V, Li M, Herbrick JA, Cui X et al. Pang L, Xue HH, Szalai G, Wang X, Wang Y, Watson DK et al. (2001a). Fusion of two novel genes, RBM15 and MKL1, in the (2006). Maturation stage-specific regulation of megakaryopoiesis by t(1;22)(p13;q13) of acute megakaryoblastic leukemia. Nat Genet pointed-domain Ets proteins. Blood 108: 2198–2206. 28: 220–221. Pardanani AD, Levine RL, Lasho T, Pikman Y, Mesa RA, Wadleigh Ma Z, Morris SW, Valentine V, Li M, Herbrick JA, Cui X et al. M et al. (2006). MPL515 mutations in myeloproliferative and (2001b). Fusion of two novel genes, RBM15 and MKL1, in the other myeloid disorders: a study of 1182 patients. Blood 108: t(1;22)(p13;q13) of acute megakaryoblastic leukemia. Nat Genet 3472–3476. 28: 220–221. Park S, Picard F, Dreyfus F. (2002). Erythroleukemia: a need for a Mahmud DL, M GA, Deb DK, Platanias LC, Uddin S, Wickrema A. new definition. Leukemia 16: 1399–1401. (2002). Phosphorylation of forkhead transcription factors by Peterson LF, Zhang DE. (2004). The 8;21 translocation in leukemo- erythropoietin and stem cell factor prevents acetylation and their genesis. Oncogene 23: 4255–4262. interaction with coactivator p300 in erythroid progenitor cells. Phillips JD, Steensma DP, Pulsipher MA, Spangrude GJ, Kushner JP. Oncogene 21: 1556–1562. (2007). Congenital erythropoietic porphyria due to a mutation in

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6814 GATA1: the first trans-acting mutation causative for a human Sun W, Downing JR. (2004). Haploinsufficiency of AML1 results in a porphyria. Blood 109: 2618–2621. decrease in the number of LTR-HSCs while simultaneously Pikman Y, Lee BH, Mercher T, McDowell E, Ebert BL, inducing an increase in more mature progenitors. Blood 104: Gozo M et al. (2006). MPLW515L is a novel somatic activating 3565–3572. mutation in myelofibrosis with myeloid metaplasia. PLoS Med 3: Szalai G, LaRue AC, Watson DK. (2006). Molecular mechanisms of e270. megakaryopoiesis. Cell Mol Life Sci 63: 2460–2476. Preudhomme C, Warot-Loze D, Roumier C, Grardel-Duflos N, Tallman MS, Neuberg D, Bennett JM, Francois CJ, Paietta E, Garand R, Lai JL et al. (2000). High incidence of biallelic point Wiernik PH et al. (2000). Acute megakaryocytic leukemia: the mutations in the Runt domain of the AML1/PEBP2aB gene in Mo Eastern cooperative oncology group experience. Blood 96: acute myeloid leukemia and in myeloid malignancies with acquired 2405–2411. trisomy 21. Blood 96: 2862–2869. Teal HE, Ni S, Xu J, Finkelstein LD, Cheng AM, Paulson RF et al. Putz G, Rosner A, Nuesslein I, Schmitz N, Buchholz F. (2006). AML1 (2006). GRB2-mediated recruitment of GAB2, but not GAB1, to deletion in adult mice causes splenomegaly and lymphomas. SF-STK supports the expansion of Friend virus-infected erythroid Oncogene 25: 929–939. progenitor cells. Oncogene 25: 2433–2443. Raffel GD, Mercher T, Shigematsu H, Williams IR, Cullen DE, Tilbrook PA, Palmer GA, Bittorf T, McCarthy DJ, Wright MJ, Sarna Akashi K et al. (2007). Ott1(Rbm15) has pleiotropic roles in MK et al. (2001). Maturation of erythroid cells and erythroleukemia hematopoietic development. Proc Natl Acad Sci USA 104: development are affected by the kinase activity of Lyn. Cancer Res 6001–6006. 61: 2453–2458. Rainis L, Toki T, Pimanda JE, Rosenthal E, Machol K, Strehl S et al. Tsang AP, Fujiwara Y, Hom DB, Orkin SH. (1998). Failure of (2005). The proto-oncogene ERG in megakaryoblastic leukemias. megakaryopoiesis and arrested erythropoiesis in mice lacking Cancer Res 65: 7596–7602. the GATA-1 transcriptional cofactor FOG. Genes Dev 12: Raslova H, Kauffmann A, Sekkai D, Ripoche H, Larbret F, 1176–1188. Robert T et al. (2006). Interrelation between polyploidization and Tubman VN, Levine JE, Campagna DR, Monahan-Earley R, megakaryocyte differentiation: a gene profiling approach. Blood Dvorak AM, Neufeld EJ et al. (2007). X-linked gray platelet 109: 3225–3234. syndrome due to a GATA1 Arg216Gln mutation. Blood 109: Raslova H, Roy L, Vourc’h C, Le Couedic JP, Brison O, 3297–3299. Metivier D et al. (2003). Megakaryocyte polyploidization Uddin S, Ah-Kang J, Ulaszek J, Mahmud D, Wickrema A. (2004). is associated with a functional gene amplification. Blood 101: Differentiation stage-specific activation of p38 mitogen-activated 541–544. protein kinase isoforms in primary human erythroid cells. Proc Natl Ravid K, Lu J, Zimmet JM, Jones MR. (2002). Roads to polyploidy: Acad Sci USA 101: 147–152. the megakaryocyte example. J Cell Physiol 190: 7–20. Vannucchi AM, Bianchi L, Cellai C, Paoletti F, Rana RA, Lorenzini Rietveld LE, Caldenhoven E, Stunnenberg HG. (2001). Avian R et al. (2002). Development of myelofibrosis in mice genetically erythroleukemia: a model for corepressor function in cancer. impaired for GATA-1 expression (GATA-1(low) mice). Blood 100: Oncogene 20: 3100–3109. 1123–1132. Rosenbauer F, Tenen DG. (2007). Transcription factors in myeloid Vyas P, Ault K, Jackson CW, Orkin SH, Shivdasani RA. (1999). development: balancing differentiation with transformation. Nat Consequences of GATA-1 deficiency in megakaryocytes and Rev Immunol 7: 105–117. platelets. Blood 93: 2867–2875. Rubnitz JE, Razzouk BI, Lensing S, Pounds S, Pui CH, Ribeiro RC. Vyas P, Crispino JD. (2007). Molecular insights into Down syndrome- (2007). Prognostic factors and outcome of recurrence in childhood associated leukemia. Curr Opin Pediatr 19: 9–14. acute myeloid leukemia. Cancer 109: 157–163. Vyas P, Roberts I. (2006). Down myeloid disorders: a paradigm for Saleque S, Cameron S, Orkin SH. (2002). The zinc-finger proto- childhood preleukaemia and leukaemia and insights into normal oncogene Gfi-1b is essential for development of the erythroid and megakaryopoiesis. Early Hum Dev 82: 767–773. megakaryocytic lineages. Genes Dev 16: 301–306. Walters DK, Mercher T, Gu TL, O’Hare T, Tyner JW, Loriaux M Samarut J, Gazzolo L. (1982). Target cells infected by avian et al. (2006). Activating alleles of JAK3 in acute megakaryoblastic erythroblastosis virus differentiate and become transformed. Cell leukemia. Cancer Cell 10: 65–75. 28: 921–929. Wechsler J, Greene M, McDevitt MA, Anastasi J, Karp JE, Shivdasani RA, Fujiwara Y, McDevitt MA, Orkin SH. (1997). A Le Beau MM et al. (2002). Acquired mutations in GATA1 in lineage-selective knockout establishes the critical role of transcrip- the megakaryoblastic leukemia of Down syndrome. Nat Genet tion factor GATA-1 in megakaryocyte growth and platelet 32: 148–152. development. EMBO J 16: 3965–3973. Weiss MJ, Orkin SH. (1995). Transcription factor GATA-1 permits Shivdasani RA, Rosenblatt MF, Zucker-Franklin D, Jackson CW, survival and maturation of erythroid precursors by preventing Hunt P, Saris CJ et al. (1995). Transcription factor NF-E2 is apoptosis. Proc Natl Acad Sci USA 92: 9623–9627. required for platelet formation independent of the actions of Wickrema A, Bondurant MC, Krantz SB. (1991). Abundance and thrombopoietin/MGDF in megakaryocyte development. Cell stability of erythropoietin receptor mRNA in mouse erythroid 81: 695–704. progenitor cells. Blood 78: 2269–2275. Song WJ, Sullivan MG, Legare RD, Hutchings S, Tan X, Kufrin D Wickrema A, Krantz SB, Winkelmann JC, Bondurant MC. (1992). et al. (1999). Haploinsufficiency of CBFA2 causes familial Differentiation and erythropoietin receptor gene expression in thrombocytopenia with propensity to develop acute myelogenous human erythroid progenitor cells. Blood 80: 1940–1949. leukaemia. Nat Genet 23: 166–175. Wickrema A, Uddin S, Sharma A, Chen F, Alsayed Y, Ahmad S et al. Sotirov N. (1981). Histone H5 in the immature blood cells of chickens (1999). Engagement of Gab1 and Gab2 in erythropoietin signaling. with leukosis induced by avian leukosis virus strain E26. J Natl J Biol Chem 274: 24469–24474. Cancer Inst 66: 1143–1149. Wiellette EL, Harding KW, Mace KA, Ronshaugen MR, Wang FY, Steensma DP, McClure RF, Karp JE, Tefferi A, Lasho TL, Powell HL McGinnis W. (1999). spen encodes an RNP motif protein et al. (2006a). JAK2 V617F is a rare finding in de novo acute myeloid that interacts with Hox pathways to repress the development of leukemia, but STAT3 activation is common and remains unex- head-like sclerites in the Drosophila trunk. Development 126: plained. Leukemia 20: 971–978. 5373–5385. Steensma DP, Pardanani A, Stevenson WS, Hoyt R, Kiu H, Grigg AP Xu Z, Huang S, Chang LS, Agulnick AD, Brandt SJ. (2003). et al. (2006b). More on Myb in myelofibrosis: molecular analyses of Identification of a TAL1 target gene reveals a positive role for the MYB and EP300 in 55 patients with myeloproliferative disorders. LIM domain-binding protein Ldb1 in erythroid gene expression and Blood 107: 1733–1735. differentiation. Mol Cell Biol 23: 7585–7599.

Oncogene Erythroid and megakaryocytic malignancies A Wickrema and JD Crispino 6815 Yamashita N, Osato M, Huang L, Yanagida M, Kogan SC, Iwasaki Zhang MY, Sun SC, Bell L, Miller BA. (1998). NF-kB transcription M et al. (2005). Haploinsufﬁciency of Runx1/AML1 promotes factors are involved in normal erythropoiesis. Blood 91: myeloid features and leukaemogenesis in BXH2 mice. Br J 4136–4144. Haematol 131: 495–507. Zhao R, Xing S, Li Z, Fu X, Li Q, Krantz SB et al. (2005). Yanagida M, Osato M, Yamashita N, Liqun H, Jacob B, Wu F et al. Identiﬁcation of an acquired JAK2 mutation in polycythemia vera. (2005). Increased dosage of Runx1/AML1 acts as a positive J Biol Chem 280: 22788–22792. modulator of myeloid leukemogenesis in BXH2 mice. Oncogene Zipursky A. (2003). Transient leukaemia—a benign form of leukaemia 24: 4477–4485. in newborn infants with trisomy 21. Br J Haematol 120: 930–938.

Oncogene