Mechanism of Leukemogenesis by the Inv(16) Chimeric Gene CBFB/PEBP2B-MHY11

Mechanism of leukemogenesis by the inv(16) chimeric gene CBFB/PEBP2B-MHY11

Katsuya Shigesada*,1, Bart van de Sluis2 and P Paul Liu*,2

1Department of Cell Biology, Institute for Virus Research, Kyoto University, Kyoto 606-8507, Japan; 2Oncogenesis and Development Section, National Human Genome Research Institute, NIH, Bethesda, MD 20892, USA

Inv(16)(p13q22) is associated with acute myeloid leukemia subsequent to which a certain mechanism(s) should act subtype M4Eo that is characterized by the presence of to bring Runx1 into a functionally incompetent state in myelomonocytic blasts and atypical eosinophils. This the end (repression).If one of these steps were lacking or chromosomal rearrangement results in the fusion of CBFB impaired, CBFb-SMMHC would fail to elicit any and MYH11 genes. CBFb normally interacts with RUNX1 substantial dominant-negative effect regardless of how to form a transcriptionally active nuclear complex. well the other step might work. The MYH11 gene encodes the smooth muscle myosin Until recently, however, most functional studies of heavy chain. The CBFb-SMMHC fusion protein is CBFb-SMMHC have centered around the mechanism capable of binding to RUNX1 and form dimers and of repression, paying relatively little attention to the first multimers through its myosin tail. Previous results from heterodimerization step.Nevertheless, evidence sugges- transgenic mouse models show that Cbfb-MYH11 is able tive of its enhanced ability for heterodimerization to inhibit dominantly Runx1 function in hematopoiesis, (hyper-heterodimerization) has come from our previous and is a key player in the pathogenesis of leukemia. In finding that CBFb-SMMHC is predominantly localized recent years, molecular and cellular biological studies in the cytoplasm in association with the actin cytoske- have led to the proposal of several models to explain the leton, and simultaneously capable of sequestering function of CBFb-SMMHC. In this review, we will first RUNX proteins to the cytoplasm in a manner over- focus our attention on the molecular mechanisms proposed riding the intrinsic or artificially modulated ability of in the recent publications. We will next examine recent Runx to localize in the nucleus (Lu et al., 1995; Adya gene expression profiling studies on inv(16) leukemia cells. et al., 1998; Kanno et al., 1998). In sharp contrast, Cbfb Finally, we will describe a recent study from one of our tends to localize in the cytoplasm in a diffuse pattern by labs on the identification of cooperating genes for itself, and can be partly, although not completely, leukemogenesis with CBFB-MYH11. translocated into the nucleus only through heterodimer- Oncogene (2004) 23, 4297–4307. doi:10.1038/sj.onc.1207748 ization with Runx protein.Furthermore, CBF b- SMMHC can stabilize RUNX1 against intracellular Keywords: CBFb-MYH11; CBFb-SMMHC; inv(16); proteasome-mediated degradation much more strongly leukemia; hematopoiesis than Cbfb (Huang et al., 2001a). To characterize the hyper-heterodimerization activity of CBFb-SMMHC more directly, we conducted extensive functional analyses using a series of CBFb- SMMHC deletions as truncated exon by exon either C- Molecular basis for the dominant inhibition of terminally or internally (Huang et al., 2003). Through in RUNX1-dependent transcription by CBFb-SMMHC vitro binding experiments by means of co-immunopre- cipitation and GST-pulldown assays as well as an Identification of dual functional domains in intracellular hyperprotection assay, it was confirmed CBFb-SMMHC that CBFb-SMMHC can heterodimerize with RUNX1 at an affinity higher than that of CBFb by one order of It has been shown that the phenotype of heterozygous magnitude or more.Parallel analyses with RUNX1 Cbfb-MYH11 knockin mice is very similar to that of deletions revealed that the region of RUNX1 required Runx1 knockout mice (Castilla et al., 1996). The results for hyper-heterodimerization is the Runt domain. imply that CBFb-SMMHC inactivates the function of Further, a minimum region of the myosin tail respon- Runx1 nearly completely despite the presence of a sible for hyper-heterodimerization was further mapped residual normal Cbfb/PEBP2b.In order for this to to exons 33–36 (residues 166–363).The myosin tail occur, CBFb-SMMHC would have to outcompete Cbfb alone, of course, did not show any detectable binding to at the step of heterodimerization with Runx1 first of all, the Runt domain.However, appreciable hetrodimeriza- tion activities became detectable when the myosin tail *Correspondence: K Shigesada; E-mail: [email protected]; was fused to PEBP2b proteins made heterodimerization- P Liu; E-mail: [email protected] defective due to double point mutations (residues 64 and Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4298 104) (Tang et al., 2000) or a short N-terminal truncation (D2-11) (Adya et al., 1998). Presumably, the myosin tail and a part of the PEBP2b protein may physically or conformationally cooperate to create a new RUNX- binding interface that functions independent of and in synergy with the original heterodimerization interface on CBFb.A minimum region of PEBP2 b involved in this second putative heterodimerization interface was narrowed down to residues 134–165. We next investigated whether and how much the hyper-heterodimerization domain could contribute to the dominant repression of RUNX1-mediated transcription by CBFb-SMMHC against PEBP2b.Upon a transcription assay using an M-CSFR promoter-based reporter system, coexpressions of CBFb-SMMHC and Figure 1 Diagrammatic summary of functional domain analyses. PEBP2b at equimolar nonsaturating doses resulted in a The top diagram represents CBFb/PEBP2b-MYH11.Red-colored segments (bE1–bE5) and blue-colored segments (E33–42) represent strong repression of transcription to a level several-fold exons encoding Cbfb/PEBP2b and SMMHC, respectively.Double- less than the control obtained by transfection of headed arrows above and below the diagram indicate the mapped PEBP2b alone.When the hyperdimerization domain locations of respective functional and structural domains as (exons 33–36) was removed from CBFb-SMMHC by annotated.ACD and ACD2: assembly competence domains internal deletion, this repression was considerably identified by Sohn et al.(1997) and Ikebe et al.(2001), respectively. The dotted red line represents an intermediate zone that was weakened to twofold or less.This confirmed that the suggested to be unpaired upon deletion of the C-terminally flanking hyperdimerization domain does have an important coiled coil region, but could be integrated at least partly into the positive impact in augmenting repression.Curiously, coiled coil domain in the intact CBFb/PEBP2b -MYH11.The blue however, a converse deletion construct retaining the double-headed arrow represents the terminal nonhelical tail that is hyper-heterodimerization domain but lacking the rest of implicated in the regulation of multimerization (Ikebe et al., 2001) the C-terminal region caused a moderate stimulation of transcription, rather than repression.A simple explanation for these seemingly paradoxical effects of the hyper- (Huang et al., 2003). In EMSA experiments, the intact heterodimerization domain may be that the hyperdi- CBFb-SMMHC mixed with RUNX1 forms large DNA- merization domain itself has no or little inhibitory bound complexes that were unable to penetrate into a influence on RUNX1-mediated transactivation, and polyacrylamide gel.When CBF b-SMMHC was deleted that another functional domain responsible for repres- from the C-terminus past exon 39 and further beyond, sion (repression domain) resides within the C-terminal the resulting DNA–protein complexes started to migrate proximal region centering around exons 39–40.In into the gel in increasing fractions at accelerated apparent coincidence with this explanation, other mobilities.Judged from their mobilities, these complexes groups previously reported that C-terminal segments represented heterodimers consisting of single RUNX1 of SMMHC overlapping exons 39–42 could show and CBFb-SMMHC molecules.On the other hand, repressive activities when artificially fused to the Gal4 internally truncated constructs lacking the hyperdimer- DNA-binding domain and assayed using a Gal4-TK luc ization domain with or without additional C-terminally reporter system (Lutterbach et al., 1999). With this extending deletions up to exon 38 still produced low- bipartite functional domain model (Figure 1), we can mobility complexes in much higher proportions than did readily predict that CBFb-SMMHC deleted by the pure C-terminal deletion constructs.Chemical cross- hyperdimerization should no longer be able to compete linking experiments using glutaraldehyde also confirmed dominantly with the normal CBFb and hence would fail that CBFb-SMMHC C-terminally truncated beyond to display its maximal possible repression potential. exon 39 tended to dissociate into monomers in a manner proportionate to their degrees of deletion.Previous glutaraldehyde crosslinking studies with a fewer vari- Molecular conformation and higher order assembly eties of C-terminal deletions also indicated similar of CBFb-SMMHC trends (Adya et al., 1998; Cao et al., 1998). In contrast, Ever since the discovery of CBFb-SMMHC (Liu et al., the above-noted internal deletions largely remained as 1993, 1995), it has implicitly been assumed that the dimers and multimers. myosin tail is able to associate into a coiled-coil On the basis of these results, we proposed a new homodimer, which in turn extensively multimerizes into structural model for CBFb-SMMHC that comprises a filaments.Furthermore, the multimerization process is Y-shaped dimer with unpaired N-terminal halves known to depend on the assembly competence domain followed by a coiled-coil homodimer in the C-terminal (ACD) located within exon 40 (Sohn et al., 1997; Ikebe region (Huang et al., 2003) (Figure 2). The hyperdimer- et al., 2001; Kummalue et al., 2002). During the ization domain and the repression domain coinciden- preceding functional analyses, however, we came upon tally fall within the unpaired region and the coiled-coil evidence suggesting that the coiled-coil rod is partially region, respectively.It may be puzzling why the hyper- unpaired around the hyper-heterodimerization domain heterodimerization domain requires such a long stretch

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4299

Figure 3 Multifarious mechanisms of transdominant repression by CBFb/PEBP2b-MYH11.A diagrammatic scheme of how Figure 2 A new bipartite model for the structure of CBFb/ CBFb/PEBP2b-MYH11 functions in the cell to cause a transdo- PEBP2b -MYH11 complexed with RUNX1.( a) A model pre- minant repression of RUNX1-promoted transcription in competi- viously proposed by Lukasik et al.(2002).( b) A new model tion with CBFb/PEBP2b through various putative mechanisms as proposed by Huang et al.(2003).Indicated are the RUNX1: CBF b/ described in the section Molecular basis for the dominant PEBP2b -MYH11 complex at three different stages of molecular inhibition of RUNX1-dependent transcription by CBFb-SMMHC assembly.The filamentous multimer is modeled after the side-polar is shown.Dotted double line, nuclear envelope; blue-colored structure proposed for smooth muscle heavy chain by Xu et al. crescent, corepressor; thick winding line, chromosomal DNA (1996).The red-colored globular pieces and the gray-colored ovals entrapped by the filamentous CBFb/PEBP2b-MYH11 multimer; represent RUNX1 and CBFb/PEBP2b, respectively.In the and TGYGGT, the consensus RUNX-binding site.The other filamentous multimer, the dimeric unit is depicted in a simplified symbols are the same as in Figure 2 form.NES: presumptive nuclear export signal, ACD: assembly competence domain, and NHT: nonhelical tail.See the text for other details domain.NMR studies identified interactions with Runx1 in the CBFb portion of the molecule, as well of the myosin tail reaching nearly 200-aa for interaction as the SMMHC portion.Moreover, they described with the Runt domain known to have a compact that the enhanced heterodimerization of b/SMMHC47 globular structure.We think that the hyper-heterodi- was hindered in the presence of DNA, which disagrees merization domain takes on a somewhat folded struc- with our current results (Huang et al., 2003). The ture, part of which makes direct contacts with the Runt discrepancy is probably attributable to differences in domain while the rest serve to keep the hyper-hetero- the analytical methods and experimental conditions dimerization domain in a proper conformation.Here, a used.Perhaps one critical factor is the concentration question may immediately arise as to how the coiled-coil of proteins used.In our in vivo and in vitro binding structure could be so fragile.In fact, the coiled-coiled assays, the Runt domain and b/SMMHC were roughly rod of SMMHC, unlike that of skeletal counterparts, is estimated to be present in nanomolar ranges.On the notoriously flexible and unstable within physiological other hand, Lukasik et al.employed isothermal titration salt concentrations, readily transforms into a folded calorimetry (ITC) as a chief analytical tool using very hairpin structure (Trybus and Lowey, 1984) or dis- high protein concentrations in the range of 20–340 mM. sociates into a monomer upon truncation (Trybus et al., Their ITC measurements yielded a stoichiometry 1997).For one illustrative example, truncated coiled-coil indicating that two Runt domain proteins bind to each tails of SMMHC containing more than 100 residues b/SMMHC47.This unusual stoichiometry might reflect were shown to be less stable than a 28-aa GCN4-derived the homodimeric interaction of the Runt domain as leucine zipper (Trybus et al., 1997). Moreover, Liu et al. previously identified by X-ray crystallographic analysis (1994) demonstrated early on that a substantial part of of the Runt domain–CBFb binary complex (Warren DNA–protein complexes produced in the presence of et al., 2000). Of note, this homodimeric interface seems RUNX1 and in vitro-translated CBFb-SMMHC mi- to be unrecognizable in the Runt domain-CBFb-DNA grated into polyacrylamide gels in gel-shift assays, ternary complex in subsequent X-ray crystallographic suggesting that even the coiled coil of intact CBFb- studies reported by the same authors’ group (Bravo SMMHC could substantially dissociate into monomers et al., 2001) as well as a few other groups (Tahirov et al., at a sufficiently low concentration. 2001; Backstrom et al., 2002). If these observations are Recently, Lukasik et al.(2002) reported that a taken into account, it is conceivable that an extra truncated CBFb-SMMHC with a 47-amino acid portion molecule of the Runt domain might well bind to each of SMMHC ending within exon 33 was sufficient to heterodimeric unit of the Runt domain and CBFb- support not only its homodimerization into a coiled-coil SMMHC in the absence of DNA, when the Runt but also its enhanced heterodimerization with the Runt domain is present in large molar excess at a sufficiently

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4300 high concentration.Apart from such mechanistic de- Given the view reached above, it becomes more tails, it is open to question whether the 2 : 1 stoichio- important to know how and with what efficiency each metry complex could be formed under in vivo conditions mode of sequestration is working in various cells, to any meaningful extent. particularly inv(16) leukemic cells.The relative distributions of CBFb-SMMHC/RUNX1 complexes between Mechanisms of repression of RUNX1-mediated the nucleus and the cytoplasm were shown to depend on transcription by CBFb-SMMHC the level of expression of CBFb-SMMHC and also the kind of cells studied.In NIH 3T3 cells, CBF b-SMMHC/ Early in the mid-1990s, Liu et al.(1994, 1995) first RUNX1 complexes were predominantly localized in the proposed two alternative basic mechanisms by which cytoplasm at a low expression level, but tended to move CBFb-SMMHC could inhibit Runx1-mediated transac- into the nucleus in increasing proportions with increas- tivation.In one mechanism, the formation of multimeric ing expression levels (Adya et al., 1998). Of note, these aggregates by CBFb-SMMHC would sequester a certain changes were accompanied by progressive deformations amount of RUNX1, making it unavailable to bind its of the cytoskeletal structure into scattered speckles. targets in the DNA (a sequestration model).The second Thus, it seems likely that the cytoskeletal actin filaments possible mechanism is that the RUNX1/CBFb- function to accommodate CBFb-SMMHC as a reservoir SMMHC complex could alter the assembly of se- that is prone to disintegrate upon saturation.In an quence-specific transcription factors on adjacent sites apparent coincidence with this speculation, ectopically in the enhancers of certain target genes by either causing overexpressed CBFb-SMMHC was preferentially loca- steric hindrance or participating in novel interactions lized in the nucleus in a lymphoid cell line, BaF3, that with other proteins (an interference model).More retains a minimal cytoplasmic space with supposedly recently, Lutterbach et al.(1999) proposed a specialized meager cytoskeletal actins (Cao et al., 1998; Kummalue variation of the second mechanism asserting that et al., 2002). In this cell line, the nuclear accumulation of b/SMMHC actively represses in the nucleus through CBFb-SMMHC and also the resultant inhibition of its ability to interact with various corepressor proteins RUNX1’s activities were both shown to be tightly (a corepressor-recruiting model).The results of our correlated with its ACD-dependent multimerization preceding functional analyses would provide fresh (Kummalue et al., 2002). The observed correlation angles to re-evaluate the relative importance of these between the nuclear accumulation of CBFb-SMMHC apparently conflicting mechanisms (Figure 3). and the ACD closely echoes with the augmented nuclear The sequestration model subsequently gained impetus export of the ACD-less mutant, b/SMMHCD95, as by the observations that CBFb-SMMHC can cause noted before.Accordingly, the results of Kummalue altered subcellular localizations of Runx1, and also et al.(2002) may well be taken as a case of enhanced Runx2, in a state closely colocalized with CBFb- intranuclear sequestration (see below for another SMMHC either in the nucleus as multimerized com- different interpretation).In this sequestration-centered plexes, which sometimes form large rod-like inclusion view, the augmentary effect of the ACD on the nuclear bodies (intranuclear sequestration) (Wijmenga et al., import of CBFb-SMMHC, per se, could be an indirect 1996), or in the cytoplasm as deposits on cytoskeletal consequence of multimerization.Judging from its filaments or aggregates (cytoplasmic sequestration) (Lu enhancednuclearexportasquotedbefore,b/SMMHCD95 et al., 1995; Adya et al., 1998). In the former case, might well harbor an NES-like sequence that would be RUNX proteins physically confined within those multi- sterically masked in the intact CBFb-SMMHC under- meric complexes only have limited access to cognate- going extensive ACD-dependent multimerization. binding sites on chromosomes, albeit they are in the Lastly, what is the situation in leukemic cells with nucleus.In addition to these mechanisms, a still inv(16)? Evidence reported in a few previous reports different type of sequestration without involving multi- consistently suggested approximately equal distributions merization was found to occur with CBFb-SMMHC between the nucleus and the cytoplasm (Liu et al., 1996; deleted from C-terminal 95 amino acids containing the Kanto et al., 2000). ACD motif (b/SMMHCD95) (Adya et al., 1998). b/ The corepressor-recruiting model was first suggested SMMHCD95 was exclusively localized together with by Lutterbach et al.(1999) based on their findings that RUNX1 in the cytoplasm in a diffuse pattern with the b/SMMHC formed ternary complexes with RUNX1 nucleus left as a complete void, suggesting that this and repressor protein mSin3A, and also that the C- protein could be actively exported out of the nucleus terminal 163-amino acid region of the myosin tail possibly by an NES-dependent mechanism (see further (approximately corresponding to exons 38–42) acted as discussion below).In our analysis, a C-terminal deletion a transcriptional repressor when it was fused to the Gal4 lacking exons 40–42, which is nearly equivalent to b/ protein and assayed using a Gal4-TK reporter system. SMMHCD95, displayed a strong repression (fourfold), Subsequently, Durst et al.(2003) demonstrated that the only slightly less than that of the intact CBFb-SMMHC putative repressor domain interacts with a corepressor (sixfold) (Huang et al., 2003). Taken together, these mSin3A and histone deacetylase 8.mSin3A itself was observations indicate that the sequestration mechanism further shown to interact with HDAC 1 and 2.As an alone can support efficient repression regardless of additional support to this model, the b/SMMHC- which mechanism among the three variations noted mediated repression in the Gal4-TK reporter system above is at work. was impaired by an HDAC inhibitor, TSA.Quite

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4301 interestingly, the ACD was implicated to play critical inhibition is not always complete.Ectopic expression of roles here again: its internal deletion resulted in CBFB-MYH11 in ES cells retaining one or two normal concurrent impairments of the repression by b/SMMHC Cbfb alleles did not inhibit definitive hematopoiesis in an as well as its ability to associate with both mSin3A and in vitro colony-forming assay (Miller et al., 2001). Ectopic HDAC8.Taken altogether, these observations have expression of CBFB-MYH11 in 32Dcl3 premyeloid cells dual rather unexpected implications.Contrary to our did not inhibit their G-CSF-induced differentiation to preceding interpretation focusing on the sequestration neutrophils nor their expression of endogenous myelo- model, RUNX1 bound by multimerized RUNX1–b/ peroxidase whose promoter has been known as RUNX1 SMMHC complexes could make productive interactions dependent, although their growth was slowed at the step with DNA carrying cognate-binding sites unless of G1 to S cell cycle transition (Cao et al.,1997). HDACs are allowed to intervene.In addition, multi- Similarly, induction of CBFB-MYH11 in Ba/F3 lym- merization of the coiled-coil domain could facilitate, phoid cells did not inhibit, but rather stimulated, rather than hinder, its association with repressor expression of mRNA from the endogenous p21WAF1 proteins.How can these events happen? Previous promoter (Cao et al., 1997), whereas the same promoter electron microscopic studies have shown that SMMHC, on a plasmid vector was contrarily inhibited (Lutterbach in its native state, has a side-polar structure in which the et al., 1999). Moreover, dominant-negative Runx2 may coiled-coil myosin tail is first dimerized antiparallelly cooperate with CBFb-SMMHC for leukemogenesis in and then the resultant dimeric units are consecutively our mouse model (see details in the section Genes stuck on their lateral sides in a manner staggered to each cooperating with CBFB-MYH11 for the pathogenesis other, thereby forming an elongated single-layered thin of AML).One potential explanation for these observed strip that is laced with paired myosin head pieces on results is chromosomal context, whether a target gene is both sides (Xu et al., 1996). Within this molecular on the cellular chromosome or on plasmids may make assemblage, the coiled-coil domain is supposed to have difference.Coincidentally, all the studies evaluating the two exposed faces on the top and bottom of the plane of inhibitory potential of CBFb-SMMHC referred to in the the strip.CBF b-SMMHC multimers associated with section Mechanisms of repression of RUNX1-mediated RUNX1 are also likely to adopt the same side-polar transcription by CBFb-SMMHC employed plasmid- structure, and hence could interact relatively freely with borne reporter systems.It is tempting to speculate that any ligand molecule coming into contact, either DNA or highly multimerized RUNX1 : CBFb-SMMHC com- protein.Thus, the ACD-mediated multimerization plexes for their large size might kinetically and probabil- could possibly play a more active role in supporting istically be able to enjoy only infrequent contacts with the CBFb-SMMHC-mediated repression by converting individual endogenous target genes embedded within RUNX1 into a dedicated transcriptional repressor, giant chromosomal bodies.On the other hand, the same rather than simply keeping the heteromeric assemblage multimeric complexes could readily capture more dif- in a functionally inert state.Since the ACD domain is fused plasmid-borne targets. required for multimerization, it is expected that b/ Nevertheless, the RUNX1 : CBFB-MYH11 multimer SMMHCD95 binds with corepressors with less affinity might have potentially hazardous consequences for and represses transcription less efficiently, even though cellular gene regulation in a more global context.Since it sequesters Runx1 with higher efficiency, as described the multimer has a dense array of RUNX1 attached on before.It is also worth noting that the HDAC- its elongated filamentous backbone, there would be a dependency of the CBFb-SMMHC-mediated repression high probability for its certain segment(s) to hit has been thus far demonstrated only in the artificial randomly one or more of numerous cognate-binding experimental setting employing the Gal4-based reporter. sites scattered all over chromosomes.On hitting such a An analogous experiment using an authentic RUNX1- site(s), it would progressively expand its zone of dependent transcription system has been hampered, and chromosomal contacts from there on.Its eventual deemed virtually impossible, because RUNX1 itself can chromosomal engagement could be tenacious and long associate with, and accordingly undergo inhibition by, lasting so that it might well interfere with various mSin3A (Lutterbach et al., 1999; Imai et al., 2004). processes operating on the chromosome including not only transcription but also replication, recombination, Perspectives pairing, and segregation, etc.To possibly make the matter worse, open chromosomal regions functioning Taken altogether, the foregoing considerations point to actively would serve preferential targets.This specula- a unifying view that sequestration and repression are tive scenario, to be called ‘a chromosomal entrapment both contributing mechanisms to the function of CBFb- model’, may provide useful clues to various hitherto SMMHC, and the relative importance of each depends poorly addressed issues regarding the function of CBFb- on the cell type and experimental or physiological SMMHC, particularly its subtle dependencies on cell conditions.These dual inhibitory mechanisms combined lineages, differentiation stages, gene contexts, and with its hyper-heterodimerization activity would make physiological or pathological conditions.A straightfor- CBFb-SMMHC an all the more efficient dominant ward initial test for this newly suggested model would be inhibitor of the RUNX1 function. to examine whether chromosomal DNA might actually Puzzlingly, however, there have also been various lines be adhering to the CBFb-SMMHC-generated nuclear of evidence suggesting that the CBFb-SMMHC-mediated inclusion bodies.

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4302 Gene expression changes in inv(16) þ AML addition, CD81 and LAMB2 were differentially expressed between t(11q23)/MLL and inv(16) cases; As discussed above, even though it is not completely ECM1 was differentially expressed between inv(16) clear how CBFb-SMMHC contributes to leukemogen- and cases with normal karyotype, complex karyotype, esis, it is generally assumed that the fusion protein or trisomy 8; SELL and HLA-DPA1 were differentially dominantly suppresses RUNX function in transcription expressed between inv(16) and t(15;17); and CDK2AP1 regulation.It is then interesting to know which target was differentially expressed between inv(16) and t(8;21). genes are affected by CBFb-SMMHC for their expres- Debernardi et al.(2003) studied five different AML sion, which are critical for leukemogenesis.Unfortu- subtypes based on their karyotype: t(8;21), t(15;17), nately, we still do not have a definitive answer to this inv16, 11q23, and normal karyotype.With a gene question.However, recently published studies on gene set of 145, they grouped the samples according to expression profiling with gene chips or cDNA micro- their cytogenetics. MYH11, SLC7A7, LRP1, PTPRM, arrays have shown that AML cases with inv(16) have a SDR1, NCF2, and TGFB1 were found to be over- pattern of gene expression that is distinctive from other expressed in inv(16) samples. HOXA9, RUNX3, RAD52, subtypes of AML. and HOXA4 were found to be underexpressed in inv(16) samples.However, only a few genes overlap between the Molecular classification and characterization of acute ‘most significant’ gene-sets of Debernardi et al., Schoch myeloid leukemia (AML) et al., and Kohlmann et al., primarily TGFB1 and MYH11 (Schoch et al., 2002; Debernardi et al., 2003; Genome-wide expression profiling has been shown Kohlmann et al., 2003). Not surprisingly, in all studies, to be useful for the classifications of many types of increased expression of MYH11 in inv(16) cases was cancer (Chung et al., 2002), and recently several most informative in acute leukemia classification.The groups demonstrated the feasibility of classifying AMLs level of expression of CBFB-MYH11 transcripts seems and acute lymphoid leukemia (ALL) by expression to have predictive value for treatment outcome (Schnitt- profiling.AML subtypes can also be distinguished ger et al., 2003). solely by gene expression profiling using microarray Presumably more overlaps would be seen if more technology. genes were included in the predictive gene-set.Schoch Golub et al.(1999) were able to distinguish ALL and et al.(2002) initially obtained 1000 differentially AML based exclusively on gene expression profile in expressed genes to specify and identify the AML bone marrow samples of patients.Only 50 of the 6817 subtypes but eventually used only 36 genes (Schoch analysed genes were necessary to divide the 27 ALL and et al., 2002). It would be interesting to combine these 11 AML patients into the different leukemia types. data sets to confirm the predictive values of these Their 50-gene predictor was able to diagnose 29 of the genes. 34 independent samples correctly. et al Schoch .(2002) specified the gene expression Candidates in the pathogenesis of AML associated with classification for the distinction of three different AML inv(16) subtypes: t(8;21)(q22;q22), inv(16)(p13q22), and t(15;17)(q22;q12).Using two independent methodolo- As mentioned above, only very few genes were in gies for class prediction in AML, Schoch et al. identified common among the published studies to have distinctive 36 genes to divide the three AML subtypes, but they changes in inv(16) AML cases.These are discussed were also able to make the discrimination with a briefly below.It is not clear if any of the reported genes minimal set of 13 genes.The most informative genes with altered expression are real targets of CBFb- for inv(16) classification were PRKAR1B (downregu- SMMHC.Further studies are needed to determine lated), MYH11, and HOXB2 (upregulated).It is if CBFb-SMMHC directly regulates their expression, interesting that inv(16) and t(8;21) AML cases have and if their expression change plays any role in leukemo- distinct gene expression patterns. MYH11 and HOXB2 genesis. are again upregulated in inv(16) samples.In addition, a One of the interesting genes, which could be relevant hypothetical gene DKFZP586N1922 is upregulated in in the pathogenesis of AML M4eo, is the HOXA9 gene. inv(16) samples relative to t(8;21) cases.On the other In two studies (Golub et al., 1999; Debernardi et al., hand, PRKAR1B expression is not significantly different 2003), the expression difference of HOXA9 was used to between the two groups. discriminate the acute leukemia types.The observation In a follow-up study, Kohlmann et al.(2003) of reduced expression of HOXA9 in all inv(16) blasts in expanded their study to discriminate eight different one study (Debernardi et al., 2003), while increased types of acute leukemia (AML and ALL).In this study, expression of the HOXA9 gene in AML compared to they included the AML expression data from their ALL cases in the other was remarkable (Golub et al., previous study (Schoch et al., 2002) and added AML 1999).It is also interesting that the expression level of samples with t(11q23)/MLL, with a normal karyotype, HOXA9 was shown to be associated with the effective- with a complex aberrant karyotype, or with trisomy 8. ness of AML treatment. Only a set of 25 genes was sufficient to subclassify the Another HOX gene, HOXB2 gene, was elevated in AML types.In this study, MYH11 was again the inv(16) samples compared with the other subtypes of most informative gene for the inv(16) classification.In myeloid leukemia (Schoch et al., 2002; Debernardi et al.,

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4303 2003).Besides the importance of the HOXB2 gene in genechips) can obviously lead to different results as well embryogenesis, especially in the development of hind- (Kohlmann et al., 2003). brain (Sham et al., 1993), expression of the HOXB2 gene Expanding the data sets by collecting more samples has been seen in different hematopoietic progenitors and combining the data sets will probably improve and (Vieille-Grosjean and Huber, 1995).The exact function enhance the reliability and reproducibility of gene of the HOXB2 gene in hematopoiesis and its potential expression profiling data with clinical samples.Alter- role in AML M4eo remain to be revealed. natively, it would be useful to develop model systems to Higher expression of the transforming growth factor generate a more homogenous condition.Cell culture (TGFB1) was observed in inv(16) blasts (Schoch et al., systems simulating hematopoiesis in the presence or 2002; Debernardi et al., 2003). Previous studies showed absence of CBFB-MYH11 expression, regulated at the pleiotropic effect of TGFB1 on cells from various different stages of hematopoiesis, are potentially useful tissues. TGFB1 has been shown to be a tumor to identify direct targets of CBFB-MYH11.Animal suppressor, since it can inhibit cell growth and induce models established to simulate leukemia development in apoptosis (Coffey et al., 1988; Perlman et al., 2001). On the presence of CBFB-MYH11will be useful to identify the other hand, expression of TGFB1 seems to enhance gene expression changes critical for leukemogenesis. migration and invasion of tumor cells (Akhurst and Derynck, 2001).Since TGFB1 plays an important role in differentiation and antiproliferation during hematopoi- Genes cooperating with CBFB-MYH11 for the esis (Fortunel et al., 2000), TGFB1 and the genes in the pathogenesis of AML TGFB1 signaling pathway (e.g. Smad family genes) are fascinating candidates as important players in the Cbfb-MYH11 knockin mouse model pathogenesis of AML M4eo and AML in general. In inv(16) blasts, downregulation of one of the three We have previously generated a mouse model to study RUNX family genes, RUNX3, was observed (Deber- the function of CBFB-MYH11 by inserting the CBFB- nardi et al., 2003). Previous studies have shown that MYH11 fusion gene into the mouse Cbfb locus in RUNX3 has an important role in the development of embryonic stem (ES) cells (Castilla et al., 1996). We dorsal root ganglia and cell proliferation of gastric demonstrated that CBFB-MYH11 dominantly inhibits epithelium (Levanon et al., 2002; Li et al., 2002; Inoue Runx1/Cbfb functions since embryos heterozygous for et al., 2003). However, RUNX3 also appears to have a the Cbfb-MYH11 allele exhibited a phenotype similar to role as a tumor suppressor gene in various cancers, Runx1- or Cbfb-null embryos (Castilla et al., 1996; including gastric, bile duct, and pancreatic cancers (Guo Okuda et al., 1996; Wang et al., 1996). In addition, we et al., 2002; Li et al., 2002; Wada et al., 2004). The role demonstrated that Cbfb-MYH11 blocks normal hema- of RUNX3 in tumorigenesis, the expression of RUNX3 topoiesis.In the embryos, definitive hematopoiesis is in hematopoietic cells, and the recent observation of the blocked at the stem cell level since c-kit þ colony- possible role of RUNX3 in hematopoiesis in zebrafish forming cells are nearly undetectable in the fetal livers (Kalev-Zylinska et al., 2003) make RUNX3 a very of E11 embryos (Castilla et al., 1996; Kundu et al., interesting gene in leukomogenesis.However, no muta- 2002).In the adult chimeric mice, Cbfb-MYH11- tions in RUNX3 have been associated with AML (Otto containing ES cells contributed to the hematopoietic et al., 2003). stem cells but not myeloid and lymphoid lineages, indicating a blockage of differentiation by Cbfb-MYH11 Summary (Castilla et al., 1999; Kundu et al., 2002). Cbfb-MYH11 chimeras did not develop myeloid As shown in the recent reports described above, only a leukemia, even though after prolonged latency they small set of genes are needed to distinguish subtype developed T-cell lymphoma at a frequency higher than M4eo from other subtypes of acute leukemia.Ob- wild-type mice (Castilla and Liu, unpublished data).No viously, overexpression of MYH11, which is part of the leukemia developed in another mouse model for Cbfb- fusion gene CBFB-MYH11, is very informative for the MYH11, in which Cbfb-MYH11 was driven by a AML classification.The importance and relevance of myeloid-specific promoter, MRP8 (Kogan et al., 1998). the other genes of the predictive gene-sets for AML Therefore, it seemed likely that Cbfb-MYH11 by itself is M4eo pathogenesis remain to be characterized. not sufficient to initiate leukemia, at least in the mouse. It is likely that true target genes are yet to be To test this hypothesis, we treated the Cbfb-MYH11 identified.As noted above, there was very little overlap knockin chimeras with ENU to induce random point in gene expression profiling data generated from mutations in the genome.AML developed in the Cbfb- different groups.This lack of data consistency has MYH11 knockin chimeras 4–7 months after ENU probably resulted from the heterogeneous nature of treatment at a high frequency.On the other hand, clinical samples (age, sex, stage of disease, percent of wild-type mice treated with ENU did not develop blasts in the sample, other chromosomal abnormalities, myeloid leukemia and only 1/20 developed T-cell etc.) as well as technical reasons. Different technologies, lymphoma within 1 year after ENU injection.This for example, SAGE vs microarray hybridization, may result supports the hypothesis that Cbfb-MYH11 pre- result in differences in data.Within a given technology, disposes mice to AML, but requires additional genetic different probe sets (e.g. different versions of Affymatrix events for full transformation (Castilla et al., 1999).

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4304 Retroviral insertional mutagenesis 2003).Therefore, some of the integrations identified above may result from random integration and do not To identify the potential cooperating genes, we treated contribute to leukemogenesis.Common integration sites, newborn Cbfb-MYH11 knockin chimeric mice with the those with more than one independent insertion in our retrovirus 4070A to induce changes in the genome by relatively small panel, would suggest nonrandom inte- insertional mutagenesis.Amphotropic murine leukemia gration and biological significance.Six genes were retrovirus (MLV) 4070A can induce myeloid leukemia inserted more than once, and they are Plag1, Plagl2, in DBA/2N mice that were undergoing an intense D6Mm5e, H2T24, Myb,andRunx2 (Castilla et al., 2004). chronic inflammatory response, but not in mice without Plag1 and Plagl2 are especially interesting since they the inflammation (Wolff et al., 1991). 4070A treatment encode highly homologous proteins that belong to a new resulted in the development of AML in 63% (27/43) of family of zinc-fingers. Plag1 was inserted eight times Cbfb-MYH11 knock-in chimeras but none of the while Plagl2 was inserted twice.Therefore, these two control wild-type mice (n 40) (Castilla et al., 2004). ¼ genes are the most frequent insertion sites in the panel The latency was 3–6 months and the phenotype of the (15% of total), suggesting that they are bona fide leukemia was similar to that of ENU-induced leukemia. cooperating genes with Cbfb-MYH11 for leukemogen- Therefore, leukemia development is dependent on the esis. PLAG1 was initially identified at the breakpoint of expression of the Cbfb-MYH11 knock-in gene and a chromosome translocation t(3;8)(p21;q12), which is independent of the mutagens used. commonly found in human pleomorphic adenoma, a Viral insertion sites were detected by Southern blot benign tumor of the salivary glands (Kas et al., 1997). hybridization and identified by inverse PCR and The translocation leads to promoter swapping between sequencing (Castilla et al., 2004). Southern blot hybri- PLAG1 and CTNNB1, which encodes the constitutively dization with a viral probe showed that there were 1–3 expressed b-catenin.As a result of the translocation, insertions per leukemia, indicating that as few as one PLAG1 is activated by the CTNNB1 promoter (Kas insertion is enough for Cbfb-MYH11 leukemogenic et al., 1997). Subsequently, it was discovered that cooperation.This supposition was supported by the PLAG1 can be activated through similar mechanisms finding that a single insertion was retained in several by the promoters of LIFR (encoding leukemia inhibitory independent secondary leukemias induced by bone factor receptor) and the elongation factor SII gene in marrow/spleen leukemic cell transplantation to isogenic pleomorphic adenomas (Voz et al., 1998; Astrom et al., recipients, which suggested that the leukemia was clonal 1999).The insertions in Plag1 and Plagl2 genes in the and that a single insertion was important for the Cbfb-MYH11 chimeras were all close to the transcrip- leukemogenesis.Using inverse PCR, the viral insertion tion start sites, which are predicted to also upregulate sites were cloned and the surrounding sequences were Plag1 and Plagl2 expression (Castilla et al., 2004). determined.A total of 67 insertion sites were identified Plag1 protein contains seven zinc-fingers near the in 20 leukemia samples, or 3.3 insertions/leukemia. This amino-terminus and a serine-rich region near the number was slightly higher than the estimation by carboxy-terminus.Two highly homologous genes have Southern blot hybridization, probably due to technical been isolated recently, named Plagl1 and Plagl2 (Kas differences between the two approaches.In addition, et al., 1998). The three proteins share 70–80% identity in viral insertions in the surrounding nonleukemia cells the zinc-finger region, but only 19–35% identity in the could have been cloned by the inverse PCR approach. serine-rich region.All three proteins transactivate gene Searching the mouse genomic databases revealed that expression in GAL4 fusion protein assays (Kas et al., 90% of the insertion sites were located within 10 kb of a 1998).Even though their zinc-finger regions are highly gene.More than half of the insertions (34/67) were conserved, the three proteins have different preferences located either within 10 kb upstream of the transcription and binding affinities for target DNA sequences, with initiation site or within 50 UTR (part of the transcript Plagl1 different from Plag1 and Plagl2 (Hensen et al., but upstream of ATG), suggesting that transcriptional 2002).Functionally, Plag1 and Plagl2 behave like activation of cellular genes is the most common classical oncogenes in NIH 3T3 transformation assays: mechanism of insertional mutagenesis in this study. they rendered the cells to grow in foci and anchorage The second most common location of viral insertion is independently, and the cells expressing either of these in the introns of the protein-coding region, accounting two genes formed tumors in nude mice (Hensen et al., for 24 of the total 67 insertions.Such events are 2002).On the other hand, Plagl1 behaves like a tumor predicted to truncate the encoded proteins.In addition, suppressor.Also known as Zac1 and Lot1, Plagl1 has one insertion disrupts a single-exon gene (Edg6), which been isolated independently as a candidate tumor should also result in protein truncation.There are only suppressor gene since it is located on 6q25, a chromo- seven insertions located 10 kb away from the flanking somal region frequently deleted in many solid tumors genes and their effect on the flanking genes is unclear. (Abdollahi et al., 1997). In cell cultures, Plagl1 is able to induce apoptosis and arrest cells at G1, leading to Candidate cooperating genes inhibition of tumor cell proliferation (Spengler et al., 1997).Plag1 and Plagl2 upregulate insulin-like growth Recent studies suggest that murine leukemia retroviruses factor II (Kas et al., 1998; Voz et al., 2000; Hensen et al., may have a propensity to insert within genes, especially 2002), while Plagl1 may serve as a coactivator of p53 when nearing the transcription start sites (Wu et al., (Huang et al., 2001b). Interestingly, we observed multi-

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4305 ple insertions in the Plag1 and Plagl2 genes, but none in the cooperation between repressive mutations in tran- the Plagl1 gene, consistent with their oncogenic and scription factors and gain-of-function mutations in tumor suppressive roles, respectively.This is the first receptor tyrosine kinases (Gilliland, 2002).Most com- time that Plag1 and Plagl2 genes are linked to mon chromosomal translocations identified in human leukemogenesis. AML generate fusion proteins involving at least one It is intriguing that three insertions were found in the transcription factor, which has some role in the Runx2 gene (Castilla et al., 2004). Runx2 is one of the regulation of hematopoietic differentiation (Look, three Runx family members that are obligate hetero- 1997).On the other hand, the most frequent mutations dimeric partners of Cbfb.Runx2 plays a key role in besides chromosomal translocations in AML have been osteogenesis and chondrocyte differentiation (Komori found in genes encoding receptor tyrosine kinases, such et al., 1997; Otto et al., 1997). In a CD2-Myc-transgenic as FLT-3 and c-kit (Gilliland and Griffin, 2002; Care model, Runx2 was found to be a frequent target for et al., 2003). Therefore, it is logical to propose that two MLV insertion to induce T-cell lymphoma (Stewart steps are required for complete leukemogenesis, one is et al., 1997). The insertions in the T-cell lymphomas block of differentiation as a result of dysfunctional were located close to the upstream promoter of the transcription factors, and the other is deregulated Runx2 gene, resulting in elevated expression (Cameron proliferation of immature blasts as a result of activated et al., 2003). Transgenic mice coexpressing Myc and receptor tyrosine kinases (Gilliland, 2002).The retro- Runx2 under the control of CD2 promoter rapidly viral insertion data in Cbfb-MYH11 mice support a developed T-cell lymphoma, suggesting that Runx2 may model that is a variation of the common theme.Most of serve as an oncogene and cooperate with Myc.The three the common insertions are close to or within genes with insertions observed in the leukemic cells from the Cbfb- demonstrated roles in oncogenesis, particularly in the MYH11 chimeras, however, are located in intron 5 process of cell cycle and proliferation regulation. (Castilla et al., 2004). These insertions are predicted to However, they are not necessarily kinases, but tran- result in the production of a truncated Runx2 protein, scription factors as well (Figure 4). retaining the Runt-homology domain for DNA and In the near future, the functional significance of these Cbfb binding but deleting the transactivation and candidate cooperating genes close to viral insertion sites repression domains located on the C-terminal side of needs to be confirmed.Transgenic mice need to be the insertions.Therefore, the truncated protein is generated, combining the expression of Cbfb-MYH11 predicted to serve as a dominant repressor of Runx with that of the candidate genes.Accelerated leukemia function, since the truncated protein can bind Runx development in these mice would prove that the DNA target sequences but do not regulate transcription. candidate genes cooperate with Cbfb-MYH11 in leuke- If verified by demonstrating cooperation between mogenesis.Once confirmed, it will be important to find coexpressing a truncated Runx2 and Cbfb-MYH11 in out if the candidate genes also cooperate with CBFB- transgenic mice for leukemia development, the result MYH11 in the pathogenesis of human AML.In would suggest that Cbfb-MYH11 is a weak inhibitor addition, the study described above was relatively small of Runx/Cbfb pathway, at least in the mice, and that scale with only 20 leukemia samples.Potential cooperat- Runx/Cbfb pathway is antileukemogenic in the ing genes may be inserted only once or not at all and myeloid lineage. therefore overlooked in this study.In addition, a Two insertions were located 80 kb upstream of the different virus may insert in a different set of genes. Myb gene, close to the so-called Mml-3 locus, which is a Therefore, it is important to carry out the retroviral frequent site for MLV insertions in published murine insertional mutagenesis again with a larger number of leukemia models (Haviernik et al., 2002). A recent study mice and using more than one strain of retrovirus.Last suggests that retroviral insertions may have long-range but not the least, it will be interesting to find out if these effects on Myb expression (Hanlon et al., 2003). candidate genes also cooperate with AML1-ETO, the Additionally, these insertions may affect the expression fusion gene involving the Runx1 gene (Miyoshi et al., of other genes in the region.The functional conse- 1991; Erickson et al., 1992). If the mechanism of quences of insertions in the D6Mm5e and H2T24 genes leukemogenesis by Cbfb-MYH11 is through dominant are less clear.The function of D6Mm5e is not known, suppression of Runx1, which is shared by AML1-ETO but insertions in the introns of this gene may affect the (Yergeau et al., 1997), at least some of the candidate expression of Dok1, a gene located 18 kb away down- genes should also be able to cooperate with AML1- stream of D6Mm5e. Dok1 encodes P62-Dok, which is an ETO. adaptor protein that interacts with SHIP1 and Ras- GAP and is involved in the pathogenesis of CML (Wisniewski et al., 1994; Dunant et al., 2000). H2T24 is part of the MHC class 1 T-gene cluster and its role in leukemogenesis has not been demonstrated.

Summary and future directions Figure 4 Two-step leukemogenesis contributed by CBFB-MYH11 In recent years, a common theme in the pathogenesis of and the candidate cooperating genes identiﬁed by retroviral human acute myeloid leukemia has emerged, which is insertional mutagenesis. dnRunx2: dominant-negative Runx2

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4306 Acknowledgements ‘Molecular basis for the dominant inhibition of RUNX1- KS expresses special gratitude to Huang Gang and dependent transcription by CBFb-SMMHC’ is essentially Hee-Jun Wee for their overwhelming contributions to detailed based.PPL thanks Lucio Castilla for the retroviral inser- functional characterization of CBFb/PEBP2b-SMMHC as tion data and John Bushweller for critical reading of the well as their valuable discussions on which the section manuscript.

References

Abdollahi A, Roberts D, Godwin AK, Schultz DC, Sonoda Golub TR, Slonim DK, Tamayo P, Huard C, Gaasenbeek G, Testa JR and Hamilton TC.(1997). Oncogene, 14, M, Mesirov JP, Coller H, Loh ML, Downing JR, Caligiuri 1973–1979. MA, Bloomfield CD and Lander ES.(1999). Science, 286, Adya N, Stacy T, Speck NA and Liu PP.(1998). Mol. Cell. 531–537. Biol., 18, 7432–7443. Guo WH, Weng LQ, Ito K, Chen LF, Nakanishi H, Akhurst RJ and Derynck R.(2001). Trends Cell Biol., 11, Tatematsu M and Ito Y.(2002). Oncogene, 21, 8351–8355. S44–S51. Hanlon L, Barr NI, Blyth K, Stewart M, Haviernik P, Wolff Astrom AK, Voz ML, Kas K, Roijer E, Wedell B, Mandahl N, L, Weston K, Cameron ER and Neil JC.(2003). J. Virol., 77, Van de Ven W, Mark J and Stenman G.(1999). Cancer Res., 1059–1068. 59, 918–923. Haviernik P, Festin SM, Opavsky R, Koller RP, Barr NI, Neil Backstrom S, Wolf-Watz M, Grundstrom C, Hard T, JC and Wolff L.(2002). J. Gen. Virol., 83, 819–827. Grundstrom T and Sauer UH.(2002). J. Mol. Biol., 322, Hensen K, Van Valckenborgh IC, Kas K, Van de Ven WJ and 259–272. Voz ML.(2002). Cancer Res., 62, 1510–1517. Bravo J, Li Z, Speck NA and Warren AJ.(2001). Nat. Struct. Huang G, Shigesada K, Ito K, Wee HJ, Yokomizo T and Ito Biol., 8, 371–378. Y.(2001a). EMBO J., 20, 723–733. Cameron ER, Blyth K, Hanlon L, Kilbey A, Mackay N, Huang G, Shigesada K, Wee HJ, Liu PP, Osato M and Ito Y. Stewart M, Terry A, Vaillant F, Wotton S and Neil JC. (2004). Blood, 103. (2003). Blood Cells Mol. Dis., 30, 194–200. Huang SM, Schonthal AH and Stallcup MR.(2001b). Cao W, Adya N, Britos-Bray M, Liu PP and Friedman AD. Oncogene, 20, 2134–2143. (1998). J. Biol. Chem., 273, 31534–31540. Ikebe M, Komatsu S, Woodhead JL, Mabuchi K, Ikebe R, Cao W, Britos-Bray M, Claxton DF, Kelley CA, Speck NA, Saito J, Craig R and Higashihara M.(2001). J. Biol. Chem., Liu PP and Friedman AD.(1997). Oncogene, 15, 1315–1327. 276, 30293–30300 Epub 2001 Jun 6. Care RS, Valk PJ, Goodeve AC, Abu-Duhier FM, Geertsma- Imai Y, Kurokawa M, Yamaguchi Y, Izutsu K, Nitta E, Kleinekoort WM, Wilson GA, Gari MA, Peake IR, Low- Mitani K, Satake M, Noda T, Ito Y and Hirai H.(2004). enberg B and Reilly JT.(2003). Br. J. Haematol., 121, 775– Mol. Cell. Biol., 24, 1033–1043. 777. Inoue K, Ozaki S, Ito K, Iseda T, Kawaguchi S, Ogawa M, Castilla LH, Garrett L, Adya N, Orlic D, Dutra A, Anderson Bae SC, Yamashita N, Itohara S, Kudo N and Ito Y.(2003). S, Owens J, Eckhaus M, Bodine D and Liu PP.(1999). Nat. Blood Cells Mol. Dis., 30, 157–160. Genet., 23, 144–146. Kalev-Zylinska ML, Horsfield JA, Flores MV, Postlethwait Castilla LH, Perrat P, Martinez NJ, Landrette SF, Keys R, JH, Chau JY, Cattin PM, Vitas MR, Crosier PS and Crosier Oikemus S, Flanegan J, Garrett L, Dutra A, Anderson S, KE.(2003). Dev. Dyn., 228, 323–336. Pihan GA, Wolff L and Liu PP.(2004).Identification of Kanno Y, Kanno T, Sakakura C, Bae SC and Ito Y.(1998). genes that synergize with Cbfb-MYH11 in the pathogenesis Mol. Cell. Biol., 18, 4252–4261. of acute myeloid leukemia. Proc. Natl. Acad. Sci. USA, Kanto S, Chiba N, Tanaka Y, Fujita S, Endo M, Kamada N, 10.1073/pnas.0400930101. Yoshikawa K, Fukuzaki A, Orikasa S, Watanabe T and Castilla LH, Wijmenga C, Wang Q, Stacy T, Speck NA, Satake M.(2000). Leukemia, 14, 1253–1259. Eckhaus M, Marin-Padilla M, Collins FS, Wynshaw-Boris Kas K, Voz ML, Hensen K, Meyen E and Van de Ven WJ. A and Liu PP.(1996). Cell, 87, 687–696. (1998). J. Biol. Chem., 273, 23026–23032. Chung CH, Bernard PS and Perou CM.(2002). Nat. Genet., 32 Kas K, Voz ML, Roijer E, Astrom AK, Meyen E, Stenman G (Suppl), 533–540. and Van de Ven WJ.(1997). Nat. Genet., 15, 170–174. Coffey Jr RJ, Bascom CC, Sipes NJ, Graves-Deal R, Weiss- Kogan SC, Lagasse E, Atwater S, Bae SC, Weissman I, Ito Y man BE and Moses HL.(1988). Mol. Cell. Biol., 8, 3088– and Bishop JM.(1998). Proc. Natl. Acad. Sci. USA, 95, 3093. 11863–11868. Debernardi S, Lillington DM, Chaplin T, Tomlinson S, Amess Kohlmann A, Schoch C, Schnittger S, Dugas M, Hiddemann J, Rohatiner A, Lister TA and Young BD.(2003). Genes W, Kern W and Haferlach T.(2003). Genes Chromosomes Chromosomes Cancer, 37, 149–158. Cancer, 37, 396–405. Dunant NM, Wisniewski D, Strife A, Clarkson B and Resh Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, MD.(2000). Cell Signal, 12, 317–326. Deguchi K, Shimizu Y, Bronson RT, Gao YH, Inada M, Durst KL, Lutterbach B, Kummalue T, Friedman AD and Sato M, Okamoto R, Kitamura Y, Yoshiki S and Kishimoto Hiebert SW.(2003). Mol. Cell. Biol., 23, 607–619. T.(1997). Cell, 89, 755–764. Erickson P, Gao J, Chang KS, Look T, Whisenant E, Kummalue T, Lou J and Friedman AD.(2002). Mol. Cell. Raimondi S, Lasher R, Trujillo J, Rowley J and Drabkin Biol., 22, 8278–8291. H.(1992). Blood, 80, 1825–1831. Kundu M, Chen A, Anderson S, Kirby M, Xu L, Castilla LH, Fortunel N, Hatzfeld J, Aoustin L, Batard P, Ducos K, Bodine D and Liu PP.(2002). Blood, 100, 2449–2456. Monier MN, Charpentier A and Hatzfeld A.(2000). Levanon D, Bettoun D, Harris-Cerruti C, Woolf E, Negreanu Hematol. J., 1, 126–135. V, Eilam R, Bernstein Y, Goldenberg D, Xiao C, Fliegauf Gilliland DG.(2002). Semin. Hematol., 39, 6–11. M, Kremer E, Otto F, Brenner O, Lev-Tov A and Groner Y. Gilliland DG and Griffin JD.(2002). Blood, 100, 1532–1542. (2002). EMBO J., 21, 3454–3463.

Oncogene Mechanism of leukemogenesis by CBFB/PEBP2B-MHY11 K Shigesada et al 4307 Li QL, Ito K, Sakakura C, Fukamachi H, Inoue K, Chi XZ, Sham MH, Vesque C, Nonchev S, Marshall H, Frain M, Lee KY, Nomura S, Lee CW, Han SB, Kim HM, Kim WJ, Gupta RD, Whiting J, Wilkinson D, Charnay P and Yamamoto H, Yamashita N, Yano T, Ikeda T, Itohara S, Krumlauf R.(1993). Cell, 72, 183–196. Inazawa J, Abe T, Hagiwara A, Yamagishi H, Ooe A, Sohn RL, Vikstrom KL, Strauss M, Cohen C, Szent-Gyorgyi Kaneda A, Sugimura T, Ushijima T, Bae SC and Ito Y. AG and Leinwand LA.(1997). J. Mol. Biol., 266, 317–330. (2002). Cell, 109, 113–124. Spengler D, Villalba M, Hoffmann A, Pantaloni C, Houssami S, Liu P, Seidel N, Bodine D, Speck N, Tarle S and Collins FS. Bockaert J and Journot L.(1997). EMBO J., 16, 2814–2825. (1994). Cold Spring Harb. Symp. Quant. Biol., 59, 547–553. Stewart M, Terry A, Hu M, O’Hara M, Blyth K, Baxter E, Liu P, Tarle´ SA, Hajra A, Claxton DF, Marlton P, Freedman Cameron E, Onions DE and Neil JC.(1997). Proc. Natl. M, Siciliano MJ and Collins FS.(1993). Science, 261, Acad. Sci. USA, 94, 8646–8651. 1041–1044. Tahirov TH, Inoue-Bungo T, Morii H, Fujikawa A, Sasaki M, Liu PP, Hajra A, Wijmenga C and Collins FS.(1995). Blood, Kimura K, Shiina M, Sato K, Kumasaka T, Yamamoto M, 85, 2289–2302. Ishii S and Ogata K.(2001). Cell, 104, 755–767. Liu PP, Wijmenga C, Hajra A, Blake TB, Kelley CA, Tang YY, Shi J, Zhang L, Davis A, Bravo J, Warren AJ, Adelstein RS, Bagg A, Rector J, Cotelingam J, Willman Speck NA and Bushweller JH.(2000). J. Biol. Chem., 275, CL and Collins FS.(1996). Genes Chromosomes Cancer, 16, 39579–39588. 77–87. Trybus KM, Freyzon Y, Faust LZ and Sweeney HL.(1997). Look AT.(1997). Science, 278, 1059–1064. Proc. Natl. Acad. Sci. U.S.A., 94, 48–52. Lu J, Maruyama M, Satake M, Bae SC, Ogawa E, Kagoshima Trybus KM and Lowey S.(1984). J. Biol. Chem., 259, 8564– H, Shigesada K and Ito Y.(1995). Mol. Cell. Biol., 15, 8571. 1651–1661. Vieille-Grosjean I and Huber P.(1995). J. Biol. Chem., 270, Lukasik SM, Zhang L, Corpora T, Tomanicek S, Li Y, 4544–4550. Kundu M, Hartman K, Liu PP, Laue TM, Biltonen RL, Voz ML, Agten NS, Van de Ven WJ and Kas K.(2000). Speck NA and Bushweller JH.(2002). Nat. Struct. Biol., 9, Cancer Res., 60, 106–113. 674–679. Voz ML, Astrom AK, Kas K, Mark J, Stenman G and Van de Lutterbach B, Hou Y, Durst KL and Hiebert SW.(1999). Ven WJ.(1998). Oncogene, 16, 1409–1416. Proc. Natl. Acad. Sci. USA, 96, 12822–12827. Wada M, Yazumi S, Takaishi S, Hasegawa K, Sawada M, Miller JD, Stacy T, Liu PP and Speck NA.(2001). Blood, 97, Tanaka H, Ida H, Sakakura C, Ito K, Ito Y and Chiba T. 2248–2256. (2004). Oncogene, 26, 26. Miyoshi H, Shimizu K, Kozu T, Maseki N, Kaneko Y Wang Q, Stacy T, Miller JD, Lewis AF, Gu TL, Huang X, and Ohki M.(1991). Proc. Natl. Acad. Sci. USA, 88, Bushweller JH, Bories JC, Alt FW, Ryan G, Liu PP, 10431–10434. Wynshaw-Boris A, Binder M, Marin-Padilla M, Sharpe AH Okuda T, van Deursen J, Hiebert SW, Grosveld G and and Speck NA.(1996). Cell, 87, 697–708. Downing JR.(1996). Cell, 84, 321–330. Warren AJ, Bravo J, Williams RL and Rabbitts TH.(2000). Otto F, Stock M, Fliegauf M, Fenaux P, Preudhomme C and EMBO J., 19, 3004–3015. Lubbert M.(2003). Leukemia, 17, 1677–1678. Wijmenga C, Gregory PE, Hajra A, Schrock E, Ried T, Eils R, Otto F, Thornell AP, Crompton T, Denzel A, Gilmour Liu PP and Collins FS.(1996). Proc. Natl. Acad. Sci. U.S.A., KC, Rosewell IR, Stamp GW, Beddington RS, Mundlos 93, 1630–1635. S, Olsen BR, Selby PB and Owen MJ.(1997). Cell, 89, Wisniewski D, Strife A, Wojciechowicz D, Lambek C and 765–771. Clarkson B.(1994). Leukemia, 8, 688–693. Perlman R, Schiemann WP, Brooks MW, Lodish HF and Wolff L, Koller R and Davidson W.(1991). J. Virol., 65, 3607– Weinberg RA.(2001). Nat. Cell Biol., 3, 708–714. 3616. Schnittger S, Weisser M, Schoch C, Hiddemann W, Haferlach Wu X, Li Y, Crise B and Burgess SM.(2003). Science, 300, T and Kern W.(2003). Blood, 102, 2746–2755 Epub 3 July 1749–1751. 2003. Xu JQ, Harder BA, Uman P and Craig R.(1996). J. Cell Biol., Schoch C, Kohlmann A, Schnittger S, Brors B, Dugas M, 134, 53–66. Mergenthaler S, Kern W, Hiddemann W, Eils R and Yergeau DA, Hetherington CJ, Wang Q, Zhang P, Sharpe Haferlach T.(2002). Proc. Natl. Acad. Sci. USA, 99, AH, Binder M, Marin-Padilla M, Tenen DG, Speck NA and 10008–10013 Epub 8 July 2002. Zhang DE.(1997). Nat. Genet., 15, 303–306.

Oncogene