Oncogene (2001) 20, 7186 ± 7203 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc

Translocations of the RARa in acute promyelocytic leukemia

Arthur Zelent*,1, Fabien Guidez1, Ari Melnick3, Samuel Waxman3 and Jonathan D Licht2

1Leukemia Research Fund Centre at the Institute of Cancer Research, Chester Beatty Laboratories, 237 Fulham Road, London SW3 6JB, UK; 2Derald H Ruttenberg Cancer Center, Department of Medicine, Mount Sinai Medical Center, New York, NY, USA; 3Division of Medical Oncology, Department of Medicine, Mount Sinai Medical Center, New York, NY, USA

Acute promyelocytic leukemia (APL) has been recognized leukemia (APL). Actions of ATRA and wide range of as a distinct clinical entity for over 40 years. Although other natural or synthetic retinoids, are mediated relatively rare among hematopoietic malignancies (ap- through binding to speci®c nuclear receptors (NRs) proximately 10% of AML cases), this disease has that regulate gene transcription (Glass and Rosenfeld, attracted a particularly good share of attention by 2000). To date, three di€erent retinoic acid becoming the ®rst human cancer in which all-trans- (RAR) and retinoid X (or rexinoid) receptor (RXR) retinoic acid (ATRA), a physiologically active derivative have been characterized, each encoding multiple N- of vitamin A, was able to induce complete remission (CR). terminal isoforms (Chambon, 1996). RXRs ATRA induced remission is not associated with rapid cell serve as obligatory heterodimerization partners for death, as in the case of conventional chemotherapy, but RARs and for a number of other NRs including those with a restoration of the `normal' granulocytic di€erentia- for thyroid hormones (TR) and vitamin D3 (VDR), thus tion pathway. With this remarkable medical success story integrating di€erent signaling pathways (Chambon, APL has overnight become a paradigm for the di€erentia- 1996; Glass and Rosenfeld, 2000). RARs, like many tion therapy of cancer. A few years later, excitement with other members of the NR superfamily (Chambon, 1996; APL was further enhanced by the discovery that a Mangelsdorf and Evans, 1995), can activate or repress cytogenetic marker for this disease, the t(15:17) recipro- transcription of their target genes. Transcriptional cal chromosomal translocation, involves a fusion between repression and activation by RARs involves recruitment the alpha (RARa) gene and a of multiprotein co-activator and co-repressor complexes previously unknown locus named promyelocytic leukemia with histone acetytransferase (HAT) and histone (PML). Consequence of this gene rearrangement is deacetylase (HDAC) activities respectively. Binding of expression of the PML ± RARa chimeric oncoprotein, ATRA to RAR provides a molecular switch that turns which is responsible for the cellular transformation as well its activity from a repressor to an activator of gene as ATRA response that is observed in APL. Since this transcription (Glass and Rosenfeld, 2000). initial discovery, a number of di€erent translocation Given its responsiveness to ATRA, the discovery partner genes of RARa have been reported in rarer cases that the t(15;17) translocation, a cytogenetic marker for of APL, strongly suggesting that disruption of RARa APL (Rowley et al., 1977), involves a fusion of the underlies its pathogenesis. This article reviews various RARa gene on 17q21 to a locus called rearrangements of the RARa gene that have so far been PML (Borrow et al., 1990; de The et al., 1990; Longo described in literature, functions of the encoded et al., 1990; Miller et al., 1990) appeared rather by the di€erent RARa partner loci, and implications that paradoxical. However, subsequent studies addressing these may have for the molecular pathogenesis of APL. the function of the resulting PML ± RARa fusion Oncogene (2001) 20, 7186 ± 7203. oncoprotein had clearly indicated that the therapeutic e€ects of ATRA in this disease are due to its direct Keywords: PLZF; NuMA; NPM; STAT; di€erentia- action on the activity of PML ± RARa (Grignani et al., tion; retinoic acid 1998; Guidez et al., 1998; He et al., 1998; Hong et al., 1997; Lin et al., 1998). In addition to the PML/RARa rearrangement, four other APL associated transloca- Introduction tions of the RARa gene have been characterized to date at the molecular level. The t(5;17)(q35;q21) (Redner et Retinoids, such as all-trans-retinoic acid (ATRA), exert al., 1996b), t(11;17)(q23;q21) (Chen et al., 1993b), a wide range of e€ects on both normal and malignant t(11;17)(q13;q21) (Wells et al., 1997) and cells. Although explored as a potential therapeutic and/ t(17;17)(q11;q21) (Arnould et al., 1999) fuse RARa to or chemopreventive agent in a number of human the Nucleophosmin (NPM), Promyelocytic Leukeia Zinc cancers, clinical use of ATRA has never been more Finger (PLZF), Nuclear Mitotic Apparatus (NuMA) successful than in the treatment of acute promyelocytic and STAT5b genes, respectively. As in the case of t(15;17) translocation, the above rearrangements of RARa lead to expression of PLZF-, NPM-, NuMA- *Correspondence: A Zelent; E-mail: [email protected] and STAT5b-RARa fusion proteins (see Figure 1 for a APL associated RARa fusions A Zelent et al 7187 schematic representation). Common features of all to generate by alternative splicing a number of C- these proteins are B through F regions of RARa, terminally divergent PML isoforms. Exact functions of which contain its DNA and ligand binding domains these isoforms are not clear, but the C-termini of PML (DBD and LBD, respectively), with non-RAR N- are rich in proline/serine residues and contain a terminal moieties that contribute additional dimeriza- number of potential phosphorylation sites (Alcalay et tion motifs to the fusion proteins. These heterologous al., 1992; de The et al., 1991; Fagioli et al., 1992; dimerization domains promote formation of chimeric Goddard et al., 1991; Kakizuka et al., 1991). receptor homodimers, thus removing requirement of Furthermore, C-terminal sequences for speci®c PML RXR for binding to DNA (Dong et al., 1996; Perez et isoforms have been shown to play a role in regulation al., 1993), and together with the DNA binding domain of PML stability (Lallemand-Breitenbach et al., 2001) of RARa appear to be required for oncogenic activities and interaction with the tumour suppressor protein of the fusion proteins (Lin and Evans, 2000; Minucci et (Ferbeyre et al., 2000; Fogal et al., 2000; Guo et al., al., 2000). 2000; Pearson et al., 2000). The PML gene is

Rearrangement between the RARa and PML genes

The human and mouse RARa genes have been cloned and partially characterized (Brand et al., 1990; Leroy et al., 1991). They each contain 10 exons and two promoters, which control expression of two major isoforms (RARa1 and RARa2) that di€er in the sequences corresponding to their 5'-untranslated (5'- UTR) and A regions. The B to F region sequences, Figure 2 Schematic representation of the RARa gene and the however, are common to both isoforms (see Figure 2 RARa1 and a2 isoforms. Exons are represented by shaded boxes and are numbered consecutively. Regions that are not shaded for a schematic representations). Although expression represent 5' and 3' untranslated sequences (UTR). Two of the RARa1 isoform is ubiquitous, the RARa2 promoters, P1 and P2, are indicated by broken arrows. Presence isoform is expressed in an ATRA-dependent and of a retinoic acid reponse element (RARE) in P2 is indicated. tissue-speci®c manner (Leroy et al., 1991). In addition, Positions of translocation breakpoints within the RARa gene are during myeloid di€erentiation of a multipotent murine indicated with horizontal arrows. Number of arrows re¯ects relative frequency of breakpoints in introns 1 (infrequent), 2 progenitor cell line (FDCPmixA4) expression of both (most frequent) and 3 (less frequent). Protein coding sequences RARa1 and 2 isoforms becomes restricted to the are indicated by shaded rectangles and are subdivided into myelomonocytic lineage (Zhu et al., 2001). conserved (B ± F) and divergent (A regions) functional domains. The PML gene encodes a nuclear protein with a Di€erent shadings are used to represent RAR isoform-speci®c sequences. Regions encoded by di€erent exons are indicated by characteristic zinc-®nger motif called the RING ®nger arrowheads underneath each diagram. DNA (DBD) and ligand (Reddy et al., 1992). The human PML gene is (LBD) binding domains (regions C and E, respectively) are comprised of 9 exons with exons 7 ± 9 having potential indicated above the scheme for each isoform

Figure 1 Schematic representation of RARa and four APL associated RARa fusion proteins. RARa functional domains (a ± f) are as indicated. Open triangles denote boundaries between sequences encoded in separate exons lying on opposite sides of a given translocation breakpoint. Di€erent patterns and colors are used to represent various functional regions of the PML, PLZF, NPM, NuMA and STAT5b proteins. Circled Zn2+ symbols and PML regions labelled as RING, B1 and B2 represent two of the nine kruÈ ppel-like Zn-®nger motifs and cystein-histidine rich domains in the PLZF and PML proteins, respectively. Light red regions (labeled Coiled-coil in PML and STAT5b, and POZ in PLZF) represent protein-protein interaction motifs present in N-termini of all the RARa chimeras. Relative positions of STAT5b DNA binding (DBD), SH3 and SH2 domains are also indicated

Oncogene APL associated RARa fusions A Zelent et al 7188 ubiquitously expressed, albeit at di€erent levels (God- stretches of homologous DNA, repair of the breaks dard et al., 1991), which is reminiscent of RARa1 and joining of the loci. expression. PML levels can be upregulated by inter- Locations of chromosomal breakpoints within the ferons (Stadler et al., 1995), higher levels of PML exon/intron structures of both genes have been expression were found in some solid tumors (Koken et extensively analysed (Alcalay et al., 1992; Chen et al., al., 1995; Terris et al., 1995) and speci®c expression of 1991; Diverio et al., 1992; Pandol® et al., 1992). Within PML was detected in myeloid cells (Daniel et al., the RARa gene, all breakpoints map to genomic 1993). PML contains a number of clearly de®ned sequences lying upstream of exon 4 which encodes structural motifs, which are chie¯y encoded by the 5'- the B-region of the RAR and N-terminal part of the exons of this gene. The N-terminal proline rich domain DBD (Figure 2). Consequently, the PML ± RARa is encoded by exon 1, the RING ®nger and the B1 and fusion gene product always contains the same regions B2 boxes, which are rich in cystein residues, are mainly of RARa (B-F) containing its DNA and ligand binding encoded by exons 2 and 3, and the a-helical coiled-coil domains. In contrast, the N-terminal PML sequences domain is essentially encoded by exon 3 (see Figure 3). joined to B-F domains of the RARa show patient to Interestingly, in addition to PML, two other RING patient variability determined by the position of the ®nger genes when translocated to other chromosomal translocation breakpoint within the PML gene and by loci are expressed as fusion proteins with transforming alternative exon splicing (see Figure 3). Pandol® et al. activities (Miki et al., 1991; Takahashi et al., 1988). (1992) described three breakpoint clusters within the The mechanism by which the t(15;17) translocation PML gene located in intron 3 (bcr3), exon 6 (bcr2) and occurs is not known. It may be that many illegitimate intron 6 (bcr1). The patients with chromosomal recombinations occur during normal cell division and breakpoints in either bcr1 or bcr3 express either the are eliminated by DNA repair systems (Tashiro et al., long (PML ± RARaL) or short (PML ± RARaS) form 1994). Those clones that contain the PML ± RARa of the PML ± RARa protein (Pandol® et al., 1992). The transcript survive and have a growth advantage. More RNA encoding the PML ± RARaL protein can be recent study found that the PML and RARa genes are alternatively spliced to give PML ± RARaM (medium), physically adjacent to each other in chromatin (Neves which is usually co-expressed with the L form in a et al., 1999), this perhaps explaining the tendency of given patient (see Figure 3 for schematic representa- the two genes to rearrange. Recent analysis also found tions). Breakpoints lying within bcr2 delete the 5'-splice short stretches of identity between the PML and RARa site from intron 6. This causes selection of a cryptic 5' genes in the breakpoint regions (Yoshida et al., 1995). splice site within the fusion gene, usually within the It was proposed that random cleavage of the RARa translocated RARa intronic sequences, and hence and PML genes is followed by limited pairing of short generation of a PML ± RARaV transcript with variable portions of PML exon 6 sequences joined in the correct reading frame with RARa exon 4 (B region). The PML ± RARaS protein contains nearly all of the structural motifs of both PML and RARa proteins (see Figures 1 ± 3 for comparison). The PML ± RARaVand L isoforms also contain additional PML coding sequences. A major potential structural motif missing in the S isoform, but present in both the L and V isoforms, is a putative nuclear localization signal (NLS) located in the proximal region of PML exon 6. In addition, part of a-helical domain encoded by exons 3 ± 6, a serine/proline (S/P) rich region, located in distal PML exon 6, are absent in all S isoform and some V isoform cases (Gallagher et al., 1995). The PML ± RARaS isoform is also missing a sumolation site and a caspase cleavage sequence. Internal splicing Figure 3 Schematic representations of the PML gene and of portions of PML exon 3 led in one patient to a protein as well as di€erent PML ± RARa oncoproteins. PML small PML ± RARa fusion protein which contained the exons 1 ± 9 are indicated by boxes (not drawn to scale), with di€erent shading corresponding to various PML functional RING ®ngers, B boxes and the ®rst two portions of the domains (as in Figure 1). Serine and proline (P/S) rich C-terminal a-helical coiled-coil domain, representing the minimal sequences of PML, which are excluded from the PML ± RARa PML moiety required for oncogenicity (Pandol® et al., fusion proteins, are as indicated. Open triangles indicate 1992). PML ± RARaL isoform is expressed in approxi- boundaries between protein regions encoded in di€erent PML exons. Black triangles denote fusion points between the PML and mately 55% of adult patients with APL, while the S RARa protein sequences. Positions of three translocation break- and V isoforms are expressed in approximately 35% point cluster regions (Bcr1 ± 3) within the PML gene are indicated. and 8% of patients, respectively (Fenaux and Cho- Bcr1, 2 and 3 lead to expression of long (PML ± RARaL), variable mienne, 1996; Kane et al., 1996; Slack et al., 1997). In (PML ± RARaV) and short (PML ± RARaS) forms of the fusion oncoprotein. PML ± RARaM and PML ± RARaV(M) arise by the pediatric population, the V isoform accounts for a alternative splicing of sequences encoded in exon 5 from the same larger proportion of cases than in adults (Kane et al., transcript as the L and V forms, respectively 1996), with a corresponding decrease in the number of

Oncogene APL associated RARa fusions A Zelent et al 7189 PML ± RARaS cases. In APL cells, PML ± RARa is patients who do or do not express the RARa ± PML present in great excess over wild-type RARa, making it transcripts (Li et al., 1997b). Transgenic mice harbor- the predominant in those cells (Jansen ing the RARa ± PML fusion did not develop leukemia et al., 1995; Pandol® et al., 1992). but when crossed with PML ± RARa mice, leukemia Each t(15;17) APL patient exhibits a unique set of developed with increased rate (Pollock et al., 1999). PML ± RARa fusion products indicating a single Hence, the RARa ± PML may modestly contribute to breakpoint with alternative splicing and highlighting the disease process, perhaps through interference with the clonal nature of the disease. Detection of the the p53 function (Fogal et al., 2000; Guo et al., 2000; PML ± RARa fusion transcripts by RT ± PCR is a Pearson et al., 2000). It is worth noting, however, that sensitive (Chen et al., 1992) and speci®c test for the PML ± RARa and RARa ± PML sequences, which were diagnosis of APL and can be used to measure minimal co-expressed in transgenic animals, were not exactly residual disease after chemotherapy, di€erentiation reciprocal (they derived from Bcr1 and Bcr3, respec- therapy and bone marrow transplantation (Miller et tively) and thus would have never been found together al., 1993). Reappearance of PML ± RARa transcripts in in a human APL. the marrow often precedes a leukemic relapse (Diverio et al., 1998; Fenaux and Chomienne, 1996). Initial studies indicated that patients treated with ATRA who The PLZF gene and its fusion with RARa harbored the PML ± RARaS had a high likelihood of early death or relapse (Huang et al., 1993; Vahdat et The promyelocytic leukemia zinc ®nger (PLZF) gene al., 1994). One in vitro study indicated that blasts from was identi®ed by its rearrangement in an APL with APL patients with the PML ± RARaV isoform had t(11;17)(q23;q21) (Chen et al., 1993a,b). Since then 16 decreased ATRA sensitivity (Gallagher et al., 1995). cases of of APL with a fusion between the PLZF and There was an association between the PML ± RARaS RARa genes were described (Culligan et al., 1998; isoform and more primitive morphology (Al-Omar et Grimwade et al., 1997, 2000; Jansen et al., 1999; Licht al., 1997) and secondary cytogenetic abnormalities, et al., 1995). A series of cases collected by our group in possibly due to dysfunctions in DNA repair or cell 1995 indicated that these patients were resistant to cycle control, suggesting a biological di€erences ATRA and had a generally poor response to between the isoforms (Slack et al., 1997). Despite these chemotherapy (Licht et al., 1995). More recently a potential di€erences among the PML ± RARa isoforms, case report indicated that one patient had a response to numerous studies reported consistently good clinical the combination of ATRA and G-CSF. The combina- outcomes in all APL patients (Fukutani et al., 1995; tion treatment led to the di€erentiation of leukemic Slack et al., 1997), probably due to the highly e€ective blasts in vitro, the appearance of mature myeloid cells nature of current therapy. in the peripheral blood of the patient and the Depending on the position of the breakpoint in the disappearance of PLZF ± RARa transcripts from bone RARa locus, the reciprocal RARa ± PML fusion gene marrow specimens for 7 weeks (Jansen et al., 1999). generated in t(15;17) can lead to the expression of both Another patient achieved a remission with concurrent RARa1- and RARa2 ± PML (breakpoints between ATRA and chemotherapy (Culligan et al., 1998). exon 4 and exon 5) or only RARa1 ± PML (break- Hence the ATRA resistance of this syndrome may points upstream of exon 4) reciprocal chimeric proteins not be absolute. A recent review of previously reported (Alcalay et al., 1992). RARa1 ± and RARa2 ± PML and new cases from the European working party on contain the A1 and A2 domains of the RARa proteins, APL found that 0.8% of APL patients harbored the rspectively, fused to a variable portion of PML, due to t(11;17)(q23;q21) translocation (Grimwade et al., alternative splicing, including the serine/proline rich C- 2000). In one case a cryptic translocation occurred terminal domain. It is hard to predict the e€ects of yielding the PLZF ± RARa fusion but not the RARa ± PML since the roles of the various C-termini reciprocal RARa ± PLZF indicating that at least in of PML are not well understood. Although all of the some patients this PLZF ± RARa is sucient to cause t(15;17) positive APLs express one of the PML ± RARa disease. Promyleocytes from the newer cases, like those species, the reciprocal RARa-PML gene is expressed originally collected, did not respond to ATRA in vitro, only in approximately 80% of cases (Alcalay et al., however many of the patients did respond to 1992). The lack of RARa-PML expression in up to conventional chemotherapy, sometimes in combination 20% of patients may re¯ect breakpoints in the RARa with ATRA, and achieved durable remissions. Hence intron 1 (excluding any coding sequences from the this disease, in contrast to prior impressions, can be fusion gene) or deletions within the derivative locus on chemotherapy sensitive. Whether marginal sensitivity chromosome 17q21. Therefore, it appears that in to ATRA aided in the response of these patients to general expression of the reciprocal gene is not treatment remains unclear. Re-review of the pheno- required for the development of APL. Consistently, types of APL blasts from t(11;17)(q23;q21) patients there were cases of APL associated with non-reciprocal indicated that these cells have a regular nucleus and fusions of PML and RARa, which did not generate abundant cytoplasm with either course granules or RARa ± PML fusion genes (Borrow et al., 1994; Hiorns many ®ne granules (Sainty et al., 2000). Few cases also et al., 1994). Furthermore, there are no di€erences in had Auer rods. Like the usual cases of APL the ATRA sensitivities or clinical outcomes between promyelocytes were CD347/CD13+/CD33+ but, as

Oncogene APL associated RARa fusions A Zelent et al 7190 reported in one of the original patients (Scott et al., culture and allowed to di€erentiate, PLZF levels 1994), three other cases were positive for the CD56 NK declined with terminal di€erentiation (M Hoatlin, cell antigen. It was proposed that due to their personal communication). PLZF expression may be characteristic regular nuclei and unusual immunophe- important for the maintenance or survival of hemato- notype, cases of APL with t(11;17)(q23;q21) should be poietic stem cells and/or early progenitors, appears given the subclassi®cation M3r. down-regulated with di€erentiation and may be re- The PLZF gene, encodes for a zinc ®nger transcrip- expressed in monocytes. tion factor (Chen et al., 1993b) of 673 amino acids with During mouse embryogenesis PLZF is expressed in a nine KruÈ ppel-like C2H2 zinc ®nger motifs. The N- segmental pattern (Cook et al., 1995) in the nervous terminal 118 amino acids comprise a POZ (Pox virus system. Dynamic and rhombomere speci®c expression and Zinc ®nger) or BTB (Broad Complex, tramtrack, of PLZF during hindbrain development suggested that Bric a Brac) domain. In addition to PLZF, at least one it may regulate transcription of the gene. A other gene encoding a POZ-domain and zinc ®nger PLZF DNA binding site was found in the hoxb2 5' protein (BCL-6) has been implicated in oncogenesis ¯anking region and PLZF could repress the hoxb2 (Kerckaert et al., 1993; Ye et al., 1993). PLZF and promoter in co-transfection assays (Ivins and Zelent, BCL6 were found to interact in vitro and in vivo in a 1998). PLZF expression is also notable in the limb MDS cell line with lymphoid and myeloid features and buds where hox genes help to guide the development PLZF could recruit BCL6 devoid of its own POZ (Cook et al., 1995) and mice lacking PLZF (Barna et domain to a speckled nuclear structures (Dhordain et al., 2000) display homeotic limb defects including the al., 2000). The POZ domain mediates homo- and presence of extra as well as transformed digits. This heterodimerization (Bardwell and Treisman, 1994), can phenotype was associated with a more anterior pattern act as a transferable transcriptional repression domain of expression of some hoxd genes in the developing (Chang et al., 1996; Numoto et al., 1993) and can be hind , suggesting that PLZF could globally involved in chromatin remodeling (Ra€ et al., 1995). inhibit the expression of this class of hox genes. The PLZF POZ domain forms a tight, highly Decreased expression of morphogenic protein 7 intertwined dimer (Ahmad et al., 1998; Li et al., (Bmp7) gene, which controls cell proliferation and 1997a, 1999; Melnick et al., 2000b). The top portion of programmed cell death in the developing limb, was the dimer structure forms a groove exposed to solvent also noted. The limb buds of PLZF null mice also and is lined with conserved charged amino acids showed increased proliferation and decreased apoptosis potentially representing a peptide binding site. consistent with our data demonstrating PLZF to be a PLZF is localized to the nucleus (Licht et al., 1996; growth suppressor (see below). Spinal homeotic Reid et al., 1995) and is phosphorylated on serine and changes were noted and were associated with a more threonine residues (Ball et al., 1999). Immuno¯uor- posterior expression boundary for the hoxc6 and escent microscopy showed PLZF in a pattern of *50 genes. PLZF null mice have not exhibited an obvious small nuclear speckles (Licht et al., 1996), which are hematopoietic phenotype, nor have they developed dependent on the presence of the POZ (Dong et al., leukemia nor other tumors. This does not rule out a 1996). In one study, PML and PLZF partially co- role for PLZF in hematopoiesis. FAZF/ROG, a gene localized in the PML NBs (Koken et al., 1997), wheras that encodes a protein with expression patterns and in another report they were found to exist in adjacent DNA binding speci®cities highly similar to those of nuclear strucures (Ruthardt et al., 1998). The PLZF ± PLZF (unpublished results), might partially compen- RARa fusion did not co-localize with PML or de- sate for the lack of PLZF during development (Hoatlin localize PML from nuclear body structures (Grimwade et al., 1999; Miaw et al., 2000). These data do indicate, et al., 2000; Koken et al., 1997) indicating that however, that as suspected and found for many other disruption of the PML NB per se is not required for leukemia-associated proteins (Look, 1997), such as the development of APL. In contrast, in t(15;17) APL, MLL (Yu et al., 1995), HOX genes may be the major PLZF appears de-localized into a microspeckled targets of PLZF and its fusion protein with RARa. pattern that is identical to that of PML ± RARa (He Like PML, PLZF can repress cell growth. 32DCL3 et al., 1998; Koken et al., 1997). cells, overexpressing the PLZF protein were highly PLZF mRNA is expressed in undi€erentiated growth inhibited, accumulated in G1, traversed S phase myeloid cell lines and at lower levels in more slowly and had an increased rate of apoptosis di€erentiated erythroleukemia, promyelocytic and (Shaknovich et al., 1998). PLZF expression inhibited monocytic cell lines as well as in peripheral blood myeloid di€erentiation induced by G-CSF or GM ± mononuclear cells (Chen et al., 1993b; Reid et al., CSF and led to the up-regulation of the early 1995). PLZF is down-regulated during di€erentiation hematopoietic marker Sca1. Acute infection of myeloid of NB4 and HL60 cells (Chen et al., 1993b), but is up- cells with a PLZF-containing retrovirus was associated regulated in the MDS cell line after treatment with with growth arrest of cells in the S phase of the cell calcium ionophore, perhaps recapitulating some aspect cycle, implicating cyclin A as a potential target. of monocyte development (Licht et al., 1996). CD34+ Fibroblasts expressing PLZF showed blunted induction human progenitor cells could be immunostained with of cyclin A when stimulated into the cell cycle. PLZF PLZF antisera in a distinct nuclear speckled pattern can bind and repress the cyclin A2 promoter (Yeyati et (Reid et al., 1995). When such cells were placed into al., 1999) and growth suppression by PLZF was

Oncogene APL associated RARa fusions A Zelent et al 7191 overcome by enforced expression of cyclin A2. This protein activates such reporters while PLZF ± RARa suggests that PLZF may inhibit cell growth by altering has no e€ect (Ivins and Zelent, 1998). Hence the the expression of cell cycle regulators. PLZF also RARa ± PLZF protein could act in a dominant protected 32D cells (Shaknovich et al., 1998) from negative manner, binding and altering transcription of apoptosis due to factor withdrawal. This has led us to PLZF target genes. The PLZF protein was found to hypothesize that PLZF plays a role in quiescence and contain two separable transcriptional repression do- resistance to apoptosis exhibited by hematopoietic stem mains (RPD1 and 2, see Figure 4), one of which cells. Down-regulation of PLZF during myeloid overlaps the POZ domain (Li et al., 1997a). Point di€erentiation may be accompanied by cycles of mutations in the POZ domain abrogate repression by committed cell division. It is possible that PLZF is a PLZF (Melnick et al., 2000b) revealing that the tumor suppressor disrupted by t(11;17)(q23;q21) in dimerization function of this domain is required for APL. The PLZF ± RARa fusion proteins resulting from full transcriptional repression of the isolated POZ this translocation may act as dominant negative region, as well as for repression by the full-length inhibitors of PLZF. In fact, heterodimerization PLZF protein. In addition, dimerization through the between PLZF ± RARa and PLZF seems preferential POZ domain was required for the ability of PLZF to to the formation of PLZF homomeric complexes form high molecular weight DNA-protein complexes (Dong et al., 1996). Hence, high-level expression of that contain cdc2 (Ball et al., 1999) and for its PLZF ± RARa in t(11;17)(q23;q21) blasts might seques- biological activity (Melnick et al., 2000b). ter PLZF from binding to its natural target genes and/ Repression by PLZF was shown to involve interac- or bind to limiting quantities of PLZF transcriptional tions with a number of co-repressors (mSin3A, N-CoR, co-factors (Ruthardt et al., 1997). Continuing our SMRT) and histone deacteylases and was enhanced by search for PLZF target genes using cDNA microarray co-expression of co-repressors and partially blocked by analyses we have found that c- expression is the HDAC inhibitor trichostatin A (David et al., 1998; dramatically inhibited in a U937 cell lines induced to Grignani et al., 1998; Guidez et al., 1998; Hong et al., express PLZF (McConnell et al., 2000). This correlates 1997; Lin et al., 1998; Wong and Privalsky, 1998). with a PLZF binding site (see below) in the c-myc PLZF was reported to co-localize to the same nuclear promoter and the ability to detected it in a chromatin sub-location as the SMRT co-repressor (Lin et al., immunoprecipitation assay using PLZF antisera. 1998). More recently histone dacetylases were found to Hence PLZF may inhibit proliferation by inhibiting co-localize with SMRT in such structures and enzy- the expression of multiple stimulators of the cell cycle. matic activity of the HDACs was required for their From site selection and other experiments (Ball et integrity (Downes et al., 2000). Together these ®ndings al., 1999; Ivins and Zelent, 1998; Sitterlin et al., 1997) a suggest that PLZF and other repressor and or co- T consensus sequence for PLZF binding of GTAC /AG- repressor proteins may be recruited through the POZ TAC can be derived. Such sites can be recognized by domain to a nuclear repressor organelle complex in the C-terminal four zinc ®ngers retained in the RARa ± order to inhibit their target genes. Recently we have PLZF fusion protein (Ivins and Zelent, 1998; Sitterlin found that the ETO protein which is fused to AML1 in et al., 1997). This suggests that the N-terminal zinc t(8;21)-associated AML and binds to PLZF primarily ®ngers of PLZF might not bind to DNA but through RPD2, may serve as an additional PLZF co- participate in protein ± protein interactions. Reporter repressor (Melnick et al., 2000b). It should be noted genes containing PLZF sites (David et al., 1998; however, that in addition to HDAC recruitment other Guidez et al., 1998) are repressed by co-expressed mechanisms of transcriptional repression by PLZF PLZF protein. In contrast, the RARa ± PLZF fusion could be at play. For example, cdc2, which seems to

Figure 4 Schematic representations of the PLZF gene and protein, indicating relative positions of the translocation breakpoints and various functional regions. Boxes labeled 1 ± 6 represent various exons of the PLZF gene (not drawn to scale) and distribution of various PLZF functional regions among them are as indicated. POZ domain is indicated as before and nine zinc-®nger motifs are represented with black circles. Relative positions of repression domains 1 and 2 (RPD1 and RPD2) and the DNA binding region of the zinc-®ngers (underlined) are indicated

Oncogene APL associated RARa fusions A Zelent et al 7192 Table 1 Comparison of RARa fusion partner proteins PML PLZF NPM NuMA STAT5b

Isoforms Up to 20 splicing One major isoform Two major isoforms. At least three Two isoforms due to products identified A third associates with proteolytic cleavage, nuclear matrix also a separate STAT5a gene

Secondary N-terminal Pro-rich RING N-terminal POZ Two acidic domains, one Central coiled-coil Coiled-coil, DNA binding structure and B box coiled coil. domain; C-terminal metal binding motif, two flanked by two domain, SH2, SH3 motifs Variable acidic C-terminal kruppel like Zn NLS, ATP-binding site globular domains region fingers

Homodimer Through the coiled-coil Through the POZ Forms hexamers through Through coiled-coil Through C-terminal domain domain N- and C-terminal domain SH2 domains and N- portion terminal coiled-coil domain

Heterologous Through the coiled-coil, Through the POZ Through the acidic Through C-terminal Through N-terminal interaction N-terminal Pro-rich domain, a second domains and C-terminal domain coiled-coil motif domain, and possibly repression domain region RING and B box and the first two Zn domains fingers

Phosphoryl- Ser and Tyr residues. Ser and Thr residues. Casein kinase II, nuclear Cdc2 kinase, cyclic By receptor tyrosine ation Substrate of cyclin A/ Possible cdc2 kinase II, PKC and cdc2 AMP kinase, PKC, kinases and Jak kinases cdk2 has casein kinase phosphorylation kinases, cyclinE/cdk2 Ca/calmodulin II sites kinase

Nuclear Mainly in nuclear bodies. Nuclear speckles, Principally nucleolar, also Interphase ± diffuse Basal state in cytoplasm, localization Also in cytoplasm and partial overlap with nucleoplasm. Shuttles to and speckled to nucleus after other nuclear regions PML cytoplasm Mitosis ± binds phosphorylation spindle poles

Nuclear Associates with nuclear Yes ± delocalized from One isoform associates, During interphase ? matrix matrix matrix by AML1- with the matrix association ETO

Expression Present in inflammatory CD34+ progenitors Ubiquitious Ubiquitious, except Widely expressed, breast, pattern tissues myeloid precursor macrophages, mouse in some terminally liver, hematopoietic cells. Induced by IFNs embryo CNS, liver differentiated cells progenitors, T cells, heart, kidney, limb NK cells and tail buds

Cell cycle Increase in G0 to G1 Blocks cells in G1/S Peaks at onset of S phase Phosphorylated in Phosphorylated by transition and decreases correlating with ± declines as cells enter G2. G2/M. Essential for cytokines that drive as cells progress to S phase down-regulation of Phosphorylated to M phase the cell into S phase cyclin A stimulate centrosome duplication

Transcription Transcriptional repression Transcriptional Modulates transcriptional Co-localizes and Transcriptional and RNA and activation. Involved repression effects of YY1 and IRF-1. co-precipitates activation metabolism in retinoid receptor Involved in ribosome with snRNPs and signaling. biogenesis splicing complexes

Protein SUMO-1, PLZF, Rb, L7 PML, LRF, SMRT, Rev, Tat, Rex, nucleolin, Tubulin and mitotic STAT5b, Nmi, p300, partners , EF-1, N-Cor, HDAC, and p120, YY1, IRF-1 spindle CBP, protein ribosomal P proteins, sin3a, sin3b, Rb, cdc2, phosphatase, Daxx and others ETO PI3Kinase

Action on Growth suppression in Growth suppression Expression highest in Required for Stimulates cell growth cell growth NB4, HeLa and CHO differentiation block tumor cells, over- proper completion required for immune cells blocks transforma- expression causes 3T3 of mitosis function tion in rat embryo cells to transform fibroblasts and 3T3 cells

Apoptosis Removal delays apoptosis; Promotes cell Becomes hypophosphoryl- Specifically targeted STAT5 activation association with SUMO-1 survival with factor ated and degraded during for proteolysis by prevents apoptosis and targeting by As203 withdrawal apoptosis caspases imply role in apoptosis

Continued

Oncogene APL associated RARa fusions A Zelent et al 7193 Table 1 (Continued ) PML PLZF NPM NuMA STAT5b Miscellan- Targeted during certain DNA binding site Nucleic acid binding Attaches to DNA Binds g interferon T eous viral infections GTAC /AGTAC matrix attachment response element regions (MAR)

Knock out K/O mice susceptible to Musculoskeletal and None published to date None published to STAT5b Null mice data infections, susceptible to limb defects, impaired date have deficiency in NK transforming agents, lack spermatogenesis, T cell function. Stat5a/b IFN-induced growth cell lymphopenia null mice have defective suppression defective cytokinase response. induction of p21 by Stat5b prevents apoptosis ATRA during myeloid maturation

Function Involved in growth Growth suppressor, Ribonucleoprotein Structural role in Transcriptional activator suppression, different- transcriptional maturation and transport interphase and in implicated in hemato- iation, and immune repressor, control of shuttle proteins between particular mitotic poietic and immune cell response pathways. developmental pro- cytoplasm and nucleolus. cells. Major target growth proliferation and Possible role in translation. grams and different- Transcriptional modulator. of apoptosis pro- function Transcriptional modulator iation, possibly Role in DNA recombin- gram through hox genes ation

associate with the high molecular PLZF ± DNA protein RARa. Curiously, inhibition of wild-type RARa complexes was previously implicated in transcriptional function was partially dependent on the presence of repression by phosphorylation of basal transcription the ®rst two PLZF zinc ®ngers, present in the fusion factors (Long et al., 1998). protein. When the POZ domain and ®rst two PLZF The t(11;17)(q23;q21) fusion yields two reciprocal zinc ®ngers were deleted from the fusion protein, transcripts (Chen et al., 1993b; Licht et al., 1995). The PLZF ± RARa became an ecient activator of ATRA- resulting PLZF ± RARa fusion contains the entire N- mediated transcription (Dong et al., 1996). As for terminal transcriptional e€ector region of PLZF and PML ± RARa, repression by PLZF ± RARa is predo- the ®rst two zinc ®ngers and in one case due to an minantly due to its high anity for the HDAC alternative 3' breakpoint in the PLZF gene, three zinc containing co-repressor compexes (David et al., 1998; ®ngers (see Figure 4). In four of seven cases tested, a Grignani et al., 1998; Guidez et al., 1998; He et al., reciprocal RARa ± PLZF transcript was detected, 1998; Hong et al., 1997; Lin et al., 1998). This appears linking the ligand independent activation domain of to be caused by the presence of four docking sites for RARa (Nagpal et al., 1992) to the last seven PLZF co-repressors on each oncoprotein homodimer rather zinc ®ngers. PML ± RARa and PLZF ± RARa have than two sites found in normal RXR/RAR hetero- some similarities as aberrant receptors. Both fusion dimers (Lin and Evans, 2000). However, due to the proteins can bind as homodimers to retinoic acid ability of the PLZF POZ domain to bind the co- response elements (RARE) (Dong et al., 1996; repressors in an ATRA insenstive manner, while Hauksdottir and Privalsky, 2001; Licht et al., 1996; PML ± RARa releases these factors in the presence of Perez et al., 1993). In PML ± RARa this is mediated by 1076 M ATRA, PLZF ± RARa does not (Grignani et the coiled-coil motif, and in PLZF by the POZ al., 1998; Guidez et al., 1998; He et al., 1998; Hong et domain. In combination with RXR PLZF ± RARa al., 1997; Lin et al., 1998). Additional binding of ETO forms multiple DNA-protein complexes, but the co-repressors to the RPD2 of PLZF may contribute to PLZF ± RARa/RXR heterodimer binds to RAREs its overall strength of transcriptional repression and with higher anity than PLZF ± RARa homodimers insensitivity to ATRA (Melnick et al., 2000a,b). (Dong et al., 1996; Licht et al., 1996). The PLZF ± HDAC inhibitors, such as trichostatin A and sodium RARa/RXR heterodimer bound to the RARE less butyrate, allowed cells harboring PLZF ± RARa to eciently than wild-type RXR/RARa perhaps due to respond to ATRA, presumably by inactivating the co- the ability of the POZ domain to multimerize and repressor complexes bound to the POZ domain preclude ecient DNA binding (Bardwell and Treis- (Grignani et al., 1998; He et al., 1998; Lin et al., 1998). man, 1994). Hence PLZF ± RARa might sequester Transgenic animals harboring PLZF ± RARa devel- limiting amounts of RXRa, an essential co-factor for oped a CML-like syndrome rather than APL (Cheng et RARa function. Furthermore, PLZF ± RAR/RXR al., 1999; He et al., 1998). Unlike PML ± RARa heteromers also bound to non-consensus RAREs, transgenic mice, the PLZF ± RARa mice did not potentially contributing to leukmogenesis by misdirect- achieve durable CR after ATRA treatment, although ing the chimeric protein to novel genes (Hauksdottir cells did show some evidence of di€erentiation and and Privalsky, 2001; So et al., 2000). Like PML ± RARa, leukemic cells from the mice treated with ATRA PLZF ± RARa is relatively weak trans-activator (Chen di€erentiated ex vivo (He et al., 1998). The relative et al., 1994b; Dong et al., 1996; Licht et al., 1996) and insensitivity of the disease correlates with the impaired serves as a dominant negative inhibitor of wild-type ability of PLZF ± RARa to transactivate due to

Oncogene APL associated RARa fusions A Zelent et al 7194 Table 2 Comparison of N-protein/RARa fusion products PML ± RARa PLZF ± RARa NPM ± RARa NuMA ± RARa STAt5b ± RARa

Breakpoint Three breakpoint Most cases include first Two fusion cDNAs Only one breakpoint One breakpoint ± variants clusters result in three two Zn fingers alternative splicing internal deletion of major forms chromosome 17

N-protein All variants contains: POZ, first two ZN Oligomerization domain, N-terminal globular Coiled-coil domain, structures RING, B-box coiled- fingers metal binding domain and central coiled DNA binding domain coil and first acidic domain region SH3 domain, part of SH2 domain

Nuclear Localizes to *100 micro- Localized to micro- Microspeckled pattern Sheetlike aggregates Nuclear microspeckled localiza- speckles, which are speckles, not to tion distinct from NBs. May nuclear bodies nor also localize in cytoplasm cytoplasm

Homo- Through coiled-coil Through the POZ Presumably through Presumably through Yes ± through coiled- dimerization domain domain oligomerization domain NuMA coiled domain coil domain

Hetero- PML, RXR, SMRT, PLZF, RXR, SMRT, Interaction with RXR, Unknown. Possible JAKs, Nmi, p300/ logous N-Cor, HDAC1 N-Corr, HDAC1, co-repressor SMRT and interaction with CBP, TRAM-1 interactions sin3A, ETO coactivators Rac-3, RXR TIF140

Transcript- Dominant negative for Dominant negative for ATRA-dependent trans- Unknown Dominant negative for tional retinoids. Avid binding retinoids. Avid binding activation and dominant retinoids. Increased effects to co-repressors, relieved to co-repressors, not negative effect on wild- binding to co-repressors. only by high dose ATRA relieved by high dose type RAR. Increased Can also interfere with ATRA affinity for co-repressors STAT-mediated tran- scription

Clinical Sensitive to ATRA Generally insensitive to Sensitive to ATRA Sensitive to ATRA ATRA as a single agent? Synergy with G-CSF

Other Re-localizes PML to ATRA degrades PLZF ± Induces differentiation Not yet reported Fusion protein responds effects of NBs, degrades PML- RARa but does not and inhibits growth to ATRA but disease ATRA RARa, upregulates induce differentiation is not possible due to RARa, differentiation ATRA+HDAC STAT interference inhibitor suppresses growth and promotes differentiation

Effects of Induces CR in APL, Fails to degrade PLZF ± Not yet reported Not yet reported Not reported arsenic rapid degradation of RARa or induce PML ± RARa and apoptosis apotosis

Transfected Inhibits differentiation, Blocks differentiation, Blocked differentiation Not yet reported Not yet reported cell models protects from apoptosis, fails to increase and enhanced enhances proliferation sensitivity to ATRA, proliferation does not transform cells murine marrow progenitors could be serially passaged

Transgenic All develop disease after Mice develop CML like APL sensitive to ATRA Not yet reported Not yet reported models a portracte prodromal syndrome earlier than period. Several models. PML ± RAR. Poor Closest to human APL response to ATRA is MRP8 promoter construct. Animals respond to ATRA

Miscell- Resistance to ATRA May bind to different Does not localize to NB, Does not localize to Could also bind to aneous caused by mutations RAREs compared to does not delocalize PML to NB, does not STAT and target genes in ligand-binding domain PML ± RARa but does delocalize de-localize PML PLZF Continued

Oncogene APL associated RARa fusions A Zelent et al 7195 Table 2 (Continued ) PML ± RARa PLZF ± RARa NPM ± RARa NuMA ± RARa STAt5b ± RARa Model Multimerization, Multimerization, Multimerization, Multimerization, Sequesttration of co- sequestration of RXR sequestrations of RXR sequestration of RXR sequestration of activators, RXR and other factors. and other factors. and other factors. RXR and other Increased affinity for Increased affinity for Transcriptional effects factors. Interferance co-repressors. co-repressors. on target genes. with apoptosis Transcriptional effects Transcriptional effects Interference with program on target genes. on target genes. PLZF actions Interference with Reciprocal fusion PLZF actions may play a role

Reciprocal Present in 70 ± 80% of Present in all cases Identified in the index Not yet reported Not present translocation cases. Unclear role in tested. Activates PLZF case. Actions still leukemogenesis target genes and induces unknown cell proliferation binding of co-repressors. This idea was solidi®ed by the months and penetrance of 100% did not change. ®nding that the histone deacetylase inhibitor and However, there was a change in the morphology of the ATRA synergistically induced di€erentiation of these leukemia that developed. There was a doubling of leukemic cells. Although ATRA induced degradation promyelocytes and blasts in the marrows of the double of the PML ± RARa and PLZF ± RARa fusion proteins transgenic mice to about 40%, associated increased (Koken et al., 1999), no response of the PLZF ± RARa expression of early di€erentiation markers c-kit and mice could be found. This could be due to a qualitative CD34. Cells from the double trangenic mice had a di€erence in the degradation of these proteins induced higher proliferative index, lower sensitivity to apoptosis by ATRA. A speci®c cleavage in the PML ± RARa induction and less di€erentiation in vitro in response to fusion is induced by ATRA that can yield a protein ATRA than cells derived from single transgenic that likely is devoid of most of the PML moiety and PLZF ± RARa mice (He et al., 2000). Remarkably an can likely transactivate RARa target genes. In contrast, almost identical phenotype was seen when PLZF ± PLZF ± RARa appears to be completely degraded by RARa trangenic mice were bred into the PLZF null RARa with no such intermediate product, hence a background. This information is in accordance with transient boost in RARa target may our earlier results suggesting that RARa ± PLZF is a not be expected in these cases. Though helping to dominant negative form of PLZF and a growth explain the response to ATRA, these animal model do activator. It is possible that through aberrant expres- not explain the poor response of some t(11;17) sion of PLZF targets such as HOX genes, cyclins and (q23;q21) patients to chemotherapy and the complete other cell cycle regulators, the presence of both fusion resistance of fresh APL cells to high doses of ATRA in genes lead to a more aggressive and undi€erentiated vitro (Guidez et al., 1994; Licht et al., 1995). These form of leukemia. ®ndings implicate the reciprocal RARa ± PLZF protein in the aggressive nature of PLZF ± RARa associated APL. Nucleophosmin and NPM ± RARa The reciprocal transcript encoding RARa ± PLZF is consistently expressed in nearly all of the In t(5;17) associated APL, RARa is translocated to a t(11;17)(q23;q21) patients (Grimwade et al., 1997; Licht region on chromosome 5q35 containing the ubiqui- et al., 1995). The seven zinc ®ngers of RARa ± PLZF tously expressed and evolutionarily conserved nucleo- can bind to the same site as full-length PLZF (Ball et phosmin (NPM) gene (Redner et al., 1996b). In al., 1999; Li et al., 1997a). While PLZF represses gene hematologic malignancies NPM is also fused to genes transcription, RARa ± PLZF activates it (Li et al., other then RARa, as in the t(2;5)(p23;q35) transloca- 1997a). Whereas PLZF is a growth suppressor, tion found in Ki-1+ anaplastic large cell lymphoma, for RARa ± PLZF activates expression of cyclin A2 example (Morris et al., 1994; Shiota et al., 1995). In (Yeyati et al., 1999) and enhances cell growth (P this situation, NPM is linked to Alk, a gene encoding a Yeyati, R Shaknovich, J Licht, unpublished results). membrane spanning tyrosine kinase (Morris et al., Hence t(11;17)(q23;q21) may be an ATRA and 1994), which is normally not expressed in lymphoid chemotherapy resistant disease due to the presence of tissue (Iwahara et al., 1997). As a result, the two oncogenes. PLZF ± RARa blocks the di€erentia- ubquitously expressed NPM gene drives the expression tion inducing action of ATRA, while RARa ± PLZF of an aberrant fusion tryosine kinases (Bischof et al., may activate cell cycle regulators, blocking the anti- 1997; Shiota et al., 1995). In the t(3;5)(q25.1;q34) proliferative e€ects of ATRA. In accordance with this translocation, found in myelodysplasia and M6-AML, idea, mice harboring the RARa ± PLZF protein a larger 175 amino acid portion of NPM is linked to develop a myeloproliferative syndrome and splenome- the MLF1 gene, which encodes an abundant cytoplas- galy (He et al., 2000). When PLZF ± RARa mice were mic protein of unknown function (Yoneda-Kato et al., crossed with RARa ± PLZF mice, leukemic latency at 6 1996).

Oncogene APL associated RARa fusions A Zelent et al 7196 NPM was initially isolated as a nucleolar phospho- nucleolar disassembly to mitotic chromosome con- protein in hepatoma cells (Hernandez-Verdun et al., densation (Peter et al., 1990). A fraction of the protein 1982; Olson et al., 1974). The human NPM gene spans can also co-localize with NuMA (see below) at the 25 kb, consists of 12 exons (Chan et al., 1997) and spindle poles during mitosis (Zatsepina et al., 1999). alternative splicing of its transcripts yields two major Recently NPM was identi®ed as the major substrate of isoforms, which di€er in their C-terminal region (Chan cyclin E/cdk2 in centrosomes. NPM associated with et al., 1997, 1989; Lee and Welch, 1997). Structural only unduplicated centrosomes and it was released features of NPM include: two Asp/Glu rich acidic from the duplicated structures after phosphorylation domains, potential binding sites for other proteins; a (Okuda et al., 2000). The function of NPM in the bipartite nuclear localization signal (NLS); a metal centrosome may relate to its chaperone activity, binding motif; an ATP binding site; phosphorylation keeping proteins in their proper conformation, and sites for cdc2 kinase and casein kinase II and a binding may regulate the duplication of centrosomes. site for proteins which contain nucleolar localization NPM is preferentially dephosphorylated and de- signals (Chan et al., 1986a,b, 1989; Li et al., 1996). graded during apoptosis and cell damage (Taw®c et al., Both isoforms reversibly multimerize to a hexameric 1995). NPM reversibly de-localizes from the nucleolus state via the N-terminal domain (Chan and Chan, to the nucleoplasm when cells are exposed to 1995). NPM is localized most prominently to areas of conditions that discourage DNA or RNA synthesis the nucleolus associated with ribonucleoprotein (RNP) (Yung et al., 1985a,b) or encourage terminal cell processing (Dumbar et al., 1989; Smetana et al., 1984). division (Chan et al., 1987). Delocalization of NPM NPM functions as part of a transport system used by from the nucleolus is an early event in drug-induced ribosomal precursors and other proteins to shuttle apoptosis (Chan and Chan, 1999). Whether the between the cytoplasm and nucleolus (Szebeni et al., delocalization is a result of or a trigger for further cell 1997). death events is not certain. However, protection of NPM levels are augmented in proliferating cells HL60 cells from apoptosis induced by drugs and (Derenzini et al., 1995) and leukemic blasts (Kondo et potentiation of cell death due to serum withdrawal by al., 1997), perhaps due to an increased requirement for overexpression of sense and anti-sense NPM tran- ribosomal precursors. However, overexpression of scripts, respectively, is consistent with its regulatory NPM transformed NIH3T3 (Kondo et al., 1997) cells, role in apoptoisis (Chou and Yung, 2001; Liu et al., while down-regulation of NPM delayed cell entry into 1999). Thus like PML, NPM underogoes changes with mitosis (Jiang and Yung, 1999), suggesting that NPM cell stress, indicating that both proteins may measure could have an active role in growth control. In support or control cell homeostatic processes. High levels of of this concept, it was found that NPM binds to the NPM found in cancer cells might promote both cell tumor suppressor IRF-1 and inhibits its ability to proliferation and prevent programmed cell death. activate genes, which mediate the anti-proliferative Recently the p14ARF protein was found to partially e€ect of IFN (Kondo et al., 1997). Furthermore, NPM co-localize with NPM in the nucleolus (Lindstrom et binds to YY1, involved in cell al., 2000). Whether the two proteins directly interact or growth, changing it from a repressor to an activator of whether NPM has any role in the ARF/MDM2/p53 transcription (Inouye and Seto, 1994; Shi et al., 1997). regulatory circuit (reviewed in Sherr and Weber, 2000) ATRA-induced di€erentiation of HL60 cells, but not is unknown. growth arrest by serum withdrawal, resulted in down The t(5;17)(q35;q21) translocation was ®rst described regulation of NPM (Hsu and Yung, 1998). When in a 2 year-old girl with APL (Corey et al., 1994) who cellular NPM levels were decreased by an antisense achieved cytogenetic remission after treatment with oligonucleotide, ATRA-induced di€erentiation was ATRA and chemotherapy. Blasts from this index augmented (Hsu and Yung, 1998) while overexpression patient, when thawed and treated with ATRA, of NPM prevented di€erentiation (Hsu and Yung, di€erentiated into mature granulocytes (Redner et al., 2000). Growth suppressive IRF-1 is up-regulated by 1997). Subsequently only two other cases were ATRA during myeloid di€erentiation, in opposition to con®rmed at the molecular level (Grimwade et al., the e€ect of NPM (Matikainen et al., 1996). These 2000), one of whom has had a prolonged survival after results support a role for NPM in both control of cell ATRA and bone marrow transplantation (Hummel et growth and cross-talk between the retinoid and IFN al., 1999) and the other remaining in CR after regulatory pathways. chemotherapy (Grimwade et al., 2000). This gene NPM expression peaks at S or G2 phase and is at rearrangement joins the NPM gene 5' to exon 3 of minimal levels in G0 (Feuerstein et al., 1988; Sirri et al., RARa, in a manner similar to the other APL fusions 1997). This might be related to the fact that NPM (Redner et al., 1997), yielding two isoforms of the speci®cally stimulates the activity of DNA polymerase chimeric oncoprotein. Both forms contain the oligo- a (Takemura et al., 1994) or may simply be a re¯ection merization domain of NPM. Like PML ± RARa and of the metabolic demand of the cell. During G2 and M PLZF ± RARa, the NPM ± RARa fusion acts as a phase, NPM is heavily phosphorylated by cdc2 (Peter ligand-dependent transcriptional activator when co- et al., 1990). During M-phase progression NPM expressed with reporter genes containing RAREs, yet associates with perichromosal regions and pre-nucleo- can act as a dominant negative inhibitor of wild-type lar bodies thus functionally linking the processes of RARa through aberrant interaction with co-repressor

Oncogene APL associated RARa fusions A Zelent et al 7197 molecules (Redner et al., 2000; So et al., 2000). Like mitotic spindle (Compton and Cleveland, 1993; Yang other APL fusions, NPM ± RARa could form homo- et al., 1992; Yang and Snyder, 1992). NuMA might dimers, thus exposing the RARa moiety in such a way play a structural role in anchoring the spindle poles in as to o€er extra binding sites for co-repressors, the cell (Dionne et al., 1999). necessitating pharmacological doses of ATRA to NuMA is regulated across the cell cycle by post- release them and recruit co-activators to the fusion translational modi®cations. For example, NuMA is proteins. Enforced expression of NPM ± RARa in phosphorylated by the cdc2/cyclin B regulatory kinase U937 cells blocked monocytic di€erentiation in at the initiation of mitosis (Hsu and Yeh, 1996; Saredi response to vitamin D3 and TGFû (Redner et al., et al., 1997) and associates with the spindle micro- 1996a). It was further found that as PML ± RARa, tubules. NuMA may be required for the formation of PLZF ± RARa and NPM ± RARa could enhance the the nuclear spindle poles and spindle aster, as well as proliferation of primitive marrow progenitor cells. for the organization of daughter nuclei towards the end Treatment of these cells with ATRA, induced di€er- of cell division (Gordon et al., 2001; Lyderson and entiation and inhibited cell growth (Du et al., 1999). A Pettijohn, 1980; Yang and Snyder, 1992). At the end of transgenic model of t(5;17)-APL was recently created mitosis, NuMA undergoes proteolytic cleavage and de- utilizing the cathepsin G promoter. These mice phosphorylation (Hsu and Yeh, 1996; Saredi et al., developed an APL like syndrome after a latent period 1997). As cells progress towards G1, the remaining and blasts derived from these animals were ATRA NuMA reverts to an insoluble form, yielding a ®brous sensitive (Cheng et al., 1999). network of 12 arm oligomers that may play a Similarly to PML ± RARa and PLZF ± RARa, structural role in forming the nuclear sca€old NPM ± RARa is expressed in a microspeckled pattern (Harborth et al., 1999). In interphase cells NuMA is throughout the nucleus (Hummel et al., 1999) and APL localized in di€use and speckled nuclear patterns blasts from t(5;17) patients exhibited a normal PML (Kempf et al., 1994; Lyderson and Pettijohn, 1980) NB con®guration (Grimwade et al., 2000; Hummel et and like the other RARa partners (see Table 1) in APL al., 1999), supplying additional evidence that disruption it is associated with the nuclear matrix (Zeng et al., of the PML NB is not required for the pathogenesis of 1994). NuMA speci®cally attaches to DNA matrix APL. However, PLZF was de-localized in t(5;17) APL attachment regions (MAR), important for chromatin cells in a pattern distinct from its wild-type distribution compaction and isolation of transcriptionally active in progenitors (Hummel et al., 1999), lending support loops of DNA (Luderus et al., 1994; Tsutsui et al., for a broader role of the PLZF protein in APL. A 1993). NuMA has been found to bind at least one reciprocal RARa ± NPM mRNA, leading to fusion putative transcriptional regulator and hence could between the A domain of RARa and the acidic participate in the regulation of transcription (Harborth domains, NLS and the rest of the C-terminus of et al., 2000). NPM, was identi®ed in the index t(5;17) patient NuMA is an early target for proteolysis by caspase-3 (Redner et al., 1996b) and one other case of t(5;17) and caspase-6 (Zweyer et al., 1997), yielding a APL (Grimwade et al., 2000). The importance of this proteolytic product (Casiano et al., 1996; Hsu and putative protein is unknown. Yeh, 1996), that could act as a dominant negative factor, disrupting normal nuclear structure. Release of NuMA from the DNA matrix attachment regions may Nuclear matrix ± Mitotic apparatus protein (NuMA) facilitate the DNA fragmentation characteristic of apoptosis (Gueth-Hallonet et al., 1997). In summary, NuMA is an abundant, conserved and ubiquitously NuMA is a structural component of the cell that expressed protein involved in the completion of responds to cell cycle signals on cue rather than a mitosis and re-formation of nuclei in the post-mitotic controlling factor in cell proliferation. It seems unlikely daughter cells (Compton and Cleveland, 1994; Comp- that inhibition of NuMA function might contribute to ton et al., 1992; He et al., 1995; Saredi et al., 1996). oncogenesis. The NuMA gene, located on chromosome 11q13 The NuMA ± RARa fusion protein was described in (Sparks et al., 1993), encodes a protein of 2115 amino a 6 month old boy with APL harboring a transloca- acids with a molecular weight of approximately tion t(11;17)(q13;q21) (Wells et al., 1997). The patient 230 kDa. Alternative splicing can also yield 1776 had ATRA therapy and was disease free at 24+ and 1763 amino acid proteins (Tang et al., 1993, months after a bone marrow transplant (Wells et al., 1994). The NuMA protein is divided into two 1996). The t(11;17)(q13;q21) results in a 2286 amino globular domains at either end of the protein, with acids protein predicted to consist of 1883 amino acids a central coiled-coil region of 1485 amino acids of NuMA, including its N-terminal globular and (Compton et al., 1992; Yang et al., 1992). The coiled-coil domains, fused to the RARa domains B- coiled-coil motifs likely mediate protein homo- and F (Wells et al., 1997). NuMA-RARa localized to hetero-association (Harborth et al., 1995; Yang et al., sheetlike nuclear aggregates in leukemic cells from the 1992). The C-terminal region contains basic sequences, patient, but the PML-NB structures remained un- motifs for phosphorylation by cdc2 and other kinases perturbed. Introduction of C-terminal mutants of (Hsu and Yeh, 1996; Yang et al., 1992), and sequences NuMA into cells completely disrupts mitosis (Comp- which confer localization to the nucleus (NLS) and ton and Cleveland, 1993; Maekawa and Kuriyama,

Oncogene APL associated RARa fusions A Zelent et al 7198 1993) and it is surprising that the existence of NuMA- to STAT5a but di€ers in the C-terminal activation RARa is even compatible with cell division. domain (Lin et al., 1996; and reviewed in Lin and The most likely mechanism of the NuMA ± RARa Leonard, 2000). STAT5 was originally identi®ed as a action is interference with function. factor important in mammary gland response to Although no functional data for the NuMA ± RARa prolactin and knockout studies showed that STAT5a fusion are yet available, it is likely that, as the other is required for breast development and lactation RARa fusion proteins, NuMA ± RARa is a dominant (Teglund et al., 1998). STAT5b null mice have a negative retinoid receptor (see Table 2). Like PML, defective response to growth hormone, no overt PLZF and NPM, NuMA can multimerize (Harborth et quantitative hematopoietic defects, but have a defect al., 1995; Yang et al., 1992). NuMA ± RARa might in NK cell killing (Imada et al., 1998). Compound null thus sequester RARa partner proteins and/or have mice have a severe defect in response to CSFs and a aberrant anity for nuclear receptor co-activators or de®ciency in T cells (Teglund et al., 1998). STAT5b co-repressors due to the formation of homodimers (Lin target genes relevant to hematopoiesis include c-myc and Evans, 2000). Like PML, NuMA is a component and the IL-2 receptor (Matikainen et al., 1999), and of the nuclear matrix and NuMA ± RARa might STAT5 activation was recently identi®ed as a crucial further inhibit RARa function by con®nement of RARa function of the bcr-abl oncoprotein (Gesbert and co-factors in a nuclear compartment di€erent from that Grin, 2000; Nieborowska-Skorska et al., 1999), usually occupied by the wild-type RARa.Itis although this latter point remains somewhat contro- unknown whether a reciprocal RARa ± NuMA tran- versial (Sexl et al., 2000). Stat5b de®cient mice show an script is expressed in this disease. In fact, there might increase in apoptosis in response to GM-CSF induced be selective pressure against this protein as it could di€erentiation (Kieslinger et al., 2000). In addition, represent a detriment to mitosis. Stat5b can induce the expression of the anti-apoptotic Bcl-x gene and therefore may play a role in preventing programmed cell death during terminal myeloid STAT5b ± RARa fusion di€erentiation. Stat5b may also promote cell survival by acting as an intermediate activating the PI3 kinase/ The newest fusion partner with RARa, identi®ed in a AKT signaling pathways (Santos et al., 2001). Stat5b patient with M1 AML but with a proportion of blasts may also promote oncogenesis, since the protein could exhibiting microgranular APL morphology (Arnould et accelerate transformation and cell cycle progression al., 1999), is the STAT5b gene. Leukemic blasts from induced by v-src, while a dominant negative form of this patient failed to respond to ATRA in vitro. Like the protein could inhibit transformation (Kazansky RARa the STAT5b gene is localized on chromosome and Rosen, 2001). Hence like PLZF, STAT5b 17q21.1-21.2 and the two genes are estimated to be represents a transcription factor fused to the RARa, about 3Mb apart. The patient with the STAT5b ± RARa which itself may play a role in normal and malignant fusion had a deletion of chromosome 17 leading to the blood development. creation of a fusion intron linking the STAT5b and The STAT5b ± RARa fusion contains the majority of RARa loci. As in the case of other forms of APL, the the STAT5b protein. The N-terminus consist of a breakpoint in the RARa locus was within the same coiled-coil region that mediates dimerization and highly localized region 5' of the third coding exon. No interaction with a protein called Nmi1, which enhances reciprocal RAR ± STAT5b transcript was detected. STAT binding to the p300/CBP co-activators (Zhu et The seven members the mammalian STAT proteins al., 1999). This domain might allow the STAT5b ± are transcription factors which are resident in the RARa to sequester co-activators from the remaining cytoplasm until they are phosphorylated on speci®c wild-type receptor in the cell. Located centrally in the tyrosine residues following activation of various protein is a DNA binding domain followed by SH3 cytokine receptors. Binding of a given cytokine to a and truncated SH2 domains. The STAT proteins speci®c receptor induces its dimerization and, for a dimerize through their SH2 domains after they become receptor with an intrinsic kinase activity, activation of phosphorylated. The C-terminal tyrosine 699, required tyrosine kinase. Receptors without intrinsic kinase for STAT5b activation, is missing in the fusion protein activities are non-covalently attached to JAK kinases, as is a more C-terminal serine residue whose which when brought into close proximity can in a phosphorylation activates STAT5b transcription func- similar way cross-phosphorylate and activate each tion. Nevertheless, a preliminary report indicates that other. STAT5b, like other STAT proteins, binds to the STAT5b ± RARa fusion can form a homodimer as the JAK kinase through an SH2 domain and then is well as a heterodimer with RXR on RAREs (Dong itself tyrosine phosphorylated at a C-terminal residue. and Tweardy, 2000) and can act as a dominant STAT5b can then homodimerize, migrate to the negative form of RAR when co-expressed with wild- nucleus and bind to a GAS element (Darnell, 1997). type receptor in a reporter assays. Deletion of the N- STAT5b is widely expressed in a number of tissues terminal coiled-coil region of the Stat5b protein including hematopoietic progenitors and can be prevented homodimer formation. Like other homo- activated by a number of cytokines including prolactin, dimeric forms of RARa in APL, the STAT5b-RARa GM-CSF, EPO and IL-3 (Chretien et al., 1996; Dong fusion protein had a high anity for the SMRT co- et al., 1998; Mui et al., 1996). STAT5b is very similar repressor, only releasing it at 1076 M ATRA. As a

Oncogene APL associated RARa fusions A Zelent et al 7199 heterodimer, the fusion had a high capacity for RARa ± N fusion protein could act either as a sequestration of RXR. Hence this chimera seems quite dominant negative factor for the N protein or sequester similar in function to the other APL fusion oncopro- RAR co-factors. teins. For the past 10 years the study of the molecular A further potential of the STAT5b ± RARa protein, pathogenesis of APL is at the forefront of molecular however, is its ability to interfere with STAT signaling. medicine, as this disease has proven the concept of STAT5b ± RARa retains the N-terminal portion of the di€erentiation therapy of leukemia (Waxman, 1978). STAT5b SH2 domain in the fusion protein and could The study of the resistant and sensitive forms of APL, compete with the wild-type STA5b for docking with has indicated the critical importance of aberrant cytokine receptors. However, in the index patient the transcriptional repression as a root cause of leukemia. STAT5b protein was localized in the nucleus while in This led to the use of the histone deacetylase inhibitor control marrow cells STAT5b was localized to the phenyl butyrate in combination with ATRA as a form of cytoplasm (Arnould et al., 1999). Nevertheless, it was targeted `transcription therapy' in a patient with resistant shown that the STAT5b ± RARa fusion could hetero- APL (Warrell et al., 1998) and to the further develop- dimerize with wild-type STAT5b in transfected cells and ment of HDAC inhibitors, such as SAHA (Richon et al., inhibited prolactin-stimulated gene expression (Dong 1998), for di€erentiation therapy. It should be noted, and Tweardy, 2000). In contrast, the STAT5b ± RARa however, that compounds without any HDAC inhibi- fusion enhanced IL6-mediated transcription from a tory activities, such as HMBA (a compound structurally Stat3-dependent promoter. The potential pleiotropic related to SAHA) or dithiophenes for example, exert the e€ects of the STAT5b ± RARa fusion might deregulate same e€ects on enhancing ATRA mediated di€erentia- the ability of STAT5 or other STATs to respond to tion of APL cells (Chen et al., 1994a; Waxman et al., cytokines and lead to altered cell proliferation and/or 1999). It remains a sound possibility, therefore, that survival. This may explain the retinoic acid resistance of some e€ects of HDAC inhibitors on cell di€erentiation the patient who had this variant form of APL, despite may be mediated through additional mechanisms, which the fact that the STAT5b ± RARa fusion appears to do not require inhibition of histone deacetylation per se. have very similar properties as PML-RARa. These More recent work indicates existence of RAR indepen- ®nding also may prompt an examination of STAT dent and RXR plus PKA agonist dependent di€erentia- signaling in other forms of APL, since its has been tion pathway for APL (Benoit et al., 1999) and activation shown that cytokines can augment the di€erentiation of this pathway could also overcome ATRA resistance. e€ects of ATRA (Nakamaki et al., 1994). The same group has very recently reported that ATRA induced di€erentiation of APL cells terminates with the induction of apoptosis through the paracrine activity of Summary Apo-2L/TRAIL death ligand (Altucci et al., 2001), suggesting the use of TRAIL in anti-APL therapies There are three axes to be investigated in under- (Altucci et al., 2001; Zelent, 2001). With the establish- standing the pathogenesis of APL. Firstly, in all cases ment of animal and cell models for sensitive and resistant of APL RARa is fused to the partner (N) proteins, forms of APL, the development of more e€ective which can multimerize (see Tables 1 and 2 for therapies can be expected to continue. It is anticipated di€erences and similarities between N and N ± RARa that therapeutic approaches that are discovered and proteins). Multimerization appears to mediate the developed through studies of APL will in the end ®nd a abnormally high anity of the fusion proteins for co- broader application in cancer treatment. repressors. In addition, PLZF itself binds to co- repressors prohibiting their full release at pharmacolo- gical doses of ATRA and leading to a resistant form of APL. These fusion proteins block the activation of Acknowledgments critical genes required for myeloid di€erentiation. The We are grateful for support from the Leukaemia Research fusion proteins may also a€ect other transcriptional Fund of Great Britain (A Zelent), the Kay Kendall Leukaemia Fund (F Guidez), National Institutes of Health pathways via protein-protein interactions. Second, loss grants CA59936 (JD Licht, A Zelent, S Waxman) and KO8 of one allele of the N partner gene could disable a CA73762 (A Melnick), as well as the American Cancer growth suppressive pathway. Furthermore, the N ± Society (JD Licht) and Samuel Waxman Cancer Research RARa fusion might sequester the normal N product. Foundation. JD Licht is a scholar of the Leukemia Society Lastly, as in the case of RARa ± PLZF, the reciprocal of America.

References

Ahmad KF, Engel CK and Prive GG. (1998). Proc. Natl. Alcalay M, Zangrilli D, Fagioli M, Pandol® PP, Mencarelli Acad. Sci. USA, 95, 12123 ± 12128. A, Lo Coco F, Biondi A, Grignani F and Pelicci PG. Al-Omar H, Khalil S, Sher G and Kamel-Reid S. (1997). (1992). Proc. Natl. Acad. Sci. USA, 89, 4840 ± 4844. Blood, 90 (Supp 1), 207b.

Oncogene APL associated RARa fusions A Zelent et al 7200 Altucci L, Rossin A, Ra€elsberger W, Reitmair A, Chretien S, Varlet P, Verdier F, Gobert S, Cartron JP, Chomienne C and Gronemeyer H. (2001). Nat. Med., 7, Gisselbrecht S, Mayeux P and Lacombe C. (1996). EMBO 680 ± 686. J., 15, 4174 ± 4181. Arnould C, Philippe C, Bourdon V, Gr goire MJ, Berger R Compton DA and Cleveland DW. (1993). J. Cell. Biol., 120, and Jonveaux P. (1999). Hum. Mol. Genet., 8, 1741 ± 1749. 947 ± 957. Ball HJ, Melnick A, Shaknovich R, Kohanski RA and Licht Compton DA and Cleveland DW. (1994). Curr. Opin. Cell. JD. (1999). Nucleic Acids Res., 27, 4106 ± 4113. Biol., 6, 343 ± 346. Bardwell VJ and Treisman R. (1994). Genes Dev., 8, 1664 ± Compton DA, Szilak I and Cleveland DW. (1992). J. Cell. 1677. Biol., 116, 1395 ± 1408. Barna M, Hawe N, Niswander L and Pandol® PP. (2000). Cook M, Gould A, Brand N, Davies J, Strutt P, Shaknovich Nat. Genet., 25, 166 ± 172. R, Licht J, Waxman S, Chen Z, Gluecksohn-Waelsch S, Benoit G, Altucci L, Flexor M, Ruchaud S, Lillehaug J, Krumlauf R and Zelent A. (1995). Proc. Natl. Acad. Sci. Ra€elsberger W, Gronemeyer H and Lanotte M. (1999). USA, 92, 2249 ± 2253. EMBO J., 18, 7011 ± 7018. Corey SJ, Locker J, Oliveri DR, Shekhter-Levin S, Redner Bischof D, Pulford K, Mason DY and Morris SW. (1997). RL, Penchansky L and Gollin SM. (1994). Leukemia, 8, Mol. Cell. Biol., 17, 2312 ± 2325. 1350 ± 1353. Borrow J, Goddard AD, Sheer D and Solomon E. (1990). Culligan DJ, Stevenson D, Chee YL and Grimwade D. Science, 249, 1577 ± 1580. (1998). Br. J. Haematol., 100, 328 ± 330. Borrow J, Shipley J, Howe K, Kiely F, Goddard A, Sheer D, Daniel MT, Koken M, Romagne O, Barbey S, Bazarbachi A, Srivastava A, Antony AC, Fioretos T, Mitelman F and Stadler M, Guillemin MC, Degos L, Chomienne C and de Solomon E. (1994). Genes Chrom. Cancer, 9, 234 ± 243. The H. (1993). Blood, 82, 1858 ± 1867. Brand NJ, Petkovich M and Chambon P. (1990). Nucleic Darnell Jr. JE. (1997). Science, 277, 1630 ± 1635. Acids Res., 18, 6799 ± 6806. David G, Alland L, Hong S-H, Wong C-W, DePinho R and Casiano C, Martin S, Green D and Tan E. (1996). Exp. Med., Dejean A. (1998). Oncogene, 16, 2549 ± 2556. 184, 765 ± 770. de The H, Chomienne C, Lanotte M, Degos L and Dejean A. Chambon P. (1996). FASEB J., 10, 940 ± 954. (1990). Nature, 347, 558 ± 561. Chan P, Aldrich M, Cook R and Busch H. (1986a). J. Biol. de The H, Lavau C, Marchio A, Chomienne C, Degos L and Chem., 261, 1868 ± 1872. Dejean A. (1991). Cell, 66, 675 ± 684. Chan P, Aldrich M and Yung B. (1987). Cancer Res., 47, Derenzini M, Sirri V, Trere D and Ochs R. (1995). Lab. 3798 ± 3801. Investig., 73, 497 ± 502. Chan P, Chan W, Yung B, Cook R, Aldrich M, Ku D, Dhordain P, Albagli O, Honore N, Guidez F, Lantoine D, Goldknopf I and Busch H. (1986b). J. Biol. Chem., 261, Schmid M, The HD, Zelent A and Koken MH. (2000). 14335 ± 14341. Oncogene, 19, 6240 ± 6250. Chan PK and Chan FY. (1995). Biochim. Biophys. Acta, Dionne MA, Howard L and Compton DA. (1999). Cell. 1262, 37 ± 42. Motil. Cytoskeleton, 42, 189 ± 203. Chan PK and Chan FY. (1999). Biochem. Pharmacol., 57, Diverio D, Lo Coco F, D'Adamo F, Biondi A, Fagioli M, 1265 ± 1273. Grignani F, Rambaldi A, Rossi V, Avvisati G, Petti MC, Chan PK, Chan FY, Morris SW and Xie Z. (1997). Nucleic Testi AM, Liso V, Specchia G, Fioritoni G, Recchia A, Acids Res., 25, 1225 ± 1232. Frassoni F, Ciolli S and Pelicci PG. (1992). Blood, 79, Chan WY, Liu QR, Borjigin J, Busch H, Rennert OM, Tease 3331 ± 3336. LA and Chan PK. (1989). Biochemistry, 28, 1033 ± 1039. Diverio D, Rossi V, Avvisati G, DeSantis S, Pistilli A, Pane Chang C-C, Ye BH, Chaganti RSK and Dalla-Favera R. F, Saglio G, Martinelli G, Petti MC, Santoro A, Pelicci (1996). Proc. Natl. Acad. Sci. USA, 93, 6947 ± 6952. PG, Mandelli F, Biondi A and Coco FL. (1998). Blood, 92, Chen A, Licht JD, Wu Y, Hellinger N, Scher W and Waxman 784 ± 789. S. (1994a). Blood, 84, 2122 ± 2129. Dong F, Liu X, de Koning JP, Touw IP, Henninghausen L, Chen S-J, Chen Z, Chen A, Tong J-H, Dong S, Wang Z-Y, Larner A and Grimley PM. (1998). J. Immunol., 161, Waxman S and Zelent A. (1992). Oncogene, 7, 1223 ± 1232. 6503 ± 6509. Chen S-J, Zelent A, Tong J-H, Yu H-Q, Wang Z-Y, Derre J, Dong S and Tweardy D. (2000). Blood, 96 (Supp 1), 457a. Berger R, Waxman S and Chen Z. (1993a). J. Clin. Invest., DongS,ZhuJ,ReidA,StruttP,GuidezF,ZhongH-J, 91, 2260 ± 2267. Wang Z-Y, Licht J, Waxman S, Chomienne C, Zelent A Chen S-J, Zhu Y-J, Tong J-H, Dong S, Huang W, Chen Y, and Chen S-J. (1996). Proc. Natl. Acad. Sci. USA, 93, Xiang W-M, Zhang L, Li X-S, Qian G-Q, Wang Z-Y, 3624 ± 3629. Chen Z, Larsen C-J and Berger R. (1991). Blood, 78, Downes M, Ordentlich P, Kao HY, Alvarez JG and Evans 2696 ± 2701. RM. (2000). Proc. Natl. Acad. Sci. USA, 97, 10330 ± Chen Z, Brand NJ, Chen A, Chen S-J, Tong J-H, Wang Z-Y, 10335. Waxman S and Zelent A. (1993b). EMBO J., 12, 1161 ± Du C, Redner RL, Cooke MP and Lavau C. (1999). Blood, 1167. 94, 793 ± 802. Chen Z, Guidez F, Rousselot P, Agadir A, Chen S-J, Wang Dumbar TS, Gentry GA and Olson MO. (1989). Biochem- Z-Y, Degos L, Zelent A, Waxman S and Chomienne C. istry, 28, 9495 ± 9501. (1994b). Proc. Natl. Acad. Sci. USA, 91, 1178 ± 1182. Fagioli M, Alcalay M, Pandol® PP, Venturini L, Mencarelli Cheng GX, Zhu XH, Men XQ, Wang L, Huang QH, Jin XL, A,SimeoneA,AcamporaD,GrignaniFandPelicciPG. Xiong SM, Zhu J, Guo WM, Chen JQ, Xu SF, So E, Chan (1992). Oncogene, 7, 1083 ± 1091. LC, Waxman S, Zelent A, Chen GQ, Dong S, Liu JX and Fenaux P and Chomienne C. (1996). Curr. Opin. Oncol., 8, Chen SJ. (1999). Proc. Natl. Acad. Sci. USA, 96, 6318 ± 3±12. 6323. Ferbeyre G, de Stanchina E, Querido E, Baptiste N, Prives C Chou CC and Yung BY. (2001). Mol. Pharmacol., 59, 38 ± 45. and Lowe SW. (2000). Genes Dev., 14, 2015 ± 2027.

Oncogene APL associated RARa fusions A Zelent et al 7201 Feuerstein N, Chan PK and Mond JJ. (1988). J. Biol. Chem., HuangW,SunGL,LiXS,CaoQ,LuY,JangGS,Zhang 263, 10608 ± 10612. FQ, Chai JR, Wang ZY, Waxman S, Chen Z and Chen S- Fogal V, Gostissa M, Sandy P, Zacchi P, Sternsdorf T, J. (1993). Blood, 82, 1264 ± 1269. Jensen K, Pandol® PP, Will H, Schneider C and Del Sal G. Hummel JL, Wells RA, Dube ID, Licht JD and Kamel-Reid (2000). EMBO J., 19, 6185 ± 6195. S. (1999). Oncogene, 18, 633 ± 641. Fukutani H, Naoe T, Ohno R, Yoshida H, Miyawaki S, Imada K, Bloom ET, Nakajima H, Horvath-Arcidiacono Shimazaki C, Miyake T, Nakayama Y, Kobayashi H, JA, Udy GB, Davey HW and Leonard WJ. (1998). J. Exp. Goto S. et al. (1995). Leukemia, 9, 1478 ± 1482. Med., 188, 2067 ± 2074. Gallagher RE, Li YP, Rao S, Paietta E, Andersen J, Etkind Inouye C and Seto E. (1994). J. Biol. Chem., 259, 6506 ± 6510. P, Bennett JM, Tallman MS and Wiernik PH. (1995). Ivins S and Zelent A. (1998). Blood, 92 (Supp 1), 308a. Blood, 86, 1540 ± 1547. Iwahara T, Fujimoto J, Wen D, Cupples R, Bucay N, Gesbert F and Grin JD. (2000). Blood, 96, 2269 ± 2276. Arakawa T, Mori S, Ratzkin B and Yamamoto T. (1997). Glass CK and Rosenfeld MG. (2000). Genes Dev., 14, 121 ± Oncogene, 14, 439 ± 449. 141. Jansen JH, de Ridder MC, Geertsma WM, Erpelinck CA, Goddard AD, Borrow J, Freemont P and Solomon E. (1991). van Lom K, Smit EM, Slater R, vd Reijden BA, de Greef Science, 254, 1371 ± 1374. GE, Sonneveld P and Lowenberg B. (1999). Blood, 94, Gordon MB, Howard L and Compton DA. (2001). J. Cell. 39 ± 45. Biol., 152, 425 ± 434. Jansen JH, Mahfoudi A, Rambaud S, Lavau C, Wahli W and Grignani F, DeMatteis S, Nervi C, Tomassoni L, Gelmetti V, Dejean A. (1995). Proc. Natl. Acad. Sci. USA, 92, 7401 ± Cioce M, Fanelli M, Ruthardt M, Ferrara FF, Zamir I, 7405. Seiser C, Grignani F, Lazar MA, Minucci S and Pelicci Jiang PS and Yung BY. (1999). Biochem. Biophys. Res. PG. (1998). Nature, 391, 815 ± 818. Commun., 257, 865 ± 870. Grimwade D, Biondi A, Mozziconacci MJ, Hagemeijer A, Kakizuka A, Miller WH, Umesono K, Warrell RP, Frankel Berger R, Neat M, Howe K, Dastugue N, Jansen J, SR, Murty VVVS, Dimitrovsky E and Evans RM. (1991). Radford-Weiss I, Lo Coco F, Lessard M, Hernandez JM, Cell, 66, 663 ± 674. Delabesse E, Head D, Liso V, Sainty D, Flandrin G, Kane JR, Head DR, Balazs L, Hulshof MG, Motroni TA, Solomon E, Birg F and Lafage-Pochitalo€ M. (2000). Raimondi SC, Carroll AJ, Behm FG, Krance RA, Blood, 96, 1297 ± 1308. Shurtle€ SA and Downing JR. (1996). Leukemia, 10, Grimwade D, Gorman P, Duprez E, Howe K, Langabeer S, 1296 ± 1302. Oliver F, Walker H, Culligan D, Waters J, Pomfret M, Kazansky AV and Rosen JM. (2001). Cell. Growth Di€er., Goldstone A, Burnett A, Freemont P, Sheer D and 12, 1±7. Solomon E. (1997). Blood, 90, 4876 ± 4885. KempfT,Bischo€FR,KaliesIandPonstinglH.(1994). Gueth-Hallonet C, Weber K and Osborn M. (1997). Exp. FEBS Lett., 354, 307 ± 310. Cell. Res., 233, 21 ± 24. Kerckaert J-P, Deweindt C, Tilly H, Quief S, Lecocq G and Guidez F, Huang W, Tong J-H, Dubois C, Balitrand N, Bastard C. (1993). Nat. Genet., 5, 66 ± 70. Waxman S, Michaux JL, Martiat P, Degos L, Chen Z and Kieslinger M, Woldman I, Moriggl R, Hofmann J, Marine Chomienne C. (1994). Leukemia, 8, 312 ± 317. JC, Ihle JN, Beug H and Decker T. (2000). Genes Dev., 14, Guidez F, Ivins S, Zhu J, Soderstrom M, Waxman S and 232 ± 244. Zelent A. (1998). Blood, 91, 2634 ± 2642. KokenMH,DanielMT,GianniM,ZelentA,LichtJ,Buzyn GuoA,SalomoniP,LuoJ,ShihA,ZhongS,GuWand A, Minard P, Degos L, Varet B and de The H. (1999). Paolo Pandol® P. (2000). Nat. Cell. Biol., 2, 730 ± 736. Oncogene, 18, 1113 ± 1118. Harborth J, Wang J, Gueth-Hallonet C, Weber K and KokenMHM,Linares-CruzG,QuignonF,VironA,Chelbi- Osborn M. (1999). EMBO J., 18, 1689 ± 1700. Alix MK, Sobczak-The pot Juhlin L, Degos L, Calvo F Harborth J, Weber K and Osborn M. (1995). EMBO J., 14, and de The H. (1995). Oncogene, 10, 1315 ± 1324. 2447 ± 2460. Koken MHM, Reid A, Quignon F, Chelbi-Alix MK, Davies Harborth J, Weber K and Osborn M. (2000). J. Biol. Chem., JM, Kabarowski JHS, Zhu J, Dong S, Chen S-J, Chen Z, 275, 31979 ± 31985. Tan CC, Licht J, Waxman S, de The H and Zelent A. Hauksdottir H and Privalsky ML. (2001). Cell. Growth (1997). Proc. Natl. Acad. Sci. USA, 94, 10255 ± 10260. Di€er., 12, 85 ± 98. Kondo T, Minamino N, Nagamura-Inoue T, Matsumoto M, He D, Zeng C and Brinkley BR. (1995). Int. Rev. Cytol., Taniguchi T and Tanaka N. (1997). Oncogene, 15, 1275 ± 162B, 1 ± 74. 1281. He L, Bhaumik M, Tribioli C, Rego EM, Ivins S, Zelent A Lallemand-Breitenbach V, Zhu J, Puvion F, Koken M, and Pandol® PP. (2000). Mol. Cell, 6, 1131 ± 1141. Honore N, Doubeikovsky A, Duprez E, Pandol® PP, He L-Z, Guidez F, Tribioli C, Peruzzi D, Ruthardt M, Zelent Puvion E, Freemont P and de The H. (2001). J. Exp. Med., A and Pandol® PP. (1998). Nat. Genet., 18, 126 ± 135. 193, 1361 ± 1372. Hernandez-Verdun D, Derenzeni M and Bouteille M. (1982). Lee JH and Welch DR. (1997). Int. J. Cancer, 71, 1035 ± Chromsoma, 85, 461 ± 473. 1044. Hiorns LR, Min T, Swansbury GJ, Zelent A, Dyer MJ and Leroy A, Nakshatri H and Chambon P. (1991). Proc. Natl. Catovsky D. (1994). Blood, 83, 2946 ± 2951. Acad.Sci.USA,88, 10138 ± 10142. Hoatlin ME, Zhi Y, Ball H, Silvey K, Melnick A, Stone S, Li JY, English MA, Ball HJ, Yeyati PL, Waxman S and Licht Arai S, Hawe N, Owen G, Zelent A and Licht J. (1999). JD. (1997a). J. Biol. Chem., 272, 22447 ± 22455. Blood, 94, 3737 ± 3747. Li X, Peng H, Schultz DC, Lopez-Guisa JM, Rauscher III FJ Hong SH, David G, Wong CW, Dejean A and Privalsky ML. and Marmorstein R. (1999). Cancer Res., 59, 5275 ± 5282. (1997). Proc. Natl. Acad. Sci. USA, 94, 9028 ± 9033. Li Y, Busch R, Valdez B and Busch H. (1996). Eur. J. Hsu CY and Yung BY. (1998). Oncogene, 16, 915 ± 923. Biochem., 237, 153 ± 158. Hsu CY and Yung BY. (2000). Int. J. Cancer, 88, 392 ± 400. Hsu HL and Yeh NH. (1996). J. Cell. Sci., 109, 277 ± 288.

Oncogene APL associated RARa fusions A Zelent et al 7202 LiYP,AndersenJ,ZelentA,RaoS,PaiettaE,TallmanMS, Nagpal S, Saunders M, Kastner P, Durand B, Nakshatri H Wiernik PH and Gallagher RE. (1997b). Blood, 90, 306 ± and Chambon P. (1992). Cell, 70, 1007 ± 1019. 312. Nakamaki T, Hino K, Yokoyama A, Hisatake J, Tomoyasu Licht JD, Chomienne C, Goy A, Chen A, Scott AA, Head S, Honma Y, Hozumi M and Tsuruoka N. (1994). DR, Michaux JL, DeBlasio A, Miller Jr W, Zelenetz AD, Anticancer Res., 14, 817 ± 823. Willman CL, Chen Z, Chen S-J, Zelent A, Macintyre E, Neves H, Ramos C, da Silva MG, Parreira A and Parreira L. Vail A, Cortes J, Kantarjian H and Waxman S. (1995). (1999). Blood, 93, 1197 ± 1207. Blood, 85, 1083 ± 1094. Nieborowska-Skorska M, Wasik MA, Slupianek A, Salo- Licht JD, Shaknovich R, English MA, Melnick A, Li J-Y, moni P, Kitamura T, Calabretta B and Skorski T. (1999). ReddyJC,DongS,ChenS-J,ZelentAandWaxmanS. J. Exp. Med., 189, 1229 ± 1242. (1996). Oncogene, 12, 323 ± 336. Numoto M, Niwa O, Kaplan J, Wong K-K, Merrell K, Lin JX and Leonard WJ. (2000). Oncogene, 19, 2566 ± 2576. Kamiya K, Yanagihara K and Calame K. (1993). Nucleic Lin JX, Mietz J, Modi WS, John S and Leonard WJ. (1996). Acids Res., 21, 3767 ± 3775. J. Biol. Chem., 271, 10738 ± 10744. Okuda M, Horn HF, Tarapore P, Tokuyama Y, Smulian Lin R, Nagy L, Inoue S, Shao W, Miller Jr W and Evans R. AG, Chan PK, Knudsen ES, Hofmann IA, Snyder JD, (1998). Nature, 391, 811 ± 814. Bove KE and Fukasawa K. (2000). Cell, 103, 127 ± 140. Lin RJ and Evans RM. (2000). Mol. Cell, 5, 821 ± 830. Olson O, Orrick L, Jones C and Busch H. (1974). J. Biol. Lindstrom MS, Klangby U, Inoue R, Pisa P, Wiman KG and Chem., 249, 2823 ± 2827. Asker CE. (2000). Exp. Cell. Res., 256, 400 ± 410. Pandol® PP, Alcalay M, Fagioli M, Zangrilli D, Mencarelli Liu WH, Hsu CY and Yung BY. (1999). Int. J. Cancer, 83, A,DiverioD,BiondiA,LoCocoF,RambaldiA, 765 ± 771. Grignani F, Rochette-Egly C, Gaube M-P, Chambon P Long JJ, Leresche A, Kriwacki RW and Gottesfeld JM. and Pelicci PG. (1992). EMBO J., 11, 1397 ± 1407. (1998). Mol. Cell. Biol., 18, 1467 ± 1476. Pearson M, Carbone R, Sebastiani C, Cioce M, Fagioli M, Longo L, Pandol® PP, Biondi A, Rambaldi A, Mencarelli A, Saito S, Higashimoto Y, Appella E, Minucci S, Pandol® Lo Coco F, Diverio D, Pegoraro L, Avanzi G, Tabilio A, PP and Pelicci PG. (2000). Nature, 406, 207 ± 210. Zangrilli D, Alcalay M, Donti E, Grignani F and Pelici Perez A, Kastner P, Sethi S, Lutz Y, Reibel C and Chambon PG. (1990). J. Exp. Med., 172, 1571 ± 1575. P. (1993). EMBO J., 12, 3171 ± 3182. Look A. (1997). Science, 278, 1059 ± 1064. Peter M, Nakagawa J, Doree M, Labbe JC and Nigg EA. Luderus ME, den Blaauwen JL, de Smit OJ, Compton DA (1990). Cell, 60, 791 ± 801. and van Driel R. (1994). Mol. Cell. Biol., 14, 6297 ± 6305. PollockJL,WesterveltP,KurichetyAK,PelicciPG, Lyderson B and Pettijohn D. (1980). Cell, 22, 489 ± 499. Grisolano JL and Ley TJ. (1999). Proc. Natl. Acad. Sci. Maekawa T and Kuriyama R. (1993). J. Cell. Sci., 105, 589 ± USA, 96, 15103 ± 15108. 600. Ra€ J, Kellum R and Alberts B. (1995). EMBO J., 13, 5977 ± Mangelsdorf DJ and Evans RM. (1995). Cell, 83, 841 ± 850. 5983. Matikainen S, Ronni T, Hurme M, Pine R and Julkunen I. Reddy BA, Etkin LD and Freemont PS. (1992). Trends (1996). Blood, 88, 114 ± 123. Biochem. Sci., 17, 344 ± 345. Matikainen S, Sareneva T, Ronni T, Lehtonen A, Koskinen Redner R, Rush L and Pollock S. (1996a). Blood, 10, S1, PJ and Julkunen I. (1999). Blood, 93, 1980 ± 1991. 551a. McConnell M, Chevallier N, Giltnane J, Carlile G, Staudt L Redner RL, Chen JD, Rush EA, Li H and Pollock SL. and Licht J. (2000). Blood, 96 (supp 1), 457a. (2000). Blood, 95, 2683 ± 2690. Melnick A, Carlile GW, McConnell MJ, Polinger A, Hiebert Redner RL, Corey SJ and Rush EA. (1997). Leukemia, 11, SW and Licht JD. (2000a). Blood, 96, 3939 ± 3947. 1014 ± 1016. Melnick AM, Westendorf JJ, Polinger A, Carlile GW, Arai Redner RL, Rush EA, Faas S, Rudert WA and Corey SJ. S, Ball HJ, Lutterbach B, Hiebert SW and Licht JD. (1996b). Blood, 87, 882 ± 886. (2000b). Mol. Cell. Biol., 20, 2075 ± 2086. Reid A, Gould A, Brand N, Cook M, Strutt P, Li J, Licht J, Miaw SC, Choi A, Yu E, Kishikawa H and Ho IC. (2000). Waxman S, Krumlauf R and Zelent A. (1995). Blood, 86, Immunity, 12, 323 ± 333. 4544 ± 4552. Miki T, Fleming TP, Crescenzi M, Molloy CJ, Blam SB, Richon VM, Emiliani S, Verdin E, Webb Y, Breslow R, Reynolds SH and Aaronson SA. (1991). Proc. Natl. Acad. Rifkind RA and Marks PA. (1998). Proc. Natl. Acad. Sci. Sci. USA, 88, 5167 ± 5171. USA, 95, 3003 ± 3007. Miller Jr WH, Levine K, DeBlasio A, Frankel SR, Rowley JD, Golomb HM and Dougherty C. (1977). Lancet, Dmitrovsky E and Warrell Jr RP. (1993). Blood, 82, 1, 549 ± 550. 1689 ± 1694. Ruthardt M, Orleth A, Tomassoni L, Puccetti E, Riganelli Miller WH, Warrell RP, Frankel SR, Jakubowski A, D, Alcalay M, Mannucci R, Nicoletti I, Grignani F, Gabrilove JG, Muindi H and Dimitrowsky E. (1990). J. Fagioli M and Pelicci PG. (1998). Oncogene, 16, 1945 ± Natl. Cancer Inst., 82, 1932 ± 1933. 1953. Minucci S, Maccarana M, Cioce M, De Luca P, Gelmetti V, Ruthardt M, Testa U, Nervi C, Ferrucci PF, Grignani F, Segalla S, Di Croce L, Giavara S, Matteucci C, Gobbi A, Puccetti E, Grignani F, Peschle C and Pelicci PG. (1997). Bianchini A, Colombo E, Schiavoni I, Badaracco G, Hu Mol. Cell. Biol., 17, 4859 ± 4869. X, Lazar MA, Landsberger N, Nervi C and Pelicci PG. Sainty D, Liso V, Cantu-Rajnoldi A, Head D, Mozziconacci (2000). Mol. Cell., 5, 811 ± 820. MJ,ArnouletC,BenattarL,FenuS,ManciniM, Morris SW, Kirstein MN, Valentine MB, Dittmer KG, N Duchayne E, Mahon FX, Gutierrez N, Birg F, Biondi A, SD, Saltman DL and Look AT. (1994). Science, 263, Grimwade D, Lafage-Pochitalo€ M, Hagemeijer A and 1281 ± 1284. Flandrin G. (2000). Blood, 96, 1287 ± 1296. Mui AL, Wakao H, Kinoshita T, Kitamura T and Miyajima A. (1996). EMBO J., 15, 2425 ± 2433.

Oncogene APL associated RARa fusions A Zelent et al 7203 Santos SC, Lacronique V, Bouchaert I, Monni R, Bernard O, Taw®c S, Olson M and Ahmed K. (1995). J. Biol. Chem., 270, Gisselbrecht S and Gouilleux F. (2001). Oncogene, 20, 21009 ± 21015. 2080 ± 2090. Teglund S, McKay C, Schuetz E, van Deursen JM, Saredi A, Howard L and Compton DA. (1996). J. Cell. Sci., Stravopodis D, Wang D, Brown M, Bodner S, Grosveld 109, 619 ± 630. G and Ihle JN. (1998). Cell, 93, 841 ± 850. Saredi A, Howard L and Compton DA. (1997). J. Cell. Sci., Terris B, Baldin V, Dubois S, Degott C, Flejou JF, Henin D 110, 1287 ± 1297. and Dejean A. (1995). Cancer Res., 55, 1590 ± 1597. Scott AA, Head DR, Kopecky KJ, Appelbaum FR, Theil Tsutsui K, Tsutsui K, Okada S, Waterai S, Seki S, Yasuda T KS, Grever MR, Chen IM, Whittaker MH, Grith BB, and Shomori T. (1993). J. Biol. Chem., 268, 12886 ± 12894. Licht JD, Waxman S, Whalen MM, Bankhurst AD, Vahdat L, Maslak P, Miller WJ, Eardley A, Heller G, Richter LC, Grogan TM and Willman CL. (1994). Blood, Scheinberg DA and Warrell RJ. (1994). Blood, 84, 3843 ± 84, 244 ± 255. 3849. Sexl V, Piekorz R, Moriggl R, Rohrer J, Brown MP, Bunting Warrell Jr. RP, He LZ, Richon V, Calleja E and Pandol® PP. KD,RothammerK,RousselMFandIhleJN.(2000). (1998). J. Natl. Cancer Inst., 90, 1621 ± 1625. Blood, 96, 2277 ± 2283. Waxman S. (1978). The Leukemia Cell. Rubin AD and Shaknovich R, Yeyati PL, Ivins S, Melnick A, Lempert C, Waxman S (eds). CRC Press, Inc.: West Palm Beach, Waxman S, Zelent A and Licht JD. (1998). Mol. Cell. Florida, pp. 169 ± 175. Biol., 18, 5533 ± 5545. Waxman S, Zelent A, Melnick A, Licht JD, Monks A, Sherr CJ and Weber JD. (2000). Curr.Opin.Genet.Dev.,10, Sausville E and Jing YK. (1999). Blood, 94 (Suppl 1), 61a. 94 ± 99. Wells RA, Catzavelos C and Kamel-Reid S. (1997). Nat. Shi Y, Lee J and Galvin K. (1997). Bioch. et Bioph. Acta, Genet., 17, 109 ± 113. 1332, F49 ± F66. Wells RA, Hummel JL, De Koven A, Zipursky A, Kirby M, Shiota M, Nakamura S, Ichinohasama R, Abe M, Akagi T, Dube I and Kamel-Reid S. (1996). Leukemia, 10, 735 ± TakeshitaM,MoriN,FujimotoJ,MiyauchiJ,MikataA, 740. Nanba K, Takami T, Yamabe H, Takano Y, Izumo T, Wong CW and Privalsky ML. (1998). J. Biol. Chem., 273, Nagatani T, Mohri N, Nasu K, Satoh H, Katano H, 27695 ± 27702. Fujimoto J, Yamamoto T and Mori S. (1995). Blood, 86, Yang CH, Lambie EJ and Snyder M. (1992). J. Cell. Biol., 1954 ± 1960. 116, 1303 ± 1317. Sirri V, Roussel P, Gendron MC and Hernandez-Verdun D. Yang CH and Snyder M. (1992). Mol. Biol. Cell., 3, 1259 ± (1997). Cytometry, 28, 147 ± 156. 1267. Sitterlin D, Tiollais P and Transy C. (1997). Oncogene, 14, Ye BH, Lista F, Lo Coco F, Knowles DM, Ot K, Chaganti 1067 ± 1074. RSK and Dalla-Favera R. (1993). Science, 262, 747 ± 750. Slack JL, Arthur DC, Lawrence D, Mrozek K, Mayer RJ, Yeyati PL, Shaknovich R, Boterashvili S, Li J, Ball HJ, Davey FR, Tantravahi R, Pettenati MJ, Bigner S, Carroll Waxman S, Nason-Burchenal K, Dmitrovsky E, Zelent A AJ, Rao KW, Schi€er CA and Bloom®eld CD. (1997). J. and Licht JD. (1999). Oncogene, 18, 925 ± 934. Clin. Oncol., 15, 1786 ± 1795. Yoneda-KatoN,LookT,KirsteinM,ValentineM, SmetanaK,OchsR,LischweM,GyorkeyF,FreireichE, Raimondi S, Cohen K, Carroll A and Morris S. (1996). Chudimael V and Busch H. (1984). Exp. Cell. Res., 152, Oncogene, 10, 265 ± 275. 195 ± 203. Yoshida H, Naoe T, Fukutani H, Kiyoi H, Kubo K and SoCW,DongS,SoCK,ChengGX,HuangQH,ChenSJ Ohno R. (1995). Genes Chrom. Cancer, 12, 37 ± 44. and Chan LC. (2000). Leukemia, 14, 77 ± 83. Yu BD, Hess JL, Horning SE, Brown GA and Korsmeyer SJ. SparksCA,BangsPL,McNeilGP,LawrenceJBandFey (1995). Nature, 378, 505 ± 508. EG. (1993). Genomics, 17, 222 ± 224. Yung B, Busch H and Chan P. (1985a). Biochimica. et Stadler M, Chelbialix MK, Koken MHM, Venturini L, Lee Biophysica. Acta, 826, 167 ± 173. C, Saib A, Quignon F, Pelicano L, Guillemin MC, Yung B, Busch R, Busch H, Mauger A and Chan P. (1985b). Schindler C and de The H. (1995). Oncogene, 11, 2565 ± Biochem. Pharm., 34, 4059 ± 4063. 2573. Zatsepina OV, Rousselet A, Chan PK, Olson MO, Jordan Szebeni A, Mehrotra B, Baumann A, Adam SA, Wing®eld EG and Bornens M. (1999). J. Cell. Sci., 112, 455 ± 466. PT and Olson MO. (1997). Biochemistry, 36, 3941 ± 3949. Zelent A. (2001). Nat. Med., 7, 662 ± 664. Takahashi M, Inaguma Y, Hiai H and Hirose F. (1988). Mol. Zeng C, He D and Brinkley BR. (1994). Cell. Motil. Cell. Biol., 8, 1853 ± 1856. Cytoskeleton, 29, 167 ± 176. Takemura M, Ohta N, Furuichi Y, Takahashi T, Yoshida S, ZhuJ,HeyworthCM,GlasowA,HuangQ-H,PetrieK, Olson M and Umekawa H. (1994). Biochem. Biophys. Res. Lanotte M, Benoit G, Gallagher R, Waxman S, Enver T Commun., 199, 46 ± 51. and Zelent A. (2001). Blood (in press). Tang TK, Tang CJ, Chao YJ and Wu CW. (1994). J. Cell. Zhu M, John S, Berg M and Leonard WJ. (1999). Cell, 96, Sci., 107, 1389 ± 1402. 121 ± 130. Tang TK, Tang CJ, Chen YL and Wu CW. (1993). J. Cell. Zweyer M, Riederer BM, Ochs RL, Fackelmayer FO, Sci., 104, 249 ± 260. Kohwi-Shigematsu T, Bareggi R, Narducci P and Martelli TashiroS,KotomuraN,TanakaK,SuzukiK,KyoT,Dohy AM. (1997). Exp. Cell. Res., 230, 325 ± 336. H, Niwa O and Kamada N. (1994). Oncogene, 9, 1939 ± 1945.

Oncogene