MOZ and MORF, Two Large Mystic Hats in Normal and Cancer Stem Cells

X-J Yang and M Ullah

Molecular Oncology Group, Department of Medicine, McGill University Health Center, Montre´al, Que´bec, Canada

Genes of the human monocytic leukemia zinc-finger protein pattern. For cancer biology, it is thus important to MOZ (HUGO symbol, MYST3) and its paralog MORF understand the fundamental mechanisms whereby chro- (MYST4) are rearranged in chromosome translocations matin structure and function are regulated. In the associated withacute myeloid leukemia and/or benign past two decades, our knowledge about regulation in uterine leiomyomata. Both proteins have intrinsic histone this field has exploded. Known regulatory mechanisms acetyltransferase activity and are components of quartet include chromatin assembly, ATP-dependent remodeling, complexes withnoncatalytic subunits containing thebromo- covalent modification, condensin-mediated condensation, domain, plant homeodomain-linked (PHD) finger and replacement with histone variants, and association of proline-tryptophan-tryptophan-proline (PWWP)-containing noncoding RNA (reviewed by Horn and Peterson, 2002; domain, three types of structural modules characteristic of Khorasanizadeh, 2004; Li et al., 2007). Covalent chromatin regulators. Although leukemia-derived fusion pro- modification can occur at both the DNA and histone teins suchas MOZ-TIF2 promote self-renewal of leukemic levels. With histones, modifications include acetylation, stem cells, recent studies indicate that murine MOZ and phosphorylation, methylation, ubiquitination, and many MORF are important for proper development of hema- others (for reviews, see Spencer and Davie, 1999; Strahl topoietic and neurogenic progenitors, respectively, thereby and Allis, 2000; Berger, 2002; Jason et al., 2002; highlighting the importance of epigenetic integrity in Kouzarides, 2007). Different modifications interplay safeguarding stem cell identity. with each other and contribute significantly to the Oncogene (2007) 26, 5408–5419; doi:10.1038/sj.onc.1210609 formation of a complex epigenetic program for generating and preserving the defined chromatin state of a given Keywords: lysine acetylation; MYST acetyltransferase; cell type. ING5; BR140; CBP; Runx1 Within this program, histone acetyltransferases (HATs) play an integral role (reviewed in Sterner and Berger, 2000; Roth et al., 2001; Yang, 2004; Lee and Workman, 2007). As a pair of such enzymes, MOZ (monocytic leukemic zinc-finger protein) and MORF Introduction (MOZ-related factor) are important for different developmental programs and have been implicated in Like many other human diseases, cancer is a pathological leukemogenic and other tumorigenic processes. Here, condition with genetic and epigenetic basis (reviewed in we start with a brief overview about the discovery of this Baylin and Herman, 2000; Jaenisch and Bird, 2003; Egger pair and related acetyltransferases. We then summarize et al., 2004). Although genetic alteration is irreversible, molecular and genetic studies about this pair of HATs epigenetic changes are sometimes reversible. Diet, life- and discuss how their multiprotein complexes may act style, social interaction and environmental cues can as functional units to process molecular information trigger the reset of epigenetic programs, opening unique from histone methylation, phosphoinositides and other windows of opportunity for potential therapeutic inter- signaling cues for acetylation-based action in normal vention (for reviews, see Egger et al., 2004; Wade and and cancer stem cells. Archer, 2006; Nuyt and Szyf, 2007). How nuclear DNA is packaged into chromatin is a major epigenetic determinant. The specific chromatin organization in a given cell type (that is, the packaging state of nuclear Identification of MOZ, MORF and other MYSTic HATs DNA) is important for its unique gene expression pattern. A cancer cell may possess an aberrant chromatin The MOZ gene was initially identified as a fusion organization and thus an abnormal gene expression partner rearranged in the recurrent chromosome translocation t(8;16)(p11;p13) (Borrow et al., 1996). Independently, elegant genetic analysis in budding yeast Correspondence: Dr X-J Yang, Molecular Oncology Group, Royal Victoria Hospital, Room H5.41, McGill University Health Center, 687 identified Sas2 (something about silencing 2) and Pine Avenue West, Montre´ al, Que´ bec H3A 1A1, Canada. Sas3 as two homologous regulators of gene silencing E-mail: [email protected] (Reifsnyder et al., 1996). Sequence comparison revealed MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5409 that the three proteins display high sequence similarity their protein complex assembly, biochemical and within a B300-residue region with TIP60 (HIV Tat- biological function, and regulation. interactive protein of 60 kDa) (Kamine et al., 1996). The homologous region was named the MYST (MOZ, Ybf2/ Sas3, Sas2 and TIP60) domain (Borrow et al., 1996; MOZ and MORF as promiscuous transcriptional Reifsnyder et al., 1996). A putative acetyl-CoA-binding co-activators motif within the domain led to the hypothesis that these proteins are HATs. It was the time when the first few As shown in Figure 1b, MOZ and MORF are highly HATs were identified, so this hypothesis generated a homologous (overall amino-acid sequence identity, considerable amount of excitement. The hypothesis 60%; similarity, 66%), but only show homology to the received additional support from the discovery that a MYST domains of TIP60, MOF and HBO1. MOZ and mutation within the putative acetyl-CoA-binding motif MORF are large proteins (B250 kDa) and possess of Drosophila Mof (male-absent on the first), the fifth intrinsic HAT activity attributable to their MYST protein found to have the MYST domain, compromises domains (Champagne et al., 1999, 2001; Thomas et al., its crucial role in gene dosage compensation (Hilfiker 2000; Kitabayashi et al., 2001a). Both histones H3 and et al., 1997). TIP60 and yeast Esa1 (essential Sas2- H4 are the substrates in vitro. Analysis of recombinant related acetyltransferase 1) were then demonstrated MORF proteins, expressed in and affinity-purified from to possess intrinsic HAT activity (Yamamoto and insect cells, indicates that the full-length protein is more Horikoshi, 1997; Smith et al., 1998). HBO1 (HAT active than the MYST domain alone (Champagne et al., bound to ORC1) was identified as a binding partner of 1999; Pelletier et al., 2003), suggesting that the HAT ORC1 (replication origin complex 1) (Iizuka and Still- activity is regulated by other domains or association man, 1999). BLAST search against expressed sequence with other proteins. Both MOZ and MORF are tag databases for additional MYST proteins led to the autoacetylated (Champagne et al., 1999), with one identification of human MORF as a MOZ paralog acetylation site located C terminus to the MYST (Champagne et al., 1999), and murine MORF was domain (Kim et al., 2006). But the functional signifi- identified independently as Querkopf (Thomas et al., cance of this modification remains to be determined. 2000). As shown in Figure 1b, other protein domains of There are three MYST proteins in budding yeast and MOZ and MORF include an NEMM (N-terminal five in Drosophila or humans (Figure 1). Notably, the part of Enok, MOZ or MORF) domain, tandem PHD MYST domain is the only structural feature common to (plant homeodomain-linked) zinc fingers, a long acidic all members of this family. Metazoan TIP60 proteins are stretch, and an SM (serine/methionine-rich) region. The orthologous to Esa1 owing to similar domain organiza- C-terminal part of the NEMM domain shows some tion, complex formation and cellular function (Doyon sequence similarity to histones H1 and H5, but the and Coˆ te´ , 2004). At the functional level, metazoan MOF function of this domain is not that clear. The H15-like proteins are somewhat similar to Sas2 (Rea et al., 2007). domain was shown to promote nuclear targeting Like TIP60 and MOF, human HBO1 is highly homo- (Kitabayashi et al., 2001a). The tandem PHD fingers logous to its counterpart in Drosophila (Hilfiker et al., are similar to those of Requiem and homologs (Figure 1) 1997; Iizuka and Stillman, 1999; Smith et al., 2005; (Borrow et al., 1996; Nabirochkina et al., 2002). Mendjan et al., 2006). By contrast, MOZ and MORF Requiem is part of an interaction network controlling only show sequence similarity to the very N-terminal pluripotency of embryonic stem cells (Wang et al., region and MYST domain of Enok (Enoki mushroom), 2006). These fingers may have important function. One the most related Drosophila protein (Figures 1a and b) possibility is to recognize methyl-lysine-containing (Scott et al., 2001; Yang, 2004). Different from motifs (Kouzarides, 2007). Drosophila, zebrafish possesses orthologs of MOZ and The SM domain possesses transcriptional activation MORF (Figure 1c). Thus, different from TIP60, MOF and potential, so it was hypothesized that MOZ and MORF HBO1, both MOZ and MORF are vertebrate-specific. are transcriptional coregulators (Champagne et al., The MYST family of HATs was not as widely studied 1999, 2001). Indeed, they interact with Runx proteins as Gcn5/PCAF and p300/CBP, but this situation has and activate transcription (Kitabayashi et al., 2001a; changed greatly in the past few years. It is now known Pelletier et al., 2002; Bristow and Shore, 2003; Kindle that MYST proteins have important roles in various et al., 2005). Upon association with corepressors such as biological processes (Ceol and Horvitz, 2004; Kusch TLE proteins and histone deacetylases (Levanon et al., et al., 2004; Dou et al., 2005; Smith et al., 2005; Mendjan 1998; Westendorf et al., 2002; Schroeder et al., 2004; et al., 2006; Sykes et al., 2006; Tang et al., 2006). These Vega et al., 2004), Runx proteins can function as enzymes have been the subjects of several excellent repressors, raising the interesting question how the reviews (Doyon and Coˆ te´ , 2004; Squatrito et al., 2006; antagonizing repressor and activator roles of Runx Avvakumov and Coˆ te´ , 2007; Lafon et al., 2007; Rea proteins are regulated. One possibility is that cell et al., 2007; Thomas and Voss, 2007). MOZ and MORF signaling controls the switch from one role to the other are quite different from other members of this family (Figure 2a). If so, this would be an important mechan- (Figure 1), so we devote the following sections to this ism for regulating function of MOZ and MORF. pair of HATs. One thematic line for the sections is how MOZ has been shown to physically and functionally the domain structures of MOZ and MORF determine interact with PU.1, an ETS transcription factor that is

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5410 a D. melanogaster C2HC dTIP60 Chromo MYST 541

Mof Chromo MYST 827 Chameau CH (dHBO1) Ser MYST 811

Enok NEMM PHD PHD MYST Neurofilament-like 2291

CG1894 MYST 421 dRequiem PHD PHD 497

b H. sapiens

TIP60 Chromo MYST 513 MOF (MYST1) Chromo MYST 458 CH HBO1 (MYST2) Ser MYST 611

MOZ (MYST3) NEMM PHD PHD MYST ED SM P 2004

MORF (MYST4) NEMM PHD PHD MYST ED SM 1781 HAT Transcriptional activation Requiem PHD PHD 371

c Danio rerio

zMOZ NEMM PHD PHD MYST ED SM P 2246

zMORF NEMM PHD PHD MYST ED SM 2008

Figure 1 Domain organization of MYST proteins from Drosophila (a), humans (b) and the zebrafish Danio rerio (c). For the zebrafish, only MOZ and MORF orthologs (GenBank accession numbers AAT11171 and XP_697383, respectively) are shown. HUGO names for four human proteins are listed in parenthesis. Domains are labeled as follows: Chromo, chromodomain; Ser, serine- rich domain; CH, cysteine/histidine-rich motif; NEMM, N-terminal part of Enok, MOZ or MORF; H15, histones H1- and H5-like domain; PHD, plant homeodomain-linked zinc finger; ED, glutamate/aspartate-rich region; and SM, serine/methionine-rich domain. Within the SM domain of MOZ, there is a proline/glutamine-stretch (labeled by the letter P). Numbers at the right correspond to total residues that each protein possesses. Solid lines underneath the MYST and SM domains of MORF denote the HAT and transcriptional activation domains, respectively. The tandem PHD fingers of Enok, MOZ and MORF are highly homolgous to those of Requiem and homologs (identity B40%, similarity B55%), so the domain organization of Drosophila and human Requiem proteins is also shown for comparison.

essential for the development of myeloid and lymphoid GSTP (glutathione S-transferase placental form), lineages (Katsumoto et al., 2006), whereas MORF a detoxiﬁcation enzyme, during hepatocarcinogenesis is present in the transcriptional co-activator complex (Ohta et al., 2007). Both MOZ and MORF are associated with the nuclear receptor peroxisome widely expressed in various mouse and human proliferator-activated receptor-a (PPARa) (Surapureddi tissues (Borrow et al., 1996; Champagne et al., 1999; et al., 2002). Studies in mouse and zebraﬁsh indicate Thomas et al., 2000, 2006; Katsumoto et al., 2006), MOZ plays a key role in controlling Hox gene so similar to GCN5/PCAF and p300/CBP, they may expression (Miller et al., 2004; Camos et al., function as co-activators for many transcription 2006; Crump et al., 2006; Katsumoto et al., 2006), factors. This interesting possibility awaits further so unknown DNA-binding transcription factors investigation. As discussed below, MOZ and MORF may recruit MOZ to achieve this goal. In addition, are part of tetrameric complexes and noncatalytic MOZ functions as a co-activator of the Nrf2- subunits modulate the roles of these two transcriptional MafK heterodimer to upregulate the expression of co-activators.

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5411 a c Cellular signaling

HDAC TLE Histone methylation Phosphoinositides ? Runx ? ? mRNA

ING5 EAF6

Signaling? BRPF MOZ/MORF

ING5 EAF6 Molecular processing unit

BRPF MOZ/MORF Histone H3 acetylation

Runx ? AcAc mRNA Ac Ac Epigenetic program Chromatin organization Gene expression pattern b

ING5 ING PHD 240

EAF6 LZ 201

BRPF1/2/3 BN IPHD Zn PHD M Bromo PWWP 1214/1058/1205

dBRPF IPHD Zn PHD M Bromo PWWP 1430

Lin49 IPHD Zn PHD M Bromo 1042 PZPM

AF10/17 PHD ZnPHD OM LZ 1068/1093

Figure 2 MOZ and MORF acetyltransferases as molecular processing units. (a) Runx proteins recruit transcriptional corepressors such as histone deacetylases (HDACs) and TLE proteins for transcriptional repression. Upon stimulation by specific signaling events, these corepressors dissociate and co-activators such as MOZ/MORF acetyltransferase complexes associate with Runx proteins to activate transcription. The signaling events involved remain to be determined. (b) Domain organization of proteins associated with MOZ and MORF. LZ, leucine zipper; ING, ING N-terminal domain; BN, BRPF N-terminal region; I and M, the N-terminal and C-terminal parts of the Epc-homology region, respectively; bromo, bromodomain; PWWP, proline-tryptophan-tryptophan-proline- containing domain; and OM, octapeptide motif. dBRPF is derived from NP_610266, and BRPF1 is also known as BR140 (Thompson et al., 1994). (c) MOZ and MORF complexes act as ‘molecular processing units’ to carry out targeted acetylation of histone H3 in response to signaling cues from other chromatin modifiers (for example, histone methyltransferases) and cellular signaling events (for example, phosphoinositides). This modification is part of an integrated epigenetic program important for defining chromatin organization and determining specific gene expression patterns. Question marks denote links awaiting further investigation.

MOZ and MORF as catalytic subunits of quartet or 3) (Figure 2b). Different from MOZ and MORF complexes alone (Champagne et al., 1999; Kitabayashi et al., 2001a), ING5-MOZ/MORF complexes acetylate only Like HAT1, Gcn5, PCAF and many other HATs histone H3 at lysine 14 (Doyon et al., 2006). Consistent (Doyon and Coˆ te´ , 2004; Lee and Workman, 2007; with this, at least BRPF1 stimulates the transcriptional Nagy and Tora, 2007; Parthun, 2007), MOZ and co-activator and HAT activities of MOZ (M Ullah, MORF are catalytic components of protein complexes. JCoˆ te´ and XJ Yang, unpublished data), further They were recently found to be stoichiometric subunits suggesting that noncatalytic subunits affect the function of ING5 (inhibitor of growth 5) complexes (Doyon of MOZ and MORF. et al., 2006). Other subunits of the complexes are As its name implies, yeast Eaf6 was initially identiﬁed EAF6 (homolog of Esa1-associated factor 6) and as a subunit of the Esa1 complex NuA4 (Doyon and BRPF1/2/3 (bromodomain-PHD finger protein 1, 2 Coˆ te´ , 2004). The only recognizable domain in EAF6 is a

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5412 leucine zipper, located at the N-terminal part. The three HAT, an ING, EAF6 and an Epc-N domain protein. BRPF proteins are highly homologous and contain Strikingly, such architecture is conserved in yeast NuA3 multiple domains (Figure 2b). Each BRPF possesses a and piccolo NuA4 complexes, which have the subunit bromodomain and a proline-tryptophan-tryptophan- compositions Sas3-Yng1-Eaf6-Not1 and Esa1-Yng2- proline-containing (PWWP) domain, with the former Eaf6-Epl1 (Epc-like 1), respectively (Carrozza et al., having the potential to recognize acetyl-lysine motifs 2003; Doyon and Coˆ te´ , 2004). Yng1 and Yng2 are two (Seet et al., 2006). The PWWP domain is a 100- to 150- yeast ING proteins. residue module with a highly conserved proline-trypto- Within NuA4, Epl1 serves as a scaffold protein phan-tryptophan-proline motif and has been proposed bridging Yng2, Eaf6 and Esa1 to form the acetyltrans- to have methyl-lysine-binding ability (Stec et al., 2000; ferase core (Boudreault et al., 2003). Similarly, Epc1/2 Maurer-Stroh et al., 2003). bridges the interaction of TIP60 with ING3 and EAF6 Domains I and M of BRPFs are similar to the to form the tetrameric acetyltransferase core (Doyon N-terminal region of Epc1/2 (enhancer of polycomb 1 or et al., 2004), so an interesting possibility is whether 2), a subunit of the TIP60 complex (Figure 2b) (Doyon BRPFs play a similar role in assembling and modulating et al., 2006). These two small domains are homologous MOZ and MORF complexes (Doyon et al., 2006). to an N-terminal region of Drosophila Epc and have Biochemical characterization revealed that BRPF1 been collectively referred to as Epc-N (Figure 2b) (Perry, serves as a scaffold bridging the association of MOZ 2006). Different from those in Epc1/2, domains I and M and MORF with ING5 and Eaf6 (Figure 2c) (M Ullah, of BRPFs are separated by a cluster of zinc fingers, JCoˆ te´ and XJ Yang, unpublished data), attesting that referred to as PZPM (PHD/Zn-knuckle/PHD motif), these complexes resemble those of HBO1 and the core containing two PHD fingers linked by a mononuclear complex of TIP60 (Doyon et al., 2006). These four zinc knuckle. PZPM domains are also present in AF10 HATs show sequence similarity only in the MYST and AF17 (Figure 2b), two frequent fusion partners in domain (Figure 1b), so this domain may be key to MLL-associated leukemia, as well as in a Jumonji- complex formation. Indeed, the MYST domains of domain histone demethylase and the NSD family of MOZ and MORF mediate the interaction with BRPF1, histone methyltransferases (Perry, 2006). Thus, such ING5 and EAF6 (M Ullah, J Coˆ te´ and XJ Yang, domains may be important for chromatin regulation. unpublished data). Unexpected from the limited sequence similarity of As shown in Figure 2b, ING5 has a PHD finger, Enok to MOZ and MORF, one authentic BRPF whose sequence is quite different from those of MOZ, ortholog is present in Drosophila (Figure 2b). Moreover, MORF and BRPFs. PHD fingers are present in many the Caenorhabditis elegans protein Lin49 is highly chromatin regulators (Saha et al., 1995). As PHD fingers homologous to BRPFs except for its PWWP domain have the potential to bind phosphoinositides and are (Figure 2b) (Chamberlin and Thomas, 2000). Thus, it is implicated in nuclear lipid signaling (Gozani et al., likely that the physical link of BRPF proteins to MOZ 2003), an interesting possibility is that phosphoinositides and MORF was acquired late during revolution. bind directly to MOZ and MORF complexes and ING5 is also a stoichiometric subunit of HBO1 regulate their activity. PHD fingers may also function complexes, which contain EAF6 and Jade1/2/3 (gene as methyl-lysine-binding modules. One PHD finger of for apoptosis and differentiation in epithelia 1, -2 or -3) NURF recognizes trimethyl lysine 4 of histone H3 to (Doyon et al., 2006). The Jade proteins are highly couple ATP-dependent chromatin remodeling to histone homologous and share the Epc-N and PZPM domains methylation (Li et al., 2006; Wysocka et al., 2006), and with BRPF proteins (Figure 2b), so these HBO1 the PHD finger of ING2 binds to the same mark to complexes are very similar to MOZ/MORF complexes. stimulate methylation-dependent deacetylation (Pena Consistent with its high sequence similarity to ING5, et al., 2006; Shi et al., 2006). Moreover, methylation of ING4 also forms tetrameric complexes with HBO1, histone H3 at lysine 4 promotes association with yeast EAF6 and Jade proteins. However, ING4 does not Yng1, a subunit of the NuA3 acetyltransferase complex appear to form complexes with MOZ and MORF under (Martin et al., 2006; Taverna et al., 2006). Like ING2, physiological conditions. ING5 binds to trimethyl lysine 4 of histone H3 through ING4 and ING5 are two members of the ING family its PHD domain (Shi et al., 2006), raising the interesting of putative tumor suppressors (Gong et al., 2005). There possibility that ING5 facilitates methylation-dependent are five known members in humans. According to their histone acetylation by MOZ or MORF. As there are sequence similarity, ING proteins can be divided into four additional PHD fingers within a MOZ or MORF three subgroups: ING4 and ING5; ING1 and ING2; complex (Figures 1 and 2b), PHD fingers may recognize and ING3. Both ING1 and ING2 are components of different histone methylation states and act as protein Sin3-HDAC complexes (Loewith et al., 2001; Kuzmi- modules to process molecular information from histone chev et al., 2002; Doyon et al., 2006), whereas ING3 is a methylation marks for modulation of HAT activity subunit of TIP60 complexes (Loewith et al., 2000; (Figure 2c). Phosphoinositides may also regulate Doyon et al., 2004; Gong et al., 2005). It is worth to note methyl-lysine-binding ability to modulate the acetyl- that a tetrameric TIP60 complex core, containing EAF6 transferase activity of MOZ and MORF complexes and Epc1/2, is very similar to ING4 and ING5 (Figure 2c). Based on different domains that they complexes (Doyon and Coˆ te´ , 2004). Thus, ING3, -4 possess, it is tempting to propose that MOZ and MORF and -5 complexes share quartet architecture: a MYST complexes process and convert molecular information

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5413 from other chromatin modifiers, as well as from cellular expression of genes required for proliferation and signaling networks (Figure 2c), for acetylation-based repopulation of potential of stem cells in the hemato- action in different developmental processes. poietic compartment. Consistent with this, MOZ interacts with Runx1 and PU.1, two important hematopoietic transcription factors (Kitabayashi et al., 2001a; Pelletier MOZ and MORF in mouse and zebrafishdevelopment et al., 2002; Katsumoto et al., 2006). Through Runx1, MOZ regulates the expression of MIP-1a,aproinflam- Initial evidence about the role of MORF in mouse matory factor that controls proliferation of hematopietic development came from elegant analysis of a hypo- progenitors (Bristow and Shore, 2003). Furthermore, morphic allele known as Querkopf (Thomas et al., MOZ expression is under direct control of Nanog, Oct4 2000). With a residual MORF mRNA level of B10%, and Sox2 in embryonic stem cells (Boyer et al., 2005). homozygous mice are small in size and display Therefore, emerging evidence suggests that MOZ and craniofacial abnormalities and severe defects in cerebral MORF play important roles in different stem cell cortex development, suggesting that MORF is impor- compartments (Figures 3 and 4). tant for neurogenesis. Neural progenitor cells from these While little is known about zMORF in zebrafish mice exhibit defects in self-renewal and differentiation (Figure 1c), function of zMOZ has been analysed (Merson et al., 2006). Related to this, Enok is required recently. Transformation of the second pharyngeal arch for mushroom body formation in fly brain (Scott et al., into a mirror-image duplicated jaw was observed in two 2001). Moreover, the BRPF ortholog Lin49 in C. mutant zebrafish lines containing translational stop elegans (Figure 2b) is part of a transcriptional hierarchy codons for the acidic region of zMOZ (Miller et al., important for the control of left/right asymmetry in 2004). These mutations are expected to produce chemosensory neurons (Chamberlin and Thomas, 2000; truncated proteins with intact NEMM and MYST Chang et al., 2003). The craniofacial abnormalities of domains (Figure 2b), although it is unclear whether Querkopf mice are in part due to defects in calvarial these proteins are stable in vivo. The zebrafish mutants bones (Thomas et al., 2000). Consistent with this, also display defects in the third and fourth pharyngeal MORF is able to interact physically and functionally arches and in facial motor neuron migration. Analysis with Runx2, a Runt-domain transcription factor with of these mutants revealed that zMOZ is important for high homology to Drosophila Runt and a key player in maintaining Hox expression in the hindbrain and post- osteoblast differentiation (Pelletier et al., 2002), suggest- migratory cranial neural crest, but not for earlier Hox ing that MORF may play a role in regulating gene expression at premigratory stages (Miller et al., 2004; expression during osteogenesis. Crump et al., 2006). zMOZ-dependent expression of Murine MOZ is widely expressed during embryonic Hox proteins such as Hoxa2a and Hoxa2b controls development and in adult tissues (Borrow et al., 1996; segment-specific fate maps of facial skeletal precursors. Katsumoto et al., 2006; Thomas et al., 2006). Two Related to this, Hoxa2 is important for development of research groups mutated the gene by homologous mouse facial somatosensory map (Oury et al., 2006) and recombination. One group introduced an insertion Lin49 controls Hox-C expression in C. elegans (Cham- mutation, resulting in truncation of the MOZ protein berlin and Thomas, 2000; Chang et al., 2003). All of at residue 1538 and deletion of the SM domain these suggest that zMOZ is important for skeletogenesis (Figure 1b) (Thomas et al., 2006). The mRNA level is in zebrafish. Taken together, studies of MOZ and normal, but the mutant protein is undetectable perhaps MORF in mice and zebrafish suggest that these two due to instability and/or insufficient translation. Homo- paralogous HATs are not functionally redundant in zygous mice die at birth, with reduced hematopoiesis vertebrate development. and profound defects in the stem cell compartment. These mice have no long-term repopulating stem cells and display substantial reduction in the number of MOZ and MORF in leukemia, leiomyomata multipotent cells able to form spleen colonies. and other malignancies The other research group replaced exon 2 (containing the first translation codon) with a neo gene cassette Consistent with its important role in hematopoiesis, the (Katsumoto et al., 2006). Although heterozygotes are MOZ gene is rearranged in chromosome translocations fertile and exhibit no obvious morphological defects, associated with acute myeloid leukemia (AML) or therapy- homozygous mutation is embryonic lethal. Homozygous related myelodysplastic syndromes. The balanced chro- embryos remain alive until E14.5, and no truncated mosome translocation t(8;16)(p11;p13) is found in o1% protein or mRNA is detectable. There are fewer of AML patients, especially prominent in French– hematopoietic stem progenitor cells in fetal livers. More- American–British M4/M5 subtypes. The patients are over, expression of c-kit, c-Mpl and HoxA9 is reduced. characterized by a block in blast differentiation at the HoxA9 is a key player in hematopoiesis and leukemogen- granulo-monocytic stage and associated with a poor esis (Grier et al., 2005), and its nuclear localization is response to chemotherapy. In this recurrent transloca- promoted by thrombopietin, the ligand of c-Mpl (Kirito tion, the MOZ gene on chromosome 8p11 is fused to the et al., 2004). Moreover, there is a genetic interaction CBP gene on 16p13, producing a transcript encoding the between c-kit and c-Mpl in repopulating hematopoietic fusion protein MOZ-CBP (Figure 3a) (Borrow et al., stem cells (Antonchuk et al., 2004). Thus, MOZ regulates 1996). The reverse fusion transcript is out of frame and

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5414 a TIF2 CBP inv(8)(p11q13) t(8;16)(p11;p13)

MOZ NEMMPHDPHD MYST ED SM P (8p11) 2004 MOZ t(8;16)(p11;p13)

CBP KIXBr HAT Q (16p13) 2442

MOZ-CBP NEMMPHD PHD MYST ED KIXBr HAT Q

Aberrant acetylation

b MOZ inv(8)(p11q13)

TIF2 bHLH-PAS NID CID (8q13) 1464

MOZ-TIF2 NEMMPHDPHD MYST CID

CBP/p300 ?

Aberrant acetylation

Figure 3 Schematic representation of leukemic fusion proteins from representative chromosome translocations. The domain structure of MOZ is as shown in Figure 1b, with breakpoints marked by vertical arrows. Domains for CBP (a) and TIF2 (b) are labeled as follows: KIX, phospho-CREB-interacting module; Br, bromodomain; Q, glutamine-rich domain; bHLH, basic helix–loop–helix; PAS, Per, aryl hydrocarbon receptor (AhR) and single-minded homology; NID, nuclear receptor-interacting domain; and CID, CBP/p300- interacting domain. While MOZ-CBP possesses double HAT domains and is a ‘super-HAT’ (a), MOZ-TIF2 associates with CBP (or p300) and forms a ‘super-HAT’ complex for aberrant acetylation (b). How the fusion proteins generate aberrant acetylation patterns in vivo is an important issue that remains to be addressed.

produces no protein. MOZ-CBP lacks the SM domain pattern of Hox gene expression, with elevated levels of of MOZ, but gains almost all of CBP (Figure 3a). HOXA9, HOXA10 and MEIS1 (Camos et al., 2006). Mirroring the homology between p300 and CBP, Increased expression of these proteins has been linked to the MOZ gene is fused to that of p300 in translocation leukemia induction (Grier et al., 2005). t(8;22)(p11;q13) (Chaffanet et al., 2000; Kitabayashi TIF2 (transcription intermediary factor 2) is a third et al., 2001b). Moreover, in t(10;16)(q22;p13) associated fusion partner identified for MOZ. In the chromosome with childhood AML or therapeutic myelodysplastic inversion inv(8)(p11;q13) associated with AML or syndromes, the MORF gene on 10q22 (Champagne mixed lineage leukemia, the MOZ gene is fused to that et al., 1999) is fused to that of CBP (Panagopoulos et al., of TIF2 on 8q13 (Carapeti et al., 1998; Liang et al., 2001; Kojima et al., 2003). MOZ-p300 and MORF- 1998). TIF2 is a member of the p160 family of nuclear CBP show similar domain organization to MOZ-CBP receptor co-activators known to interact with p300 and (Figure 3a). CBP (Leo and Chen, 2000). The CBP/p300-interacting In addition to leukemia and myelodysplastic synd- domain (CID) of TIF2 remains intact in MOZ-TIF2 romes, the MORF gene is disrupted in multiple cases of resulting from inv(8)(p11;q13), suggesting that this uterine leiomyomata with t(10;17)(p11;q21) (Moore fusion protein acts through p300 and CBP and that et al., 2004). The fusion partner on 17q21 remains to aberrant acetylation plays a role in leukemogenesis be identified, but an attractive candidate is GCN5. (Figure 3b). MOZ-TIF2 displays intrinsic transforming Thus, like the translocations described above, two ability in vitro and induces AML in murine bone HATs may be involved in this chromosomal abnorm- marrow transplant assays (Deguchi et al., 2003). In ality, thereby causing aberrant histone acetylation and support of the involvement of p300 and CBP, the CID abnormal gene expression (Figure 4). Consistent with of TIF2 is crucial for these activities. Unlike BCR- this, gene expression profiling of samples from AML ABL, MOZ-TIF2 confers self-renewal properties to patients with t(8;16)(p11;p13) revealed a characteristic committed murine hematopoietic progenitors and its

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5415 a b ING5 EAF6 ING5 EAF6

BRPF BRPF MOZ/MORF MOZ/MORF-X

Normal epigenetic programs Oncogenic epigenetic programs

MOZ-CBP? MOZ-p300? MORF-CBP? Aberrant MOZ MOZ MORF MOZ or MORF MOZ-TIF2 MORF-GCN5?? or MORF

Hematopoietic Neural Other? Leukemic Leiomyoma Other cancer?

Type of stem cells Figure 4 Schematic illustration of roles of MOZ and MORF in normal and cancer stem cells. (a) Tetrameric MOZ or MORF complexes are proposed to carry out histone H3 acetylation for generating and maintaining proper epigenetic programs in hematopoietic, neural and other stem cells. MOZ and MORF have been shown to be important for proper development of hematopoietic and neural progenitors, respectively, but it is unclear whether they also function in other stem cell compartments. (b) Chromosome translocations lead to the production of fusion proteins such as MOZ-CBP, MOZ-p300, MORF-CBP and MOZ- TIF2 (collectively referred to as MOZ/MORF-X), which possess intact MYST domains for tetrameric complex formation. These abnormal HAT complexes alter normal epigenetic programs and change chromatin organization and gene expression patterns. Among the fusion proteins, only MOZ-TIF2 has been shown to promote self-renewal of leukemic stem cells, so additional studies are needed to determine whether other fusion proteins can also do so. It remains to be determined whether GCN5 is involved in leiomyomata with t(10;17); different from other known chromosome abnormalities involving MOZ or MORF, the MYST domain is disrupted in this translocation (Moore et al., 2004). Needless to say, additional investigation is needed to establish whether MOZ and MORF play a role in other types of cancer.

expression results in AML that can be serially trans- expression of MOZ and MORF may also cause planted, so MOZ-TIF2 is a unique leukemic gene malignancies. MOZ and MORF are catalytic subunits product (Huntly et al., 2004). At the molecular level, of quartet complexes (Doyon et al., 2006), so it is temp- MOZ-TIF2 depletes CBP from PML bodies to inhibit ting to speculate that genes of the associated subunits transcriptional activity of p53 and the retinoic acid may be mutated in cancers. In addition to genetic receptor RAR (Kindle et al., 2005; Collins et al., 2006). abnormalities in AML, therapy-related myelodysplastic Through p300 or CBP, MOZ-TIF2 may deregulate syndromes, and leiomyomata, function of MOZ and chromatin acetylation. Thus, leukemic MOZ and MORF may be impaired in other types of cancer MORF fusion proteins have the potential to induce (Figure 4). Notably, future investigation is needed to aberrant epigenetic alterations (Figure 3). The MYST examine interesting possibilities. domain is intact in most of the fusion proteins (Figure 3) and sufﬁcient for mediating tetrameric complex formation (M Ullah, J Coˆ te´ and XJ Yang, unpublished data), Conclusions and perspectives so the leukemic fusion proteins may form tetrameric complexes with ING5, EAF6 and a BRPF protein, Mammalian GCN5/PCAF and p300/CBP are two pairs resulting in aberrant acetylation (Figure 4). of HATs well known to have important roles in diverse One interesting question is whether MOZ and MORF cellular processes (reviewed in Goodman and Smolik, also affect other malignancies. It has been suggested that 2000; Sterner and Berger, 2000; Roth et al., 2001; Lee both play a role in maintaining acetylation of histone and Workman, 2007). MOZ and MORF constitute a H4 at lysine 16, a hallmark of cancer (Fraga et al., third pair (Figure 1). Like GCN5/PCAF and p300/CBP, 2005). Moreover, the RUNX1 gene is frequently MOZ and MORF are transcriptional co-activators with involved in human acute leukemia, RUNX2 has intrinsic HAT activity. In their quartet complexes, oncogenic potential, and Runx3 expression is frequently noncatalytic subunits such as BRPF proteins modulate silenced in different types of cancer (reviewed by Ito, HAT activity and co-activator potential of MOZ and 2004). Leukemia-associated Runx1 proteins often lack MORF (Doyon et al., 2006). GCN5/PCAF and p300/ the C-terminal domain, a region important for associa- CBP also acetylate many non-histone proteins (for a tion with MOZ and MORF. Reduced expression of recent review, see Yang and Gre´ goire, 2007), so it will be Runx3 may affect its normal partnership with MOZ and interesting to determine whether MOZ and MORF are MORF. It is also possible that in addition to chromo- able to do so. It will also be interesting to investigate some translocation and inversion, other abnormalities whether they act through ING5 to regulate activities of such as small deletion, point mutation, and deregulated p53 and nuclear factor-kB, which are known to bind

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5416 ING proteins (Gong et al., 2005). MOZ is important for regulates hematopoiesis and proliferation of related expression of Hox proteins, so one possible mechanism progenitor cells (Katsumoto et al., 2006; Thomas et al., is that DNA-binding transcription factors recruit MOZ 2006), MORF has a key role in the neural stem cell to the Hox gene promoters. Thus, MOZ and MORF compartment (Merson et al., 2006), raising an interest- may function as co-activators for multiple transcription ing possibility as to whether MOZ and MORF play a factors. In addition, they may have a role in other role in other stem cells during normal development nuclear processes such as DNA replication (Doyon (Figure 4a). The leukemic fusion protein MOZ-TIF2 et al., 2006). has been shown to promote self-renewal of leukemic Different epigenetic regulators interplay with each stem cells (Huntly et al., 2004), so one question is other to form an integrated epigenetic program for whether other leukemia-derived MOZ and MORF coordinated regulation, so an important question is how proteins have a similar role (Figure 4b). These aberrant MOZ and MORF may interact with other epigenetic proteins may be valuable molecular targets of cancer regulators. Within one MOZ or MORF complex ‘epigenetic therapy,’ so their molecular and functional (Doyon et al., 2006), there are one bromodomain, five analysis will also help develop novel therapeutics. In PHD fingers and one PWWP domain (Figures 1b and conclusion, MOZ and MORF have just emerged as a 2), which have the potential to recognize acetyl or novel pair of HATs important for regulating chromatin methyl histones, thereby leading to crosstalk with other dynamics, preserving epigenetic integrity, and safe- chromatin modifiers and regulators. PHD fingers may guarding stem cell identity in human and other also recognize phosphoinositides. Moreover, cellular vertebrates. signaling networks may directly modify MOZ and MORF complexes to regulate their activities. Thus, it will be interesting and also important to determine Acknowledgements how these complexes process and integrate molecular We thank Jacques Coˆ te´ for sharing unpublished results about information from other epigenetic regulators and MYST protein complexes and two anonymous reviewers for cellular signaling networks (Figure 2c). constructive comments on a previous version of this paper. MOZ and MORF play a key role in regulating Related research was supported by funds from the National different developmental programs, including skeletogen- Cancer Institute of Canada and the Canadian Institutes for esis, neurogenesis and hematopoiesis. Although MOZ Health Research (to XJY).

References

Antonchuk J, Hyland CD, Hilton DJ, Alexander WS. (2004). Carapeti M, Aguiar RC, Goldman JM, Cross NC. (1998). A Synergistic effects on erythropoiesis, thrombopoiesis, and novel fusion between MOZ and the nuclear receptor stem cell competitiveness in mice deficient in thrombopoietin coactivator TIF2 in acute myeloid leukemia. Blood 91: and steel factor receptors. Blood 104: 1306–1313. 3127–3133. Avvakumov N, Coˆ te´ J. (2007). Histone acetyltransferases of Carrozza MJ, Utley RT, Workman JL, Coˆ te´ J. (2003). The the MYST family and their roles in cancer. Oncogene 26: diverse functions of histone acetyltransferase complexes. 5395–5407. Trends Genet 19: 321–329. Baylin SB, Herman JG. (2000). DNA hypermethylation in Ceol CJ, Horvitz HR. (2004). A new class of C. elegans tumorigenesis: epigenetics joins genetics. Trends Genet 16: synMuv genes implicates a Tip60/NuA4-like HAT complex 168–174. as a negative regulator of Ras signaling. Dev Cell 6: 563–576. Berger SL. (2002). Histone modifications in transcriptional Chaffanet M, Gressin L, Preudhomme C, Soenen-Cornu V, regulation. Curr Opin Genet Dev 12: 142–148. Birnbaum D, Pebusque MJ. (2000). MOZ is fused to p300 in Borrow J, Stanton Jr VP, Andresen JM, Becher R, Behm FG, an acute monocytic leukemia with t(8;22). Genes Chromo- Chaganti RS et al. (1996). The translocation somes Cancer 28: 138–144. t(8;16)(p11;p13) of acute myeloid leukaemia fuses a putative Chamberlin HM, Thomas JH. (2000). The bromodomain acetyltransferase to the CREB-binding protein. Nat Genet protein LIN-49 and trithorax-related protein LIN-59 affect 14: 33–41. development and gene expression in Caenorhabditis elegans. Boudreault AA, Cronier D, Selleck W, Lacoste N, Utley RT, Development 127: 713–723. Allard S et al. (2003). Yeast enhancer of polycomb defines Champagne N, Bertos NR, Pelletier N, Wang AH, Vezmar M, global Esa1-dependent acetylation of chromatin. Genes Dev Yang Y et al. (1999). Identification of a human histone 17: 1415–1428. acetyltransferase related to monocytic leukemia zinc finger Boyer LA, Lee TI, Cole MF, Johnstone SE, Levine SS, Zucker protein. J Biol Chem 274: 28528–28536. JP et al. (2005). Core transcriptional regulatory circuitry in Champagne N, Pelletier N, Yang XJ. (2001). The monocytic human embryonic stem cells. Cell 122: 947–956. leukemia zinc finger protein MOZ is a histone acetyltrans- Bristow CA, Shore P. (2003). Transcriptional regulation of the ferase. Oncogene 20: 404–409. human MIP-1alpha promoter by RUNX1 and MOZ. Chang S, Johnston Jr RJ, Hobert O. (2003). A transcrip- Nucleic Acids Res 31: 2735–2744. tional regulatory cascade that controls left/right asymmetry Camos M, Esteve J, Jares P, Colomer D, Rozman M, Villamor in chemosensory neurons of C. elegans. Genes Dev 17: N et al. (2006). Gene expression profiling of acute myeloid 2123–2137. leukemia with translocation t(8;16)(p11;p13) and MYST3- Collins HM, Kindle KB, Matsuda S, Ryan C, Troke PJ, CREBBP rearrangement reveals a distinctive signature with Kalkhoven E et al. (2006). MOZ-TIF2 alters cofactor a specific pattern of HOX gene expression. Cancer Res 66: recruitment and histone modification at the RARbeta2 6947–6954. promoter: differential effects of MOZ fusion proteins on

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5417 CBP- and MOZ-dependent activators. J Biol Chem 281: Kamine J, Elangovan B, Subramanian T, Coleman D, 17124–17133. Chinnadurai G. (1996). Identification of a cellular protein Crump JG, Swartz ME, Eberhart JK, Kimmel CB. (2006). that specifically interacts with the essential cysteine region of Moz-dependent Hox expression controls segment-specific the HIV-1 Tat transactivator. Virology 216: 357–366. fate maps of skeletal precursors in the face. Development Katsumoto T, Aikawa Y, Iwama A, Ueda S, Ichikawa H, 133: 2661–2669. Ochiya T et al. (2006). MOZ is essential for maintenance of Deguchi K, Ayton PM, Carapeti M, Kutok JL, Snyder CS, hematopoietic stem cells. Genes Dev 20: 1321–1330. Williams IR et al. (2003). MOZ-TIF2-induced acute myeloid Khorasanizadeh S. (2004). The nucleosome: from genomic leukemia requires the MOZ nucleosome binding motif and organization to genomic regulation. Cell 116: 259–272. TIF2-mediated recruitment of CBP. Cancer Cell 3: 259–271. Kim SC, Sprung R, Chen Y, Xu Y, Ball H, Pei J et al. (2006). Dou Y, Milne TA, Tackett AJ, Smith ER, Fukuda A, Substrate and functional diversity of lysine acetylation Wysocka J et al. (2005). Physical association and coordinate revealed by a proteomics survey. Mol Cell 23: 607–618. function of the H3 K4 methyltransferase MLL1 and the H4 Kindle KB, Troke PJ, Collins HM, Matsuda S, Bossi D, K16 acetyltransferase MOF. Cell 121: 873–885. Bellodi C et al. (2005). MOZ-TIF2 inhibits transcription by Doyon Y, Cayrou C, Ullah M, Landry AJ, Cote V, Selleck W nuclear receptors and p53 by impairment of CBP function. et al. (2006). ING tumor suppressors are critical regulators Mol Cell Biol 25: 988–1002. of chromatin acetylation required for genome expression Kirito K, Fox N, Kaushansky K. (2004). Thrombopoietin and perpetuation. Mol Cell 21: 51–64. induces HOXA9 nuclear transport in immature hemato- Doyon Y, Coˆ te´ J. (2004). The highly conserved and multi- poietic cells: potential mechanism by which the hormone functional NuA4 HAT complex. Curr Opin Genet Dev 14: favorably affects hematopoietic stem cells. Mol Cell Biol 24: 147–154. 6751–6762. Doyon Y, Selleck W, Lane WS, Tan S, Coˆ te´ J. (2004). Kitabayashi I, Aikawa Y, Nguyen LA, Yokoyama A, Ohki M. Structural and functional conservation of the NUA4 histone (2001a). Activation of AML1-mediated transcription by acetyltransferase complex from yeast to humans. Mol Cell MOZ and inhibition by the MOZ-CBP fusion protein. Biol 24: 1884–1896. EMBO J 20: 7184–7196. Egger G, Liang G, Aparicio A, Jones PA. (2004). Epigenetics Kitabayashi I, Aikawa Y, Yokoyama A, Hosoda F, Nagai M, in human disease and prospects for epigenetic therapy. Kakazu N et al. (2001b). Fusion of MOZ and p300 histone Nature 429: 457–463. acetyltransferases in acute monocytic leukemia with a Fraga MF, Ballestar E, Villar-Garea A, Boix-Chornet M, t(8;22)(p11;q13) chromosome translocation. Leukemia 15: Espada J, Schotta G et al. (2005). Loss of acetylation at 89–94. Lys16 and trimethylation at Lys20 of histone H4 is a Kojima K, Kaneda K, Yoshida C, Dansako H, Fujii N, common hallmark of human cancer. Nat Genet 37: 391–400. Yano T et al. (2003). A novel fusion variant of the MORF Gong W, Suzuki K, Russell M, Riabowol K. (2005). Function and CBP genes detected in therapy-related myelodysplastic of the ING family of PHD proteins in cancer. Int J Biochem syndrome with t(10;16)(q22;p13). Br J Haematol 120: Cell Biol 37: 1054–1065. 271–273. Goodman RH, Smolik S. (2000). CBP/p300 in cell growth, Kouzarides T. (2007). Chromatin modifications and their transformation, and development. Genes Dev 14: 1553–1577. function. Cell 128: 693–705. Gozani O, Karuman P, Jones DR, Ivanov D, Cha J, Kusch T, Florens L, Macdonald WH, Swanson SK, Glaser Lugovskoy AA et al. (2003). The PHD finger of the RL, Yates III JR et al. (2004). Acetylation by Tip60 is chromatin-associated protein ING2 functions as a nuclear required for selective histone variant exchange at DNA phosphoinositide receptor. Cell 114: 99–111. lesions. Science 306: 2084–2087. Grier DG, Thompson A, Kwasniewska A, McGonigle GJ, Kuzmichev A, Zhang Y, Erdjument-Bromage H, Tempst P, Halliday HL, Lappin TR. (2005). The pathophysiology of Reinberg D. (2002). Role of the Sin3-histone deacetylase HOX genes and their role in cancer. J Pathol 205: 154–171. complex in growth regulation by the candidate tumor Hilfiker A, Hilfiker-Kleiner D, Pannuti A, Lucchesi JC. (1997). suppressor p33(ING1). Mol Cell Biol 22: 835–848. MOF, a putative acetyl transferase gene related to the Tip60 Lafon A, Chang CS, Scott EM, Jacobson SJ, Pillus L. and MOZ human genes and to the SAS genes of yeast, is (2007). MYST opportunities for growth control: yeast required for dosage compensation in Drosophila. EMBO J genes illuminate human cancer gene functions. Oncogene 16: 2054–2060. 26: 5373–5384. Horn PJ, Peterson CL. (2002). Chromatin higher order folding Lee KK, Workman JL. (2007). Histone acetyltransferase – wrapping up transcription. Science 297: 1824–1827. complexes: one size doesn’t fit all. Nat Rev Mol Cell Biol Huntly BJ, Shigematsu H, Deguchi K, Lee BH, Mizuno S, 8: 284–295. Duclos N et al. (2004). MOZ-TIF2, but not BCR-ABL, Leo C, Chen JD. (2000). The SRC family of nuclear receptor confers properties of leukemic stem cells to committed coactivators. Gene 245: 1–11. murine hematopoietic progenitors. Cancer Cell 6: 587–596. Levanon D, Goldstein RE, Bernstein Y, Tang H, Goldenberg Iizuka M, Stillman B. (1999). Histone acetyltransferase HBO1 D, Stifani S et al. (1998). Transcriptional repression by interacts with the ORC1 subunit of the human initiator AML1 and LEF-1 is mediated by the TLE/Groucho protein. J Biol Chem 274: 23027–23034. corepressors. Proc Natl Acad Sci USA 95: 11590–11595. Ito Y. (2004). Oncogenic potential of the RUNX gene family: Li B, Carey M, Workman JL. (2007). The role of chromatin ‘overview’. Oncogene 23: 4198–4208. during transcription. Cell 128: 707–719. Jaenisch R, Bird A. (2003). Epigenetic regulation of gene Li H, Ilin S, Wang W, Duncan EM, Wysocka J, Allis CD et al. expression: how the genome integrates intrinsic and (2006). Molecular basis for site-specific read-out of histone environmental signals. Nat Genet 33(Suppl): 245–254. H3K4me3 by the BPTF PHD finger of NURF. Nature 442: Jason LJ, Moore SC, Lewis JD, Lindsey G, Ausio J. (2002). 91–95. Histone ubiquitination: a tagging tail unfolds? Bioessays 24: Liang J, Prouty L, Williams BJ, Dayton MA, Blanchard KL. 166–174. (1998). Acute mixed lineage leukemia with an inv(8)(p11q13)

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5418 resulting in fusion of the genes for MOZ and TIF2. Blood 92: Runt-domain transcription factor Runx2. Oncogene 21: 2118–2122. 2729–2740. Loewith R, Meijer M, Lees-Miller SP, Riabowol K, Young D. Pena PV, Davrazou F, Shi X, Walter KL, Verkhusha VV, (2000). Three yeast proteins related to the human candidate Gozani O et al. (2006). Molecular mechanism of histone tumor suppressor p33(ING1) are associated with H3K4me3 recognition by plant homeodomain of ING2. histone acetyltransferase activities. Mol Cell Biol 20: Nature 442: 100–103. 3807–3816. Perry J. (2006). The Epc-N domain: a predicted protein- Loewith R, Smith JS, Meijer M, Williams TJ, Bachman N, protein interaction domain found in select chromatin Boeke JD et al. (2001). Pho23 is associated with the Rpd3 associated proteins. BMC Genomics 7:6. histone deacetylase and is required for its normal function in Rea S, Xouri G, Akhtar A. (2007). Male absent on the first regulation of gene expression and silencing in Saccharo- (MOF): from flies to humans. Oncogene 26: 5385–5394. myces cerevisiae. J Biol Chem 276: 24068–24074. Reifsnyder C, Lowell J, Clarke A, Pillus L. (1996). Yeast SAS Martin DG, Grimes DE, Baetz K, Howe L. (2006). Methyla- silencing genes and human genes associated with AML and tion of histone H3 mediates the association of the NuA3 HIV-1 Tat interactions are homologous with acetyltrans- histone acetyltransferase with chromatin. Mol Cell Biol 26: ferases. Nat Genet 14: 42–49. 3018–3028. Roth SY, Denu JM, Allis CD. (2001). Histone acetyltrans- Maurer-Stroh S, Dickens NJ, Hughes-Davies L, Kouzarides ferases. Annu Rev Biochem 70: 81–120. T, Eisenhaber F, Ponting CP. (2003). The Tudor domain Saha V, Chaplin T, Gregorini A, Ayton P, Young BD. (1995). ‘Royal Family’: Tudor, plant Agenet, Chromo, PWWP and The leukemia-associated-protein (LAP) domain, a cysteine- MBT domains. Trends Biochem Sci 28: 69–74. rich motif, is present in a wide range of proteins, including Mendjan S, Taipale M, Kind J, Holz H, Gebhardt P, Schelder MLL, AF10, and MLLT6 proteins. Proc Natl Acad Sci USA M et al. (2006). Nuclear pore components are involved in 92: 9737–9741. the transcriptional regulation of dosage compensation in Schroeder TM, Kahler RA, Li X, Westendorf JJ. (2004). Drosophila. Mol Cell 21: 811–823. Histone deacetylase 3 interacts with Runx2 to repress the Merson TD, Dixon MP, Collin C, Rietze RL, Bartlett PF, osteocalcin promoter and regulate osteoblast differentiation. Thomas T et al. (2006). The transcriptional coactivator J Biol Chem 279: 41998–42007. Querkopf controls adult neurogenesis. J Neurosci 26: Scott EK, Lee T, Luo L. (2001). Enok encodes a Drosophila 11359–11370. putative histone acetyltransferase required for mushroom Miller CT, Maves L, Kimmel CB. (2004). moz regulates Hox body neuroblast proliferation. Curr Biol 11: 99–104. expression and pharyngeal segmental identity in zebrafish. Seet BT, Dikic I, Zhou MM, Pawson T. (2006). Reading Development 131: 2443–2461. protein modifications with interaction domains. Nat Rev Moore SD, Herrick SR, Ince TA, Kleinman MS, Cin PD, Mol Cell Biol 7: 473–483. Morton CC et al. (2004). Uterine leiomyomata with t(10;17) Shi X, Hong T, Walter KL, Ewalt M, Michishita E, Hung T disrupt the histone acetyltransferase MORF. Cancer Res 64: et al. (2006). ING2 PHD domain links histone H3 lysine 4 5570–5577. methylation to active gene repression. Nature 442: 96–99. Nabirochkina E, Simonova OB, Mertsalov IB, Kulikova DA, Smith ER, Cayrou C, Huang R, Lane WS, Coˆ te´ J, Lucchesi Ladigina NG, Korochkin LI et al. (2002). Expression JC. (2005). A human protein complex homologous to the pattern of dd4, a sole member of the d4 family of tran- Drosophila MSL complex is responsible for the majority scription factors in Drosophila melanogaster. Mech Dev 114: of histone H4 acetylation at lysine 16. Mol Cell Biol 25: 119–123. 9175–9188. Nagy Z, Tora L. (2007). Distinct GCN5/PCAF-containing Smith ER, Eisen A, Gu W, Sattah M, Pannuti A, Zhou J et al. complexes function as coactivators and are involved in tran- (1998). ESA1 is a histone acetyltransferase that is essential scription factor and global histone acetylation. Oncogene 26: for growth in yeast. Proc Natl Acad Sci USA 95: 3561–3565. 5341–5357. Spencer VA, Davie JR. (1999). Role of covalent modifications Nuyt AM, Szyf M. (2007). Developmental programming of histones in regulating gene expression. Gene 240: 1–12. through epigenetic changes. Circ Res 100: 425–452. Squatrito M, Gorrini C, Amati B. (2006). Tip60 in DNA Ohta K, Ohigashi M, Naganawa A, Ikeda H, Sakai M, damage response and growth control: many tricks in one Nishikawa J et al. (2007). Histone acetyltransferase MOZ acts HAT. Trends Cell Biol 16: 433–442. as a co-activator of Nrf2-MafK and induces tumour marker Stec I, Nagl SB, van Ommen GJ, den Dunnen JT. (2000). The gene expression during hepatocarcinogenesis. Biochem J 402: PWWP domain: a potential protein–protein interaction 559–566. domain in nuclear proteins influencing differentiation? Oury F, Murakami Y, Renaud JS, Pasqualetti M, Charnay P, FEBS Lett 473: 1–5. Ren SY et al. (2006). Hoxa2- and rhombomere-dependent Sterner DE, Berger SL. (2000). Acetylation of histones and development of the mouse facial somatosensory map. transcription-related factors. Microbiol Mol Biol Rev 64: Science 313: 1408–1413. 435–459. Panagopoulos I, Fioretos T, Isaksson M, Samuelsson U, Strahl BD, Allis CD. (2000). The language of covalent histone Billstrom R, Strombeck B et al. (2001). Fusion of the modifications. Nature 403: 41–45. MORF and CBP genes in acute myeloid leukemia with the Surapureddi S, Yu S, Bu H, Hashimoto T, Yeldandi AV, t(10;16)(q22;p13). Hum Mol Genet 10: 395–404. Kashireddy P et al. (2002). Identification of a transcriptionally Parthun MR. (2007). Hat1: The emerging cellular roles of a active peroxisome proliferator-activated receptor alpha-inter- type B histone acetyltransferase. Oncogene 26: 5319–5328. acting cofactor complex in rat liver and characterization of Pelletier N, Champagne N, Lim H, Yang XJ. (2003). PRIC285 as a coactivator. Proc Natl Acad Sci USA 99: Expression, purification, and analysis of MOZ and MORF 11836–11841. histone acetyltransferases. Methods 31: 24–32. Sykes SM, Mellert HS, Holbert MA, Li K, Marmorstein R, Pelletier N, Champagne N, Stifani S, Yang XJ. (2002). MOZ Lane WS et al. (2006). Acetylation of the p53 DNA-binding and MORF histone acetyltransferases interact with the domain regulates apoptosis induction. Mol Cell 24: 841–851.

Oncogene MOZ and MORF histone acetyltransferases X-J Yang and M Ullah 5419 Tang Y, Luo J, Zhang W, Gu W. (2006). Tip60-dependent Wade PA, Archer TK. (2006). Epigenetics: environmental acetylation of p53 modulates the decision between cell-cycle instructions for the genome. Environ Health Perspect 114: arrest and apoptosis. Mol Cell 24: 827–839. A140–141. Taverna SD, Ilin S, Rogers RS, Tanny JC, Lavender H, Li H et al. Wang J, Rao S, Chu J, Shen X, Levasseur DN, Theunissen (2006). Yng1 PHD finger binding to H3 trimethylated at K4 TW et al. (2006). A protein interaction network promotes NuA3 HAT activity at K14 of H3 and transcription for pluripotency of embryonic stem cells. Nature 127: at a subset of targeted ORFs. Mol Cell 24: 785–796. 1137–1150. Thomas T, Corcoran LM, Gugasyan R, Dixon MP, Brodnicki T, Westendorf JJ, Zaidi SK, Cascino JE, Kahler R, van Wijnen Nutt SL et al. (2006). Monocytic leukemia zinc finger protein is AJ, Lian JB et al. (2002). Runx2 (Cbfa1, AML-3) interacts essential for the development of long-term reconstituting with histone deacetylase 6 and represses the p21CIP1/WAF1 hematopoietic stem cells. Genes Dev 20: 1175–1186. promoter. Mol Cell Biol 22: 7982–7992. Thomas T, Voss AK. (2007). The diverse biological roles of Wysocka J, Swigut T, Xiao H, Milne TA, Kwon SY, Landry J MYST histone acetyltransferase family proteins. Cell Cycle et al. (2006). A PHD finger of NURF couples histone H3 6: 696–704. lysine 4 trimethylation with chromatin remodelling. Nature Thomas T, Voss AK, Chowdhury K, Gruss P. (2000). 442: 86–90. Querkopf, a MYST family histone acetyltransferase, is Yamamoto T, Horikoshi M. (1997). Novel substrate required for normal cerebral cortex development. Develop- specificity of the histone acetyltransferase activity of ment 127: 2537–2548. HIV-1-Tat interactive protein Tip60. J Biol Chem 272: Thompson KA, Wang B, Argraves WS, Giancotti FG, 30595–30598. Schranck DP, Ruoslahti E. (1994). BR140, a novel zinc- Yang XJ. (2004). The diverse superfamily of lysine acetyl- finger protein with homology to the TAF250 subunit of transferases and their roles in leukemia and other diseases. TFIID. Biochem Biophys Res Commun 198: 1143–1152. Nucleic Acids Res 32: 959–976. Vega RB, Matsuda K, Oh J, Barbosa AC, Yang X, Meadows Yang XJ, Gre´ goire S. (2007). Metabolism, cytosketon and E et al. (2004). Histone deacetylase 4 controls chondrocyte cellular signaling in grip of protein Ne- and O-acetylation. hypertrophy during skeletogenesis. Cell 119: 555–566. EMBO Rep 8: 556–562.

Oncogene