Transcriptional Control of Erythropoiesis: Emerging Mechanisms and Principles

Home , MAFG, NFE2, Precursor cell, RGS18, Small Maf, TAL1

Oncogene (2007) 26, 6777–6794 & 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc REVIEW Transcriptional control of erythropoiesis: emerging mechanisms and principles

S-I Kim and EH Bresnick

Department of Pharmacology, University of Wisconsin School of Medicine and Public Health, Medical Sciences Center, Madison, WI, USA

Transcriptional networks orchestrate fundamental bio- leads to the development and progression of specific logical processes, including hematopoiesis, in which leukemias (Gilliland et al., 2004; Rosenbauer and hematopoietic stem cells progressively differentiate into Tenen, 2007). Whereas the focus of this review is on specific progenitors cells, which in turn give rise to the transcriptional mechanisms that underlie red blood cell diverse blood cell types. Whereas transcription factors development, the process termed erythropoiesis, the recruit coregulators to chromatin, leading to targeted fundamental principles emerging from these studies chromatin modification and recruitment of the transcrip- have broad relevance in diverse systems. tional machinery, many questions remain unanswered Since a host of transcriptional regulators and signal- regarding the underlying molecular mechanisms. Further- ing pathways that control erythropoiesis have already more, how diverse cell type-specific transcription factors been identified, major efforts are focused on elucidating function cooperatively or antagonistically in distinct the underlying molecular mechanisms. Canonical trans- cellular contexts is poorly understood, especially since criptional mechanisms involve sequence-specific binding genes in higher eukaryotes commonly encompass broad of trans-acting factors (transcription factors) to DNA chromosomal regions (100 kb and more) and are littered motifs termed cis-elements in chromatin, followed by with dispersed regulatory sequences. In this article, we recruitment of additional regulatory proteins (coregula- describe an important set of transcription factors and tors) via direct protein–protein interactions (Kadonaga, coregulators that control erythropoiesis and highlight 2004). Coregulators typically exist as large multiprotein emerging transcriptional mechanisms and principles. It is complexes and either mediate activation (coactivators) not our intent to comprehensively survey all factors or repression (corepressors) (Bresnick et al., 2006; Lee implicated in the transcriptional control of erythropoiesis, and Workman, 2007). Certain coregulator complexes but rather to underscore specific mechanisms, which have mediate both activation and repression in a context- potential to be broadly relevant to transcriptional control dependent manner (Crispino et al., 1999; Rogatsky in diverse systems. et al., 2002). It is instructive to classify coregulators as Oncogene (2007) 26, 6777–6794; doi:10.1038/sj.onc.1210761 chromatin remodeling or chromatin modifying enzymes, based on whether they lack or have the capacity, Keywords: erythropoiesis; chromatin; transcription; respectively, to post-translationally modify histones that epigenetic mark form the octameric core of the nucleosome. Chromatin remodeling enzymes utilize ATP in a biochemical reaction that modifies nucleosome structure and alters nucleosome positioning (Saha et al., 2006). Since chromatin can be a formidable impediment to trans- Introduction cription factor access to nucleosomal DNA (Hager et al., 1993), remodeling enzymes regulate transcription The differentiation of hematopoietic stem cells (HSCs) factor access to chromatin. In addition, as nucleosomal into specific progenitor cells, and ultimately into diverse filaments condense into higher-order structures (Felsen- blood cell types, is intricately controlled by intercellular feld and Groudine, 2003), remodeling enzyme-depen- and intracellular signaling mechanisms (Kaushansky, dent chromatin structure transitions almost certainly 2006; Mikkola and Orkin, 2006). These mechanisms regulate higher-order chromatin folding. commonly target transcriptional regulators, which in In contrast to remodeling enzymes, chromatin mod- turn establish complex transcriptional networks. Dysre- ifying enzymes catalyse a plethora of histone post- gulation of signaling and transcriptional mechanisms translational modifications, including acetylation, methylation, phosphorylation, ubiquitination, sumoylation and ADP ribosylation, which are termed epigenetic Correspondence: Dr EH Bresnick, Department of Pharmacology, marks (Allfrey et al., 1964; Fischle et al., 2003). Such University of Wisconsin School of Medicine and Public Health, 1300 University Avenue, 385 Medical Sciences Center, Madison, WI 53706, modifications are particularly prevalent within the USA. conserved N-terminal core histone tails, but also occur E-mail: [email protected] within the central globular domain (Berger, 2007; Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6778 Bernstein et al., 2007; Kouzarides, 2007). In certain subdomains (Forsberg et al., 2000b; Kiekhaefer et al., cases, epigenetic marks regulate chromatin structure 2002; Bulger et al., 2003). directly. For example, histone acetylation counteracts Recent work with the b-globin system has provided higher-order chromatin folding with purified, reconsti- important insights into perhaps the most fundamental tuted chromatin templates in vitro (Tse et al., 1998). In aspect of transcriptional control, how trans-acting addition, specific epigenetic marks function as ligands to factors recognize and occupy functional sites in chro- attract additional regulatory factors to the chromatin matin (Bresnick et al., 2006). cis-elements that mediate (Marmorstein, 2001; Fischle et al., 2003). Protein modules transcription factor binding to chromatin are typically present in transcriptional regulatory factors, including the short sequence motifs of 5–10 base pairs, and such bromodomain that recognizes acetyl-lysine (Dhalluin motifs occur at a high frequency throughout genomes, et al., 1999; Jacobson et al., 2000; Owen et al., 2000), simply based on the statistical distribution of nucleo- bind histones bearing specific epigenetic marks. tides. The development of the chromatin immunopreci- In addition to the canonical mechanisms noted above, pitation (ChIP) assay to measure protein occupancy at certain transcription factors retain functionality upon endogenous chromatin sites in living cells has provided a disabling their sequence-specific DNA binding activity, powerful technology for analysing how transcription indicating the importance of DNA binding-independent factors select functional sites in the genome (Orlando mechanisms in certain contexts (Reichardt et al., 1998; et al., 1997; Johnson and Bresnick, 2002; Im et al., 2004; Porcher et al., 1999; Tuckermann et al., 1999). Given Kirmizis and Farnham, 2004). that transcriptional complexes assembled at promoters Extensive analyses of chromatin occupancy by the and distal regulatory elements, such as enhancers and hematopoietic zinc-finger protein GATA-1, which is locus control regions (LCRs), often contain a large discussed below in detail, revealed that only a small cohort of factors that engage in a multitude of protein– fraction of high-affinity GATA motifs are occupied in protein interactions (Bresnick et al., 2006; Dean, 2006), chromatin (Johnson et al., 2002, 2007; Grass et al., 2003, it seems reasonable that certain transcription factors can 2006; Pal et al., 2004b; Im et al., 2005; Martowicz et al., integrate into such complexes without a critical DNA 2005). Intriguingly, the cell type-specific coregulator binding activity requirement. Such tethering might also Friend of GATA-1 (FOG-1), which mediates certain permit stable transcription factor interactions at low- biological functions of GATA-1 (Tsang et al., 1997, affinity motifs, which otherwise would not be occupied. 1998), facilitates GATA-1 occupancy at certain, but not However, considerably less is known about transcrip- all, chromatin target sites (Letting et al., 2004; Pal et al., tion factor tethering mechanisms vs DNA binding- 2004a). We refer to this FOG-1 activity as chromatin dependent mechanisms. occupancy facilitator (COF) activity. Thus, mechanisms Important principles underlying transcriptional con- responsible for the selective recognition of a small subset trol in higher eukaryotes have regularly emerged from of motifs represent a crucial primary mode of transcrip- mechanistic studies on the regulation of the b-like globin tional control, and specific protein–protein interactions gene cluster, a commonly used model system to influence this decision-making process. Subsequent to elucidate cell type-specific and developmental-stage transcription factor occupancy of chromatin, a multi- specific transcriptional mechanisms (Bank, 2006; Bres- tude of regulatory events, including coregulator recruit- nick et al., 2006). Studies with this system have led to ment, coregulator-dependent chromatin structure multiple seminal discoveries including (1) the concept of transitions, dynamics of transcription factor and cor- an LCR, a transcriptional regulatory element that egulator interactions with the template, and interactions activates a linked gene proportional to template copy between these components and RNA Polymerase II (Pol number and independent of the chromosomal integra- II) dictate the magnitude and kinetics of the transcription site (Forrester et al., 1987; Grosveld et al., 1987); (2) tional response. identification of the founding member of the GATA Transcriptional control at endogenous loci in higher transcription factor family (GATA-1), which regulates eukaryotes requires numerous cell type-specific tran- the development of specific blood cell lineages (Evans scription factors, and therefore, not surprisingly, multi- and Felsenfeld, 1989; Tsai et al., 1989); (3) identification ple factors are implicated in regulating b-like globin of a novel chromatin insulator element (chicken gene transcription (Bank, 2006; Bresnick et al., 2006). hypersensitive site 4 (HS4)) (Chung et al., 1993), which Studies in cells genetically modified to lack specific is efficacious as a biotechnology tool in gene therapy factors have revealed that individual factors can have vectors and various expression cassettes (Emery et al., crucial non-redundant functions (Lu et al., 1994; Weiss 2000, 2002; Puthenveetil et al., 2004); (4) the demonstra- et al., 1997; Coghill et al., 2001). Despite these examples tion that cell- and tissue-specific chromatin domains can of non-redundant function, transcription factors often be characterized by broad regions of enhanced sensitiv- share a limited number of broadly expressed coregula- ity to DNaseI, which is considered to be a hallmark of tors. Even though broadly expressed coregulators unfolded chromatin domains (Hebbes et al., 1988; Litt mediate transcriptional control of a plethora of genes et al., 2001); and (5) the demonstration that cell- and in diverse cell and tissue types, three such coregulators, tissue-specific active chromatin domains can have the protein/histone acetyltransferase CREB-binding discontinuous histone modification patterns in which protein (CBP)/p300 (Chrivia et al., 1993; Arias et al., irregularly distributed epigenetic marks associated with 1994), the thyroid hormone receptor-associated protein active or repressed chromatin delineate functional 220 (TRAP220) component of the ‘Mediator’ complex

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6779 (Ito et al., 2000), and the Brahma-related gene1 (BRG1) 1995; Simon et al., 1992; Weiss et al., 1994). GATA-1 component of the SWI/SNF remodeling complex arrests cellular proliferation (Rylski et al., 2003; Munuga- (Khavari et al., 1993) are crucial for erythropoiesis lavadla et al., 2005), and GATA-1-mediated survival of and are discussed below in detail. erythroid precursors involves induction of Bcl-xL expres- A major challenge in dissecting transcriptional sion (Weiss and Orkin, 1995; Gregory et al., 1999). mechanisms that choreograph cellular differentiation is The C-terminal GATA-1 zinc finger recognizes to understand the basis for context-dependent mechan- GATA motifs (Martin and Orkin, 1990), whereas istic permutations operational at distinct loci and at a the N-terminal zinc finger stabilizes DNA binding at single locus during different stages of differentiation. certain palindromic motifs (Trainor et al., 1996). The The organization of a cis-element near one or more N-terminal finger also directly binds FOG-1 (Crispino additional cis-elements in a specific architecture yields a et al., 1999) and TRAP220 (Stumpf et al., 2006) composite motif, which represents an important com- (Figure 1). GATA-1 also binds CBP/p300 (Blobel ponent of context-dependent transcriptional mechan- et al., 1998) and several transcription factors including isms. This common regulatory mode diversifies the PU.1 (Rekhtman et al., 1999; Nerlov et al., 2000), Sp1 spectrum of cell type-specific transcriptional outputs (Merika and Orkin, 1995) and erythroid Kruppel-like achieved by a finite number of factors. Subsequent to factor (EKLF) (Merika and Orkin, 1995); (Figure 1). reviewing several important transcription factors and Furthermore, proteomics analysis of GATA-1 interac- coregulators that control erythropoiesis, emerging me- tors identified diverse GATA-1-containing multiprotein chanisms and principles will be discussed, specifically complexes (Rodriquez et al., 2005). chromatin target site selection, mechanistic permuta- Analyses of GATA-1 function, specifically to eluci- tions that diversify factor functionality, the role of date mechanisms underlying chromatin target site selec- composite elements in combinatorial control, similar tion, transcriptional regulatory activity, and protein– and distinct functions of transcription factor family protein interaction logic have been greatly facilitated by members, novel coregulator mechanisms, and Pol II the development of a facile genetic complementation function at extragenic sites. assay. G1E cells were derived from Gata1-targeted mouse embryonic stem cells (Weiss et al., 1997) and have been engineered to stably express a ligand- Transcription factors that control erythropoiesis: a brief activated estrogen receptor hormone-binding domain primer fusion to GATA-1 (ER-GATA-1) (Gregory et al., 1999). In response to estradiol or tamoxifen, G1E–ER-GATA- GATA-1 1 cells differentiate from proerythroblast-like cells to a Following the recognition that a common sequence motif more mature phenotype. Transcriptional profiling pro- [(A/T)GATA(A/G)] (GATA motif) exists within transcrip- vided evidence that the G1E system recapitulates a tional regulatory regions at most, if not all, erythroid cell- normal window of erythropoiesis (Welch et al., 2004). specific genes (Evans et al., 1988; Reitman and Felsenfeld, Furthermore, the ER-GATA-1 chromatin occupancy 1988), a dual zinc-finger transcription factor, GATA-1, pattern is indistinguishable from that of endogenous that binds this motif (Ko and Engel, 1993; Merika GATA-1 in mouse erythroleukemia (MEL) cells (Johnson and Orkin, 1993) was purified and cloned (Evans and et al., 2002; Im et al., 2005) and in human embryonic stem Felsenfeld, 1989; Tsai et al., 1989). This founding member cell-derived erythroid cells (unpublished), further justify- of the GATA transcription factor family is expressed in ing the use of G1E cells as a physiologically relevant erythroid, megakaryocytic, eosinophil, and mast cell system to dissect mechanisms regulating erythropoiesis. lineages (Cantor and Orkin, 2005). Targeted disruption Prior to the rise in GATA-1 expression during erythro- of Gata1 in mice provided evidence for its essential poiesis, GATA-2 is expressed in certain hematopoietic function in stimulating erythropoiesis (Pevny et al., 1991, precursors, including HSCs (Tsai et al., 1994; Tsai and

GATA-1 204 228 258 282 NCN-finger C-finger FOG-1 TRAP220 SP-1 PU.1 ? CBP ? EKLF

Figure 1 Structural organization of GATA-1. Schematic representation of GATA-1. Two zinc fingers are depicted as colored bars. The amino-terminal (N) zinc finger of GATA-1 binds FOG-1, TRAP220 and Sp1, whereas the C-terminal zinc finger binds PU.1, CBP/ p300 and EKLF. Evidence for binding is described in the following studies: Crispino et al. (1999) (FOG-1); Stumpf et al. (2006) (TRAP220); Merika and Orkin, (1995) (Sp1); Nerlov et al. (2000) and Rekhtman et al. (1999) (PU.1); Blobel et al. (1998) (CBP); and Merika and Orkin (1995) (EKLF). CBP, CREB-binding protein; EKLF, erythroid Kruppel-like factor; FOG-1, Friend of GATA-1; TRAP220, thyroid hormone receptor-associated protein 220.

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6780 Orkin, 1997; Minegishi et al., 1999). GATA-2 is crucial Pol II has functions at or near the LCR prior to for the survival and proliferation of HSCs (Tsai et al., promoter activation. Such functions might include long- 1994; Tsai and Orkin, 1997). Studies in G1E cells revealed range Pol II transfer, in which the LCR provides that, in the absence of GATA-1, GATA-2 occupies Pol II to the promoter (Johnson et al., 2001), transcrip- conserved regulatory sequences dispersed over B80 kb of tion-dependent chromatin remodeling (Travers, 1999) or its own locus, and as GATA-1 levels rise, GATA-1 generation of functional intergenic transcripts (Gribnau displaces GATA-2 from these sites (Grass et al., 2003, et al., 2000). Subsequent to the report of Pol II 2006; Martowicz et al., 2005). This ‘GATA switch’ is occupying the LCR prior to bmajor activation (Johnson tightly coupled to Gata2 repression (Grass et al., 2003) et al., 2001), other examples emerged, in which Pol II and also occurs during murine embryonic stem (ES) cell occupancy of distal regulatory regions precedes activa- differentiation into erythroid cells (Lugus et al., 2007). As tion (Szutorisz et al., 2005b). FOG-1 facilitates GATA switches (Pal et al., 2004a), it is While the function of LCR-bound Pol II is not yet attractive to propose that GATA switches are an established, kinetic analyses in G1E–ER-GATA-1 cells, important component of the repression mechanism. graded activation of ER-GATA-1 by titrating estradiol, Whereas the precise mechanism by which FOG-1 and comparison of locus structure/function in G1E–ER- facilitates GATA switches is unclear, studies with a GATA-1 cells cultured at 37 vs 25 1C provided important GATA-1 mutant defective in FOG-1 binding (V205G) insights into the multistep activation mechanism and in FOG-1-null cells (Cantor et al., 2002) revealed (Im et al., 2005; Kim et al., 2007). These experiments COF activity at certain chromatin sites (Letting et al., revealed that GATA-1 occupies the LCR prior to the 2004; Pal et al., 2004a). Upon engaging FOG-1, GATA-1 promoter. An LCR subcomplex also assembles prior to might acquire a competitive advantage for chromatin promoter complex assembly and establishment of a occupancy at GATA-2-bound chromatin sites. Thus, chromatin loop, which increases the proximity of the FOG-1-dependent GATA switches at Gata2 and addi- LCR relative to the adult b-like globin gene promoters tional loci might represent an important process driving (Kim et al., 2007). erythropoiesis (Bresnick et al., 2005). High-affinity GATA motifs are scattered throughout Once GATA switches at the Gata2 locus occur, how the locus (Im et al., 2005), and therefore the mere does GATA-1 instigate Gata2 repression? Whereas presence of such motifs is insufficient to determine the fundamental biochemical differences between GATA-1 preferential LCR vs promoter occupancy. Of potential and GATA-2 are not established, Gata2 repression is relevance to the preferential LCR occupancy, GATA-1 dependent on FOG-1. Analysis of GST–FOG-1 inter- target genes are differentially sensitive to changes in actions with proteins in MEL cell nuclear extracts GATA-1 levels/activities (Johnson et al., 2006). Thus, as identified the nucleosome remodeling and histone GATA-1 levels rise during erythropoiesis, one would deacetylase (NuRD) complex (Hong et al., 2005), which predict that distinct subsets of target genes will be mediates transcriptional repression and chromatin activated, which has important implications for under- remodeling (Ayer, 1999; Zhang et al., 1999b). The standing the progression of events that culminate in N-terminal repression domain of FOG-1, and also that red blood cell development. As GATA-1 mutations in of FOG-2 (Lin et al., 2004), which functions in the megakaryoblastic leukemia result in truncation of the cardiovascular system (Svensson et al., 1999; Tevosian GATA-1 N terminus (Wechsler et al., 2002; Crispino, et al., 1999, 2000; Crispino et al., 2001), binds the NuRD 2005), and this short form of GATA-1 (Bourquin et al., complex containing Mi-2b (CHD4), RbAp46, RbAp48, 2006) and a larger N-terminal deletion (Johnson MTA-1, MTA-2, p66, histone deacetylase 1 and histone et al., 2006), have reduced transactivation activity, it is deacetylase 2 subunits (Hong et al., 2005). The FOG-1– attractive to propose that genes requiring maximal NuRD complex persists over multiple purification steps, GATA-1 levels are preferentially dysregulated by the indicative of a high stability in vitro (Hong et al., 2005). leukemogenic GATA-1 mutant. Components of the NuRD complex occupy several GATA-1 is subject to multi-site phosphorylation GATA-1 target genes including c-Myc, c-Myb, Gata2 (Crossley and Orkin, 1994; Zhao et al., 2006) and and Tac-2 in vivo (Hong et al., 2005; Johnson et al., acetylation (Boyes et al., 1998; Hung et al., 1999), 2006). Whether GATA-2 differs from GATA-1 in its and such modifications could obviously contribute to capacity to recruit and utilize the FOG-1–NuRD key steps in GATA-1 function. However, the role of complex at the Gata2 locus is unknown. Nevertheless, phosphorylation in regulating GATA-1 activity has NuRD components and GATA-1 occupancy at GATA- been elusive, as a triple phosphorylation site mutant has 1 target genes do not strictly correlate (Hong et al., 2005; indistinguishable activity vs wild-type GATA-1 in vivo Johnson et al., 2006), and therefore either additional (Rooke and Orkin, 2006). corepressors are crucial for repression or NuRD We have discussed GATA factor mechanisms complex activity is regulated post-recruitment. independent of additional hematopoietic transcription In contrast to GATA-1-mediated Gata2 repression, factors, but GATA-1 engages in important functional GATA-1 activates many genes, including the adult interactions with other transcription factors, which need b-like globin gene bmajor. In G1E cells, GATA-1 to be considered. For example, the GATA-1 interaction recruits Pol II to the bmajor promoter (Johnson et al., with PU.1, a member of the Ets family of transcription 2002). As LCR-bound Pol II can precede promoter- factors (Klemsz et al., 1990) illustrates such an interac- bound Pol II (Johnson et al., 2001, 2003), it is likely that tion. PU.1 is expressed predominantly in blood cells and

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6781 is required for the development of hematopoietic occupies HS1–HS3of the b-globin LCR and the bmajor progenitor cells into B and T lymphocytes, monocytes promoter in G1E–ER-GATA-1 cells (Im et al., 2005). and granulocytes (Hromas et al., 1993; Scott et al., In primitive and definitive erythroid cells from E10.5 1994; McKercher et al., 1996). By contrast, PU.1 blocks murine yolk sac, E14.5 fetal liver and Ter119 þ bone erythropoiesis (Schuetze et al., 1992, 1993), and this marrow, EKLF also occupies HS1–HS4 and the b-like repressive activity interfaces with GATA-1-dependent globin promoters in a differentiation stage-specific transcriptional mechanisms. manner, and EKLF levels differ in primitive vs definitive The Ets domain of PU.1 binds the C-terminal zinc cells (Zhou et al., 2006). The gene disruption results, finger of GATA-1 (Rekhtman et al., 1999; Liew et al., analyses with EKLF mutants, and EKLF occupancy at 2006) and antagonizes GATA-1 DNA binding (Zhang the endogenous b-globin locus provide strong evidence et al., 1999a, 2000). PU.1 also inhibits CBP/p300- that EKLF is an essential regulator of adult b-like mediated acetylation of GATA-1 (Hong et al., 2002). globin gene regulation. As a GATA-1 mutant lacking acetylation sites is EKLF-interacting coregulators include CBP/p300, impaired in occupying chromatin target sites in vivo (Zhang et al., 2001) and BRG1 (Armstrong et al., (Lamonica et al., 2006), inhibition of GATA-1 acety- 1998; Tewari et al., 1998; Brown et al., 2002). Targeted lation might underlie PU.1-mediated repression of disruption of Eklf abrogates DNaseI hypersensitivity at GATA-1 activity. PU.1-mediated inhibition of erythro- HS3and the b-globin promoter (Tewari et al., 1998), poiesis involves recruitment of a corepressor complex, indicating that EKLF induces complex assembly at consisting of retinoblastoma protein (pRb) (Rekhtman these sites, and analogous to conventional transcription et al., 2003), histone methyltransferase, Suv39H and factors, regulates chromatin structure (Armstrong heterochromatin protein 1a, to create a repressive et al., 1998). EKLF resembles GATA-1 and FOG-1 is chromatin structure at GATA-1 target genes, at least in promoting the formation of a higher-order chromatin certain contexts (Stopka et al., 2005). Given the loop at the b-globin locus (Drissen et al., 2004). EKLF functional importance of the PU.1–GATA-1 interaction, also functions as a transcriptional repressor by recruit- and the balance between GATA-1 and GATA-2 levels, ing Sin3A and histone deacetylase 1 (Chen and Bieker, an important determinant of GATA switches, it will be 2001, 2004). It will be crucial to determine the relative crucial to develop assays to quantitate the absolute levels importance of EKLF activating vs repressing functions and specific activities of these factors, as well as their in regulating the b-like globin genes and in promoting affinities for targets (protein-protein and protein–chro- both primitive and definitive erythropoiesis. matin interactions) in primary hematopoietic precursors during development. In aggregate, this will allow one to model how differentiation stage-specific changes in the p45/NF-E2 levels and activities of these factors reconfigure transcrip- The tandem AP-1-like motifs within HS2 of the b-globin tional networks that drive erythropoiesis. LCR confer exceptionally strong, erythroid cell-specific enhancer activity in transfection assays (Mignotte et al., 1990; Moi and Kan, 1990; Ney et al., 1990a, b; Moon EKLF and Ley, 1991). Studies to elucidate the molecular basis Differential cloning in the MEL cell system led to the of this activity led to the discovery of the heterodimeric identification of a novel erythroid cell-specific cDNA transcription factor nuclear factor-erythroid 2 (NF-E2). encoding a Kruppel-like transcription factor (KLF), NF-E2, which consists of a hematopoietic-specific subunit, which was designated as EKLF (Miller and Bieker, p45/NF-E2 (Ney et al., 1993; Andrews et al., 1993a), 1993). EKLF binds CACCC motifs, analogous to other and a member of the small Maf protein family (Andrews KLFs that have diverse roles during cellular differentia- et al., 1993b), is expressed in erythroid and megakar- tion and development (Bieker, 2001). The targeted yocytic cells. Despite the enhancer activity mediated by disruption of Eklf in mice demonstrates that EKLF is NF-E2-binding sites, targeted disruption of p45/NF-E2 crucial for erythropoiesis and for adult b-like globin in mice did not appear to significantly perturb erythro- gene transcription (Nuez et al., 1995; Perkins et al., poiesis or b-globin transcription. Rather, these studies 1995). Whereas EKLF was initially considered to solely demonstrated that p45/NF-E2 is crucial for megakar- regulate definitive erythropoiesis (Nuez et al., 1995), yopoiesis (Shivdasani and Orkin, 1995; Shivdasani recent work has demonstrated that EKLF also pro- et al., 1995b). motes primitive erythropoiesis (Hodge et al., 2006). Paradoxically, CB3murine erythroleukemia cells, Transcriptional profiling studies of EklfÀ/À fetal liver and which lack p45/NF-E2 expression due to retroviral ER-EKLF-expressing EKLF-null cells revealed that insertion within the p45/NF-E2 locus, do not express EKLF regulates diverse genes, including those encoding bmajor, and p45/NF-E2 expression reactivates bmajor heme biosynthesis enzymes and cytoskeletal proteins expression (Lu et al., 1994; Kotkow and Orkin, 1995; (Drissen et al., 2005; Hodge et al., 2006; Nilson et al., Bean and Ney, 1997; Mosser et al., 1998; Kiekhaefer 2006; Pilon et al., 2006). Whether the bulk of these genes et al., 2004). Endogenous p45/NF-E2 occupies the LCR, are direct EKLF targets is unknown. and to a lesser extent, the bmajor promoter in erythroid EKLF is an important determinant of hemoglobin cells (Daftari et al., 1999; Forsberg et al., 2000a; Johnson switching (Donze et al., 1995; Perkins et al., 1996; et al., 2001; Sawado et al., 2001; Im et al., 2005). The Wijgerde et al., 1996; Gillemans et al., 1998). EKLF finding that Pol II occupies the LCR, but not the bmajor

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6782 promoter, and p45/NF-E2 expression rescues Pol II acute lymphoblastic leukemia led to the discovery of a occupancy at the promoter in CB3cells, led to the model transcription factor, stem cell leukemia (SCL) or T-cell that p45/NF-E2 induces long-range Pol II transfer, acute lymphocytic leukemia-1, which functions at multi- resulting in considerably more Pol II at the promoter ple stages of hematopoiesis (Begley et al., 1989a, b, 1991; and transcriptional activation (Johnson et al., 2001). Aplan et al., 1990a, b). SCL is a member of the basic Moreover, Pol II occupancy at the promoter decreased helix–loop–helix family of transcription factors, which 30–40% in p45/NF-E2-null mice, whereas Pol II occu- dimerize with other E-proteins (for example, E47) and pancy at the LCR remained unchanged (Kooren et al., bind E-box motifs (CANNTG) (Begley et al., 1989b; 2007). Biochemical analysis supports the feasibility of Chen et al., 1990). Mice lacking SCL are incapable of long-range Pol II transfer (Vieira et al., 2004). generating red blood cells in the yolk sac, and SclÀ/À ES Analogous to GATA-1 and EKLF, p45/NF-E2 also cells fail to give rise to hematopoietic cells (Robb et al., binds CBP/p300. CBP/p300-mediated acetylation of the 1995; Shivdasani et al., 1995a; Porcher et al., 1996). SCL p45/NF-E2 heterodimeric partner MafG enhances has important functions to stimulate the generation of NF-E2 DNA binding activity (Hung et al., 2001). hemangioblasts, which differentiate into both blood and Unlike GATA-1 and EKLF, the b-globin locus in endothelial cells (Chung et al., 2002; Ema et al., 2003; E14.5 fetal liver from wild-type and p45/NF-E2-null Gering et al., 2003; Patterson et al., 2007). SCL is mice exhibits the chromatin loop, in which the bmajor required to generate HSCs, although conditional knock- promoter resides in proximity to the LCR (Kooren out analysis indicates that SCL is not required for HSC et al., 2007). This analysis revealed an B35% reduction function (Mikkola et al., 2003). in bmajor transcription, and as noted above, 30–40% In addition to heterodimerizing with an E-protein less Pol II occupancy at the promoter in E14.5 fetal partner, SCL can exist as a multiprotein complex livers of mutant mice. consisting of LMO2, E47, Ldb1 and GATA-1 (Wadman p45/NF-E2 contains two PPXY motifs within its N- et al., 1997). This complex is found in cells representing terminal activation domain, which bind certain WW different stages of erythropoiesis (Vyas et al., 1999; Xu domains in vitro and are required for activation of et al., 2003; Anguita et al., 2004; Brand et al., 2004; Lahlil endogenous bmajor in CB3cells (Mosser et al., 1998; et al., 2004). DNA site selection studies indicated that the Kiekhaefer et al., 2004). Whereas the Yes-associated SCL complex preferentially binds a composite motif protein (Yagi et al., 1999) and transcriptional coactivator consisting of a GATA motif and a neighboring E-box with PDZ-binding motif (Kanai et al., 2000) coregulators (Wadman et al., 1997), which is implicated in mediating are WW domain-containing proteins, the endogenous SCL actions at distinct stages of hematopoiesis. WW domain partner of p45/NF-E2 is unknown. The Ldb1 component of the SCL complex interacts p45/NF-E2-mediated transactivation in a chromatin with single-stranded DNA-binding proteins, which context is enhanced by mitogen-activated protein kinase appear to be important regulators of SCL function signaling (Versaw et al., 1998; Forsberg et al., 1999b). (Chen et al., 2002; Xu et al., 2007). Single-stranded DNA- Furthermore, lysine 368 and serine 169 of p45/NF-E2 binding protein 2 increases steady-state levels of endo- are sumoylated and phosphorylated, respectively, and genous Ldb1 and LMO2, GATA-E-box DNA binding both modifications are important for bmajor transcrip- activity, and facilitates expression of a direct target of the tional activation in CB3cells (Casteel et al., 1998; Shyu GATA-E-box-binding complex Protein 4.2 (P4.2) in et al., 2005). erythroid precursor cells (Xu et al., 2007). Proteomic The p45/NF-E2 occupancy at the endogenous b-globin studies identified Eto-2 and cyclin-dependent kinase locus in erythroid cells, the p45/NF-E2-dependence for Cdk9 as Ldb1-interacting proteins (Meier et al., 2006). bmajor transcription in CB3cells, and the significant Ldb1 forms multiple complexes during erythroid differ- decreases in Pol II occupancy at the bmajor promoter and entiation, and morpholino-dependent knockdown of bmajor transcription in p45/NF-E2 knockout mice these components in zebrafish revealed that they are indicate that NF-E2 is an important regulator of hemo- essential for definitive hematopoiesis (Meier et al., 2006). globin synthesis. The lack of a large defect in b-like SCL activity is regulated via standard mechanisms globin expression in p45/NF-E2-null mice might be involving post-translational modifications and interac- related to the existence of NF-E2-related factors, such tions with coregulators, although many questions as NF-E2-related factors 1 and 2 (Nrf1 and Nrf2) (Chan remain unanswered regarding the precise mechanisms. et al., 1993, 1996; Caterina et al., 1994) or Bach1 and SCL phosphorylation increases DNA binding activity Bach 2 (Igarashi et al., 1998). However, p45/NF-E2–Nrf2 (Prasad et al., 1995; Prasad and Brandt, 1997), while double knockout mice do not exhibit a more severe SCL acetylation mediated by CBP/p300 (Huang et al., erythroid phenotype than p45/NF-E2-null mice (Martin 1999) and p300/CBP-associated factor (Huang et al., et al., 1998). Alternatively, compensatory mechanisms in 2000) appears to be required for transactivation. SCL the mutant mice might suppress the magnitude of a b-like transactivation activity is inhibited by an mSin3A- and globin gene expression defect. histone deacetylase 1/2-containing corepressor complex (Huang and Brandt, 2000; Xu et al., 2006), and BRG1 overexpression increases occupancy of this complex at SCL (TAL-1) the P4.2 promoter (Xu et al., 2006). Coincident with the purification and cloning of Although canonical transcriptional mechanisms GATA-1, studies on mechanisms underlying T-cell involve sequence-specific DNA binding activity of

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6783 transcription factors, SCL mutants defective in DNA means by which FOG-1 activates transcription, whether binding rescue hematopoiesis from SclÀ/À murine FOG-1 has novel biochemical activities that specifically embryonic stem cells (Porcher et al., 1999). Thus, SCL regulate chromatin loop formation, and how the activa- DNA binding activity might be dispensable in certain tion vs repression functions of FOG-1 are selectively contexts. Given the apparent SCL complex heterogene- utilized. ity, regulation via multiple post-translational modifications, the novel non-DNA binding function, and CBP/p300 functions early in hematopoiesis to generate hemangio- CBP and its paralog p300 are acetyltransferases that blasts and to regulate HSC function, as well as later acetylate histones and a plethora of non-histone during erythropoiesis, considerably more work is proteins (Goldman et al., 1997; Mayr and Montminy, required to understand SCL structure/function. 2001). Targeted disruption of Cbp in mice leads to embryonic death due to developmental retardation and multiple organ defects, including reduced numbers of primitive and definitive hematopoietic precursor cells Coregulators that control erythropoiesis and defective vasculogenesis (Yao et al., 1998; Oike et al., 1999; Kung et al., 2000). Targeted gene disruption FOG-1 studies also provided evidence for a role of CBP, but not The nine-zinc-finger protein FOG-1 was identified by p300, in HSC self-renewal, whereas p300, but not CBP, yeast two-hybrid screening with the zinc-finger DNA- is required for differentiation (Rebel et al., 2002). binding domain of GATA-1 as bait (Tsang et al., 1997). Furthermore, CBP and p300 have distinct functions in FOG-1 deficiency causes severe anemia and death during T-cell lymphopoiesis (Kasper et al., 2006). mid-embryonic development (Tsang et al., 1998). As the CBP/p300 physically and functionally interact with expression of FOG-1 mimics that of GATA-1, FOG-1 many transcription factors, including GATA-1, p45/NF- exemplifies a cell type-specific coregulator, in contrast to E2 and EKLF (Blobel et al., 1998; Zhang and Bieker, the common broadly expressed coregulators. Unlike 1998; Forsberg et al., 1999a). Studies with the CBP/p300 many protein–protein interactions, in which the func- inhibitor adenoviral E1A provided evidence that CBP/ tional relevance is assessed via correlative analyses, a p300 facilitates GATA-1-mediated transcriptional activity rigorous approach using altered-specificity mutants of and promotes erythroid differentiation (Blobel et al., 1998). GATA-1 provided definitive evidence that the GATA-1– CBP acetylates GATA-1 (Hung et al., 1999) and increases FOG-1 interaction is crucial for GATA-1-dependent NF-E2 DNA binding activity (Hung et al., 2001). Studies erythropoiesis (Crispino et al., 1999). Moreover, this with E1A also indicated that CBP/p300 is crucial for LCR- same mutation in the N-terminal finger of GATA-1 was mediated transactivation (Forsberg et al., 1999a). CBP/ detected in patients with familial dyserythropoietic anemia p300 acetylates EKLF, which enhances EKLF binding to and thrombocytopenia (Nichols et al., 2000). the SWI/SNF complex (Zhang et al., 2001; Letting et al., Despite the established FOG-1 function to mediate 2003; Im et al., 2005; Kim et al., 2007). GATA-1 recruits important GATA-1 activities, molecular mechanisms CBP/p300 at the LCR and bmajor promoters, and underlying FOG-1-dependent transcriptional activation consistent with this result, GATA-1 increases histone H3 and repression are poorly understood. FOG-1 facilitates and H4 acetylation at these sites (Kiekhaefer et al., 2002; GATA-1 chromatin occupancy at certain chromatin Letting et al., 2003; Im et al., 2005). However, whether sites (Letting et al., 2004; Pal et al., 2004a) and is CBP/p300 are the principle acetyltransferases mediating required for GATA switches (Pal et al., 2004a). Thus, these epigenetic events is unknown. FOG-1 coactivator activity might solely involve facilitated It has been proposed that the concentration of CBP/ GATA-1 chromatin occupancy. FOG-1 also facilitates p300 is limiting in certain cellular contexts, and there- GATA-1-mediated looping that brings the LCR in fore activation of one or more pathways that utilize proximity to the adult b-globin genes (Vakoc et al., CBP/p300 could in principle sequester CBP/p300 away 2005). However, the FOG-1 requirement for looping from additional CBP/p300-dependent pathways (Kamei might be secondary to FOG-1 facilitation of GATA-1 et al., 1996). Thus, CBP/p300 might be crucial for chromatin occupancy. FOG-1 interacts with transforming establishing signaling crosstalk and/or transcriptional acidic coiled-coil protein 3, which regulates FOG-1 sub- networks that drive erythropoiesis. It will be important cellular localization (Garriga-Canut and Orkin, 2004). to determine whether dynamic changes in CBP/p300 FOG-1 also binds the corepressor C-terminal-binding levels/activities occur during erythropoiesis, and protein (Fox et al., 1999) and the NuRD complex (Hong whether such changes similarly or distinctly affect the et al., 2005; Johnson et al., 2006). Knock-in mice expressing activities of hematopoietic transcription factors. a FOG-1 mutant defective in C-terminal-binding protein binding revealed that the FOG-1–C-terminal-binding protein interaction is dispensable for erythropoiesis TRAP220 in vivo (Katz et al., 2002). Mediator is a broadly expressed coregulator complex Many questions remain unanswered regarding me- that participates in transcriptional activation or repres- chanisms of FOG-1 function, including whether FOG-1 sion by diverse transcription factors and functions by zinc fingers not involved in GATA-1 binding have specific regulating recruitment of Pol II and the basal transcrip- functions, whether COF activity is the predominant tional machinery (Kornberg, 2005; Belakavadi and

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6784 Fondell, 2006). Targeted disruption of TRAP220 in in a MafK complex in MEL cells (Brand et al., 2004), mice, which encodes an important Mediator compo- and also in a b-globin LCR-associated chromatin nent, perturbs various physiological processes, including remodeling complex (LARC) (Mahajan et al., 2005). placental, cardiac, neural and hepatic development, LARC binds the HS2-Maf recognition element, occu- which causes lethality during early gestation (Ito et al., pies HS2 in K562 erythroleukemia cells, and has 2000; Zhu et al., 2000; Crawford et al., 2002; Landles chromatin remodeling activity (Mahajan et al., 2005). et al., 2003). TRAP220 mutant embryos also exhibit BRG1 also binds EKLF, and the SWI/SNF complex is abnormalities in megakaryocyte and erythroid differ- required for EKLF-mediated transcriptional activation entiation (Crawford et al., 2002; Landles et al., 2003). in vitro (Armstrong et al., 1998; Kadam et al., 2000). Hematopoietic precursor cells isolated from E10.5 Early embryonic lethality has been a significant obstacle TRAP220-deficient embryos fail to form erythroid for analysing specific BRG1 functions in vivo (Bultman burst-forming units-erythroid and colony-forming et al., 2000). However, mice bearing a hypomorphic Brg1 units-erythroid (Stumpf et al., 2006). These hemato- allele, isolated via an ENU mutagenesis screen, survive poietic defects appear to be related in part to perturbed until E14.5 (Bultman et al., 2005). Embryos containing GATA-1 function. the hypomorphic allele exhibit a block in erythropoiesis, TRAP220 binds GATA factors (Crawford et al., leading to anemia. BRG1 mutant erythroid progenitor 2002), and the GATA-1 N-finger contacts TRAP220 cells fail to develop into polychromatic erythroblasts, (Stumpf et al., 2006). TRAP220 enhances GATA-1- resembling the loss of GATA-1 (Bultman et al., 2005). mediated transactivation in a transient transfection Reduced BRG1 activity in E12.5 fetal livers from context (Stumpf et al., 2006). As TRAP220 co-occupies hypomorph embryos compromises Pol II and serine numerous chromatin sites with GATA-1 (unpublished), 5-phosphorylated Pol II occupancy at the bmajor promoter, TRAP220 might be a generally important GATA-1 but not at the LCR (Kim et al., 2007). These defects coregulator. resemble those seen upon deletion of the b-globin LCR Interestingly, different Mediator complexes exist, in mice (Sawado et al., 2003), consistent with a model in which can function distinctly (Uhlmann et al., 2007). which the LCR critically requires BRG1 to maximally Accordingly, it will be important to determine whether all recruit Pol II to the promoter. Alternatively, maximal or a subset of Mediator components occupy GATA-1 BRG1 activity might be required for erythroid precursor target sites in chromatin, and whether Mediator cells to progress to a specific stage of erythropoiesis, in subcomplexes function uniquely at different chromatin which full Pol II occupancy occurs. Nevertheless, the target sites. Finally, the biochemistry of how GATA-1 BRG1 hypomorphic mice represent a novel tool for elects to utilize TRAP220, CBP/p300 and FOG-1, and dissecting the role of this chromatin remodeling whether these coregulators are utilized simultaneously, component in erythropoiesis. sequentially or in a context-dependent manner is entirely unclear.

Emerging mechanisms and principles BRG1 Transcriptional activation often requires ATP-depen- Chromatin target site selection: exquisite discrimination dent chromatin remodeling complexes, such as the SWI/ among abundant transcription factor-binding motifs in SNF complex (Saha et al., 2006). SWI/SNF uses ATP to the genome disrupt histone–DNA interactions, which can increase As indicated above, GATA-1 occupies a small subset of factor accessibility to nucleosomal DNA (Narlikar regions containing GATA motifs within endogenous et al., 2002). Chromatin remodeling and modifying chromatin domains, despite GATA motif abundance mechanisms can be intricately linked. For example, throughout the genome (Johnson et al., 2002, 2007; Grass histone acetylation favors retention of SWI/SNF on et al., 2003, 2006; Pal et al., 2004b; Im et al., 2005; reconstituted chromatin in vitro (Hassan et al., 2001). Martowicz et al., 2005). Whereas it is unclear whether this Not surprisingly, the BRG1 catalytic subunit of the exquisite discrimination among binding motifs can be SWI/SNF complex is essential in developmental and extrapolated to unrelated transcription factors, based on physiological processes (Narlikar et al., 2002; Bultman the impact of chromatin on cis-element accessibility, it et al., 2005; de la Serna et al., 2006). As BRG1 occupies would not be surprising if stringent selection mechanisms HS4, HS3, HS2 and the bmajor promoter at the endo- are common. Of particular interest will be to determine genous b-globin locus, and GATA-1 increases BRG1 whether factors recognizing considerably longer motifs occupancy at a subset of these sites (Im et al., 2005), that occur much less frequently in the genome also select a SWI/SNF functions directly to regulate b-like globin small subset of the motifs. Furthermore, as certain gene expression. SWI/SNF is also implicated in regulat- factors, such as the glucocorticoid receptor, have the ing human g-tob-globin gene switching (O’Neill et al., capacity to interact with nucleosomal DNA, whereas 1999, 2000; Lopez et al., 2002). other factors that bind naked DNA with exceedingly high À12 Although the precise mechanisms of how SWI/SNF afﬁnity, such as nuclear factor-1 (Kd ¼ 10 M), lack this functions at the b-globin locus are unresolved, BRG1 capacity (Pina et al., 1990; Archer et al., 1991; Hager has been reported to associate with several factors et al., 1993), parameters dictating whether a motif is implicated in b-like globin gene regulation. BRG1 exists accessible or not can be factor-speciﬁc.

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6785 With the application of chromatin immunoprecipita- ER-GATA-1 represses these genes in FOG-1-null cells tion-chip, in which chromatin immunoprecipitation (Johnson et al., 2007). Thus, GATA-1 can also repress samples are analysed on genomic microarrays (Kirmizis certain target genes in a FOG-1-independent manner. and Farnham, 2004; Huebert et al., 2006), a vast ER-GATA-1 and ER-GATA-1(V205G) activate Tac-2 amount of factor binding data is currently being accu- expression in FOG-1-null and G1E cells, respectively mulated. Thus, the development of chromatin target site (Johnson et al., 2006), indicating that GATA-1 does not selection rules, both factor-specific rules and those with require FOG-1 for Tac-2 activation. However, as a broader applicability, will be forthcoming. Since chro- function of the ER-GATA-1 residency time at a Tac-2 matin target site selection represents a very early step in intronic regulatory element (Pal et al., 2004b), ER-GATA- transcription factor function, defining these rules is 1 acquires the capacity to repress Tac-2, which requires of great importance. With respect to GATA factor FOG-1 (Johnson et al., 2006). Interestingly, Tac-2 encodes function, we proposed a GATA recognition code in a neurokinin-B precursor (Pal et al., 2004b), and neuroki- which the intrinsic features of GATA motifs, nearest- nin-B acts on receptors expressed by endothelial cells to neighbor cis-elements and the chromatin microenviron- confer antiangiogenic activity in diverse systems (Pal et al., ment are crucial (Bresnick et al., 2006). However, 2006). Thus, GATA-1 regulates transcription via FOG-1- elucidating the relative importance of these parameters dependent and -independent activation (Figures 2a and c), requires considerable additional experimental analysis. FOG-1-dependent and -independent repression (Figures 2b and e), and FOG-1-independent activation coupled to FOG-1-dependent repression (Figure 2d). Mechanistic permutations in transcriptional control: While the role of FOG-1 as a GATA-1 coregulator to how many mechanisms can a single transcription stimulate erythropoiesis is established (Crispino et al., factor utilize? 1999), whether GATA-1 has crucial FOG-1-indepen- GATA-1 utilizes FOG-1 to both activate and repress dent physiological functions remains to be determined. target genes (Crispino et al., 1999). However, expression Since FOG-1 facilitates GATA-1 chromatin occupancy of ER-GATA-1(V205G), defective in FOG-1 binding, in at a subset of chromatin target sites (Letting et al., 2004; G1E cells induces expression of heme-regulated eIF-a- Pal et al., 2004a), in principle, certain GATA-1- kinase (HRI), Eklf, Fog1 and Tac-2 (Crispino et al., activated genes might be intrinsically FOG-1-indepen- 1999; Johnson et al., 2006; Kim et al., 2007). Further- dent, based on the specific chromatin configuration of more, ER-GATA-1 expression in FOG-1-null hemato- their regulatory regions. Diverse nuclear receptor co- poietic precursors can activate certain GATA-1 target regulators seem to be differentially utilized in different genes (Johnson et al., 2007). Thus, GATA-1 can contexts (Perissi et al., 2004), suggesting a similar activate certain target genes in a FOG-1-independent mechanistic complexity. However, many studies with manner. ER-GATA-1(V205G) represses several genes nuclear receptor coregulators rely upon transient trans- including lymphoblastic leukemia (Lyl1) and regulator fection assays, and given the abundant copies of hetero- of G-protein signaling 18 (Rgs18) in G1E cells, and geneous plasmid templates per cell, it will be important

abFOG-1-Dependent Activation FOG-1-Dependent Repression cFOG-1-Independent Activation (βmajor) (Gata2) (Fog1)

CoR

FOG-1 FOG-1

N N N C C + C deFOG-1-Independent Activation/FOG-1-Dependent Repression FOG-1-Independent Repression (Tac-2) (Lyl1)

CoR

FOG-1

N N N C + C C +

Figure 2 Diverse modes of GATA-1-mediated transcriptional regulation. (a) GATA-1 utilizes FOG-1 to activate certain target genes. (b) GATA-1 utilizes FOG-1 to repress certain target genes. (c and e) Without FOG-1 activity, GATA-1 retains the capacity to regulate certain activated (c) and repressed (e) genes. (d) GATA-1-mediated Tac-2 activation is FOG-1-independent, but FOG-1 subsequently mediates repression. An example of a gene regulated via each mode is indicated in the parentheses. FOG-1, Friend of GATA-1.

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6786 to analyse these mechanisms at endogenous target genes represses the promoter. Thus, Ets motifs are an important with physiological levels of receptors and coregulators. determinant of context-dependent GATA factor activity. As GATA target genes can differ significantly in their DNA site selection experiments, using MEL cell sensitivities to altered GATA-1 levels, which does not nuclear extracts that contain GATA-1, revealed the appear to correlate with FOG-1 dependency (Johnson assembly of a multiprotein complex, consisting of SCL, et al., 2006), it will be important to define the target LMO2, E47, Ldb1 and GATA-1, on oligonucleotides gene sensitivity determinants. Such determinants are containing a GATA motif and an adjacent E-box relevant to both chromatin target site selection and (Wadman et al., 1997). While the GATA-E-box compo- post-chromatin occupancy regulatory steps. Never- site element is implicated in the activation of certain theless, since GATA factor levels dynamically change GATA-1 target genes (Vyas et al., 1999; Anderson et al., during erythropoiesis (Weiss et al., 1994), during the 2000; Lahlil et al., 2004; Khandekar et al., 2007; Wozniak development of distinct hematopoietic lineages (Bres- et al., 2007; Xu et al., 2007) (Figure 3b), the mechanism of nick et al., 2005) and in specialized non-hematopoietic the apparent synergism between GATA-1 and the cell and tissue types, including the heart and nervous additional factors, and whether other GATA factors system (Charron and Nemer, 1999; Molkentin, 2000), share this mechanism are unknown. it is essential to ascertain how changes in factor Dissecting mechanisms underlying the enhancer levels/activities translate into target gene transcriptional activity of the þ 9.5 kb intronic GATA switch site of changes. the Gata2 locus revealed a GATA-E-box composite element, which confers enhancer activity in primary human endothelial cells and erythroid precursor cell Mechanistic diversification via combinatorial regulation lines. Furthermore, this element activates a LacZ Chromatin target site selection is predicted to be regu- reporter gene in the vasculature of transgenic mouse lated in part by additional cis-elements in the neighbor- embryos (Khandekar et al., 2007; Wozniak et al., 2007) hood of GATA motifs. Such elements might regulate and a subset of cells within E11.5 fetal liver (Wozniak GATA factor occupancy or confer a competitive advan- et al., 2007) of E12.5 transgenic mouse embryos. tage post-GATA factor occupancy. An example of the Although the GATA-E-box motif is sufficient for former mechanism would be if factors binding to the enhancer activity in G1E cells, additional regulatory neighboring motifs recruit coregulators that influence modules are required for enhancer activity in human GATA motif accessibility, or if such factors physically endothelial cells (Wozniak et al., 2007). These studies interact with GATA factors, thereby influencing GATA provide evidence that the combinatorial usage of factor conformation and potentially altering DNA GATA-E-box motif with additional regulatory modules binding affinity. An example of the latter mechanism is an important determinant of context-dependent would be if factors binding the neighboring cis-elements GATA factor-mediated transcriptional regulation recruit coregulators that synergize with GATA factor- (Figure 3c). It would not be surprising if multiple recruited coregulators. scenarios exist in which GATA motifs merge with GATA-Ets composite elements have been identified in additional cis-elements under stringent architectural megakaryocyte-specific regulatory regions (Lemarchandel constraints to yield composite elements. Studies to et al., 1993), the þ 19 Scl enhancer region (Gottgens et al., define such composite motifs, and importantly to define 2002), and the Lyl1 proximal promoter region (Chan the underlying mechanisms, are in their infancy. et al., 2007; Figure 3a). Mutations in GATA and Ets motifs abolish Scl þ 19 enhancer activity (Gottgens et al., 2002), and Lyl1 promoter activity (Chan et al., 2007). Transcription factor families: common vs distinct GATA and Ets motifs are required for GATA-1-mediated determinants of chromatin target site selection activation of the megakaryocyte aIIB promoter (Wang and post-chromatin occupancy function et al., 2002). Importantly, without the Ets motif, which Transcription factors often exist in families based on binds the tissue-specific Ets factor Fli-1, GATA-1 sequence homology, and family members can have both

a GATA-Ets Composite Element bcGATA-E-box Composite Element GATA-E-box Combinatorial

N N E N E C C C

Figure 3 Combinatorial function of GATA motifs with neighboring cis-elements. (a) GATA-Ets composite elements are found at regulatory regions of certain megakaryocyte-speciﬁc genes, as well as the Scl þ 19 enhancer and the Lyl1 proximal promoter region. (b) A GATA-E-box composite element, which assembles a multiprotein complex containing GATA-1, SCL and other factors, is implicated in activation of certain GATA-1 target genes. (c) The GATA-E-box composite element is sufﬁcient for enhancer activity of the Gata2 þ 9.5 kb GATA switch site in erythroid precursor cells (G1E), whereas its enhancer activity in human endothelial cells requires additional regulatory modules (GATA-E-box combinatorial). SCL, stem cell leukemia.

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6787 overlapping and unique expression patterns. Family Intriguingly, both GATA-1 and FOG-1 promote a members often share DNA binding specificities, at least higher-order chromatin structure transition, in which based on naked DNA binding assays. In the case of the proximity of the LCR relative to the bmajor factors with overlapping expression in a specific target promoter increases (Vakoc et al., 2005; Kim et al., cell type, and those that do not function redundantly in 2007). Thus, GATA-1-mediated recruitment of FOG-1 all contexts (for example, GATA-1 and GATA-2), an might instigate or facilitate chromatin looping, or underlying logic must dictate unique transcriptional by virtue of its COF activity, FOG-1 might establish outputs for individual family members. Elucidating such maximal GATA-1 occupancy, which would suffice for logic will require information on the absolute concen- chromatin looping. While the exact mechanism is trations of the respective factors and an assessment of unknown, four zinc fingers of FOG-1 contact GATA-1, their specific activities, the latter posing serious experi- and disabling this protein–protein interaction requires mental challenges. mutational disruption of all four fingers (Cantor et al., The discovery of GATA switches in which GATA-1 2002; Liew et al., 2005). Although nothing is known replaces GATA-2 from certain chromatin sites during about the stoichiometry of GATA–FOG-1 complexes in erythropoiesis (Grass et al., 2003), provides a system to chromatin, FOG-1 and/or other GATA-1-bound cor- analyse how related factors function both similarly and egulators might have the capacity to simultaneously distinctly. Since GATA-1 and GATA-2 function redun- interact with more than one GATA-1 molecule, for dantly to confer survival and/or proliferation to primitive example, GATA-1 bound to dispersed sites. Alterna- erythroblasts (Fujiwara et al., 2004), but also function tively, since histone acetylation can unfold chromatin distinctly to regulate erythroid cell-specific genes in G1E (Forsberg and Bresnick, 2001), coregulator-dependent cells (Weiss et al., 1997; Grass et al., 2003), the cellular acetylation and the establishment of other epigenetic context dictates GATA factor subtype-specific actions. marks at strategic sites within a chromatin domain might Although GATA-1 and GATA-2 share an almost serve as a primary step in loop formation. Of course, such indistinguishable DNA-binding domain, their N- and a mechanism would not be mutually exclusive with C- termini are considerably different, and therefore it mechanisms involving transcription factor or coregula- would be not surprising if fundamental differences exist tor-dependent multivalent interactions. in their biochemical mechanisms of action. Identifying Parameters governing chromatin loop formation are such differences has the potential to yield broad insights largely unknown, and therefore the relationship between relevant to GATA factors and other transcription such parameters and those relevant to DNA looping factor families. over short distances of a 1000 bp or less without chromatin (Schleif, 1992), is unclear. Since entire chromatin domains cannot be reconstituted in vitro, Novel coregulator mechanisms: facilitation of chromatin and many components of such domains are not even occupancy and looping known, it will be important to dissect the underlying As discussed above, conventional coregulators function as mechanisms at endogenous loci using genetic and chromatin remodeling or modifying enzymes, and in the molecular approaches. case of Mediator, interact with Pol II and/or associated components. While FOG-1 binds the NuRD complex and is therefore associated with both chromatin remodeling Pol II function at extragenic sites: a specialized or and modifying activities (Hong et al., 2005), FOG-1 has common mode of transcriptional control? the unique activity of facilitating GATA-1 chromatin While traditional transcriptional mechanisms focus on occupancy at certain sites (Letting et al., 2004; Pal et al., how activators and repressors utilize coregulators to 2004a). This COF activity is likely relevant to its ability to regulate Pol II recruitment and function at promoters, a promote GATA switches. Whether the COF activity considerably more complex picture has emerged, where- requires the NuRD complex and/or other coregulators is by Pol II resides at regulatory regions of genes, often unknown, as assays are not established to study before it appears at the associated promoter. Evidence FOG-1 function in cell free system. Furthermore, the exists for intergenic transcription broadly throughout genetic complementation assay in FOG-1-null cells is the genome (Kapranov et al., 2007), and given the complex, as the cells do not respond synchronously to barrage of new insights regarding small RNA-based FOG-1 expression (Cantor et al., 2002). Thus, consider- regulatory mechanisms (Chang and Mendell, 2007; able additional work is required to dissect mechanisms Grewal and Elgin, 2007), it is likely that intergenic underlying COF activity. It will be instructive to consider transcripts provide precursors for the generation of whether the FOG-1 zinc fingers not involved in GATA-1 regulatory RNA molecules. However, as Pol II com- binding are important in this regard. As factors such plexes can contain coregulators that modify or remodel as nuclear factor-1 (NF-1) cannot bind motifs within chromatin (Cho et al., 1998; Krogan et al., 2003a, b; Li nucleosomal DNA, despite very high-affinity DNA et al., 2007), the processive movement of Pol II along binding activity (Pina et al., 1990; Archer et al., 1991), chromatin within an intergenic region might also serve one might predict that mechanisms in which coregulators to regulate chromatin structure independent of RNA facilitate chromatin occupancy by the interacting tran- generation (Travers, 1999). Finally, delivery mechan- scription factor are extremely useful in higher eukaryotic isms have been proposed in which Pol II tracks along genomes. chromatin from a regulatory element to a promoter

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6788 (Kong et al., 1997; Ling et al., 2004), or Pol II bound to elements (Szutorisz et al., 2005a; Levings et al., 2006), a distal regulatory element is brought into proximity of the requirement for cell type-speciﬁc transcription a promoter (Gribnau et al., 2000; Carter et al., 2002; factors for Pol II recruitment to distal elements Tolhuis et al., 2002; Vakoc et al., 2005; Kim et al., 2007), (Johnson et al., 2001, 2002, 2003; Vieira et al., 2004; which in principle would allow long-range Pol II Szutorisz et al., 2005a), and the compelling mechanistic transfer to the promoter (Johnson et al., 2001). possibilities, future work in this area promises to be Although evidence supporting the mechanisms noted rewarding vis-a` -vis elucidating the transcriptional con- above exist, studies to deﬁne Pol II function at trol of erythropoiesis, and more broadly, transcriptional intergenic sites are in their infancy. The challenge will mechanisms that regulate diverse biological processes. be to devise innovative strategies to dissect mechanisms in the context of endogenous loci, rather than synthetic Acknowledgements templates in which promiscuous Pol II access and function might occur. Given the presence of Pol II at We thank the Bresnick laboratory members for a critical reading certain distal regulatory elements prior to promoter of this review. We acknowledge support from NIH Grants activation (Szutorisz et al., 2005b; Bresnick et al., 2006), DK50107 (EHB), DK55700 (EHB), DK68634 (EHB), and an developmental alterations in Pol II occupancy at such American Heart Association Predoctoral Fellowship (S-I K).

References

Allfrey V, Faulkner RM, Mirsky AE. (1964). Acetylation and hematopoiesis and encodes a differentiation-related DNA-binding methylation of histones and their possible role in the regulation of motif. Proc Natl Acad Sci USA 86: 10128–10132. RNA synthesis. Proc Natl Acad Sci USA 51: 786–794. Begley CG, Visvader J, Green AR, Aplan PD, Metcalf D, Kirsch IR Anderson KP, Crable SC, Lingrel JB. (2000). The GATA–E-box motif et al. (1991). Molecular cloning and chromosomal localization of in the EKLF promoter is required for in vivo expression. Blood 95: the murine homolog of the human helix–loop–helix gene SCL. Proc 1652–1655. Natl Acad Sci USA 88: 869–873. Andrews NC, Erdjument-Bromage H, Davidson MB, Tempst P, Orkin Belakavadi M, Fondell JD. (2006). Role of the mediator complex in SH. (1993a). Erythroid transcription factor NF-E2 is a haemato- nuclear hormone receptor signaling. Rev Physiol Biochem Pharma- poietic-specific basic-leucine zipper protein. Nature 362: 722–728. col 156: 23–43. Andrews NC, Kotkow KJ, Ney PA, Erdjument-Bromage H, Tempst Berger SL. (2007). The complex language of chromatin regulation P, Orkin SH. (1993b). The ubiquitous subunit of erythroid during transcription. Nature 447: 407–412. transcription factor NF-E2 is a small basic-leucine zipper protein Bernstein BE, Meissner A, Lander ES. (2007). The mammalian related to the v-maf oncogene. Proc Natl Acad Sci U SA 90: 11488– epigenome. Cell 128: 669–681. 11492. Bieker JJ. (2001). Kruppel-like factors: three fingers in many pies. Anguita E, Hughes J, Heyworth C, Blobel GA, Wood WG, Higgs DR. J Biol Chem 276: 34355–34358. (2004). Globin gene activation during hematopoiesis is driven by Blobel GA, Nakajima T, Eckner R, Montminy M, Orkin SH. (1998). protein complexes nucleated by GATA-1 and GATA-2. EMBO J CREB-binding protein cooperates with transcription factor GATA- 23: 2841–2852. 1 and is required for erythroid differentiation. Proc Natl Acad Sci Aplan PD, Begley CG, Bertness V, Nussmeier M, Ezquerra A, Coligan J USA 95: 2061–2066. et al. (1990a). The SCL gene is formed from a transcriptionally Bourquin JP, Subramanian A, Langebrake C, Reinhardt D, Bernard complex locus. Mol Cell Biol 10: 6426–6435. O, Ballerini P et al. (2006). Identification of distinct molecular Aplan PD, Lombardi DP, Ginsberg AM, Cossman J, Bertness VL, phenotypes in acute megakaryoblastic leukemia by gene expression Kirsch IR. (1990b). Disruption of the human SCL locus by profiling. Proc Natl Acad Sci USA 103: 3339–3344. ‘illegitimate’ V–(D)–J recombinase activity. Science 250: 1426–1429. Boyes J, Byfield P, Nakatani Y, Ogryzko V. (1998). Regulation of Archer TK, Cordingley MG, Wolford RG, Hager GL. (1991). activity of the transcription factor GATA-1 by acetylation. Nature Transcription factor access is mediated by accurately positioned 396: 594–598. nucleosomes on the mouse mammary tumor virus promoter. Mol Brand M, Ranish JA, Kummer NT, Hamilton J, Igarashi K, Cell Biol 11: 688–698. Francastel C et al. (2004). Dynamic changes in transcription factor Arias J, Alberts AS, Brindle P, Claret FX, Smeal T, Karin M et al. complexes during erythroid differentiation revealed by quantitative (1994). Activation of cAMP and mitogen responsive genes relies on proteomics. Nat Struct Mol Biol 11: 73–80. a common factor. Nature 370: 226–229. Bresnick EH, Johnson KD, Kim S-I, Im H. (2006). Establishment and Armstrong JA, Bieker JJ, Emerson BM. (1998). A SWI/SNF-related regulation of chromatin domains: mechanistic insights from studies chromatin remodeling complex, E-RC1, is required for tissue- of hemoglobin synthesis. Prog Nucl Acids Res Mol Biol 81: 435–471. specific transcriptional regulation by EKLF in vitro. Cell 95: 93–104. Bresnick EH, Martowicz ML, Pal S, Johnson KD. (2005). Develop- Ayer DE. (1999). Histone deacetylases: transcriptional repression with mental control via GATA factor interplay at chromatin domains. SINers and NuRDs. Trends Cell Biol 9: 193–198. J Cell Physiol 205: 1–9. Bank A. (2006). Regulation of human fetal hemoglobin: new players, Brown RC, Pattison S, van Ree J, Coghill E, Perkins A, Jane SM et al. new complexities. Blood 107: 435–443. (2002). Distinct domains of erythroid Kruppel-like factor modulate Bean TL, Ney PA. (1997). Multiple regions of p45 NF-E2 are required chromatin remodeling and transactivation at the endogenous beta- for beta-globin gene expression in erythroid cells. Nucleic Acids Res globin gene promoter. Mol Cell Biol 22: 161–170. 25: 2509–2515. Bulger M, Schubeler D, Bender MA, Hamilton J, Farrell CM, Begley CG, Aplan PD, Davey MP, Nakahara K, Tchorz K, Kurtzberg Hardison RC et al. (2003). A complex chromatin landscape revealed J et al. (1989a). Chromosomal translocation in a human leukemic by patterns of nuclease sensitivity and histone modification within stem-cell line disrupts the T-cell antigen receptor delta-chain the mouse beta-globin locus. Mol Cell Biol 23: 5234–5244. diversity region and results in a previously unreported fusion Bultman SJ, Gebuhr TC, Magnuson T. (2005). A Brg1 mutation that transcript. Proc Natl Acad Sci USA 86: 2031–2035. uncouples ATPase activity from chromatin remodeling reveals an Begley CG, Aplan PD, Denning SM, Haynes BF, Waldmann TA, essential role for SWI/SNF complexes in beta-globin expression and Kirsch IR. (1989b). The gene SCL is expressed during early erythroid development. Genes Dev 19: 2849–2861.

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6789 Bultman SJ, Gebuhr TC, Yee D, La Mantia C, Nicholson J, Gilliam A Crispino JD. (2005). GATA-1 in normal and malignant hematopoiesis. et al. (2000). A Brg1 null mutation in the mouse reveals functional Semin Cell Dev Biol 16: 137–147. differences among mammalian SWI/SNF complexes. Mol Cell 6: Crispino JD, Lodish MB, MacKay JP, Orkin SH. (1999). Use of 1287–1295. altered specificity mutants to probe a specific protein–protein Cantor AB, Orkin SH. (2005). Coregulation of GATA factors by the interaction in differentiation: the GATA-1:FOG complex. Mol Cell Friend of GATA (FOG) family of multitype zinc finger proteins. 3: 219–228. Sem Cell Devel Biol 16: 117–128. Crispino JD, Lodish MB, Thurberg BL, Litovsky SH, Collins T, Cantor AB, Katz SG, Orkin SH. (2002). Distinct domains of the Molkentin JD et al. (2001). Proper coronary vascular development GATA-1 cofactor FOG-1 differentially influence erythroid versus and heart morphogenesis depend on interaction of GATA-4 with megakaryocytic maturation. Mol Cell Biol 22: 4268–4279. FOG cofactors. Genes Dev 15: 839–844. Carter D, Chakalova L, Osborne CS, Dai YF, Fraser P. (2002). Crossley M, Orkin SH. (1994). Phosphorylation of the erythroid Long-range chromatin regulatory interactions in vivo. Nat Genet 32: transcription factor GATA-1. J Biol Chem 269: 16589–16596. 623–626. Daftari P, Gavva NR, Shen CK. (1999). Distinction between AP1 and Casteel D, Suhasini M, Gudi T, Naima R, Pilz RB. (1998). Regulation NF-E2 factor-binding at specific chromatin regions in mammalian of the erythroid transcription factor NF-E2 by cyclic adenosine cells. Oncogene 18: 5482–5486. monophosphate-dependent protein kinase. Blood 91: 3193–3201. de la Serna IL, Ohkawa Y, Imbalzano AN. (2006). Chromatin remodel- Caterina JJ, Donze D, Sun CW, Ciavatta DJ, Townes TM. (1994). ling in mammalian differentiation: lessons from ATP-dependent Cloning and functional characterization of LCR-F1: a bZIP remodellers. Nat Rev Genet 7: 461–473. transcription factor that activates erythroid-specific, human globin Dean A. (2006). On a chromosome far, far away: LCRs and gene gene expression. Nucleic Acids Res 22: 2383–2391. expression. Trends Genet 22: 38–45. Chan JY, Han XL, Kan YW. (1993). Cloning of Nrf1, an NF-E2- Dhalluin C, Carlson JE, Zeng L, He C, Aggarwal AK, Zhou MM. related transcription factor, by genetic selection in yeast. Proc Natl (1999). Structure and ligand of a histone acetyltransferase bromo- Acad Sci USA 90: 11371–13375. domain. Nature 399: 491–496. Chan K, Lu R, Chang JC, Kan YW. (1996). NRF2, a member of the Donze D, Townes TM, Bieker JJ. (1995). Role of erythroid Kruppel- NFE2 family of transcription factors, is not essential for murine like factor in human gamma- to beta-globin gene switching. J Biol erythropoiesis, growth, and development. Proc Natl Acad Sci USA Chem 270: 1955–1959. 93: 13943–13948. Drissen R, Palstra RJ, Gillemans N, Splinter E, Grosveld F, Philipsen Chan WY, Follows GA, Lacaud G, Pimanda JE, Landry JR, Kinston S et al. (2004). The active spatial organization of the beta-globin S et al. (2007). The paralogous hematopoietic regulators Lyl1 and locus requires the transcription factor EKLF. Genes Dev 18: SCL are co-regulated by Ets and GATA factors yet Lyl1 cannot 2485–2490. rescue the early SCLÀ/À phenotype. Blood 109: 1908–1916. Drissen R, von Lindern M, Kolbus A, Driegen S, Steinlein P, Beug H Chang TC, Mendell JT. (2007). The roles of microRNAs in vertebrate et al. (2005). The erythroid phenotype of EKLF-null mice: defects in physiology and human disease. Ann Rev Genomics Hum Genet e-pub hemoglobin metabolism and membrane stability. Mol Cell Biol 25: ahead of print. 5205–5214. Charron F, Nemer M. (1999). GATA transcription factors and cardiac Ema M, Faloon P, Zhang WJ, Hirashima M, Reid T, Stanford WL development. Semin Cell Dev Biol 10: 85–91. et al. (2003). Combinatorial effects of Flk1 and Tal1 on vascular Chen L, Segal D, Hukriede NA, Podtelejnikov AV, Bayarsaihan D, and hematopoietic development in the mouse. Genes Dev 17: Kennison JA et al. (2002). Ssdp proteins interact with the LIM- 380–393. domain-binding protein Ldb1 to regulate development. Proc Natl Emery DW, Yannaki E, Tubb J, Nishino T, Li Q, Stamatoyanno- Acad Sci USA 99: 14320–14325. poulos G. (2002). Development of virus vectors for gene therapy of Chen Q, Yang CY, Tsan JT, Xia Y, Ragab AH, Peiper SC et al. beta chain hemoglobinopathies: flanking with a chromatin insulator (1990). Coding sequences of the tal-1 gene are disrupted by reduces gamma-globin gene silencing in vivo. Blood 100: 2012–2019. chromosome translocation in human T cell leukemia. J Exp Med Emery DW, Yannaki E, Tubb J, Stamatoyannopoulos G. (2000). A 172: 1403–1408. chromatin insulator protects retrovirus vectors from chromosomal Chen X, Bieker JJ. (2001). Unanticipated repression function linked to position effects. Proc Natl Acad Sci USA 97: 9150–9155. erythroid Kruppel-like factor. Mol Cell Biol 21: 3118–3125. Evans T, Felsenfeld G. (1989). The erythroid-specific transcription Chen X, Bieker JJ. (2004). Stage-specific repression by the EKLF factor Eryf1: a new finger protein. Cell 58: 877–885. transcriptional activator. Mol Cell Biol 24: 10416–10424. Evans T, Reitman M, Felsenfeld G. (1988). An erythrocyte-specific Cho H, Orphanides G, Sun X, Yang XJ, Ogryzko V, Lees E et al. DNA-binding factor recognizes a regulatory sequence common to (1998). A human RNA polymerase II complex containing factors all chicken globin genes. Proc Natl Acad Sci USA 85: 5976–5980. that modify chromatin structure. Mol Cell Biol 18: 5355–5363. Felsenfeld G, Groudine M. (2003). Controlling the double helix. Chrivia JC, Kwok RP, Lamb N, Hagiwara M, Montminy MR, Nature 421: 448–453. Goodman RH. (1993). Phosphorylated CREB binds specifically to Fischle W, Wang Y, Allis CD. (2003). Histone and chromatin the nuclear protein CBP. Nature 365: 855–859. crosstalk. Curr Opin Cell Biol 15: 172–183. Chung JH, Whiteley M, Felsenfeld G. (1993). A 50 element of the Forrester WC, Takegawa S, Papayannopoulou T, Stamatoyannopou- chicken beta-globin domain serves as an insulator in human los G, Groudine M. (1987). Evidence for a locus activation region: erythroid cells and protects against position effect in Drosophila. the formation of developmentally stable hypersensitive sites in Cell 74: 505–514. globin-expressing hybrids. Nucleic Acids Res 15: 10159–10177. Chung YS, Zhang WJ, Arentson E, Kingsley PD, Palis J, Choi K. Forsberg EC, Bresnick EH. (2001). Histone acetylation beyond (2002). Lineage analysis of the hemangioblast as defined by FLK1 promoters: long-range acetylation patterns in the chromatin world. and SCL expression. Development 129: 5511–5520. Bioessays 23: 820–830. Coghill E, Eccleston S, Fox V, Cerruti L, Brown C, Cunningham J Forsberg EC, Downs KM, Bresnick EH. (2000a). Direct interaction of et al. (2001). Erythroid Kruppel-like factor (EKLF) coordinates NF-E2 with hypersensitive site 2 of the beta-globin locus control erythroid cell proliferation and hemoglobinization in cell lines region in living cells. Blood 96: 334–339. derived from EKLF null mice. Blood 97: 1861–1868. Forsberg EC, Downs KM, Christensen HM, Im H, Nuzzi PA, Crawford SE, Qi C, Misra P, Stellmach V, Rao MS, Engel JD et al. Bresnick EH. (2000b). Developmentally dynamic histone acetyla- (2002). Defects of the heart, eye, and megakaryocytes in peroxisome tion pattern of a tissue-specific chromatin domain. Proc Natl Acad proliferator activator receptor-binding protein (PBP) null embryos Sci USA 97: 14494–14499. implicate GATA family of transcription factors. J Biol Chem 277: Forsberg EC, Johnson K, Zaboikina TN, Mosser EA, Bresnick EH. 3585–3592. (1999a). Requirement of an E1A-sensitive coactivator for

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6790 long-range transactivation by the beta-globin locus control region. Hong W, Nakazawa M, Chen YY, Kori R, Vakoc CR, Rakowski C J Biol Chem 274: 26850–26859. et al. (2005). FOG-1 recruits the NuRD repressor complex to Forsberg EC, Zaboikina TN, Versaw WK, Ahn NG, Bresnick EH. mediate transcriptional repression by GATA-1. EMBO J 24: 67–78. (1999b). Enhancement of beta-globin locus control region-mediated Hromas R, Orazi A, Neiman RS, Maki R, Van Beveran C, Moore J transactivation by mitogen-activated protein kinases through et al. (1993). Hematopoietic lineage- and stage-restricted expression stochastic and graded mechanisms. Mol Cell Biol 19: 5565–5575. of the ETS oncogene family member PU.1. Blood 82: 2998–3004. Fox AH, Liew C, Holmes M, Kowalski K, MacKay JP, Crossley M. Huang S, Brandt SJ. (2000). mSin3A regulates murine erythroleuke- (1999). Transcriptional cofactors of the FOG family interact with mia cell differentiation through association with the TAL1 (or SCL) GATA proteins by means of multiple zinc fingers. EMBO J 18: transcription factor. Mol Cell Biol 20: 2248–2259. 2812–2822. Huang S, Qiu Y, Shi Y, Xu Z, Brandt SJ. (2000). P/CAF-mediated Fujiwara Y, Chang AN, Williams AM, Orkin SH. (2004). Functional acetylation regulates the function of the basic helix–loop–helix overlap of GATA-1 and GATA-2 in primitive hematopoietic transcription factor TAL1/SCL. EMBO J 19: 6792–6803. development. Blood 103: 583–585. Huang S, Qiu Y, Stein RW, Brandt SJ. (1999). p300 functions as a Garriga-Canut M, Orkin SH. (2004). Transforming acidic coiled-coil transcriptional coactivator for the TAL1/SCL oncoprotein. Onco- protein 3(TACC3)controls friend of GATA-1 (FOG-1) subcellular gene 18: 4958–4967. localization and regulates the association between GATA-1 and Huebert DJ, Kamal M, O’Donovan A, Bernstein BE. (2006). Genome- FOG-1 during hematopoiesis. J Biol Chem 279: 23597–23605. wide analysis of histone modifications by ChIP-on-chip. Methods Gering M, Yamada Y, Rabbitts TH, Patient RK. (2003). Lmo2 and 40: 365–369. Scl/Tal1 convert non-axial mesoderm into haemangioblasts which Hung HL, Kim AY, Hong W, Rakowski C, Blobel GA. (2001). differentiate into endothelial cells in the absence of Gata1. Stimulation of NF-E2 DNA binding by CREB-binding protein Development 130: 6187–6199. (CBP)-mediated acetylation. J Biol Chem 276: 10715–10721. Gillemans N, Tewari R, Lindeboom F, Rottier R, de Wit T, Wijgerde M Hung HL, Lau J, Kim AY, Weiss MJ, Blobel GA. (1999). CREB- et al. (1998). Altered DNA-binding specificity mutants of EKLF binding protein acetylates hematopoietic transcription factor and Sp1 show that EKLF is an activator of the beta-globin locus GATA-1 at functionally important sites. Mol Cell Biol 19: control region in vivo. Genes Dev 12: 2863–2873. 3496–3505. Gilliland DG, Jordan CT, Felix CA. (2004). The molecular basis of Igarashi K, Hoshino H, Muto A, Suwabe N, Nishikawa S, Nakauchi leukemia. Hematology Am Soc Hematol Educ Program, 80–97. H et al. (1998). Multivalent DNA binding complex generated by Goldman PS, Tran VK, Goodman RH. (1997). The multifunctional small Maf and Bach1 as a possible biochemical basis for beta-globin role of the co-activator CBP in transcriptional regulation. Recent locus control region complex. J Biol Chem 273: 11783–11790. Prog Horm Res 52: 103–119. Im H, Grass JA, Johnson KD, Boyer ME, Wu J, Bresnick EH. (2004). Gottgens B, Nastos A, Kinston S, Piltz S, Delabesse EC, Stanley M Measurement of protein–DNA interactions in vivo by chromatin et al. (2002). Establishing the transcriptional programme for immunoprecipitation. Methods Mol Biol 284: 129–146. blood: the SCL stem cell enhancer is regulated by a multi- Im H, Grass JA, Johnson KD, Kim S-I, Boyer ME, Imbalzano AN protein complex containing Ets and GATA factors. EMBO J 21: et al. (2005). Chromatin domain activation via GATA-1 utilization 3039–3050. of a small subset of dispersed GATA motifs within a broad Grass JA, Boyer ME, Pal S, Wu J, Weiss MJ, Bresnick EH. (2003). chromosomal region. Proc Natl Acad Sci USA 102: 17065–17070. GATA-1-dependent transcriptional repression of GATA-2 via Ito M, Yuan CX, Okano HJ, Darnell RB, Roeder RG. (2000). disruption of positive autoregulation and domain-wide chromatin Involvement of the TRAP220 component of the TRAP/SMCC remodeling. Proc Natl Acad Sci USA 100: 8811–8816. coactivator complex in embryonic development and thyroid Grass JA, Jing H, Kim S-I, Martowicz ML, Pal S, Blobel GA et al. hormone action. Mol Cell 5: 683–693. (2006). Distinct functions of dispersed GATA factor complexes at Jacobson RH, Ladurner AG, King DS, Tjian R. (2000). Structure and an endogenous gene locus. Mol Cell Biol 26: 7056–7067. function of a human TAFII250 double bromodomain module. Gregory T, Yu C, Ma A, Orkin SH, Blobel GA, Weiss MJ. (1999). Science 288: 1422–1425. GATA-1 and erythropoietin cooperate to promoter erythroid cell Johnson KD, Bresnick EH. (2002). Dissecting long-range transcrip- survival by regulating bcl-xl expression. Blood 94: 87–96. tional mechanisms by chromatin immunoprecipitation. Methods 26: Grewal SI, Elgin SC. (2007). Transcription and RNA interference in 27–36. the formation of heterochromatin. Nature 447: 399–406. Johnson KD, Boyer ME, Kim S-I, Kang SY, Wickrema A, Cantor AB Gribnau J, Diderich K, Pruzina S, Calzolari R, Fraser P. (2000). et al. (2007). Friend of GATA-1-independent transcriptional Intergenic transcription and developmental remodeling of chroma- repression: a novel mode of GATA-1 function. Blood 109: tin subdomains in the human beta-globin locus. Mol Cell 5: 5230–5233. 377–386. Johnson KD, Christensen HM, Zhao B, Bresnick EH. (2001). Distinct Grosveld F, van Assendelft GB, Greaves DR, Kollias G. (1987). mechanisms control RNA polymerase II recruitment to a tissue- Position-independent, high-level expression of the human beta- specific locus control region and a downstream promoter. Mol Cell globin gene in transgenic mice. Cell 51: 975–985. 8: 465–471. Hager GL, Archer TK, Fragoso G, Bresnick EH, Tsukagoshi Y, John Johnson KD, Grass JA, Im H, Park C, Choi K, Bresnick EH. (2003). S et al. (1993). Influence of chromatin structure on the binding of Highly restricted localization of RNA polymerase II to the transcription factors to DNA. Cold Spring Harb Symp Quant Biol hypersensitive site cores of a tissue-specific locus control region. 58: 63–71. Mol Cell Biol 23: 6468–6493. Hassan AH, Neely KE, Workman JL. (2001). Histone acetyltransfer- Johnson KD, Grass JD, Boyer ME, Kiekhaefer CM, Blobel GA, ase complexes stabilize swi/snf binding to promoter nucleosomes. Weiss MJ et al. (2002). Cooperative activities of hematopoietic Cell 104: 817–827. regulators recruit RNA polymerase II to a tissue-specific chromatin Hebbes TR, Thorne AW, Crane-Robinson C. (1988). A direct link domain. Proc Natl Acad Sci USA 99: 11760–11765. between core histone acetylation and transcriptionally active Johnson KD, Kim S-I, Bresnick EH. (2006). Differential sensitivities chromatin. EMBO J 7: 1395–1402. of transcription factor target genes underlie cell type-specific gene Hodge D, Coghill E, Keys J, Maguire T, Hartmann B, McDowall A expression patterns. Proc Natl Acad Sci USA 103: 15939–15944. et al. (2006). A global role for EKLF in definitive and primitive Kadam S, McAlpine GS, Phelan ML, Kingston RE, Jone KA, erythropoiesis. Blood 107: 3359–3370. Emerson BM. (2000). Functional selectivity of recombinant Hong W, Kim AY, Ky S, Rakowski C, Seo SB, Chakravarti D et al. mammalian SWI/SNF subunits. Genes Dev 14: 2441–2451. (2002). Inhibition of CBP-mediated protein acetylation by the Ets Kadonaga JT. (2004). Regulation of RNA polymerase II transcription family oncoprotein PU.1. Mol Cell Biol 22: 3729–3743. by sequence-specific DNA binding factors. Cell 116: 247–257.

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6791 Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B et al. and hematologic tumor suppression by CBP. Genes Dev 14: (1996). A CBP integrator complex mediates transcriptional regula- 272–277. tion and AP-1 inhibition by nuclear receptors. Cell 85: 403–414. Lahlil R, Lecuyer E, Herblot S, Hoang T. (2004). SCL assembles a Kanai F, Marignani PA, Sarbassova D, Yagi R, Hall RA, Donowiwtz multifactorial complex that determines glycophorin A expression. M et al. (2000). TAZ: a novel transcriptional coactivator regulated Mol Cell Biol 24: 1439–1452. by interactions with 14-3-3 and PDZ domain proteins. EMBO J 19: Lamonica JM, Vakoc CR, Blobel GA. (2006). Acetylation of GATA-1 6778–6791. is required for chromatin occupancy. Blood 108: 3736–3738. Kapranov P, Willingham AT, Gingeras TR. (2007). Genome-wide Landles C, Chalk S, Steel JH, Rosewell I, Spencer-Dene B, Lalani el N transcription and the implications for genomic organization. Nat et al. (2003). The thyroid hormone receptor-associated protein Rev Genet 8: 413–423. TRAP220 is required at distinct embryonic stages in placental, Kasper LH, Fukuyama T, Biesen MA, Boussouar F, Tong C, de Pauw cardiac, and hepatic development. Mol Endocrinol 17: 2418–2435. A et al. (2006). Conditional knockout mice reveal distinct functions Lee KK, Workman JL. (2007). Histone acetyltransferase complexes: for the global transcriptional coactivators CBP and p300 in T-cell one size doesn’t fit all. Nat Rev Mol Cell Biol 8: 284–295. development. Mol Cell Biol 26: 789–809. Lemarchandel V, Ghysdael J, Mignotte V, Rahuel C, Romeo PH. Katz SG, Cantor AB, Orkin SH. (2002). Interaction between FOG-1 (1993). GATA and Ets cis-acting sequences mediate megakaryocyte- and the corepressor C-terminal bindng protein is dispensible for specific expression. Mol Cell Biol 13: 668–676. normal erythropoiesis in vivo. Mol Cell Biol 22: 3121–3128. Letting DL, Chen YY, Rakowski C, Reedy S, Blobel GA. (2004). Kaushansky K. (2006). Lineage-specific hematopoietic growth factors. Context-dependent regulation of GATA-1 by friend of GATA-1. N Engl J Med 354: 2034–2045. Proc Natl Acad Sci USA 101: 476–481. Khandekar M, Brandt W, Zhou Y, Dagenais S, Glover TW, Suzuki N Letting DL, Rakowski C, Weiss MJ, Blobel GA. (2003). Formation of et al. (2007). A Gata2 intronic enhancer confers its pan-endothelia- a tissue-specific histone acetylation pattern by the hematopoietic specific regulation. Development 134: 1703–1712. transcription factor GATA-1. Mol Cell Biol 23: 1334–1340. Khavari PA, Peterson CL, Tamkun JW, Mendel DB, Crabtree GR. Levings PP, Zhou Z, Vieira KF, Crusselle-Davis VJ, Bungert J. (2006). (1993). BRG1 contains a conserved domain of the SWI2/SNF2 Recruitment of transcription complexes to the beta-globin locus family necessary for normal mitotic growth and transcription. control region and transcription of hypersensitive site 3prior to Nature 366: 170–174. erythroid differentiation of murine embryonic stem cells. FEBS J Kiekhaefer CM, Grass JA, Johnson KD, Boyer ME, Bresnick EH. 273: 746–755. (2002). Hematopoietic activators establish an overlapping pattern of Li B, Carey M, Workman JL. (2007). The role of chromatin during histone acetylation and methylation within a tissue-specific chro- transcription. Cell 128: 707–719. matin domain. Proc Natl Acad Sci USA 99: 14309–14314. Liew CW, Rand KD, Simpson RJ, Yung WW, Mansfield RE, Crossley Kiekhaefer CM, Grass JD, Bresnick EH. (2004). WW domain binding M et al. (2006). Molecular analysis of the interaction between motif within the activation domain of the hematopoietic activator the hematopoietic master transcription factors GATA-1 and PU.1. p45/NF-E2 is essential for establishment of a tissue-specific histone J Biol Chem 281: 28296–28306. modification pattern. J Biol Chem 279: 7456–7461. Liew CK, Simpson RJY, Kwan AHY, Crofts LA, Loughlin FE, Kim S-I, Bultman SJ, Jing H, Blobel GA, Bresnick EB. (2007). Matthews JM et al. (2005). Zinc fingers as protein recognition Dissecting molecular steps in chromatin domain activation during motifs: structural basis for the GATA-1/Friend of GATA interac- hematopoietic differentiation. Mol Cell Biol 27: 4551–4565. tion. Proc Natl Acad Sci USA 102: 583–588. Kirmizis A, Farnham PJ. (2004). Genomic approaches that aid in the Lin AC, Roche AE, Wilk J, Svensson EC. (2004). The N termini of identification of transcription factor target genes. Exp Biol Med 229: Friend of GATA (FOG) proteins define a novel transcriptional 705–721. repression motif and a superfamilyl of transcriptional repressors. Klemsz MJ, McKercher SR, Celada A, Van Beveren C, Maki RA. J Biol Chem 279: 55017–55023. (1990). The macrophage and B cell-specific transcription factor Ling J, Ainol L, Zhang L, Yu X, Pi W, Tuan D. (2004). HS2 enhancer PU.1 is related to the ets oncogene. Cell 61: 113–124. function is blocked by a transcriptional terminator inserted between Ko LJ, Engel JD. (1993). DNA-binding specificities of the GATA the enhancer and the promoter. J Biol Chem 279: 51704–51713. transcription factor family. Mol Cell Biol 13: 4011–4022. Litt MD, Simpson M, Gaszner M, Allis CD, Felsenfeld G. (2001). Kong S, Bohl D, Li C, Tuan D. (1997). Transcription of the HS2 Correlation between histone lysine methylation and developmental enhancer toward a cis-linked gene is independent of the orientation, changes at the chicken {beta}-globin locus. Science 9: 2453–2455. position, and distance of the enhancer relative to the gene. Mol Cell Lopez RA, Schoetz S, DeAngelis K, O’Neill D, Bank A. (2002). Biol 17: 3955–3965. Multiple hematopoietic defects and delayed globin switching in Kooren J, Palstra RJ, Klous P, Splinter E, von Lindern M, Grosveld F Ikaros null mice. Proc Natl Acad Sci USA 99: 602–607. et al. (2007). Beta-globin active chromatin hub formation in Lu SJ, Rowan S, Bani MR, Ben-David Y. (1994). Retroviral differentiating erythroid cells and in p45 NF-E2 knockout mice. integration within the Fli-2 locus results in inactivation of the J Biol Chem 282: 16544–16552. erythroid transcription factor NF-E2 in Friend erythroleukemias: Kornberg RD. (2005). Mediator and the mechanism of transcriptional evidence that NF-E2 is essential for globin expression. Proc Natl activation. Trends Biochem Sci 30: 235–239. Acad Sci USA 91: 8398–8402. Kotkow KJ, Orkin SH. (1995). Dependence of globin gene expression Lugus JJ, Chung YS, Mills JC, Kim SI, Grass JA, Kyba M et al. in mouse erythroleukemia cells on the NF-E2 heterodimer. Mol Cell (2007). GATA2 functions at multiple steps in hemangioblast Biol 15: 4640–4647. development and differentiation. Development 134: 393–405. Kouzarides T. (2007). Chromatin modifications and their function. Mahajan MC, Narlikar GJ, Boyapaty G, Kingston RE, Weissman Cell 128: 693–705. SM. (2005). Heterogeneous nuclear ribonucleoprotein C1/C2, Krogan NJ, Dover J, Wood J, Schneider J, Heidt J, Boateng MA et al. MeCP1, and SWI/SNF form a chromatin remodeling complex at (2003a). The Paf1 complex is required for histone H3 methylation the beta-globin locus control region. Proc Natl Acad Sci USA 102: by COMPASS and Dot1p: linking transcriptional elongation to 15012–15017. histone methylation. Mol Cell 11: 721–729. Marmorstein R. (2001). Protein modules that manipulate histone tails Krogan NJ, Kim M, Tong A, Golshani A, Cagney G, Canadien V for chromatin regulation. Nat Rev Mol Cell Biol 2: 422–432. et al. (2003b). Methylation of histone H3 by Set2 in Saccharomyces Martin DI, Orkin SH. (1990). Transcriptional activation and DNA cerevisiae is linked to transcriptional elongation by RNA polymer- binding by the erythroid factor GF-1/NF-E1/Eryf 1. Genes Dev 4: ase II. Mol Cell Biol 12: 4207–4218. 1886–1898. Kung AL, Rebel VI, Bronson RT, Ch’ng LE, Sieff CA, Livingston Martin F, van Deursen JM, Shivdasani RA, Jackson CW, Troutman DM et al. (2000). Gene dose-dependent control of hematopoiesis AG, Ney PA. (1998). Erythroid maturation and globin gene

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6792 expresion in mice with combined deficiency of NF-E2 and Nrf-2. Ney PA, Sorrentino BP, McDonagh KT, Nienhuis AW. (1990b). Blood 91: 3459–3466. Tandem AP-1-binding sites within the human beta-globin dominant Martowicz ML, Grass JA, Boyer ME, Guend H, Bresnick EH. (2005). control region function as an inducible enhancer in erythroid cells. Dynamic GATA factor interplay at a multi-component regulatory Genes Dev 4: 993–1006. region of the GATA-2 locus. J Biol Chem 280: 1724–1732. Nichols KE, Crispino JD, Poncz M, White JG, Orkin SH, Maris JM Mayr B, Montminy M. (2001). Transcriptional regulation by the et al. (2000). Familial dyserythropoietic anaemia and thrombocytope- phosphorylation-dependent factor CREB. Nat Rev Mol Cell Biol 2: nia due to an inherited mutation in GATA-1. Nat Genet 24: 266–270. 599–609. Nilson DG, Sabatino DE, Bodine DM, Gallagher PG. (2006). Major McKercher SR, Torbett BE, Anderson KL, Henkel GW, Vestal DJ, erythrocyte membrane protein genes in EKLF-deficient mice. Exp Baribault H et al. (1996). Targeted disruption of the PU.1 gene Hematol 34: 705–712. results in multiple hematopoietic abnormalities. EMBO J 15: Nuez B, Michalovich D, Bygrave A, Ploemacher R, Grosveld F. 5647–5658. (1995). Defective haematopoiesis in fetal liver resulting from Meier N, Krpic S, Rodriguez P, Strouboulis J, Monti M, Krijgsveld J inactivation of the EKLF gene. Nature 375: 316–318. et al. (2006). Novel binding partners of Ldb1 are required for Oike Y, Takakura N, Hata A, Kaname T, Akizuki M, Yamaguchi Y haematopoietic development. Development 133: 4913–4923. et al. (1999). Mice homozygous for a truncated form of CREB- Merika M, Orkin SH. (1993). DNA-binding specificity of GATA binding protein exhibit defects in hematopoiesis and vasculo- family transcription factors. Mol Cell Biol 13: 3999–4010. angiogenesis. Blood 93: 2771–2779. Merika M, Orkin SH. (1995). Functional synergy and physical O’Neill D, Yang J, Erdjument-Bromage H, Bornschlegel K, Tempst P, interactions of the erythroid transcription factor GATA-1 Bank A. (1999). Tissue-specific and developmental stage-specific with the Kruppel family proteins Sp1 and EKLF. Mol Cell Biol DNA binding by a mammalian SWI/SNF complex associated with 15: 2437–2447. human fetal-to-adult globin gene switching. Proc Natl Acad Sci Mignotte V, Navarro S, Eleouet JF, Zon LI, Romeo PH. (1990). The USA 96: 349–354. extinction of erythroid genes after tetradecanoylphorbol acetate O’Neill DW, Schoetz SS, Lopez RA, Castle M, Rabinowitz L, Shor E treatment of erythroleukemic cells correlates with down-regulation et al. (2000). An ikaros-containing chromatin-remodeling complex of the tissue-specific factors NF-E1 and NF-E2. J Biol Chem 265: in adult-type erythroid cells. Mol Cell Biol 20: 7572–7582. 22090–22092. Orlando V, Strutt H, Paro R. (1997). Analysis of chromatin structure Mikkola HK, Klintman J, Yang H, Hock H, Schlaeger TM, Fujiwara by in vivo formaldehyde cross-linking. Methods 11: 205–214. Y et al. (2003). Hematopoietic stem cells retain long-term Owen DJ, Ornaghi P, Yang JC, Lowe N, Evans PR, Ballario P et al. repopulating activity and multipotency in the absence of stem-cell (2000). The structural basis for the recognition of acetylated histone leukemia SCL/tal-1 gene. Nature 421: 547–551. H4 by the bromodomain of histone acetyltransferase gcn5p. EMBO J Mikkola HK, Orkin SH. (2006). The journey of developing 19: 6141–6149. hematopoietic stem cells. Development 133: 3733–3744. Pal S, Cantor AB, Johnson KD, Moran T, Boyer ME, Orkin SH et al. Miller IJ, Bieker JJ. (1993). A novel, erythroid cell-specific murine (2004a). Coregulator-dependent facilitation of chromatin occu- transcription factor that binds to the CACCC element and is pancy by GATA-1. Proc Natl Acad Sci USA 101: 980–985. related to the Kruppel family of nuclear proteins. Mol Cell Biol 13: Pal S, Nemeth MJ, Bodine DM, Miller JL, Svaren J, Thein SL et al. 2776–2786. (2004b). Neurokinin-B transcription in erythroid cells: direct Minegishi N, Ohta J, Yamagiwa H, Suzuki N, Kawauchi S, Zhou Y activation by the hematopoietic transcription factor GATA-1. et al. (1999). The mouse GATA-2 gene is expressed in the para- J Biol Chem 279: 31348–31356. aortic splanchnopleura and aorta-gonads and mesonephros region. Pal S, Wu J, Murray JK, Gellman SH, Wozniak MA, Keely PJ et al. Blood 93: 4196–4207. (2006). An antiangiogenic neurokinin-B/thromboxane A2 regula- Moi P, Kan YW. (1990). Synergistic enhancement of globin gene tory axis. J Cell Biol 174: 1047–1058. expression by activator protein-1-like proteins. Proc Natl Acad Sci Patterson LJ, Gering M, Eckfeldt CE, Green AR, Verfaille CM, Ekker USA 87: 9000–9004. SC et al. (2007). The transcription factors Scl and Lmo2 act together Molkentin JD. (2000). The zinc finger-containing transcription factors during development of the hemangioblast in zebrafish. Blood 109: GATA-4, -5, and -6. Ubiquitously expressed regulators of tissue- 2389–2398. specific gene expression. J Biol Chem 275: 38949–38952. Perissi V, Aggarwal A, Glass CK, Rose DW, Rosenfeld MG. (2004). A Moon AM, Ley TJ. (1991). Functional properties of the beta- corepressor/coactivator exchange complex required for transcrip- globin locus control region in K562 erythroleukemia cells. Blood tional activation by nuclear receptors and other regulated transcrip- 77: 2272–2284. tion factors. Cell 116: 511–526. Mosser EA, Kasanov JD, Forsberg EC, Kay BK, Ney PA, Bresnick Perkins AC, Gaensler KM, Orkin SH. (1996). Silencing of human fetal EH. (1998). Physical and functional interactions between the globin expression is impaired in the absence of the adult beta- transactivation domain of the hematopoietic transcription factor globin gene activator protein EKLF. Proc Natl Acad Sci USA 93: NF-E2 and WW domains. Biochemistry 37: 13686–13695. 12267–12271. Munugalavadla V, Dore LC, Tan BL, Hong L, Vishnu M, Weiss MJ Perkins AC, Sharpe AH, Orkin SH. (1995). Lethal beta-thalassaemia et al. (2005). Repression of c-kit and its downstream substrates by in mice lacking the erythroid CACCC- transcription factor EKLF. GATA-1 inhibits cell proliferation during erythroid maturation. Nature 375: 318–322. Mol Cell Biol 25: 6747–6759. Pevny L, Lin CS, D’Agati V, Simon MC, Orkin SH, Costantini F. Narlikar GJ, Fan H-Y, Kingston RE. (2002). Cooperation between (1995). Development of hematopoietic cells lacking transcription complexes that regulate chromatin structure and transcription. Cell factor GATA-1. Development 121: 163–172. 108: 475–487. Pevny L, Simon MC, Robertson E, Klein WH, Tsai SF, D’Agati V Nerlov C, Querfurth E, Kulessa H, Graf T. (2000). GATA-1 interacts et al. (1991). Erythroid differentiation in chimaeric mice blocked by with the myeloid PU.1 transcription factor and represses PU.1- a targeted mutation in the gene for transcription factor GATA-1. dependent transcription. Blood 95: 2543–2551. Nature 349: 257–260. Ney PA, Andrews NC, Jane SM, Safer B, Purucker ME, Weremowicz Pilon AM, Nilson DG, Zhou D, Sangerman J, Townes TM, Bodine S et al. (1993). Purification of the human NF-E2 complex: cDNA DM et al. (2006). Alterations in expression and chromatin configua- cloning of the hematopoietic cell-specific subunit and evidence for tion of the alpha hemoglobin-stabilizing protein gene in erythroid an associated partner. Mol Cell Biol 13: 5604–5612. Kruppel-like factor-deficient mice. Mol Cell Biol 26: 4368–4377. Ney PA, Sorrentino BP, Lowrey CH, Nienhuis AW. (1990a). Pina B, Bruggemeier U, Beato M. (1990). Nucleosome positioning Inducibility of the HS II enhancer depends on binding of an modulates accessibility of regulatory proteins to the mouse erythroid specific nuclear protein. Nucleic Acids Res 18: 6011–6017. mammary tumor virus promoter. Cell 60: 719–731.

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6793 Porcher C, Liao EC, Fujiwara Y, Zon LI, Orkin SH. (1999). Schuetze S, Stenberg PE, Kabat D. (1993). The Ets-related factor PU.1 Specification of hematopoietic and vascular development by the immortalizes erythroblasts. Mol Cell Biol 13: 5670–5678. bHLH transcription factor SCL without direct DNA binding. Scott EW, Simon MC, Anastasi J, Singh H. (1994). Requirement of Development 126: 4603–4615. transcription factor PU.1 in the development of multiple hemato- Porcher C, Swat W, Rockwell K, Fujiwara Y, Alt FW, Orkin SH. poietic lineages. Science 265: 1573–1577. (1996). The T cell leukemia oncoprotein SCL/tal-1 is essential for Shivdasani RA, Mayer EL, Orkin SH. (1995a). Absence of blood development of all hematopoietic lineages. Cell 86: 47–57. formation in mice lacking the T-cell leukaemia oncoprotein tal-1/ Prasad KS, Brandt SJ. (1997). Target-dependent effect of phosphor- SCL. Nature 373: 432–434. ylation on the DNA binding activity of the TAL1/SCL oncoprotein. Shivdasani RA, Orkin SH. (1995). Erythropoiesis and globin gene J Biol Chem 272: 11457–11462. expression in mice lacking the transcription factor NF-E2. Proc Natl Prasad KS, Jordan JE, Koury MJ, Bondurant MC, Brandt SJ. (1995). Acad Sci USA 92: 8690–8694. Erythropoietin stimulates transcription of the TAL1/SCL gene Shivdasani RA, Rosenblatt MF, Zucker-Franklin D, Jackson CW, and phosphorylation of its protein products. J Biol Chem 270: Hunt P, Saris CJ et al. (1995b). Transcription factor NF-E2 is 11603–11611. required for platelet formation independent of the actions of Puthenveetil G, Scholes J, Carbonnell D, Qureshi N, Xia P, Zeng L thrombopoietin/MGDF in megakaryocyte development. Cell 81: et al. (2004). Successful correction of the human beta-thalassemia 695–704. major phenotype using a lentiviral vector. Blood 104: 3445–3453. Shyu YC, Lee TL, Ting CY, Wen SC, Hsieh LJ, Li YC et al. (2005). Rebel VI, Kung AL, Tanner EA, Yang H, Bronson RT, Livingston Sumoylation of p45/NF-E2: nuclear positioning and transcriptional DM. (2002). Distinct roles for CREB-binding protein and p300 in activation of the mammalian beta-like globin locus. Mol Cell Biol hematopoietic stem cell self-renewal. Proc Natl Acad Sci USA 99: 25: 10365–10378. 14789–14794. Simon MC, Pevny L, Wiles MV, Keller G, Costantini F, Orkin SH. Reichardt HM, Kaestner KH, Tuckermann J, Kretz O, Wessely O, (1992). Rescue of erythroid development in gene targeted GATA-1- Bock R et al. (1998). DNA binding of the glucocorticoid receptor is mouse embryonic stem cells. Nat Genet 1: 92–98. not essential for survival. Cell 93: 531–541. Stopka T, Amanatullah DF, Papetti M, Skoultchi AI. (2005). PU.1 Reitman M, Felsenfeld G. (1988). Mutational analysis of the chicken inhibits the erythroid program by binding to GATA-1 on DNA and beta-globin enhancer reveals two positive-acting domains. Proc Natl creating a repressive chromatin structure. EMBO J 24: 3712–3723. Acad Sci USA 85: 6267–6271. Stumpf M, Waskow C, Krotschel M, van Essen D, Rodriguez P, Rekhtman N, Choe KS, Matushansky I, Murray S, Stopka T, Zhang X et al. (2006). The mediator complex functions as a Skoultchi AI. (2003). PU.1 and pRB interact and cooperate to coactivator for GATA-1 in erythropoiesis via subunit Med1/ repress GATA-1 and block erythroid differentiation. Mol Cell Biol TRAP220. Proc Natl Acad Sci USA 103: 18504–18509. 23: 7460–7474. Svensson EC, Tufts RL, Polk CE, Leiden JM. (1999). Molecular Rekhtman N, Radparvar F, Evans T, Skoultchi AI. (1999). Direct cloning of FOG-2: a modulator of transcription factor GATA-4 in interaction of hematopoietic transcription factors PU.1 and cardiomyocytes. Proc Natl Acad Sci USA 96: 956–961. GATA-1: functional antagonism in erythroid cells. Genes Dev 13: Szutorisz H, Canzonetta C, Georgiou A, Chow CM, Tora L, Dillon N. 1398–1411. (2005a). Formation of an active tissue-specific chromatin domain Robb L, Lyons I, Li R, Hartley L, Kontgen F, Harvey RP et al. (1995). initiated by epigenetic marking at the embryonic stem cell stage. Absence of yolk sac hematopoiesis from mice with a Mol Cell Biol 25: 1804–1820. targeted disruption of the scl gene. Proc Natl Acad Sci USA 92: Szutorisz H, Dillon N, Tora L. (2005b). The role of enhancers as 7075–7079. centres for general transcription factor recruitment. Trends Biochem Rodriquez P, Bonte E, Krijgsveld J, Kolodziej KE, Guyot B, Heck AJ Sci 30: 593–599. et al. (2005). GATA-1 forms distinct activating and repressive Tevosian SG, Deconinck AE, Cantor AB, Rieff HI, Fujiwara Y, complexes in erythroid cells. EMBO J 24: 2354–2366. Corfas G et al. (1999). FOG-2: a novel GATA-family cofactor Rogatsky I, Luecke HF, Leitman DC, Yamamoto KR. (2002). related to multitype zinc-finger protein Friend of GATA-1 and U- Alternate surfaces of transcriptional coregulator GRIP1 function shaped. Proc Natl Acad Sci USA 96: 950–955. in different glucocorticoid receptor activation and repression Tevosian SG, Deconinck AE, Tanaka M, Schinke M, Litovsky SH, contexts. Proc Natl Acad Sci USA 99: 16701–16706. Izumo S et al. (2000). FOG-2, a cofactor for GATA transcription Rooke HM, Orkin SH. (2006). Phosphorylation of Gata1 at serine factors, is essential for heart morphogenesis and development of residues 72, 142, and 310 is not essential for hematopoiesis in vivo. coronary vessels from epicardium. Cell 101: 729–739. Blood 107: 3527–3530. Tewari R, Gillemans N, Wijgerde M, Nuez B, von Lindern M, Rosenbauer F, Tenen DG. (2007). Transcription factors in myeloid Grosveld F et al. (1998). Erythroid Kruppel-like factor (EKLF) is development: balancing differentiation with transformation. Nat active in primitive and definitive erythroid cells and is required for Rev Immunol 7: 105–117. the function of 50HS3of the beta-globin locus control region. Rylski M, Welch JJ, Chen YY, Letting DL, Diehl JA, Chodosh LA EMBO J 17: 2334–2341. et al. (2003). GATA-1-mediated proliferation arrest during ery- Tolhuis B, Palstra RJ, Splinter E, Grosveld F, de Laat W. (2002). throid maturation. Mol Cell Biol 23: 5031–5042. Looping and interaction between hypersensitive sites in the active Saha A, Wittmeyer J, Cairns BR. (2006). Chromatin remodeling: the beta-globin locus. Mol Cell 10: 1453–1475. industrial revolution of DNA around histones. Nat Rev Mol Cell Trainor CD, Omichinski JG, Vandergon TL, Gronenborn AM, Biol 7: 437–447. Clore GM, Felsenfeld G. (1996). A palindromic regulatory site Sawado T, Halow J, Bender MA, Groudine M. (2003). The beta- within vertebrate GATA-1 promoters requires both zinc fingers of globin locus control region (LCR) functions primarily by enhancing the GATA-1 DNA-binding domain for high-affinity interaction. the transition from transcription initiation to elongation. Genes Dev Mol Cell Biol 16: 2238–2247. 17: 1009–1118. Travers A. (1999). Chromatin modification by DNA tracking. Proc Sawado T, Igarashi K, Groudine M. (2001). Activation of beta-major Natl Acad Sci USA 96: 13634–13637. globin gene transcription is associated with recruitment of NF-E2 to Tsai FY, Keller G, Kuo FC, Weiss M, Chen J, Rosenblatt M et al. the beta-globin LCR and gene promoter. Proc Natl Acad Sci USA (1994). An early haematopoietic defect in mice lacking the 98: 10226–10231. transcription factor GATA-2. Nature 371: 221–226. Schleif R. (1992). DNA looping. Annu Rev Biochem 61: 199–223. Tsai F-Y, Orkin SH. (1997). Transcription factor GATA-2 is required Schuetze S, Paul R, Gliniak BC, Kabat D. (1992). Role of the PU.1 for proliferation/survival of early hematopoietic cells and mast cell transcription factor in controlling differentiation of Friend ery- formation, but not for erythroid and myeloid terminal differentia- throleukemia cells. Mol Cell Biol 12: 2967–2975. tion. Blood 89: 3636–3643.

Oncogene Transcriptional control of erythropoiesis S-I Kim and EH Bresnick 6794 Tsai SF, Martin DI, Zon LI, D’Andrea AD, Wong GG, Orkin SH. Welch JJ, Watts JA, Vakoc CR, Yao Y, Wang H, Hardison RC et al. (1989). Cloning of cDNA for the major DNA-binding protein of the (2004). Global regulation of erythroid gene expression by transcrip- erythroid lineage through expression in mammalian cells. Nature tion factor GATA-1. Blood 104: 3136–3147. 339: 446–451. Wijgerde M, Gribnau J, Trimborn T, Nuez B, Philipsen S, Grosveld F Tsang AP, Fujiwara Y, Hom DB, Orkin SH. (1998). Failure of et al. (1996). The role of EKLF in human beta-globin gene megakaryopoiesis and arrested erythropoiesis in mice lacking the competition. Genes Dev 10: 2894–2902. GATA-1 transcriptional cofactor FOG. Genes Dev 12: 1176–1188. Wozniak RJ, Boyer ME, Grass JA, Lee Y-S, Bresnick EH. (2007). Tsang AP, Visvader JE, Turner CA, Fujuwara Y, Yu C, Weiss MJ Context-dependent GATA factor function: combinatorial require- et al. (1997). FOG, a multitype zinc finger protein as a cofactor for ments for transcriptional control in hematopoietic and endothelial transcription factor GATA-1 in erythroid and megakaryocytic cells. J Biol Chem 282: 14665–14674. differentiation. Cell 90: 109–119. Xu Z, Huang S, Chang LS, Agulnick AD, Brandt SJ. (2003). Tse C, Sera T, Wolffe AP, Hansen JC. (1998). Disruption of higher- Identification of a TAL1 target gene reveals a positive role for the order folding by core histone acetylation dramatically enhances LIM domain-binding protein (Ldb1) in erythroid gene expression transcription of nucleosomal arrays by RNA polymerase III. Mol and differentiation. Mol Cell Biol 23: 7585–7599. Cell Biol 18: 4629–4638. Xu Z, Meng X, Cai Y, Koury MJ, Brandt SJ. (2006). Recruitment Tuckermann JP, Reichardt HM, Arribas R, Richter KH, Schutz G, of the SWI/SNF protein Brg1 by a multiprotein complex Angel P. (1999). The DNA binding-independent function of the effects transcriptional repression in murine erythroid progenitors. glucocorticoid receptor mediates repression of AP-1-dependent Biochem J 399: 297–304. genes in skin. J Cell Biol 147: 1365–1370. Xu Z, Meng X, Cai Y, Liang H, Nagarajan L, Brandt SJ. (2007). Uhlmann T, Boeing S, Lehmbacher M, Meisterernst M. (2007). The Single-stranded DNA-binding proteins regulate the abundance of VP16 activation domain establishes an active mediator lacking LIM domain and LIM domain-binding proteins. Genes Dev 21: CDK8 in vivo. J Biol Chem 282: 2163–2173. 942–955. Vakoc CR, Letting DL, Gheldof N, Sawado T, Bender MA, Groudine Yagi R, Chen LF, Shigesada K, Murakami Y, Ito Y. (1999). A WW M et al. (2005). Proximity among distant regulatory elements at domain-containing yes-associated protein (YAP) is a novel tran- the beta globin locus requires GATA-1 and FOG-1. Mol Cell 17: scriptional coactivator. EMBO J 18: 2551–2562. 453–462. Yao T-P, Oh SP, Fuchs M, Zhou N-D, Ch’ng L-E, Newsome D et al. Versaw WK, Blank V, Andrews NM, Bresnick EH. (1998). Mitogen- (1998). Gene dosage-dependent embryonic development and pro- activated protein kinases enhance long-range activation by the beta- liferation defects in mice lacking the transcriptional integrator p300. globin locus control region. Proc Natl Acad Sci USA 95: 8756–8760. Cell 93: 361–372. Vieira KF, Levings PP, Hilll MA, Cruselle VJ, Kang SH, Engel JD Zhang P, Behre G, Pan J, Iwama A, Wara-aswapati N, Radomska HS et al. (2004). Recruitment of transcription complexes to the et al. (1999a). Negative cross-talk between hematopoietic regu- beta-globin gene locus in vivo and in vitro. J Biol Chem 279: lators: GATA proteins repress PU.1. Proc Natl Acad Sci USA 96: 50350–50357. 8705–8710. Vyas P, McDevitt MA, Cantor AB, Katz SG, Fujiwara Y, Orkin SH. Zhang PJ, Zhang X, Iwama A, Yu C, Smith KA, Mueller BU et al. (1999). Different sequence requirements for expression in erythroid (2000). PU.1 inhibits GATA-1 function and erythroid differentia- and megakaryocytic cells within a regulatory element upstream of tion by blocking GATA-1 DNA binding. Blood 96: 2641–2648. the GATA-1 gene. Development 126: 2799–2811. Zhang W, Bieker JJ. (1998). Acetylation and modulation of erythroid Wadman IA, Osada H, Grutz GG, Agulnick AD, Westphal H, Forster Kruppel-like factor (EKLF) activity by interaction with histone A et al. (1997). The LIM-only protein Lmo2 is a bridging molecule acetyltransferases. Proc Natl Acad Sci USA 95: 9855–9860. assembling an erythroid, DNA-binding complex which includes Zhang W, Kadam S, Emerson BM, Bieker JJ. (2001). Site-specific the TAL1, E47, GATA-1 and Ldb1/NLI proteins. EMBO J 16: acetylation by p300 or CREB binding protein regulates erythroid 3145–3157. Kruppel-like factor transcriptional activity via its interaction with Wang X, Crispino JD, Letting DL, Nakazawa M, Poncz M, Blobel the SWI–SNF complex. Mol Cell Biol 21: 2413–2422. GA. (2002). Control of megakaryocyte-specific gene expression by Zhang Y, Ng HH, Erdjument-Bromage H, Tempst P, Bird A, GATA-1 and FOG-1: role of Ets transcription factors. EMBO J 21: Reinberg D. (1999b). Analysis of the NuRD subunits reveals a 5225–5234. histone deacetylase core complex and a connection with DNA Wechsler J, Greene M, McDevitt MA, Anastasi J, Karp JE, LeBeau methylation. Genes Dev 13: 1924–1935. MM et al. (2002). Acquired mutations in GATA-1 in the Zhao W, Kitidis C, Fleming MD, Lodish HF, Ghaffari S. (2006). megakaryoblastic leukemia of Down syndrome. Nat Genet 32: Erythropoietin stimulates phosphorylation and activation of 148–152. GATA-1 via the PI3-kinase/AKT signaling pathway. Blood Weiss MJ, Keller G, Orkin SH. (1994). Novel insights into erythroid 107: 907–915. development revealed through in vitro differentiation of GATA-1 Zhou D, Pawlik KM, Ren J, Sun CW, Townes TM. (2006). embryonic stem cells. Genes Dev 8: 1184–1197. Differential binding of erythroid Krupple-like factor to embryo- Weiss MJ, Orkin SH. (1995). Transcription factor GATA-1 permits nic/fetal globin gene promoters during development. J Biol Chem survival and maturation of erythroid precursors by preventing 281: 16052–16057. apoptosis. Proc Natl Acad Sci USA 92: 9623–9627. Zhu Y, Qi C, Jia Y, Nye JS, Rao MS, Reddy JK. (2000). Deletion of Weiss MJ, Yu C, Orkin SH. (1997). Erythroid-cell-specific properties PBP/PPARBP, the gene for nuclear receptor coactivator peroxi- of transcription factor GATA-1 revealed by phenotypic rescue of a some proliferator-activated receptor-binding protein, results in gene-targeted cell line. Mol Cell Biol 17: 1642–1651. embryonic lethality. J Biol Chem 275: 14779–14782.

Oncogene