(Journal of Leukocyte Biology. 2002;71:545-556.)
© 2002
by Society for Leukocyte Biology
CBP and p300: versatile coregulators with important roles in hematopoietic gene expression
Gerd A. Blobel
Division of Hematology, Children’s Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia
Correspondence: Gerd A. Blobel, M.D., Ph.D., Children’s Hospital of Philadelphia, Abramson Research Center #316A, 3615 Civic Center Blvd., Philadelphia, PA 19104. E-mail: blobel{at}email.chop.edu
Key Words: CBP/p300 • chromatin • coactivators • hematopoiesis • transcription
 |
INTRODUCTION
|
|---|
According to recent estimates, a human cell has between 30,000 and 40,000 coding genes, but only a fraction of these are expressed in any cell lineage. Gene expression profiles of a given cell type undergo dramatic changes throughout development, differentiation, and the cell cycle, presenting the gene regulatory apparatus of a cell with a phenomenal degree of complexity. Cells deal efficiently with this enormous task by using transcriptional regulators in a combinatorial way. Thus, unique combinations of transcription factors convey specificity of gene expression and allow the use of each transcription factor at multiple gene loci. One advantage of this strategy is that even widely expressed transcription factors or cofactors can contribute to tissue-specific gene expression. The related transcription coactivators, CBP [cyclic adenosine monophosphate (cAMP) response element-binding protein (CREB)-binding protein] and p300, are expressed widely and interact with a surprisingly large number of transcription factors via dedicated domains. As such, they could be viewed as general transcription regulators. Yet, loss of function and aberrant regulation or expression of CBP and p300 are associated with distinct phenotypes and diseases in humans and experimental organisms, indicating that certain genes/tissues are more sensitive to changes in CBP and p300 activity than others. Several thorough reviews provide a guide to the multitude of transcription factors regulated by CBP and p300 in various organisms [1
2
3
4
5
6
]. Another recent review concentrates on the function of p300 and CBP during transcriptional regulation of hematopoietic development [7
]. This review summarizes the functions of CBP and p300 in hematopoietic cells and the diseases associated with them. Special attention is given to erythroid gene expression, because it provides novel insights into lineage-specific gene expression and the function of enhancers and locus control regions (LCRs). Mechanistic aspects of CBP and p300 function are discussed in the context of their function in hematopoietic cells.
|
A BRIEF HISTORY
|
|---|
CBP was identified through its ability to interact with the cAMP-regulated transcription factor CREB [8
], and p300 was identified as a protein targeted by the adenovirus E1A oncoprotein [9
] (for review, see ref. [10
]). In the years following their cloning, a panoply of transcriptional regulators was shown to interact with CBP and p300. Although most transcription factors appear to bind CBP and p300 constitutively, in some cases this interaction is regulated; for example, phosphorylation of certain transcription factors (e.g., CREB) [8
] or CBP itself [11
] can increase their binding affinity, and acetylation of some transcription factors has been shown to stabilize their interaction with CBP [12
, 13
]. Multiple mechanisms have been proposed to explain CBP and p300 functions. Several studies showed interactions of CBP and p300 with the general transcription factors TFIIB, TBP, and RNA polymerase II, suggesting that CBP and p300 might function by physically connecting activators with components of the basal transcription machinery (see below). The discovery that CBP and p300 are enzymes that catalyze the acetylation of histones [14
, 15
] and that they associate with other acetyltransferases, including PCAF [16
], GCN5 [17
], and members of the nuclear hormone-receptor coactivators (for review, see ref. [18
]), suggests that CBP and p300 might act in part through regulating chromatin structure. Gene activation is generally associated with hyperacetylated histones, and repressed genes are usually found in hypoacetylated chromatin. In support of this view, several transcriptional corepressors associate with deacetylases (for review, see ref. [19
]).
Recent evidence suggests that CBP and p300 activities are modulated through protein-protein interactions. For example, adenoviral E1A has been shown to inhibit CBP and p300 activities, thereby elucidating the mechanism of E1A function in gene regulation further [20
, 21
]. Later studies demonstrated that acetyltransferase activities of CBP and p300 are regulated dynamically. A growing number of molecules have been found to inhibit [22
23
24
] or facilitate [25
, 26
] CBP- and p300-mediated histone acetylation (see below). In addition, post-translational modifications regulate CBP and p300 acetyltransferase activities [27
, 28
]. Therefore, CBP and p300 are not passive effectors of the transcription factors to which they bind but serve as additional relay stations where cellular signals are integrated into transcriptional output. Furthermore, it has been suggested that the ability of CBP and p300 to form multiple protein contacts mediates synergy between transcriptional regulators. Conversely, competition for limiting amounts of CBP and p300 has been invoked as a potential mechanism for mutual repression between activators. The ability to mediate this kind of transcriptional cross-talk earned CBP and p300 the description as transcriptional integrators.
|
THE STRUCTURE OF CBP/p300
|
|---|
CBP and p300 are encoded by different genes but are similar structurally and functionally. Therefore, throughout this review, they will subsequently be referred to as CBP/p300. However, notable differences exist between them, some of which will be discussed below. CBP/p300 are large molecules that contain multiple, functionally important domains (Fig. 1
). Three cysteine/histidine-rich (CH) domains serve as docking modules for numerous transcriptional regulators. In addition, the KIX domain that binds to CREB, the archetypical CBP-binding transcription factor, serves as an anchor for several other transcription factors. Additional sites for protein-protein contacts include the glutamine-rich C-terminus that forms contacts with other coactivators, most notably those involved in nuclear hormone-receptor signaling (for review, see ref. [18
]). Although few obvious signature motifs have been identified in transcriptional regulators that would predict with certainty their ability to bind to a given domain within CBP/p300, a short motif (FXE/DXXXL) has been identified in E1A, p53, and E2F that forms specific contacts with a defined site (called the TRAM motif) in the CH3 domain of CBP [29
].
View larger version (17K):
[in this window]
[in a new window]
|
Figure 1. Domain structure of CBP. Hematopoietic transcription factors that bind CBP are indicated. The binding site for EKLF has not been determined yet. Numbers represent approximate domain boundaries. CH1, CH2, and CH3, CH-rich domains; KIX, CREB-binding domain; Bromo, bromodomain; AT, acetyltransferase domain; Q-rich, glutamine-rich domain. Note that some proteins can interact with more than one domain in CBP.
|
|
The acetyltransferase domain catalyzes acetylation of all four core histones and of several nonhistone nuclear proteins (see below). The acetylation function of CBP/p300 is required for transcriptional activation by some but not all CBP/p300-bound transcription factors. This suggests that CBP can mediate activation in an acetyltransferase-independent way, perhaps through contacts with components of the basal transcription machinery. In addition, it is possible that other CBP-bound cofactors with acetyltransferase activity such as PCAF might provide the necessary chromatin-modifying function in cases where CBP acetyltransferase activity is dispensable.
The bromodomain that resides N-terminal to the acetyltransferase domain is found in most acetyltransferases and in components of adenosine 5'-triphosphate (ATP)-dependent chromatin-remodeling complexes (for review, see ref. [30
]). Bromodomains bind specifically to acetylated lysine [31
], which might contribute to acetylation-dependent contacts with the N-terminal tails of histones. The bromodomain of CBP/p300 is required for histone acetylation and activation of transcription on chromatin templates in vitro [32
]. Recently, it was shown that acetylation of the myogenic transcription factor MyoD stimulates its affinity for the bromodomain of CBP/p300 [13
]. Thus, acetylation of histones and/or transcription factors might stabilize the interaction of CBP/p300 with transcriptionally active genes.
|
CBP AND p300 IN DISEASE
|
|---|
Haploinsufficiency of CBP in humans leads to Rubinstein-Taybi syndrome (RTS) [33
], which is characterized by mental retardation, craniofacial abnormalities, and broad thumbs and big toes [34
]. In addition, RTS patients exhibit a propensity for malignancies of mostly neural but also hematopoietic origin [35
]. The hematopoietic malignancies found in RTS patients include acute lymphocytic leukemia, acute myelogenous leukemia, and non-Hodgkin’s lymphoma. Heterozygous deletion of CBP in mice causes defects in multiple tissues including the hematopoietic system (see also below). Strikingly, a fraction of CBP+/- mice developed hematological malignancies including histiocytic sarcomas and myelogenous and lymphocytic leukemia [36
]. Transplantation of spleen or bone marrow cells from grossly tumor-free CBP+/- donor mice into sublethally irradiated, wild-type recipient mice caused a similar spectrum of malignancies, indicating that a substantial fraction of CBP+/- mice harbored cells with tumorigenic potential even in the absence of overt disease. Of note, one of the recipient mice developed plasmacytoma. Examination of some tumors revealed loss of the wild-type CBP allele, consistent with a role of CBP as tumor suppressor. In light of the limited number of mice analyzed, it remains possible that loss of CBP function causes tumors of other cell lineages as well. Although p300 mutations and truncations have been observed in a variety of human tumors, some of which had even lost the second allele [37
, 38
], so far no neoplastic disease has been observed in p300+/- mice. Whether loss of heterozygosity contributes to formation of malignancies in human RTS patients remains to be determined.
A role for CBP/p300 as tumor suppressors is supported further by the observation that they are bound and inhibited by certain viral oncoproteins that interfere with cellular differentiation and promote cell-cycle progression. The adenoviral protein E1A is the most well-studied among these and is widely used to examine the requirement of CBP/p300 in cellular functions (for review, see ref. [39
]). Although E1A interacts with several proteins, the analysis of various E1A mutants suggests that its ability to interact with CBP/p300 accounts in part for its biological effects. A plethora of transcription factors are inhibited by forced expression of E1A but not by mutant forms of E1A deficient for CBP/p300 binding, implying that CBP/p300 are essential for the function of these factors. However, CBP/p300 requirement cannot be inferred solely from studies using E1A, because E1A engages in multiple protein interactions (for a critical discussion of this subject, see ref. [3
]). This point is illustrated by a recent study showing that the N-terminus of E1A binds and perturbs a newly identified multiprotein complex with ATPase and helicase activity [40
]. This complex is suspected to play a role in the regulation of chromatin structure, although direct evidence for this function is still missing. The domain of E1A that mediates this interaction is required for its full transforming activity and resides adjacent to the N-terminal CBP/p300-binding site [40
]. Thus, an implementation of CBP/p300 as coactivator for a given transcription factor requires additional functional and biochemical evidence.
Altered function of CBP and p300 that results from chromosomal translocations also contributes to the formation of leukemias. This topic has been covered by two thoughtful, recent reviews [3
, 5
]. The first example of an involvement of the CBP gene in the formation of leukemias in humans came from the discovery that the CBP gene is fused in-frame to the MOZ (monocytic leukemia zinc finger) gene in acute myeloid leukemias (AML) of the M4/M5 subtype carrying a t(8;16)(p11;p13) translocation [41
]. Like CBP, MOZ is an acetyltransferase [42
], suggesting that this fusion event generates a powerful chromatin-modifying activity. More recently, the p300 gene has also been found as fusion partner for MOZ in patients with AML bearing the t(8;22)(p11;q13) translocation, suggesting that p300 can fulfill the same role as CBP during leukemogenesis [43
, 44
].
Additional studies identified a chromosomal translocation [t(11;16)(q23;p13)] in patients that had previously undergone chemotherapy with DNA topoisomerase inhibitors. This translocation fuses the CBP gene in-frame to the MLL (mixed lineage leukemia) gene (also called ALL-1 and HRX) [45
46
47
48
]. The MLL gene, which is similar to the Drosophila trithorax genes and thought to regulate chromatin structure, is involved in numerous chromosomal translocations found in patients with acute lymphoblastic leukemia and AML (for review, see ref. [49
]). MLL gene translocations involve more than 30 different partner genes, which show no common theme that would hint at a mechanism by which they contribute to malignant transformation. Initially, this suggested that perturbation of MLL activity, rather than its fusion partners, is the cause of the diseases. However, all fusion partners are in frame, with MLL suggesting that they do play a role in leukemogenesis. Furthermore, in the case of the MLL-CBP fusion, neither the MLL nor the CBP portion was sufficient by itself to induce myeloid leukemia in mice [50
]. Instead, leukemia induction required fusion of MLL to the bromodomain and acetyltransferase domain of CBP. This demonstrates the functional importance of CBP as a MLL fusion partner and suggests that MLL-CBP-induced transformation results in part through aberrant regulation of chromatin structure. Following this theme, the t(11;22)(q23;q13) translocation found in a patient with therapy-induced AML fused the MLL gene to p300 [51
], again suggesting that CBP and p300 might be functionally interchangeable when contributing to the transformed state. Finally, a MOZ-related acetyltransferase, MORF [52
], was fused to the CBP gene in an AML patient with a t(10;16)(q22;p13) translocation [53
]. Together, these findings suggest that dysregulation of chromatin structure through CBP/p300 can lead to altered cellular differentiation and proliferation of hematopoietic cells. So far, no alterations have been observed in the nontranslocated CBP and p300 alleles in human leukemias, suggesting that transformation is not simply a result of loss of CBP/p300 function, as might have been inferred from the high incidence of leukemias in CBP+/- mice. Rather, the translocation events likely produce molecules with novel functions that act in a dominant way to trigger transformation in vivo.
|
CBP AND p300 FUNCTION IN MICE
|
|---|
Mice lacking one allele of CBP display skeletal abnormalities, mostly of the skull and rib cage, with some structures displaying growth retardation and delayed ossification [54
]. This phenotype has been likened to that of RTS patients [54
], although defects in other organ systems found in RTS patients were not reproduced. The penetrance varied widely, depending on the genetic background, and led to embryonic lethality in the most severe cases. In another study, the CBP gene was disrupted by insertional mutagenesis such that a truncated form of CBP was left behind [55
]. The phenotype of mice heterozygous for this defect was more severe than that observed in mice heterozygous for a CBP null allele and included cardiac defects and various neurological deficiencies. This suggests that the truncated form of CBP acted as a dominant, interfering allele that perturbed the function of the normal CBP protein as well as p300. Impaired p300 function might be inferred from the observation that p300-/- mice displayed obvious cardiac abnormalities [56
], and CBP-/- mice did not [57
]. These results suggest further that CBP and p300 have overlapping functions in vivo and that the combined dose of CBP and p300 is essential for normal development. In a more recent study, it was noticed that most but not all CBP+/- mice have substantial hematological abnormalities [36
]. These include diminished bone marrow cellularity resulting from reduced numbers of erythroid, myeloid, and lymphoid cells. Extramedullary hematopoiesis was present as reflected in dramatic splenomegaly. Peripheral blood contained reduced numbers of B cells and increased numbers in myeloid cells. No overt hematological defects were observed in p300+/- mice [36
], although reduced vascularity was found in the yolk sac [56
]. As described above, haploinsufficiency of CBP, but not p300, leads to various hematopoietic malignancies that become overt with the increasing age of the mice [36
].
Homozygous deficiency of CBP in mice has been described by three groups and leads to embryonic death between days 9.5 and 10.5 [56
57
58
]. Prior to death, embryos exhibit developmental retardation and defects in multiple organ systems. These include failed neural tube closure [56
57
58
], reduced numbers of primitive and definitive hematopoietic precursor cells, and defects in blood-vessel formation [57
, 58
]. Brain hemorrhages were only observed in one study [57
] and are likely the result of defective vasculature. Mice homozygous for the CBP truncation appeared to have a slightly more pronounced phenotype [58
]. Erythroid cells in CBP-/- mice were reduced in number but were able to mature, suggesting that CBP deficiency affects proliferation or survival of precursor cells rather than their differentiation [58
].
Homozygous loss of p300 revealed defects similar but distinct from that found in CBP-/- mice [56
]. The most obvious features were developmental retardation, reduced size, neural tube defects, and reduced cardiac trabeculation. The cardiac defects are likely the reason for embryonic lethality between days 9 and 11. In light of the long-held assumption that the proliferative effects of E1A are in part a result of inhibition of CBP/p300 function, an unexpected finding was that cultured fibroblasts from p300-/- mice grew at a reduced rate and ceased to proliferate beyond three to four generations [56
]. It is possible that CBP/p300 promote cell proliferation and that E1A augments this activity further. It is also conceivable that the effects of p300 and CBP are dependent on the cell type. For example, loss of heterozygosity of the CBP gene was found in hematopoietic malignancies in mice (see above), suggesting that CBP might normally slow growth of the hematopoietic cells.
|
CBP AND p300 IN HEMATOPOIETIC GENE REGULATION
|
|---|
CBP/p300 interact with numerous transcription factors that function in the hematopoietic system (for review, see ref. [7
]). Figure 1 shows a subset of these factors and their binding sites in CBP. As is apparent from this diagram, CBP/p300 interact with transcriptional regulators with critical functions in virtually all hematopoietic lineages. Although most of these transcription factors play important roles during normal hematopoietic development, some have oncogenic potential when aberrantly expressed or when rearranged as a result of chromosomal translocations. It is perplexing that CBP/p300 cooperate with transcription factors required for normal differentiation and cell-cycle arrest (e.g., GATA-1; see below) but also with factors that promote cell-cycle progression and transformation (e.g., Myb and PU.1). It is possible that CBP/p300 recruitment to specific promoters depends on the abundance and availability of DNA-binding proteins and on promoter context. Thus, it is conceivable that an overexpressed oncoprotein might compete with differentiation-inducing factors for CBP/p300 function. Similarly, during normal development, changes in the partitioning of CBP/p300 between transcriptional regulators with opposing functions might in part control the balance of differentiation and proliferation. For example, maturation of the erythroleukemic cell line MEL is accompanied by down-modulation of the proto-oncoproteins PU.1 and c-Myb among others. Because PU.1 and c-Myb bind to CBP/p300 [59
60
61
], this might increase availability of CBP/p300 for differentiation-associated factors such as GATA-1, NF-E2, and EKLF (erythroid Krüppel-like factor; see below). In addition, post-translational modifications might determine CBP/p300 use at a given promoter.
An important role of CBP/p300 in establishing lymphoid- and myeloid-specific gene expression is supported by the interaction of CBP/p300 with transcription factors that are critical for the formation and differentiation of myeloid and lymphoid cells and by the observation that CBP-deficient mice display altered myeloid and lymphoid development (see above). A meaningful discussion of all myeloid and lymphoid transcription factors that interact with CBP (including the AML, C/EBP, E2A, GATA-2, GATA-3, Myb, and Ets families) is beyond the scope of this review, and the reader is directed toward another review on this subject [7
]. Instead, CBP/p300 function will be discussed in the context of erythroid gene expression. The molecular interactions of CBP/p300 with erythroid transcription factors and their mechanistic consequences have been studied thoroughly and therefore can be used to illustrate general concepts.
Given the highly specialized function of erythroid cells, which is largely limited to production of hemoglobin, the globin genes have long served as a model system for high-level, tissue-specific, developmentally controlled gene expression. Thus, principles gleaned from studies of the globin-gene locus are likely to apply at other gene loci as well. The human ß-like globin genes are arranged along the chromosome in the same order in which they are expressed throughout development (Fig. 2
). The embryonic 
-globin gene is expressed during the embryonic stages of development when erythropoiesis takes place in the yolk sac, the 
-globin genes are expressed during fetal liver erythropoiesis, and the 
- and ß-globin genes are expressed after birth when hematopoiesis shifts to the bone marrow. The promoters of these genes contain most of the regulatory elements required for stage-specific gene expression. However, high-level transcription requires the action of a far-upstream enhancer, the LCR, which contains five DNaseI hypersensitive sites (HS). These sites contain elements for broadly expressed, as well as tissue-restricted transcription factors. The latter include GATA-1, NF-E2, EKLF, and SCL/TAL1. In transgenic animals, the LCR mediates position-independent, copy number-dependent expression of linked transgenes, suggesting that it can trigger and/or maintain an open chromatin configuration (for review, see refs. [62
, 63
]). However, gene targeting in mice showed that loss of the LCR does not lead to profound changes in chromatin structure, despite the loss of globin-gene expression, leading to the suggestion that sequences outside the LCR might substitute for its chromatin-opening functions [64
, 65
]. It has been known for some time that the globin-gene locus resides in a domain characterized by hyperacetylated histones that coincides with general DNaseI sensitivity, indicating that the locus is within an open chromatin domain [66
]. More recent studies showed that in avian and mammalian erythroid cells, acetylation of histone H3 and to a lesser extent, H4 is increased further at actively transcribed globin genes and at the LCR [67
68
69
]. Of note, histone hyperacetylation can occur even in the absence of transcription, strongly suggesting that it is a prerequisite for transcription rather than simply a consequence thereof [67
, 68
, 70
, 71
]. Together, these studies are consistent with a model in which transcriptional regulators that govern globin-gene expression modulate chromatin structure in part through histone acetylation. Among the candidate acetyltransferases that contribute to histone acetylation at the globin locus are CBP/p300. Indeed, inhibition of CBP/p300 function by expression of a conditional form of E1A leads to a block in erythroid differentiation and globin-gene expression [72
]. In addition, several transcription factors involved in erythroid differentiation and globin-gene expression, including GATA-1, NF-E2, EKLF, and SCl/Tal1, associate with CBP/p300 and/or PCAF and are themselves regulated by acetylation [73
74
75
76
77
78
79
80
81
]. Binding sites for all of these factors are present at the LCR, and it has been suggested that they synergize in the recruitment of CBP/p300 and other cofactors to the globin-gene locus [7
]. In the following paragraphs, the interplay between these factors and their coactivators will be discussed.
View larger version (6K):
[in this window]
[in a new window]
|
Figure 2. Structure of the human globin-gene cluster. Chromosomal arrangement of the globin genes (shaded boxes) reflects the order of their expression. The LCR is defined by five DNaseI HS 1-5 (black triangles). More recent studies identified additional HS further upstream (HS 6 and 7). Each HS contains numerous binding sites for transcriptional regulators.
|
|
|
GATA-1
|
|---|
GATA-1 is a zinc-finger transcription factor essential for the formation of normal, mature erythroid cells and blood platelets (for review, see ref. [82
]). Functionally important GATA-1-binding sites are present in virtually all erythroid-restricted genes, including the promoters of the globin genes and the LCR. GATA-1 nullizygous mice succumb to anemia between days 10.5 and 11.5 [83
] as a result of impaired differentiation and survival of erythroid precursor cells [84
, 85
]. CBP is one of several transcriptional regulators that binds to GATA-1 in vitro and in vivo and stimulates its transcriptional activity [72
]. CBP acetylates GATA-1 at conserved lysine-rich domains near the zinc fingers in vitro [73
, 74
]. In transfection-based assays, CBP stimulates GATA-1 acetylation in vivo [74
]. Mutations at the major acetylation sites impair the ability of GATA-1 to restore erythroid differentiation of a GATA-1-deficient erythroid cell line, suggesting that acetylation of GATA-1 is important for its biological activity in vivo [74
]. However, the molecular mechanism by which acetylation of GATA-1 stimulates its activity is unresolved. Although acetylation of chicken GATA-1 by p300 markedly increased GATA-1’s affinity for DNA [73
], CBP-mediated acetylation of murine GATA-1 did not lead to changes in GATA-1 DNA-binding activity [74
]. Mutations at the major acetylation sites of murine GATA-1 abrogated its biological activity without affecting DNA binding, suggesting alternative functions associated with GATA-1 acetylation [74
]. Similar findings were reported normally for GATA-3, which is expressed in T cells [86
]. One possibility is that acetylation stimulates the interaction with accessory proteins that might aid in transcriptional activation and/or chromatin remodeling. An example for this possibility is provided in the case of EKLF, where acetylation stimulates interaction with a component of the SWI/SNF complex (see below). Alternatively, acetylation might disrupt interaction between GATA-1 and a repressor molecule.
|
NF-E2
|
|---|
NF-E2 is a heterodimeric basic-leucine zipper transcription factor consisting of p45 NF-E2 and a member of the small Maf family of proteins (for review, see refs. [87
, 88
]). Expression of p45NF-E2 is limited mostly to erythroid cells and megakaryocytes [89
], and Maf proteins are expressed more widely. Of particular interest are MafK (also called p18 NF-E2) and MafG, which are the predominant p45 NF-E2 partners in erythroid cells and megakaryocytes [90
, 91
]. NF-E2-binding sites (also called MAREs for Maf recognition elements) are present in the regulatory regions of select erythroid and megakaryocytic genes, most notably in HS2 of the ß-globin LCR, where they are essential for enhancer function [92
93
94
]. In vivo occupancy of HS2 by p45 NF-E2 has been demonstrated in living cells using chromatin immunoprecipitation experiments [95
96
97
]. Homozygous p45 NF-E2 knockout mice display defects in megakaryocyte maturation with profound thrombocytopenia and hemorrhaging [98
]. The defect in erythroid development is surprisingly mild, and ß-globin mRNA levels are virtually normal, suggesting that other factors might compensate for p45 NF-E2 function in vivo [99
]. Loss of MafG leads to a phenotype similar to that of p45 NF-E2 deficiency, with thrombocytopenia being the most striking defect [90
]. Mice deficient for MafK are overtly normal and, like the MafG knockout mice, have normal red cells [100
]. A role for CBP/p300 in mediating NF-E2 function was suggested by in vitro-binding studies showing a direct physical interaction between the N-terminal activation domain of p45 NF-E2 and CBP [76
]. The functional importance of this interaction was supported by the observation that E1A inhibits the enhancer function of HS2 of the ß-globin LCR and that inhibition was conferred to a large extent by NF-E2 elements [77
]. A more recent study showed that p45 NF-E2 and MafG interact with CBP in vitro and in vivo [75
]. Furthermore, this interaction leads to potent acetylation of MafG but not p45. The acetylation sites were mapped to four conserved lysine residues in the DNA-binding domain of MafG. Acetylation of MafG stimulates DNA binding of NF-E2 in vitro, and mutations of the MafG acetylation sites reduce DNA binding and transcriptional activation by NF-E2 [75
]. Thus, the interaction between CBP and NF-E2 might serve two functions, namely modulation of NF-E2 activity and regulation of chromatin structure, similar to what has been proposed for GATA-1. However, a recent study called a role for NF-E2 in establishing histone acetylation at the LCR into question [97
]. Specifically, the authors carefully analyzed histone acetylation patterns at the LCR and the ß-globin-gene promoters in MEL and CB3 erythroleukemia cells. CB3 cells lack both alleles of p45 NF-E2 and in contrast to primary murine p45 NF-E2-deficient erythroid cells, fail to express high levels of ß-globin mRNA [101
]. Reintroduction of p45 NF-E2 restores ß-globin gene expression [101
, 102
], thus providing a functional assay for p45 NF-E2 function. Surprisingly, high levels of histone H3 and H4 acetylation at HS2 and HS3 of the LCR were independent of p45 NF-E2, suggesting that it is dispensable for histone acetylation at the LCR [97
]. However, histone acetylation at the ß-globin promoter was reduced in CB3 cells and was increased to normal levels upon expression of p45 NF-E2. This finding is especially remarkable in light of the absence of NF-E2-binding sites at the ß-globin gene promoter, suggesting that LCR-bound NF-E2 might interact with the promoter indirectly through other promoter-bound transcription factors or via an LCR-coupled looping mechanism [96
].
The lack of p45 NF-E2 requirement for histone acetylation at the LCR might reflect compensation by p45 NF-E2-related factors. Alternatively, it is possible that p45 NF-E2 is required for establishment of an acetylated chromatin domain but not for its maintenance. CB3 cells were derived from infection of normal erythroid precursor cells with the Friend virus complex [101
]. One of the p45 NF-E2 alleles was inactivated as a result of retroviral integration, and the other was lost for unknown reasons. Because the cells were fully committed, normal erythroid precursor cells prior to viral infection, it is likely that the globin locus was already in an open and acetylated chromatin domain [103
]. Thus, one cannot completely rule out a requirement of p45 NF-E2 for establishing histone hyperacetylation at the LCR.
Several proteins have been described to bind p45 NF-E2 and might contribute to its function in vivo. Of note, NF-E2 has been shown to trigger chromatin remodeling at HS2 in vitro by recruiting an ATP-dependent chromatin-remodeling complex [104
, 105
]. It is conceivable that histone acetyltransferases and ATP-dependent chromatin-remodeling machines act in a sequential way in regulating chromatin structure, similar to what has been described for certain genes in yeast [106
, 107
].
|
EKLF
|
|---|
EKLF (also known as KLF1) is an erythroid-specific zinc-finger-containing transcription factor that binds to CACCC elements located in the proximal promoter region of the adult-expressed ß-globin-gene in mice and humans (for review, see ref. [108
]). Consistent with its essential role in ß-globin gene expression, loss of EKLF in gene-targeted mice leads to lethal ß-thalassemia [109
, 110
]. EKLF deficiency is associated with reduced DNaseI hypersensitivity at the ß-globin promoter and HS3 of the LCR, suggesting an important role of EKLF in regulating chromatin structure [111
]. Indeed, EKLF interacts with acetyltransferases and ATP-dependent chromatin-remodeling complexes, and these interactions appear to be coordinated [78
, 79
, 112
]. A mammalian SWI/SNF chromatin-remodeling complex called E-RC1 was shown to facilitate EKLF-dependent formation of a DNaseI HS and transcriptional activation on chromatinized DNA templates in vitro [112
]. These data are supported further by the observation that the BRG1 and BAF170 subunits of SWI/SNF are associated with the ß-globin promoter in vivo in transfected cells [113
]. BRG1 recruitment requires cooperation of the LCR and the ß-globin promoter and depends on an intact EKLF-binding site.
Although EKLF can interact with PCAF and CBP/p300, only CBP/p300 acetylate EKLF and stimulate its transcriptional activity [78
]. Acetylation occurs predominantly at two lysine residues, one in the activation domain (lysine 288) and one in the zinc-finger domain (lysine 302) [79
]. Mutation of the former but not the latter site results in loss of CBP/p300-stimulated transcription in transfected cells. Interestingly, the transcriptional activity of EKLF on chromatinized DNA templates in vitro was enhanced further upon prior acetylation by p300, and mutations at the acetylation sites reduced the p300 response markedly [79
]. A potential link between EKLF acetylation and SWI/SNF-complex recruitment was discovered when it was found that acetylation of EKLF stimulated its interaction with BRG1 in vitro [79
]. This result imparts a hierarchy on the mechanism by which EKLF activates transcription at the ß-globin promoter. Accordingly, EKLF binds to CBP/p300, which leads to acetylation of histones and EKLF itself. Acetylation in turn stimulates SWI/SNF recruitment, followed by chromatin remodeling and transcriptional activation. However, evidence for a direct interaction between EKLF and SWI/SNF components in vivo is still lacking, leaving the possibility that association of SWI/SNF with the ß-globin promoter might occur via locally hyperacetylated histones rather than with EKLF [114
]. It is conceivable that other transcription factors such as NF-E2 and GATA-1 similarly recruit SWI/SNF components upon acetylation.
|
SCL/TAL1
|
|---|
The basic helix-loop-helix (bHLH) transcription factor SCL, also called TAL1, is expressed in multiple hematopoietic lineages and plays an important role in erythroid cell differentiation (for review, see ref. [115
]). SCL forms heterodimers with E12, E47, E2-2, and HEB to bind DNA and activate or repress transcription. SCL is part of multimeric transcription-factor complexes that in erythroid cells, contain LMO1, LMO2, Lbd, and GATA-1 [116
117
118
]. Mice, in which the SCL gene has been disrupted, show a complete lack of primitive and definitive hematopoiesis [119
120
121
122
] and display blood-vessel defects in the embryonic yolk sac [119
, 123
]. The broad nature of this phenotype suggests that SCL is required at the stem-cell stage of hematopoiesis or even earlier during formation of the putative hemangioblast, the precursor of hematopoietic and endothelial cells. SCL plays a positive role during erythroid development because its expression increases during chemically or erythropoietin-induced differentiation of erythroid cells (for review, see ref. [115
]). In addition, forced SCL expression enhances differentiation of erythroid cell lines [124
, 125
]. The mechanism by which SCL regulates erythroid differentiation is unknown. Although the requirement of the basic (DNA-binding) region of SCL suggests that its activity depends on interaction with DNA, SCL can also regulate gene expression indirectly through association with other proteins, including LMO2, Ldb, and GATA-1. Furthermore, SCL can activate or inhibit transcriptional activation depending on the gene context, and it is unclear to what extent these functions contribute to erythroid differentiation. The key SCL target genes in erythroid cells remain unknown. Although DNA binding of SCL appears surprisingly dispensable for the formation of primitive erythroid cells [126
], complete maturation of definitive precursor cells and MEL cells requires a functional DNA-binding domain [125
, 126
]. Two studies found that mutations at a conserved E-box (which binds bHLH proteins) in HS2 of the LCR lead to reduced HS2-enhancer activity [127
, 128
], and SCL has been shown to bind to this site in vitro [128
]. However, whether SCL regulates globin-gene expression in vivo remains an open question because SCL-/- erythroid-precursor cells expressing a DNA-binding defective form of SCL have close-to-normal ß-globin levels [126
].
SCL was found to interact with p300 and PCAF in vitro and in vivo, leading to a modest increase in SCL transcriptional activity [80
, 81
]. PCAF and p300 acetylate SCL at distinct sites [81
]. Acetylation by PCAF, which occurs in the bHLH region, leads to stimulation of DNA binding by SCL, whereas acetylation by p300 has no effect. Acetylation of SCL increases during MEL cell differentiation, paralleling PCAF protein levels, suggesting that in vivo acetylation is mediated by PCAF. It is interesting that acetylation by PCAF also leads to reduced affinity of SCL for the corepressor molecule Sin3A, which associates with histone deacetylases, and might mediate the repressive function of SCL in certain promoter contexts [129
]. SCL binding to Sin3A and PCAF occurs in undifferentiated and differentiated cells, respectively. Thus, acetylation of SCL is highly regulated during erythroid-cell differentiation and might contribute to a switch from repressing to activating function. This example further illustrates the importance of balancing acetylation and deacetylation during gene expression and cellular differentiation.
|
MECHANISMS OF CBP/p300 FUNCTION
|
|---|
Communication with the basal transcription machinery
Although the repertoire of CBP/p300 activities is broad, and their role(s) must be evaluated in the context of a given gene, certain general and recurring functions can be discussed. Shortly after its initial discovery, it was found that CBP binds to the general transcription factor TFIIB [130
]. Additional studies showed direct or indirect contacts of CBP/p300 with TBP and RNA polymerase II (for references, see ref. [7
]). Together, these findings suggest that CBP/p300 might link transcription factors with components of the basal transcription machinery, thereby establishing and/or maintaining transcription-preinitiation complex formation (Fig. 3A
). This bridging function, which is similar to that proposed for numerous other transcription factors, implies that CBP/p300 act in a stoichiometric fashion on some promoters.
View larger version (34K):
[in this window]
[in a new window]
|
Figure 3. Proposed models of CBP function. (A) Tissue-specific transcription factors bound to erythroid-regulatory elements synergize in recruiting CBP, leading to contacts with the basal transcription machinery. GTFs, General transcription factors such as TFIIB and TBP; pol II, RNase polymerase II. (B) Recruitment of an acetyltransferase complex containing CBP, PCAF, and the nuclear hormone receptor coactivator SRC-1 leads to acetylation of histones with resulting changes in chromatin structure. In addition, acetylation of transcription factors regulates their activities. ac, Acetyl residues.
|
|
Acetylation of histone and nonhistone nuclear proteins
In addition to their proposed stoichiometric action, CBP/p300 function as enzymes, as reflected in their ability to catalyze acetylation of all four core histones [14
, 15
]. Furthermore, CBP/p300 bind to proteins with acetyltransferase activity, including PCAF [16
], GCN5 [17
], and certain nuclear hormone-receptor coactivators [131
132
133
134
], leading to the formation of a large acetyltransferase complex with broad substrate specificity. Recruitment of acetyltransferases by transcriptional activators leads to local increases in histone acetylation, which is believed to aid in opening up chromatin (Fig. 3B)
. Fusion of the acetyltransferase domain of CBP to the DNA-binding domain of GAL4 (GAL4-CBP-AT) leads to transcriptional activation of a GAL4-dependent reporter gene, suggesting that histone acetylation is sufficient for transcriptional activation [135
]. However, activation by GAL4-CBP-AT varied between reporter genes, suggesting that acetyltransferase activity depends on promoter context/architecture for proper function or that additional activation functions are required for gene expression. Also, it remains possible that activation by GAL4-CBP-AT results from acetylation of components of the basal transcription machinery [136
, 137
].
In general, acetylation of transcription factors (Fig. 3B)
can alter their activities at various levels, including DNA binding, transcriptional activity, interactions with other proteins, nuclear transport, and protein turnover. Examples for some of these activities are provided by the acetylation of GATA-1, NF-E2, EKLF, SCL, and certain nonhematopoietic transcription factors (see above). Thus, the effects of protein acetylation parallel those observed with protein phosphorylation (for review, see ref. [138
]). Acetylation can also inhibit transcription-factor activity. For example, in some cases, protein acetylation leads to disassembly of higher-order transcription factor/cofactor complexes, thereby imposing a temporal constraint on transcriptional activation (for examples, see refs. [139
, 140
]). Requirement for acetylation functions of CBP/p300 and PCAF varies between transcription factors. For example, although stimulation of the retinoic acid receptor (RAR) requires acetyltransferase activity of PCAF but not CBP, the converse is the case for CREB-mediated transcriptional activation [141
].
Regulation of CBP/p300 activities
Several recent studies suggest that the enzymatic activities of CBP/p300 and PCAF can be modulated by protein contacts. The first example for this was provided by the viral oncoprotein E1A, which can inhibit CBP/p300 and PCAF acetyltransferase activities at high concentrations [22
23
24
]. In addition, some cellular proteins with inhibitory activities have been identified, including Twist and the E1A-like inhibitor of differentiation (EID-1), both of which inhibit myogenic differentiation [22
, 142
, 143
]. These findings link inhibition of histone acetylation to inhibition of cellular differentiation. In erythroid cells, E1A inhibits differentiation, which led to the speculation that cellular oncoproteins that block erythroid maturation might also target CBP/p300 acetyltransferase activity [72
]. Indeed, the Ets family oncoprotein PU.1, which can transform erythroid precursor cells and inhibit their differentiation, is a potent inhibitor of CBP/p300 acetyltransferase activity in vitro and in vivo (unpublished results). Conversely, it can be speculated that CBP/p300 activities might be stimulated by interactions with transcription factors that promote cellular differentiation. Consistent with this possibility, NF-E2 has been shown to augment CBP acetyltransferase activity in vitro when assayed on nucleosomal histones [25
]. In addition, acetylation of free histones by CBP is stimulated by the hepatocyte nuclear factor-1
(HNF-1) in vitro and in vivo [144
]. Thus, CBP/p300 are engaged in a controlled, bidirectional cross-talk with transcriptional regulators.
A clinically relevant example illustrating the importance of a regulated equilibrium between acetylation and deacetylation comes from studies of chromosomal translocations associated with acute promyelocytic leukemia (APL) that fuse the PLZF or PML gene to the RAR gene. In the absence of RA, the normal RAR and the PML-RAR and PLZF-RAR fusion proteins bind a transcriptional repressor complex with histone-deacetylase activity. When exposed to physiological levels of RA, the normal RAR sheds the deacetylase complex, followed by association with an acetyltransferase-containing complex that includes CBP/p300. In contrast, PML-RAR has a higher affinity for the deacetylase complex, requiring pharmacological doses of RA for release, and PLZF-RAR is RA-resistant [145
146
147
]. The response of leukemic cells to RA treatment correlates inversely with the affinity of their translocation fusion proteins for the repressor complex. Thus, APL patients with the PML-RAR translocation undergo remission upon treatment with pharmacological doses of RA, and patients with the PLZF-RAR translocation do not.
Further regulation of CBP/p300 activities occurs via post-translational modifications, as exemplified by the observation that phosphorylation of CBP by cyclin-dependent kinases and mitogen-activated protein kinases stimulates enzymatic activity [27
, 28
]. CBP/p300 are subject to additional regulation by phosphorylation in response to various cellular-signaling pathways (reviewed in ref. [2
]). It is noteworthy that CBP/p300 and PCAF undergo efficient autoacetylation, which can be observed in virtually all acetylation reactions. However, the sites of acetylation are not fully mapped, and the biological significance of autoacetylation remains unresolved.
CBP/p300 as mediators of transcriptional synergy and antagonism
Several observations relating to communication between transcription factors have been linked to CBP/p300 function. For example, transcriptional synergy between nuclear factors bound to the same regulatory region has been proposed to result from cooperative assembly of a multicomponent complex (called the enhanceosome) in which multiple activation domains contribute to CBP/p300 recruitment (for example, see ref. [148
]). Conversely, mutual inhibition of certain transcription factors can result from competition for limiting amounts of CBP/p300. This model has been proposed to account for inhibition between AP-1 and nuclear-hormone receptors and is supported by the finding that overexpression of CBP/p300 alleviates mutual inhibition [132
]. Finally, inhibition between nuclear factors has been suggested to result from direct competition for the same binding site in CBP/p300. This mechanism has been invoked to explain the antagonistic effects between GATA-1 and c-Myb during erythroid gene expression [149
]. In summary, CBP/p300 coordinate transcriptional regulation by mediating communication between transcription factors, by regulating transcription-factor activity, and by translating cellular signals into a transcriptional response.
|
SUMMARY AND PERSPECTIVE
|
|---|
CBP/p300 have been linked at a biochemical or functional level with a daunting number of transcription factors, some of which play opposing roles in cellular functions. For example, although CBP/p300 cooperate with tumor-suppressor proteins, such as p53, they also stimulate the activities of oncoproteins such as SCL. Moreover, CBP/p300 can enhance and inhibit transcription-factor activities. Thus, the phenotypes associated with loss of CBP/p300 function in mammalian organisms likely reflect global changes in gene-expression patterns that result from direct and secondary effects of CBP/p300 on transcription factor function. Perhaps the biggest challenge will be to link a given cellular function directly to a specific set of CBP/p300-regulated transcription factors. Although it is tempting to suggest that the tumor-suppressor function of CBP/p300 is mediated through interaction with p53, it is likely that additional pathways are involved that modulate cell growth and viability. This point is illustrated by the observation that loss of p53 is associated with a much broader spectrum of tumors than loss of CBP function. One way to tackle this problem is to use organisms that lend themselves to efficient genetic analysis. In this regard, several studies in Drosophila confirmed, by genetic means, a link between CBP and specific transcription factors in certain developmental pathways (for review, see ref. [3
]). In mammalian cells, the function of CBP/p300 must be studied individually at any given promoter under defined conditions of cell-cycle progression and differentiation. Domain analysis of CBP/p300 might be best performed using the labor-intensive approach of knocking CBP or p300 mutations into their respective native loci. This will help elucidate the biological roles of various domains, including the protein-docking sites, the bromodomain, and the acetyltransferase domain in vivo at physiological expression levels. These studies can be complemented by micorarray analysis comparing gene expression profiles between cells expressing various forms of CBP or p300. In vivo studies of CBP/p300 benefit greatly from chromatin immunoprecipitation experiments, which are useful for determining whether CBP/p300 associate with a given promoter in vivo. It is equally important to measure the duration of CBP/p300 recruitment to regulatory regions in vivo. For example ligand-induced activation of nuclear hormone receptors was found to trigger a transient interaction of CBP and p300 with hormone-inducible genes in vivo [139
, 150
]. Although it will take tremendous efforts to dissect the complexity surrounding CBP/p300-regulated pathways, the rewards will include deep insight into transcriptional control. In addition, they will likely lead to the identification of novel drug targets for treatment of disorders affecting the hematopoietic system and other cell lineages. Because CBP/p300 are enzymes, compounds regulating their enzymatic activity and substrate specificity might be useful in modulating cell growth and differentiation. Inhibitors of deacetylases are already being explored for treatment of leukemias and other malignancies (for review, see refs. [151
, 152
]), and it is likely that inhibitors of acetyltransferases might find use for some medical conditions as well.
|
ACKNOWLEDGEMENTS
|
|---|
Work in my laboratory was supported by grants from the National Institutes of Health (1RO1 DK54937). I apologize to colleagues whose work could not be included because of space limitations. I am grateful for critical comments by Margaret Chou, Richard Eckner, and Mitch Weiss.
Received October 30, 2001;
revised December 12, 2001;
accepted December 17, 2001.
|
REFERENCES
|
|---|
-
McManus, K. J., Hendzel, M. J. (2001) CBP, a transcriptional coactivator and acetyltransferase Biochem. Cell Biol. 79,253-266[Medline]
-
Vo, N., Goodman, R. H. (2001) CREB-binding protein and p300 in transcriptional regulation J. Biol. Chem. 276,13505-13508[Free Full Text]
-
Goodman, R. H., Smolik, S. (2000) CBP/p300 in cell growth, transformation, and development Genes Dev 14,1553-1577[Free Full Text]
-
Shikama, N., Lyon, J., LaThangue, N. B. (1997) The p300/CBP family: integrating signals with transcription factors and chromatin Trends Cell Biol 7,230-236
-
Giles, R. H., Peters, D. J. M., Breuning, M. H. (1998) Conjunction dysfunction: CBP/p300 in human disease Trends Genet 14,178-183[Medline]
-
Eckner, R. (1996) p300 and CBP as transcriptional regulators and targets of oncogenic events Biol. Chem. 377,685-688[Medline]
-
Blobel, G. A. (2000) CBP/p300: molecular integrators of hematopoietic transcription Blood 95,745-755[Free Full Text]
-
Chrivia, J. C., Kwok, R. P. S., Lamb, N., Hagiwara, M., Montminy, M. R., Goodman, R. G. (1993) Phosphorylated CREB binds specifically to the nuclear protein CBP Nature (London) 365,855-859[Medline]
-
Eckner, R., Ewen, M. E., Newsome, D., Gerdes, M., DeCaprio, J. A., Lawrence, J. B., Livingston, D. M. (1994) Molecular cloning and functional analysis of the adenovirus E1A-associated 300-kd protein (p300) reveals a protein with properties of a transcriptional adaptor Genes Dev 8,869-884[Abstract/Free Full Text]
-
Moran, E. (1993) DNA tumor transforming proteins and the cell cycle Curr. Opin. Genet. Dev. 3,63-70[Medline]
-
Zanger, K., Radovick, S., Wondisford, F. E. (2001) CREB binding protein recruitment to the transcription complex requires growth factor-dependent phosphorylation of its GF box Mol. Cell 7,551-558[Medline]
-
Soutoglou, E., Katrakili, N., Talianidis, I. (2000) Acetylation regulates transcription factor activity at multiple levels Mol. Cell 5,745-751[Medline]
-
Polesskaya, A., Naguibneva, I., Duquet, A., Bengal, E., Robin, P., Harel-Bellan, A. (2001) Interaction between acetylated MyoD and the bromodomain of CBP and/or p300 Mol. Cell. Biol. 21,5312-5320[Abstract/Free Full Text]
-
Bannister, A. J., Kouzarides, T. (1996) The CBP co-activator is a histone acetyltransferase Nature (London) 384,641-643[Medline]
-
Ogryzko, V. V., Schiltz, L. R., Russanova, V., Howard, B. H., Nakatani, Y. (1996) The transcriptional coactivators p300 and CBP are histone acetyltransferases Cell 87,953-959[Medline]
-
Yang, X-J., Ogryzko, V. V., Nishikawa, J., Howard, B. H., Nakatani, Y. (1996) A p300/CBP-associated factor that competes with the adenoviral oncoprotein E1A Nature (London) 382,319-324[Medline]
-
Xu, W., Edmondson, D. G., Roth, S. Y. (1998) Mammalian GCN5 and P/CAF acetyltransferases have homologous amino- terminal domains important for recognition of nucleosomal substrates Mol. Cell. Biol. 18,5659-5669[Abstract/Free Full Text]
-
Torchia, J., Glass, C., Rosenfeld, M. G. (1998) Co-activators and co-repressors in the integration of transcriptional responses Curr. Opin. Cell Biol. 10,373-383[Medline]
-
Kuo, M. H., Allis, C. D. (1998) Roles of histone acetyltransferases and deacetylases in gene regulation Bioessays 20,615-626[Medline]
-
Lundblad, J. R., Kwok, R. P., Laurance, M. E., Harter, M. L., Goodman, R. H. (1995) Adenoviral E1A-associated protein p300 as a functional homologue of the transcriptional co-activator CBP Nature 374,85-88[Medline]
-
Arany, Z., Newsome, D., Oldread, E., Livingston, D. M., Eckner, R. (1995) A family of transcriptional adaptor proteins targeted by the E1A oncoprotein Nature 374,81-84[Medline]
-
Hamamori, Y., Sartorelli, V., Ogryzko, V., Puri, P. L., Wu, H. Y., Wang, J. Y., Nakatani, Y., Kedes, L. (1999) Regulation of histone acetyltransferases p300 and PCAF by the bHLH protein twist and adenoviral oncoprotein E1A Cell 96,405-413[Medline]
-
Chakravarti, D., Ogryzko, V., Kao, H. Y., Nash, A., Chen, H., Nakatani, Y., Evans, R. M. (1999) A viral mechanism for inhibition of p300 and PCAF acetyltransferase activity Cell 96,393-403[Medline]
-
Perissi, V., Dasen, J. S., Kurokawa, R., Wang, Z., Korzus, E., Rose, D. W., Glass, C. K., Rosenfeld, M. G. (1999) Factor-specific modulation of CREB-binding protein acetyltransferase activity Proc. Natl. Acad. Sci. USA 96,3652-3657[Abstract/Free Full Text]
-
Chen, C. J., Deng, Z., Kim, A. Y., Blobel, G. A., Lieberman, P. M. (2001) Stimulation of CREB binding protein nucleosomal histone acetyltransferase activity by a class of transcriptional activators Mol. Cell. Biol. 21,476-487[Abstract/Free Full Text]
-
Zhang, Q., Vo, N., Goodman, R. H. (2000) Histone binding protein RbAp48 interacts with a complex of CREB binding protein and phosphorylated CREB Mol. Cell. Biol. 20,4970-4978[Abstract/Free Full Text]
-
Ait-Si-Ali, S., Ramirez, S., Barre, F. X., Dkhissi, F., Magnaghi-Jaulin, L., Girault, J. A., Robin, P., Knibiehler, M., Pritchard, L. L., Ducommun, B., Trouche, D., Harel-Bellan, A. (1998) Histone acetyltransferase activity of CBP is controlled by cycle-dependent kinases and oncoprotein E1A Nature (London) 396,184-186[Medline]
-
Ait-Si-Ali, S., Carlisi, D., Ramirez, S., Upegui-Gonzalez, L. C., Duquet, A., Robin, P., Rudkin, B., Harel-Bellan, A., Trouche, D. (1999) Phosphorylation by p44 MAP Kinase/ERK1 stimulates CBP histone acetyl transferase activity in vitro Biochem. Biophys. Res. Commun. 262,157-162[Medline]
-
O’Connor, M. J., Zimmermann, H., Nielsen, S., Bernard, H. U., Kouzarides, T. (1999) Characterization of an E1A-CBP interaction defines a novel transcriptional adapter motif (TRAM) in CBP/p300 J. Virol. 73,3574-3581[Abstract/Free Full Text]
-
Horn, P. J., Peterson, C. L. (2001) The bromodomain: a regulator of ATP-dependent chromatin remodeling? Front Biosci 6,D1019-D1023[Medline]
-
Dhalluin, C., Carlson, J. E., Zeng, L., He, C., Aggarwal, A. K., Zhou, M-M. (1999) Structure and ligand of a histone acetyltransferase bromodomain Nature (London) 399,491-496[Medline]
-
Kraus, W. L., Manning, E. T., Kadonaga, J. T. (1999) Biochemical analysis of distinct activation functions in p300 that enhance transcription initiation with chromatin templates Mol. Cell. Biol. 19,8123-8135[Abstract/Free Full Text]
-
Petrij, F., Giles, R. H., Dauwerse, H. G., Saris, J. J., Hennekam, R. C., Masuno, M., Tommerup, N., van Ommen, G. B., Goodman, R. H., Peters, D. J. M., Breuning, M. H. (1995) Rubinstein-Taybi syndrome caused by mutations in the transcriptional co-activ
TOP
INTRODUCTION
A BRIEF HISTORY
