DNA methyltransferase Dnmt1 associates with histone deacetylase activity.

The DNA methyltransferase Dnmt1 is responsible for cytosine methylation in mammals and has a role in gene silencing. DNA methylation represses genes partly by recruitment of the methyl-CpG-binding protein MeCP2, which in turn recruits a histone deacetylase activity. Here we show that Dnmt1 is itself associated with histone deacetylase activity in vivo. Consistent with this association, we find that one of the known histone deacetylases, HDAC1, has the ability to bind Dnmt1 and can purify methyltransferase activity from nuclear extracts. We have identified a transcriptional repression domain in Dnmt1 that functions, at least partly, by recruiting histone deacetylase activity and shows homology to the repressor domain of the trithorax-related protein HRX (also known as MLL and ALL-1). Our data show a more direct connection between DNA methylation and histone deacetylation than was previously considered. We suggest that the process of DNA methylation, mediated by Dnmt1, may depend on or generate an altered chromatin state via histone deacetylase activity.

Mutations truncating the EP300 acetylase in human cancers.

The EP300 protein is a histone acetyltransferase that regulates transcription via chromatin remodelling and is important in the processes of cell proliferation and differentiation. EP300 acetylation of TP53 in response to DNA damage regulates its DNA-binding and transcription functions. A role for EP300 in cancer has been implied by the fact that it is targeted by viral oncoproteins, it is fused to MLL in Leukaemia and two missense sequence alterations in EP300 were identified in epithelial malignancies. Nevertheless, direct demonstration of the role of EP300 in tumorigenesis by inactivating mutations in human cancers has been lacking. Here we describe EP300 mutations, which predict a truncated protein, in 6(3%) of 193 epithelial cancers analysed. Of these six mutations, two were in primary tumours (a colorectal cancer and a breast cancer) and four were in cancer cell lines (colorectal, breast and pancreatic). In addition, we identified a somatic in-frame insertion in a primary breast cancer and missense alterations in a primary colorectal cancer and two cell lines (breast and pancreatic). Inactivation of the second allele was demonstrated in five of six cases with truncating mutations and in two other cases. Our data show that EP300 is mutated in epithelial cancers and provide the first evidence that it behaves as a classical tumour-suppressor gene.

Functions of pRb and p53: what's the connection?

The pRb and p53 proteins have tumour suppressor functions and both are regulators of transcription. Their mechanisms of transcriptional regulation are unrelated in many ways--in contrast, it would appear, to their biological functions. This review highlights their biological connections in the light of recent advances in understanding the mechanisms of pRb and p53 function.

Transcriptional regulation by the retinoblastoma protein.

The retinoblastoma protein (RB) plays a key role in the control of cell proliferation and mediates the terminal differentiation of certain cell types. Increasing evidence suggests that RB functions by contacting and modifying the behaviour of transcription factors. RB can form complexes with E2F and MyoD in vivo, and complexes with a number of other transcription factors have also been demonstrated in vitro. The interaction of regulatory transcription factors with RB may be explained by sequence similarity between RB and two general transcription factors: TBP and TFIIB. Here I review the evidence for a role of RB in the regulation of transcription and highlight some of the likely mechanisms of RB function.

Retinoblastoma protein meets chromatin.

The retinoblastoma (RB) protein exerts its tumour-suppressor function by repressing the transcription of cellular genes required for DNA replication and cell division. Recent investigations into the mechanism of RB repression have revealed that RB can regulate transcription by effecting changes in chromatin structure. These findings point towards a link between chromatin regulation and cancer.

Viral oncoproteins target the DNA methyltransferases.

Small DNA tumour viruses have evolved a number of mechanisms to drive nondividing cells into S phase. Virally encoded oncoproteins such as adenovirus E1A and human papillomavirus (HPV) E7 can bind an array of cellular proteins to override proliferation arrest. The DNA methyltransferase Dnmt1 is the major mammalian enzyme responsible for maintaining CpG methylation patterns in the cell following replication. One of the hallmarks of tumour cells is disrupted DNA methylation patterns, highlighting the importance of the proper regulation of DNA methyltransferases in normal cell proliferation. Here, we show that adenovirus 5 E1A and HPV-16 E7 associate in vitro and in vivo with the DNA methyltransferase Dnmt1. Consistent with this interaction, we find that E1A and E7 can purify DNA methyltransferase activity from nuclear extracts. These associations are direct and mediated by the extreme N-terminus of E1A and the CR3 zinc-finger domain of E7. Furthermore, we find that a point mutant at leucine 20 of E1A, a residue known to be critical for its transformation functions, is unable to bind Dnmt1 and DNA methyltransferase activity. Finally, both E1A and E7 can stimulate the methyltransferase activity of Dnmt1 in vitro. Our results provide the first indication that viral oncoproteins bind and regulate Dnmt1 enzymatic activity. These observations open up the possibility that this association may be used to control cellular proliferation pathways and suggest a new mechanism by which small DNA tumour viruses can steer cells through the cell cycle.

p300 is required for orderly G1/S transition in human cancer cells.

The role of the transcriptional coactivator p300 in cell cycle control has not been analysed in detail due to the lack of appropriate experimental systems. We have now examined cell cycle progression of p300-deficient cancer cell lines, where p300 was disrupted either by gene targeting (p300(-) cells) or knocked down using RNAi. Despite significant proliferation defects under normal growth conditions, p300-deficient cells progressed rapidly through G1 with premature S-phase entry. Accelerated G1/S transition was associated with early retinoblastoma (RB) hyperphosphorylation and activation of E2F targets. The p300-acetylase activity was dispensable since expression of a HAT-deficient p300 mutant reversed these changes. Co-immunoprecipitation showed p300/RB interaction occurs in vivo during G1, and this interaction has two peaks: in early G1 with unphosphorylated RB and in late G1 with phosphorylated RB. In vitro kinase assays showed that p300 directly inhibits cdk6-mediated RB phosphorylation, suggesting p300 acts in early G1 to prevent RB hyperphosphorylation and delay premature S-phase entry. Paradoxically, continued cycling of p300(-) cells despite prolonged serum depletion was observed, and this occurred in association with persistent RB hyperphosphorylation. Altogether, these results suggest that p300 has an important role in G1/S control, possibly by modulating RB phosphorylation.

Histone acetylases and deacetylases in cell proliferation.

There are several enzymes, acetylases and deacetylases, that can regulate transcription by modifying the acetylation state of histones or other promoter-bound transcription factors. Some of these enzymes are present in multisubunit complexes. Recent efforts to understand the biological role of these enzymes reveals their involvement in cell-cycle regulation and differentiation. Furthermore, accumulating evidence suggests that deregulation of acetylase and deacetylase activity plays a causative role in the generation of cancer.

Acetylation of importin-alpha nuclear import factors by CBP/p300.

Histone acetylases were originally identified because of their ability to acetylate histone substrates [1] [2] [3]. Acetylases can also target other proteins such as transcription factors [4] [5] [6] [7]. We asked whether the acetylase CREB-binding protein (CBP) could acetylate proteins not directly involved in transcription. A large panel of proteins, involved in a variety of cellular processes, were tested as substrates for recombinant CBP. This screen identified two proteins involved in nuclear import, Rch1 (human importin-alpha) and importin-alpha7, as targets for CBP. The acetylation site within Rch1 was mapped to a single residue, Lys22. By comparing the context of Lys22 with the sequences of other known substrates of CBP and the closely related acetylase p300, we identified G/SK (in the single-letter amino acid code) as a consensus acetylation motif. Mutagenesis of the glycine, as well as the lysine, severely impaired Rch1 acetylation, supporting the view that GK is part of a recognition motif for acetylation by CBP/p300. Using an antibody raised against an acetylated Rch1 peptide, we show that Rch1 was acetylated at Lys22 in vivo and that CBP or p300 could mediate this reaction. Lys22 lies within the binding site for a second nuclear import factor, importin-beta. Acetylation of Lys22 promoted interaction with importin-beta in vitro. Collectively, these results demonstrate that acetylation is not unique to proteins involved in transcription. Acetylation may regulate a variety of biological processes, including nuclear import.

Temporal recruitment of the mSin3A-histone deacetylase corepressor complex to the ETS domain transcription factor Elk-1.

The transcriptional status of eukaryotic genes is determined by a balance between activation and repression mechanisms. The nuclear hormone receptors represent classical examples of transcription factors that can regulate this balance by recruiting corepressor and coactivator complexes in a ligand-dependent manner. Here, we demonstrate that the equilibrium between activation and repression via a single transcription factor, Elk-1, is altered following activation of the Erk mitogen-activated protein kinase cascade. In addition to its C-terminal transcriptional activation domain, Elk-1 contains an N-terminal transcriptional repression domain that can recruit the mSin3A-histone deacetylase 1 corepressor complex. Recruitment of this corepressor is enhanced in response to activation of the Erk pathway in vivo, and this recruitment correlates kinetically with the shutoff of one of its target promoters, c-fos. Elk-1 therefore undergoes temporal activator-repressor switching and contributes to both the activation and repression of target genes following growth factor stimulation.

c-Fos-induced activation of a TATA-box-containing promoter involves direct contact with TATA-box-binding protein.

Transcriptional activation in eukaryotes involves protein-protein interactions between regulatory transcription factors and components of the basal transcription machinery. Here we show that c-Fos, but not a related protein, Fra-1, can bind the TATA-box-binding protein (TBP) both in vitro and in vivo and that c-Fos can also interact with the transcription factor IID complex. High-affinity binding to TBP requires c-Fos activation modules which cooperate to activate transcription. One of these activation modules contains a TBP-binding motif (TBM) which was identified through its homology to TBP-binding viral activators. This motif is required for transcriptional activation, as well as TBP binding. Domain swap experiments indicate that a domain containing the TBM can confer TBP binding on Fra-1 both in vitro and in vivo. In vivo activation experiments indicate that a GAL4-Fos fusion can activate a promoter bearing a GAL4 site linked to a TATA box but that this activity does not occur at high concentrations of GAL4-Fos. This inhibition (squelching) of c-Fos activity is relieved by the presence of excess TBP, indicating that TBP is a direct functional target of c-Fos. Removing the TBM from c-Fos severely abrogates activation of a promoter containing a TATA box but does not affect activation of a promoter driven only by an initiator element. Collectively, these results suggest that c-Fos is able to activate via two distinct mechanisms, only one of which requires contact with TBP. Since TBP binding is not exhibited by Fra-1, TBP-mediated activation may be one characteristic that discriminates the function of Fos-related proteins.

Differential localization of HDAC4 orchestrates muscle differentiation.

The class II histone deacetylases HDAC4 and HDAC5 interact specifically with the myogenic MEF2 transcription factor and repress its activity. Here we show that HDAC4 is cytoplasmic during myoblast differentiation, but relocates to the nucleus once fusion has occurred. Inappropriate nuclear entry of HDAC4 following overexpression suppresses the myogenic programme as well as MEF2-dependent transcription. Activation of the Ca(2+)/calmodulin signalling pathway via constitutively active CaMKIV prevents nuclear entry of HDAC4 and HDAC4-mediated inhibition of differentiation. Consistent with a role of phosphorylation in HDAC4 cytoplasmic localisation, HDAC4 binds to 14-3-3 proteins in a phosphorylation-dependent manner. Together these data establish a role for HDAC4 in muscle differentiation. Recently, HDAC5 has also been implicated in muscle differentiation. However, despite the functional similarities of HDAC4 and HDAC5, their intracellular localisations are opposed, suggesting a distinct role for these enzymes during muscle differentiation.

Transcriptional repression by the insulator protein CTCF involves histone deacetylases.

The highly conserved zinc-finger protein, CTCF, is a candidate tumor suppressor protein that binds to highly divergent DNA sequences. CTCF has been connected to multiple functions in chromatin organization and gene regulation including chromatin insulator activity and transcriptional enhancement and silencing. Here we show that CTCF harbors several autonomous repression domains. One of these domains, the zinc-finger cluster, silences transcription in all cell types tested and binds directly to the co-repressor SIN3A. Two distinct regions of SIN3A, the PAH3 domain and the extreme C-terminal region, bind independently to this zinc-finger cluster. Analysis of nuclear extract from HeLa cells revealed that CTCF is also capable of retaining functional histone deacetylase activity. Furthermore, the ability of regions of CTCF to retain deacetylase activity correlates with the ability to bind to SIN3A and to repress gene activity. We suggest that CTCF driven repression is mediated in part by the recruitment of histone deacetylase activity by SIN3A.

BRCA2 associates with acetyltransferase activity when bound to P/CAF.

Predisposition to hereditary breast cancer has been attributed in part to inherited mutations in the BRCA2 gene. The large protein it encodes is still poorly characterized with respect to functions. We have previously shown that BRCA2 has transcriptional activation potential conferred by its amino-terminal third exon. Here, we show that BRCA2 interacts with a transcriptional co-activator protein, P/CAF, which possesses histone acetyltransferase activity. The interaction with P/CAF is demonstrated in vitro as well as in vivo and is shown to be mediated by residues 290-453 of BRCA2. Consistent with the binding to an acetyltransferase, BRCA2 is shown to associate with acetyltransferase activity in nuclear extracts. Contrary to a recent report, we find no evidence in support of an intrinsic HAT activity in BRCA2 amino-terminus. Our results further substantiate the notion that BRCA2 has transcriptional activation function and suggest that one mechanism by which BRCA2 regulates transcription may be through the recruitment of histone-modifying activity of the P/CAF co-activator.

In vitro DNA binding activity of Fos/Jun and BZLF1 but not C/EBP is affected by redox changes.

The leucine zipper family of proteins have a DNA binding domain composed of a leucine zipper dimerisation interface and a basic DNA binding structure. We show here that redox changes affect the in vitro DNA binding ability of a select subset of leucine zipper proteins. The bacterially expressed DNA binding domains of Fos/Jun and BZLF1 are unable to bind DNA under non-reducing conditions whereas binding of the C/EBP DNA binding domain is unaffected. Sensitivity to redox state is due to the presence of a conserved cysteine residue in the basic DNA binding motif of Fos, Jun and BZLF1 but not C/EBP. Under non-reducing conditions an intermolecular disulphide bridge is formed between the cysteine residues of each basic motif within a dimer, which prevents DNA binding. We show that oxidation of these C residues can be achieved enzymatically, using glutathione peroxidase, and that DNA binding protects them from oxidation. These data raise the possibility that intracellular changes in the redox state may differentially regulate the activity of leucine zipper family members. In addition the loss of DNA binding activity under non-reducing conditions has implications for the purification methods used to isolate proteins of the leucine zipper family for structural analysis.

The putative tumour suppressor Fus-2 is an N-acetyltransferase.

Acetyltransferases are essential enzymes for a wide variety of cellular processes and mutations in acetyltransferase genes have been associated with the development of certain cancers. For this reason, we conducted a computerized sequence homology search for novel acetyltransferases. Here, we show that the putative tumour suppressor protein Fus-2 has homology to the catalytic domain of acetyltransferases. We demonstrate that Fus-2 can acetylate the N-terminus of proteins using a ping-pong mechanism and that it has a specificity for substrates. Consistent with other N-acetyltransferases, Fus-2 localizes to the cytoplasm, as shown by GFP-tag experiments. Since the Fus-2 gene maps to the chromosomal region 3p21.3, which contains at least one tumour suppressor gene, the N-acetyltransferase functions of Fus-2 may be relevant to its potential role in cancer.

The BRCA2 activation domain associates with and is phosphorylated by a cellular protein kinase.

A substantial proportion of familial breast cancers have mutations within the BRCA2 gene. The product of this gene has been implicated in DNA repair and in the regulation of transcription. We have previously identified at the amino-terminus of BRCA2 a transcriptional activation domain whose importance is highlighted by the presence of predisposing mutations and in-frame deletions in breast cancer families. This activation domain shows sequence similarity to a region of c-Jun which has been defined as a binding site for the c-Jun N-terminal kinase. Here, we show that the analogous region in BRCA2 is also a binding site for a cellular kinase, although this kinase is distinct from JNK. The BRCA2 associated enzyme is able to phosphorylate residues within the BRCA2 activation domain. Consistent with this observation, we find that the activation domain of BRCA2 is phosphorylated in vivo. Our results indicate that the BRCA2 activation domain possesses a binding site for a kinase that may regulate BRCA2 activity by phosphorylation.

The HMG-box transcription factor HBP1 is targeted by the pocket proteins and E1A.

A yeast two-hybrid screen has identified HBP1 as a transcription factor capable of interacting with the pocket protein family. We show that HBP1 can interact with one of these, RB, both in vitro and in mammalian cells. Two distinct RB binding sites are present within HBP1--a high affinity binding site, mediated by an LXCXE motif and a separate low affinity binding site present within an activation domain. GAL4-fusion experiments indicate that HBP1 contains a masked activation domain. Deletion of two independent N- and C-terminal inhibitor domains unmasks an activation domain which is 100-fold more active than the full length protein. The released activation capacity is repressed by RB, p130 and p107. In addition, E1A can repress the activity of HBP1 via conserved region 1 sequences in a manner independent of the CBP co-activator. We show by stable expression in NIH3T3 cells that HBP1 has the capacity to induce morphological transformation of cells in culture.

Stimulation of c-Jun activity by CBP: c-Jun residues Ser63/73 are required for CBP induced stimulation in vivo and CBP binding in vitro.

The CBP protein mediates PKA induced transcription by binding to the PKA phosphorylated activation domain of CREB. Here we show that CBP also stimulates the activity of both c-Jun and v-Jun in vivo. The CREB binding domain of CBP is sufficient to contact to c-Jun in vitro. When this domain of CBP is linked to the activation domain of VP16 and expressed in vivo it stimulates c-Jun dependent transcription. Deletion analysis of c-Jun indicate that the CBP binding site is within the N-terminal activation domain. Loss of binding to CBP in vitro correlates with severely reduced transactivation capacity in vivo. Mutation of Ser63/73 in c-Jun, or the corresponding position in v-Jun (Ser36/46) leads to reduced binding to CBP in vitro and abolishes augmentation of transcription in vivo. These data are consistent with a mechanism by which CBP acts as a co-activator protein for Jun dependent transcription by interacting with the Jun N-terminal activation domain.

The BZLF1 protein of EBV has a coiled coil dimerisation domain without a heptad leucine repeat but with homology to the C/EBP leucine zipper.

The EBV transactivator protein BZLF1 can bind many sites in the EBV genome, most of which have homology to a consensus AP-1 site, the binding site for the fos/jun family of transcription factors. Here we present evidence that BZLF1 can also recognise the binding site for the CCAAT/enhancer binding protein C/EBP and that a BZLF1 binding site within the BZLF1 promoter is recognised by the C/EBP protein. Analysis of the BZLF1 DNA binding domain suggests that the BZLF1 protein binds to DNA as a dimer using sequences adjacent to a basic DNA binding motif. The BZLF1 dimerisation domain does not have a heptad repeat of leucine residues common to leucine zipper proteins but does have characteristics of a coiled coil structure, as judged by site directed mutagenesis. We therefore propose that the dimerisation domain of BZLF1 is structurally related to the coiled coil structure of leucine zippers but lacks the highly conserved leucine repeat. We show that the PZLF1 dimerisation domain has residues in common with the C/EBP leucine zipper and discuss the possible implications of this relationship.