The bromodomain mediates transcriptional intermediary factor 1alpha -nucleosome interactions.

Nuclear histone acetyltransferases, DNA-dependent ATPases, and transcriptional intermediary factors (TIFs) all harbor a distinct structural module known as the bromodomain (BrD). Although the BrD can interact with histones H3 and H4 and their acetylated N-terminal tails in vitro, its function in a chromosomal environment remains elusive. We used the nuclear receptor coregulator TIF1alpha, a protein kinase that associates tightly with euchromatin, to analyze the properties of the BrD in a nucleosomal environment in vitro. Here, we report that TIF1alpha-chromatin association is direct and involves DNA and nucleosome interactions mediated by the BrD. Mutation of the BrD signature peptide, PMDL, abolishes DNA binding and disrupts BrD-nucleosome interactions. Based on our results, we propose that the BrD plays a critical role in vivo by directing transregulators to their cognate location on nucleosomal DNA.

TIF1␣ is an abundant nuclear protein present in embryonic stem cells and mouse embryos (20). Upon cell differentiation at the onset of organogenesis, the expression levels of TIF1␣ decrease dramatically (20,21); TIF1␣ then becomes restricted to the developing central nervous system and to selected cell populations in the adult brain and germ-line tissues (21). In the nucleus, TIF1␣ is tightly bound to euchromatin, preferentially on sites of RNA polymerase II transcription and at the borders between euchromatin and heterochromatin where TIF1␣ may act as a docking molecule for liganded nuclear receptors (20), such as the androgen receptor (22). Thus, TIF1␣ defines a novel class of chromatin-associated TIFs that facilitate access of transregulators to target genes during development and differentiation ("chromatin access model") (20).
Here, we report the molecular properties of the BrD in a nucleosomal environment in vitro. We demonstrate that the BrD is sufficient to mediate direct DNA and nucleosome interactions. Four residues, PMDL, that represent a signature present in helix A of several BrDs, are crucial both for DNA and for nucleosomal binding. Our results reveal a novel DNA-binding activity for the BrD and suggest a new role for this domain as a nucleosomal DNA anchor.

EXPERIMENTAL PROCEDURES
Chromatin Assembly-Drosophila post-blastoderm extracts (S-190) were prepared and chromatin was assembled essentially as presented in Ref. 44 with minor modifications also described in Ref. 45. Chromatin was assembled with 15 mg of S-190 extract, 5 g of calf thymus core histones (Roche), 5 g of mRAR␤2-LacZ (10.2 kb) plasmid (46), and an * This work was supported in part by grants from CNRS, INSERM, Hopitaux Universitaires de Strasbourg, BMS, and a grant from the Fondation pour la Recherche Médicale. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM 1 The abbreviations used are: TIF, transcriptional intermediary factor; BrD, bromodomain; PMSF, phenylmethylsulfonyl fluoride; MNase, micrococcal nuclease; EMSA, electrophoretic mobility-shift assay; PIC, protease inhibitor cocktail. ATP-regenerating system (30 mM creatine phosphate, 3 mM ATP, 4.2 mM MgCl 2 , and 6 g of creatine kinase) in 500 l of final volume supplemented with buffer R (45). The assembly factors in the extract were first charged with core histones on ice for 30 min, and the plasmid and the ATP regenerating system were added and the samples were further incubated for 5 h at 27°C. At the end of the incubation period, chromatin was either analyzed by microccocal nuclease (MNase) (Sigma) digestion (45) or purified for further analysis by sucrose gradient sedimendation as described previously (20,44).
Sucrose Gradient Sedimentation-10.2 ml of 30 -50% linear sucrose gradients were prepared in 10 mM Hepes KOH buffer pH 7.5 containing 1 mM EDTA, 35 mM NaCl, 1 mM dithiothreitol, protease inhibitor cocktail (2.5 g/ml each leupeptin, pepstatin, aprotinin, antipain, chymostatin) (1ϫ PIC), and 1 mM phenylmethylsulfonyl fluoride (PMSF) in Beckman tubes (44) and overlaid with 0.4 ml of the same buffer containing 5% sucrose (20). 0.8 ml of chromatin was applied on the top, and the sample was centrifuged in a SW41 rotor (Beckman) at 26,000 rpm for 16 h at 4°C. At the end of the centrifugation, 1-ml fractions were recovered from the top of the gradient and 50 l were used either directly for immunoblotting or deproteinized for plasmid DNA isolation (47).
Chromatin Preparation by Salt Dialysis-Chromatin reconstruction was performed on pBluescript plasmid (2.9 kbp, Stratagene) using a core histone:DNA molar ratio of 1:1, by step salt dialysis from 2 M to 0.05 M NaCl in 100 mM Tris HCl buffer, pH 8.0, 1 mM EDTA, 1 mM ␤-mercaptoethanol, and 0.05% Nonidet P-40 (buffer A) in dialysis bags presaturated with 25 g/ml bovine serum albumin. The samples were dialyzed in buffer A (above) containing 2, 1.5, 1, 0.5 M NaCl for 2 h and 0.05 M NaCl overnight at 4°C. Chromatin was purified by sucrose gradient sedimentation and prepared for electron microscopy.
Electron Microscopy-Chromatin, prepared either by chromatin assembly using the S190 extract or by salt dialysis, was purified by sucrose gradient sedimentation and dialyzed in 5 mM triethanolamine, 1 mM EDTA, and 1 mM PMSF before it was diluted to an estimated DNA concentration of 1 ng/l. Grids were prepared and observed under a Philips CM12 electron microscope (48). Size measurements for individual nucleosomes were performed using the NIH Image program; particles were circled, and the program was fitted an ellipse to define a long and short axis.
Genomic DNA Fragment Isolation-1.5 ϫ 10 7 P19 nuclei were incubated with 24 units of MNase (Roche) in 200 l of buffer N for 30 min at 37°C as described (20). After centrifugation at 6,000 rpm for 10 min at 4°C, the supernatant (S1 fraction, euchromatin) was collected (20,49), and genomic DNA fragments were purified after Proteinase K treatment (40 g/ml), phenol-chloroform extraction, and ethanol precipitation. DNA fragments were cloned into the EcoRV site of pBluescript and sequenced. One of the clones contained a 158-bp insert with a unique PstI site whereas a 145-bp fragment, including an 8-bp multiple cloning site from pBluescript in the 5Ј-flanking region of the insert, was released by HindIII-PstI digestion (for core nucleosome reconstitution). This fragment was designated M2. M2 sequence (158 bp) is shown below: 5Ј-GAGTTTGCCATTGAAACTTAAAGTAGGGGAAGAAGGGC-TGAGGAGGTGACTCATGGATAAGAGCATTCCTTGCTCTTTGAGA-GTAAATGACAGTTCACAATTGCCTGTAACCCTAACTCCATGGCAT-CTGATCCTTCTGCAGACCTCTGTGGCCTCCT-3Ј. Another clone contained a 173-bp insert and a 189-bp fragment, including an 8-bp multiple cloning site both in the 5Ј-and 3Ј-flanking region of the insert, was released by HindIII-EcoRI digestion (for linker-contained nucleosome reconstitution). This fragment, designated M4.M4 sequence (173 bp) is: 5Ј-CAGCCGGCCTTAGAAAGGCCATCTGATTCTTTGAGTTGCTTGT-GGTCGACGCAGAGTCGCCACCGTTTCTGGTTTCTTTTTTGTCTTA-GTCTCGTGTCCGCTCTTGTAGTGTCTACTGTTTTTCTAGAAATGG-GACAATCGGTGTCCACTCCCCTTTCTCTGAATCTGGAGCA-3Ј. The M2 and M4 sequences have been submitted to Genbank data base with accession numbers AY155467 (M2) and AY155468 (M4).
Chromatin Immunoprecipitation-Proteins and DNA from P19 cells (3 ϫ 10 6 per plate) were crosslinked by the addition of formaldehyde directly to the culture medium to a final concentration of 1% and incubated for 10 min at 37°C. The medium was then aspirated, and the cells were washed twice using ice-cold phosphate-buffered saline in the presence of protease inhibitors as described above. 200 l of SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl pH, and a protease inhibitor mixture) were added to each plate, the cells were scraped off, placed into an Eppendorf tube, and incubated on ice for 10 min before they were sonicated to generate 200-to 1000-bp fragments. After centrifugation, the cleared supernatant was diluted 10-fold with immunoprecipitation buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5%Nonidet P-40, and 1ϫ PIC). The cell lysate was precleared by incubation with 40 l of salmon sperm DNA/protein A beads (Upstate Biotechnology) supplemented with 40 l of either protein A (for polyclonal antibodies) or protein G (for monoclonal antibodies) beads for 2 h at 4°C. At the same time, with anti-acetylated histone H3 polyclonal antibody (Upstate Biotechnology) or no antibody (negative control), the following was pre-absorbed on 80 l of salmon sperm DNA/protein A beads:Protein A beads (1:1 ratio) and anti-TIF1␣ monoclonal antibody (5T1E8) on 80 l of salmon sperm DNA/protein A beads:protein G beads (1:1 ratio) in 500 l of immunoprecipitation buffer for 2 h at room temperature. At the end of the incubation period, the antibody-coupled beads were washed once in immunoprecipitation buffer before they were further incubated with the cleared lysates overnight at 4°C. The immune complexes were precipitated on the beads by centrifugation, washed twice in immunoprecipitation buffer, and eluted in 1% SDS/100 mM sodium carbonate. DNA-protein crosslinks were reversed by heating the samples at 65°C for 4 h, and the DNA was purified by phenolchloroform extraction and ethanol precipitation. Levels of M2 and M4 genomic fragments were determined by PCR using specific sets of oligonucleotides that would generate 130-bp and 160-bp DNA fragments, respectively. The primers used for PCR were: for the M2 fragment, 5Ј-GTAGGGGAAGAAGGGCTGAGGAGGTGACTC-3Ј (primer A, 5Ј 3 3Ј) and 5Ј-CCACAGAGGTCTGCAGAAGGATCAGATGCC-3Ј (primer B, 3Ј 3 5Ј); and for the M4 fragment, 5Ј-TAGAAAGGCCATCT-GATTCTTTGAGTTGCT-3Ј (primer A, 5Ј 3 3Ј) and 5Ј-TCCAGATTCA-GAGAAAGGGGAGTGGACACC-3Ј (primer B, 3Ј 3 5Ј). Donor Nucleosome Preparation and Histone H1 Depletion-1.5 ϫ 10 7 P19 nuclei were incubated with 24 units of MNase (Roche) in 100 l of buffer N for 30 min at 37°C as described (20,49). After centrifugation at 6,000 rpm for 10 min at 4°C, the pellet was resuspended in 100 l of 2 mM EDTA, 1 mM dithiothreitol, 1ϫ PIC, and 1 mM PMSF solution and incubated for 10 min on ice (20). The supernatant (S2 fraction, heterochromatin) was collected by centrifugation for 15 min at 4°C, and the NaCl concentration was adjusted to 0.35 M. To deplete histone H1, an equal volume of 80% slurry of CM Sephadex C-25 (in 0.35 M NaCl) was added and incubated for 2 h at 4°C with rocking. The supernatant was recovered by centrifugation at 10,000 rpm for 5 min at 4°C. H1 depletion was confirmed by immunoblotting using a monoclonal antibody specific for mouse histone H1 (data not shown).
Nucleosome Reconstitution by Octamer Transfer-M2 and M4 DNA fragments were end-labeled with [␣ 32 P]dCTP using Klenow DNA polymerase and incubated in a 1:1 ratio with histone H1-depleted donor nucleosomes in 10 mM Tris-HCl buffer, pH 7.5, 1 M NaCl, 1 mM EDTA, and 1 mM ␤-mercaptoethanol for 1 h. TE buffer (10 mM Tris-HCl, pH 7.5, 1 mM EDTA) was added to adjust the NaCl concentration to 0.8 M. After 1 h of incubation, the NaCl concentration was adjusted to 0.66 M, followed by a 1-h incubation before NaCl was subsequently diluted to 0.2, 0.1, and 0.05 M with successive additions of TE every 10 min. All incubations were performed at room temperature. Nucleosome particles were purified by sedimentation in sucrose as described (20)  For competition experiments, 25 or 100 g/ml poly(dI⅐dC) was included in the reaction. Samples were loaded onto 4% (29:1) native polyacrylamide gels and electrophoresed at room temperature in 0.5ϫ TBE for 1 h at 10 V/cm. Gels were dried and exposed to autoradiography.

RESULTS
The fact that TIF1␣ associates with euchromatin in vivo (20) prompted us to analyze its molecular properties on reconstituted chromatin in vitro. For our study, we used an in vitro chromatin assembly system prepared from Drosophila postblastoderm embryos (S-190 extract) (44,45,47) (Fig. 1A). Micrococcal nuclease (MNase) digestion analysis (Fig. 1E), supercoiling assays (data not shown), and electron microscopy (Fig. 1, A and F) demonstrated that the extract efficiently assembled regularly spaced nucleosomal arrays (50) on various length templates in the presence of exogenously added core histones as previously described (44). Because the S-190 extract appeared to be devoid of any detectable TIF1␣ cross-reacting activity (Fig. 1C, Western Blot), purified recombinant mouse TIF1␣ (Fig. 1C, Coomassie Staining) was included in the assembly reaction to determine whether TIF1␣ can be incorporated efficiently onto chromatin (Fig.  1B). At the end of the chromatin assembly reaction, chromatin (assembled on mRAR␤2-LacZ, (Ref. 46), 10.2 kb) was purified by sucrose gradient sedimentation and the DNA was isolated and analyzed by agarose gel electrophoresis and ethidium bromide staining (44,45,47) (Fig. 1D). From immunoblotting analysis of gradient fractions and by comparison to the sedimentation profile of the recombinant protein, it was evident that TIF1␣ was readily incorporated onto chromatin assembled in vitro (Fig. 1B). A ratio of approximately one TIF1␣ molecule per 3 nucleosomes was estimated. When chromatin generated with or without TIF1␣ was subjected to time-dependent MNase digestion (45), we found that TIF1␣ incorporation had no apparent effect on the regularity of nucleosomal spacing (Fig. 1E). Moreover, analysis of purified chromatin by electron microscopy, where fully assembled chromatin was indeed observed both with and without TIF1␣ (Fig. 1, A and F), indicated that TIF1␣ does not affect the efficiency of the chromatin assembly (Fig. 1, A, E, and F). However, chromatin prepared with TIF1␣ consistently presented a statistically significant enlargement in core nucleosome dimensions (Fig. 1F). Nucleosomes assembled in the presence of TIF1␣ presented on average an increase of 16 -19 Å when compared with core nucleosomes (50). Similar results were obtained when chromatin was packaged either using postblastoderm extracts (Fig. 1, A and D) or by salt dialysis using purified core histones ( Fig. 1D; data not shown), suggesting an interaction between TIF1␣ and core nucleosomes. These results demonstrate that TIF1␣ associates with chromatin in vitro and moreover indicate that no additional factors are required to mediate TIF1␣-chromatin interactions.
In the nucleus, TIF1␣ is tightly bound to euchromatin (20); however, the chromosomal sites that bind TIF1␣ in vivo are still unknown. Thus, to investigate whether TIF1␣ binds to nucleosomal DNA directly, genomic DNA fragments from the euchromatic fraction of the genome, fractionated as described in Refs. 20 and 49, were isolated and tested. Fractions enriched in DNA and proteins from either euchromatin or heterochromatin can be prepared from intact nuclei subjected to mild MNase digestion (20,49). This leads to the initial release of primarily mononucleosomes that contain a subset of the chromosomal DNA (S1 fraction), which is enriched in TIF1␣ (20) and other non-histone chromosomal proteins (49) but is devoid of histone H1 (49) and HP1␣ (45), contains actively transcribed genes (49), and therefore represents euchromatin (20,49). Subsequent hypotonic lysis of the nuclei pellet releases a soluble fraction (S2) that contains transcriptionally inactive polynucleosomes that are enriched in histone H1 (49) and HP1␣ (20) and therefore represent heterochromatin (20,49). Two different genomic DNA fragments (M2 and M4) were isolated from the S1 fraction of P19 nuclei (20), cloned, and sequenced. To verify whether TIF1␣ binds to these sequences in vivo, chromatin immunoprecipitations were performed ( Fig. 2A). Cross-linked chromatin from P19 EC cells was precipitated with TIF1␣ ( Fig. 2B, a-TIF1␣) and acetylated histone H3 (Fig. 2B, a-AcH3) antibodies or without an antibody (Fig. 2B, no Ab) (negative control). Crosslinking was then reversed, and the DNA was purified and subjected to PCR analysis using oligonucleotide primers that generated a 130-bp fragment specific for M2 and a 160-bp fragment specific for M4 sequences (Fig. 2B). The length of the PCR products was compared with DNA fragments obtained when plasmids bearing the M2 and M4 fragments (Fig. 2B, lane 1) and chromatin DNA before immunoprecipitation (Fig. 2B,  lane 2) were used as templates. No specific binding of either of the two analyzed genomic sequences was revealed in chromatin immunoprecipitates in the absence of a primary antibody (Fig. 2B, lane 3). However, TIF1␣ was recruited either alone on M2 chromatin (Fig. 2B, lane 5) or together with acetylated histone H3 (Fig. 2B, lane 4) on M4 chromatin (Fig.  2B, lane 5). These results indicate that TIF1␣ is recruited onto M2 and M4 chromatin in vivo.
To investigate whether TIF1␣ binds directly to DNA and nucleosomes, probes from M2 and M4 fragments were generated. Nucleosomes isolated from the S2 fraction of P19 nuclei (20) were stripped of histone H1 (data not shown) and used as donor nucleosomes in octamer transfer experiments to prepare the corresponding nucleosomal probes. Then, electrophoretic mobility shift assay (EMSA) was performed with naked DNA and nucleosomes (51,52). EMSA demonstrated that TIF1␣ (0.8 M) binds both to naked M2 (145 bp) and M4 (189 bp) DNA fragments and to their corresponding nucleosomes (Fig. 2C). Similar to histone H1 and its variants (52-54), TIF1␣ binds directly to core nucleosomes (Fig. 2C). When increasing amounts of protein (0.1, 0.2, 0.4, and 0.8 M) were incubated with an equal ratio of nucleosome to naked DNA, a preference for naked DNA at low TIF1␣ levels was revealed (0.2-0.4 M for naked DNA compared with 0.8 M for nucleosomes) (Fig.  2D). TIF1␣ binding to nucleosomes and naked DNA in vitro appears to be sequence-independent as it can be competed out by nonspecific DNA competitor (poly(dI⅐dC)) (TIF1␣ at 0.8 M; poly(dI⅐dC at 25 and 100 g/ml) ( Fig. 2E; data not shown). These results demonstrate that TIF1␣ binds directly to DNA and nucleosomes.
We then used EMSA to find out whether the PHD/TTC finger and/or the BrD were involved in DNA and nucleosome interactions (Fig. 3C). The PHD/TTC finger (TIF1␣(734 -853) (0.8, 1.6, 3.2, and 6.3 M)) binds to naked DNA-resulting in the formation of large nucleoprotein complexes-but not to the nucleosomes (Fig. 3C). Only the BrD (TIF1␣(854 -1017) (0.6, 1.2, 2.3, and 4.6 M)) associated both with DNA and nucleosomes, resulting in a single stable BrD-DNA and BrD-nucleosome com-plex, respectively (Fig. 3C). Moreover, multimeric complexes appear to be formed as the mobility of BrD-DNA and BrDnucleosome complexes is dramatically reduced when the concentration of the BrD is increased (Fig. 3D). Thus, the BrD represents both a stable DNA binding domain and a nucleosome interaction motif.
A stretch of four residues, namely PMDL, is highly conserved in helix A of BrDs (spanning from the end of the ZA loop to the first residue of the ␣ A ) in the majority of BrD-containing factors (28,38), and it represents a signature for this domain (Fig. 4A). Hence, these residues were selectively mutated to TAQA (mt-BrD) (Fig. 4A), the recombinant protein was expressed as a His-tag protein in E. coli and purified by affinity chromatography. Its ability to interact with DNA and nucleosomes was then analyzed by EMSA and compared with the wild-type BrD (Fig.   FIG. 2. TIF1␣ binds to DNA and core nucleosomes. A, outline of the chromatin immunoprecipitation assay. B, in vivo recruitment of TIF1␣ to target sequences. Crosslinked chromatin was extracted from P19 EC cells and subjected to immunoprecipitation with either acetylated histone H3 or TIF1␣ antibodies or without any antibody. The length of the amplified fragment is indicated on the left. Control, M2-and M4-containing plasmids (lane 1); input, chromatin before immunoprecipitation (lane 2); no Ab, no primary antibody (lane 3); a-AcH3, chromatin immunoprecipitated with acetylated histone H3 polyclonal antibodies (lane 4); a-TIF1␣, chromatin immunoprecipitated with a TIF1␣ monoclonal antibody (lane 5). C, TIF1␣ associates with DNA and core nucleosomes. DNA was isolated and nucleosomes were reconstituted and purified as described under "Experimental Procedures." EMSA was performed at room temperature with 0.8 M TIF1␣ and 32 P-labeled naked M2(145 bp) or M4(189 bp) DNA and nucleosomes. D, TIF1␣ exhibits a preference for naked DNA. Quantitative EMSA was performed with 0.1, 0.2, 0.4, and 0.8 M TIF1␣ with 32 P-labeled naked M4 DNA and nucleosomes. E, TIF1␣ associates with DNA and nucleosomes in a sequence-independent manner. EMSA was performed with 0.8 M TIF1␣ and 32 P-labeled naked M4 DNA and nucleosomes. For competition experiments, 25 or 100 g/ml poly(dI⅐dC) were used. The samples were subjected to electrophoresis at room temperature on a 5% native polyacrylamide gel in 0.5ϫ TBE. The gels were dried and exposed to autoradiography. 4B, 4.6 M). The disruption of both BrD-DNA and BrD-nucleosome interactions was evident (Fig. 4B). Because abolition of DNA binding correlates with the disruption of BrD-nucleosome interactions, we believe that association of BrD with nucleosomes is mediated primarily via BrD-DNA interactions. DISCUSSION We have used in vitro chromatin assembly assays and nucleosomal bandshifts to analyze the role of the bromodomain (BrD) on TIF1␣-chromatin interactions. In our experiments, we found that the BrD alone is sufficient to bind to nucleosomes; . The samples were subjected to electrophoresis at room temperature on a 5% native polyacrylamide gel in 0.5ϫ TBE. The gels were dried and exposed to autoradiography. moreover, the BrD binds directly to nucleosomal DNA.
TIF1␣ Binds to DNA and Nucleosomes-We have examined the functional association of nuclear receptor coactivator TIF1␣ with chromatin templates. We find that TIF1␣ binds directly and forms a stable complex with chromatin in vivo and in vitro and therefore acts as an integral component of the protein-DNA complex in euchromatin. This is achieved through TIF1␣nucleosome interactions and depends on the ability of TIF1␣ to interact with DNA in a way that does not necessitate the presence of additional factors. The mode of TIF1␣ selection of genomic sites in vivo is unknown. Although binding of TIF1␣ to nucleosomal DNA in vitro appears to be sequence-independent, the exclusive residence of TIF1␣ in euchromatin suggests that a selection mechanism for TIF1␣ binding to chromosomal sites must exist in vivo. Presumably, there are TIF1␣-interacting factors that confer specificity on TIF1␣-binding, functioning to target and/or stabilize TIF1␣ association with euchromatin, in a similar way proposed for the association of HP1 with heterochromatin (63). Because TIF1␣ is a phosphoprotein (19) and can undergo SUMO-lation (18), post-translational modification events may well be required for efficient euchromatin targeting in vivo.
A role for TIF1␣ as a docking platform, for liganded nuclear receptors to bind in order to enhance the efficiency of transcription by selectively scanning the chromatin (20,22) would require the constant presence of TIF1␣ on chromatin. Indeed, in mouse embryos, where TIF1␣ associates strongly with euchromatin (20,21), the extensive TIF1␣ surface on chromosomal DNA is consistent with an activation-competent environment where all TIF1␣ molecules are associated with chromosomal sites. TIF1␣ may act over some distance and with a degree of flexibility, binding to various sites within a genomic region that could consistently keep a promoter or enhancer region competent to respond; consequently, the degree of accessibility to transcription factors experienced by a given genomic region would depend on the concentration of TIF1␣ within a given cell (20). To assess the role of TIF1␣ as a docking molecule, further analyses of TIF1␣ chromatin targets and their potential to associate with various TIF1␣ deletion mutants in vivo are currently in progress.
BrD Is a DNA-binding Motif-TIF1␣ association with chromatin is mediated by BrD-DNA and BrD-nucleosome interactions. This is consistent with the BrD acting as a targeting module (29) for TIF1␣ on selected chromosomal sites. Current models of transcriptional activation via the BrD have involved the sequence-specific binding of transcription factors to DNA that recruit complexes of BrD-containing factors to promoters via protein-protein interactions. However, we have already demonstrated that in vivo interactions between TIF1␣ and chromatin prefigure the recruitment of transcription factors such as liganded nuclear receptors and/or other coactivators or chromatin proteins such as HP1s (20). This result argues for a role for TIF1␣ as a chromatin-recruiting molecule rather than a recruited one. Presumably, to be able to tether itself tightly to chromatin, a DNA-binding mechanism has been employed and interactions between the BrD and nucleosomal DNA have evolved. Whether this is a property of the TIF1␣ BrD alone or is shared by other BrD-containing factors remains to be determined. Nevertheless, it is noteworthy that the BrD signature motif, PMDL, which plays a crucial role both in BrD-DNA and BrD-nucleosome interactions, is highly conserved throughout the BrD family (28,42), indicating that DNA recognition may be a more general feature of BrDs, than has previously been realized. The importance of this motif is underlined by the fact that the first two residues, PM, are essential for GCN5-dependent transcriptional activation in yeast (34,38). In fact, our observation that the BrD is a DNA-and nucleosome-binding motif explains why the BrD is indispensable for GCN5-mediated nucleosomal acetylation, (31,55) and for subsequent activation events on selected promoters in yeast (34).
It has been proposed that the BrD is a protein-protein interaction domain that selectively recognizes amino acid sequences in which lysines are acetylated, as for histone H3 and H4 (29) and for MyoD (36) and Tat (39,40) transactivators, suggesting a more general role for the BrD as an acetyl-lysine binding motif (36 -38). Also, it has been recently shown that the BrD is necessary for p300 association with chromatin in vitro (43), thus accounting for why it is critical for efficient acetylation of nucleosomal histones and transcriptional activation (64). However, the BrD alone did not appear to be sufficient to mediate p300-chromatin interactions (43).
Here, we provide the first experimental evidence that the BrD binds to nucleosomes and DNA, adding BrD-DNA interactions to the protein-protein interactions that are known to be involved in selecting the sites that enable chromatin opening to a wide range of transcription factors. The DNA-binding activity of the BrD could be critical for tumorigenesis because, in acute myeloid leukemia, fusions exist between genes such as MORF and MOZ with CBP regions that include the BrD, thus creating chimeric proteins that could now be recruited to chromosomal DNA (65)(66)(67). Whether DNA binding or a histone-BrD interaction is the first step in the cascade of events that determine the accessibility of a specific chromosomal site remains to be investigated. FIG. 4. The signature motif, PMDL, mediates BrD-DNA interactions. A, transcriptional regulators harboring a PMDL motif. Four proteins are selectively shown: mmTIF1␣, hsP300, mmCBP, and scGCN5. Other factors include hsCBP, CeYNJ1, hsCCG1-1, maCCG1-1, dmTAFII50 -1, hsCCG1-2, maCCG1-2, and dmTAFII250 -2 (28). B, the PMDL motif mediates BrD-DNA and BrDnucleosome interactions. Mutation in BrD signature, PMDL to TAQA. EMSA was performed at room temperature with 4.6 M either wild-type BrD or PMDL signature mutant (mtBrD) and 32 P-labeled naked M4 DNA or nucleosomes. The samples were subjected to electrophoresis at room temperature on a 5% native polyacrylamide gel in 0.5ϫ TBE. The gels were dried and exposed to autoradiography.
Our results suggest that the BrD targets TIF1␣, and possibly other transcriptional regulators, to nucleosomal DNA at those chromatin sites where they exert their function in vivo. Remarkably, many different protein domains can be linked to BrD, including histone acetyltransferase, kinase, ATPase, and helicase activities (27,28), emphasizing its role as a multipurpose chromatin adaptor in vivo.