Limits...
Molecular basis of histone tail recognition by human TIP5 PHD finger and bromodomain of the chromatin remodeling complex NoRC.

Tallant C, Valentini E, Fedorov O, Overvoorde L, Ferguson FM, Filippakopoulos P, Svergun DI, Knapp S, Ciulli A - Structure (2014)

Bottom Line: TIP5 domains that recognize posttranslational modifications on histones are essential for recruitment of NoRC to chromatin, but how these reader modules recognize site-specific histone tails has remained elusive.PHD finger functions as an independent structural module in recognizing unmodified H3 histone tails, and the bromodomain prefers H3 and H4 acetylation marks followed by a key basic residue, KacXXR.Further low-resolution analyses of PHD-bromodomain modules provide molecular insights into their trans histone tail recognition, required for nucleosome recruitment and transcriptional repression of the NoRC complex.

View Article: PubMed Central - PubMed

Affiliation: Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK; Structural Genomics Consortium, Nuffield Department of Clinical Medicine, University of Oxford, Old Road Campus Research Building, Roosevelt Drive, Oxford OX3 7DQ, UK; Target Discovery Institute, Nuffield Department of Clinical Medicine, University of Oxford, NDM Research Building, Roosevelt Drive, Oxford OX3 7FZ, UK.

Show MeSH
Structures of TIP5 and BAZ2B BRDs in the Free State and in Complex with H3 and H4 Peptides(A) Crystal structures superposition on ribbon representation of novel TIP5 (PDB accession number 4LZ2) and BAZ2B (PDB accession number 3G0L) BRDs on Cα positions (rmsd = 0.488).(B) Structure-based sequence alignment of single BRDs from the human TIP5 and BAZ2B. The conserved asparagine involved in the histone recognition is highlighted in red.(C, F, and I) Surface representations of the peptide-bound structures of single BRDs. (C) Corresponds to TIP5-BRD in complex with H4K16acK20ac (GAKacRHRKacVL), with the acetylation recognized by the conserved asparagine highlighted in bold. (F) BAZ2B-BRD in complex with H3K14ac (TGGKacAPRKQ). (I) BAZ2B-BRD in complex with H4K8acK12ac (GGKacGLGKacG). Electrostatic surface potentials are shown between +5 V (blue) and −5 V (red).(D, G, and J) Detailed interactions between BRD of TIP5 with H4K16acK20ac and BRD of BAZ2B with H3K14ac and H4K8acK12ac.(E, H, and K) View of the peptide-binding pocket. Simulated-annealing 2Fo-Fc maps showing the electron density for the H3 and H4 peptides. Maps were calculated at 1.65, 1.99, and 1.60 Å resolution and contoured between 1.2 and 1.5 σ above the mean.See also Figure S5.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4291147&req=5

fig3: Structures of TIP5 and BAZ2B BRDs in the Free State and in Complex with H3 and H4 Peptides(A) Crystal structures superposition on ribbon representation of novel TIP5 (PDB accession number 4LZ2) and BAZ2B (PDB accession number 3G0L) BRDs on Cα positions (rmsd = 0.488).(B) Structure-based sequence alignment of single BRDs from the human TIP5 and BAZ2B. The conserved asparagine involved in the histone recognition is highlighted in red.(C, F, and I) Surface representations of the peptide-bound structures of single BRDs. (C) Corresponds to TIP5-BRD in complex with H4K16acK20ac (GAKacRHRKacVL), with the acetylation recognized by the conserved asparagine highlighted in bold. (F) BAZ2B-BRD in complex with H3K14ac (TGGKacAPRKQ). (I) BAZ2B-BRD in complex with H4K8acK12ac (GGKacGLGKacG). Electrostatic surface potentials are shown between +5 V (blue) and −5 V (red).(D, G, and J) Detailed interactions between BRD of TIP5 with H4K16acK20ac and BRD of BAZ2B with H3K14ac and H4K8acK12ac.(E, H, and K) View of the peptide-binding pocket. Simulated-annealing 2Fo-Fc maps showing the electron density for the H3 and H4 peptides. Maps were calculated at 1.65, 1.99, and 1.60 Å resolution and contoured between 1.2 and 1.5 σ above the mean.See also Figure S5.

Mentions: The crystal structure of BAZ2B BRD with acetylated H3K14 revealed sequence-specific interactions of the BRD. The histone peptide was well defined in the electron density map, allowing unambiguous tracing of residues 12 to 19 interacting with the acetyl-lysine binding site (Figure 3H). As expected, the carbonyl group of H3K14ac formed the canonical hydrogen bond with the conserved N2140 as well as a water-mediated hydrogen bond with Y2097 (Figure S5B). The complex was further stabilized by additional hydrogen bonds mediated between the BRD backbone and H3 peptide amide bonds (Figures 3F and 3G). The typical network of water molecules located at the bottom of the acetyl-lysine binding pocket was conserved (Figure S5B) (Filippakopoulos et al., 2012). Interestingly, two key interactions were observed comprising a hydrophobic stacking interaction between H3P16 with F2139 from the end of helix αB, as well as a sequence-specific interaction involving H3R17 and residues located in the BC loop. The BAZ2B BC loop contains two acidic residues (E2141 and D2142) that formed electrostatic interactions with the side chain of H3R17. These salt bridges are particularly important in determining the binding specificity. An additional hydrogen bond was formed between Nε-H of the arginine side chain and the carbonyl of the E2137 backbone (Figure 3F). We observed only little structural changes comparing the BAZ2B BRD in its free (Protein Data Bank [PDB] accession number 3G0L) and peptide-bound states, suggesting a rigid interaction between the BRD and its targeted acetylated substrate. The most notable difference comparing both structures was the ordering of the side chain of E2141 in the peptide complex. The key residues involved in the H3K14ac recognition (F2139, E2141, and D2142) are conserved in the TIP5 BRD but not in the related BRD proteins BAZ1A and BAZ1B. Sequence alignment (Figure 3B) and thermodynamic binding data (Table 1) suggested a similar molecular recognition of the H3 tail by the TIP5 BRD. Interestingly, moderate to weak binding affinities have been reported for other BRDs to recognize the H3K14ac mark, such as SMARCA2 (KD = 285 μM by ITC) (Filippakopoulos et al., 2012), SMARCA4 (KD = 1.2 mM by saturation transfer difference nuclear magnetic resonance [STD-NMR]) (Shen et al., 2007), PCAF (KD = 128 μM by STD-NMR, KD = 188 μM by ITC) (Zeng et al., 2008; Filippakopoulos et al., 2012), or PB1 BRD 2 (KD = 500 μM by STD-NMR) (Charlop-Powers et al., 2010). Our binding studies and molecular characterization with the histone peptide interaction exhibit, unexpectedly, much stronger binding with NoRC BRDs to the acetylated K14 in the low micromolar range (KD = 10 and 33 μM for BAZ2B and TIP5 BRDs, respectively). A structural comparison with the available NMR structure of the human polybromo BRD 2 in complex with H3K14ac uncovers a diverse peptide-binding mode compared with the BAZ2B complex (Figure S5D). The superposed structures (Cα rmsd = 2.30 Å) exhibit distinct binding cavities. The shallower surface area for the PB1 BRD displays how the peptide extends through the ZA loop, forming weak protein-protein interactions with selectivity driven through H3A15 specific contacts, whereas the deeper hydrophobic binding cavity of BAZ2B accommodates more efficiently the ε-acetyl-lysine 14.


Molecular basis of histone tail recognition by human TIP5 PHD finger and bromodomain of the chromatin remodeling complex NoRC.

Tallant C, Valentini E, Fedorov O, Overvoorde L, Ferguson FM, Filippakopoulos P, Svergun DI, Knapp S, Ciulli A - Structure (2014)

Structures of TIP5 and BAZ2B BRDs in the Free State and in Complex with H3 and H4 Peptides(A) Crystal structures superposition on ribbon representation of novel TIP5 (PDB accession number 4LZ2) and BAZ2B (PDB accession number 3G0L) BRDs on Cα positions (rmsd = 0.488).(B) Structure-based sequence alignment of single BRDs from the human TIP5 and BAZ2B. The conserved asparagine involved in the histone recognition is highlighted in red.(C, F, and I) Surface representations of the peptide-bound structures of single BRDs. (C) Corresponds to TIP5-BRD in complex with H4K16acK20ac (GAKacRHRKacVL), with the acetylation recognized by the conserved asparagine highlighted in bold. (F) BAZ2B-BRD in complex with H3K14ac (TGGKacAPRKQ). (I) BAZ2B-BRD in complex with H4K8acK12ac (GGKacGLGKacG). Electrostatic surface potentials are shown between +5 V (blue) and −5 V (red).(D, G, and J) Detailed interactions between BRD of TIP5 with H4K16acK20ac and BRD of BAZ2B with H3K14ac and H4K8acK12ac.(E, H, and K) View of the peptide-binding pocket. Simulated-annealing 2Fo-Fc maps showing the electron density for the H3 and H4 peptides. Maps were calculated at 1.65, 1.99, and 1.60 Å resolution and contoured between 1.2 and 1.5 σ above the mean.See also Figure S5.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4291147&req=5

fig3: Structures of TIP5 and BAZ2B BRDs in the Free State and in Complex with H3 and H4 Peptides(A) Crystal structures superposition on ribbon representation of novel TIP5 (PDB accession number 4LZ2) and BAZ2B (PDB accession number 3G0L) BRDs on Cα positions (rmsd = 0.488).(B) Structure-based sequence alignment of single BRDs from the human TIP5 and BAZ2B. The conserved asparagine involved in the histone recognition is highlighted in red.(C, F, and I) Surface representations of the peptide-bound structures of single BRDs. (C) Corresponds to TIP5-BRD in complex with H4K16acK20ac (GAKacRHRKacVL), with the acetylation recognized by the conserved asparagine highlighted in bold. (F) BAZ2B-BRD in complex with H3K14ac (TGGKacAPRKQ). (I) BAZ2B-BRD in complex with H4K8acK12ac (GGKacGLGKacG). Electrostatic surface potentials are shown between +5 V (blue) and −5 V (red).(D, G, and J) Detailed interactions between BRD of TIP5 with H4K16acK20ac and BRD of BAZ2B with H3K14ac and H4K8acK12ac.(E, H, and K) View of the peptide-binding pocket. Simulated-annealing 2Fo-Fc maps showing the electron density for the H3 and H4 peptides. Maps were calculated at 1.65, 1.99, and 1.60 Å resolution and contoured between 1.2 and 1.5 σ above the mean.See also Figure S5.
Mentions: The crystal structure of BAZ2B BRD with acetylated H3K14 revealed sequence-specific interactions of the BRD. The histone peptide was well defined in the electron density map, allowing unambiguous tracing of residues 12 to 19 interacting with the acetyl-lysine binding site (Figure 3H). As expected, the carbonyl group of H3K14ac formed the canonical hydrogen bond with the conserved N2140 as well as a water-mediated hydrogen bond with Y2097 (Figure S5B). The complex was further stabilized by additional hydrogen bonds mediated between the BRD backbone and H3 peptide amide bonds (Figures 3F and 3G). The typical network of water molecules located at the bottom of the acetyl-lysine binding pocket was conserved (Figure S5B) (Filippakopoulos et al., 2012). Interestingly, two key interactions were observed comprising a hydrophobic stacking interaction between H3P16 with F2139 from the end of helix αB, as well as a sequence-specific interaction involving H3R17 and residues located in the BC loop. The BAZ2B BC loop contains two acidic residues (E2141 and D2142) that formed electrostatic interactions with the side chain of H3R17. These salt bridges are particularly important in determining the binding specificity. An additional hydrogen bond was formed between Nε-H of the arginine side chain and the carbonyl of the E2137 backbone (Figure 3F). We observed only little structural changes comparing the BAZ2B BRD in its free (Protein Data Bank [PDB] accession number 3G0L) and peptide-bound states, suggesting a rigid interaction between the BRD and its targeted acetylated substrate. The most notable difference comparing both structures was the ordering of the side chain of E2141 in the peptide complex. The key residues involved in the H3K14ac recognition (F2139, E2141, and D2142) are conserved in the TIP5 BRD but not in the related BRD proteins BAZ1A and BAZ1B. Sequence alignment (Figure 3B) and thermodynamic binding data (Table 1) suggested a similar molecular recognition of the H3 tail by the TIP5 BRD. Interestingly, moderate to weak binding affinities have been reported for other BRDs to recognize the H3K14ac mark, such as SMARCA2 (KD = 285 μM by ITC) (Filippakopoulos et al., 2012), SMARCA4 (KD = 1.2 mM by saturation transfer difference nuclear magnetic resonance [STD-NMR]) (Shen et al., 2007), PCAF (KD = 128 μM by STD-NMR, KD = 188 μM by ITC) (Zeng et al., 2008; Filippakopoulos et al., 2012), or PB1 BRD 2 (KD = 500 μM by STD-NMR) (Charlop-Powers et al., 2010). Our binding studies and molecular characterization with the histone peptide interaction exhibit, unexpectedly, much stronger binding with NoRC BRDs to the acetylated K14 in the low micromolar range (KD = 10 and 33 μM for BAZ2B and TIP5 BRDs, respectively). A structural comparison with the available NMR structure of the human polybromo BRD 2 in complex with H3K14ac uncovers a diverse peptide-binding mode compared with the BAZ2B complex (Figure S5D). The superposed structures (Cα rmsd = 2.30 Å) exhibit distinct binding cavities. The shallower surface area for the PB1 BRD displays how the peptide extends through the ZA loop, forming weak protein-protein interactions with selectivity driven through H3A15 specific contacts, whereas the deeper hydrophobic binding cavity of BAZ2B accommodates more efficiently the ε-acetyl-lysine 14.

Bottom Line: TIP5 domains that recognize posttranslational modifications on histones are essential for recruitment of NoRC to chromatin, but how these reader modules recognize site-specific histone tails has remained elusive.PHD finger functions as an independent structural module in recognizing unmodified H3 histone tails, and the bromodomain prefers H3 and H4 acetylation marks followed by a key basic residue, KacXXR.Further low-resolution analyses of PHD-bromodomain modules provide molecular insights into their trans histone tail recognition, required for nucleosome recruitment and transcriptional repression of the NoRC complex.

View Article: PubMed Central - PubMed

Affiliation: Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK; Structural Genomics Consortium, Nuffield Department of Clinical Medicine, University of Oxford, Old Road Campus Research Building, Roosevelt Drive, Oxford OX3 7DQ, UK; Target Discovery Institute, Nuffield Department of Clinical Medicine, University of Oxford, NDM Research Building, Roosevelt Drive, Oxford OX3 7FZ, UK.

Show MeSH