Crystal structure of a binary complex between human GCN5 histone acetyltransferase domain and acetyl coenzyme A

Histone acetylation leads to modification of the chromatin structure and function. This process is directly involved in the DNA recognition by transcription factors, regulates the transcription of many important genes, and affects DNA replication and repair.1-3 Because of this, enzymes catalyzing histone acetylation, the histone acetyltransferases (HATs, EC 2.3.1.48), have a major role in the cell fate, proliferation, and differentiation. The histone acetyltransferase GCN5 acts as a supervisor in the normal cell cycle progression having comprehensive control over expressions of these cell cycle-related genes, as well as apoptosis-related genes, via alterations in the chromatin structure, mimicked by changing acetylation status of core histones, which surround these genes.4 Since GCN5 is a critical regulator of cell cycle and c-myc, it has a potential role in cancer.5 GCN5 is also involved in developmental processes and the loss of the GCN5 gene (referred as GCN5L2, general control of amino acid synthesis protein 5-like 2), was shown to lead to increased apoptosis and mesodermal defects during mouse development.6 Recently, it was reported that GCN5 controls glucose metabolism through transcriptional repression of PGC-1α, and thus GCN5 represents a pharmacological target in the treatment of metabolic disorders occurring in diabetes and aging.7 GCN5 was originally identified as a gene required for amino acid biosynthesis in yeast.8 The protein belongs to the GCN5-related N-acetyltransferase (GNAT) superfamily that comprises the HATs, the aminoglycoside N-acetyltransferases, serotonin N-acetyltransferase, glucoseamine-6-phosphate N-acetyltransferase, mycothiol synthase, protein N-myristoyltransferase, and the Fem family of amino acyl transferases.9 The HAT family itself can be subdivided into five families that, besides the GNATs, includes the MYST (for MOZ, Ybf2/Sas3, Sas2, and Tip60)-related HATs, the p300/CBP HATs, the general transcription factor HATs, and the nuclear hormone-related HATs.10 Human GCN5L2 gene encodes a protein containing three functional domains: the N-terminal P/CAF domain, the C-terminal bromodomain, and the HAT domain. The protein is classified as a HAT enzyme, catalyzing the transfer of the acetyl group from the cofactor acetyl coenzyme A (AcCoA) to the side chain amino group of lysine residues in the N-terminus of core histone proteins.11, 12 Human GCN5 is also capable of acetylation of nonhistone proteins, like the tumor suppressor p53,13 the c-MYC oncoprotein,14 and the metabolic coactivator PGC-1α.7 In the cell, GCN5 is a component of large transcriptional coactivator complexes. Human GCN5 was shown to be part of the STAGA (SPT3-TAFII31-GCN5-L acetylase) and the TFTC (TATA-binding protein-free TAFII-containing) complex.15, 16 TFTC-type HAT complexes, containing GCN5 and TRRAP (transformation/transcription domain-associated protein), for example, are involved in the estrogen receptor α transactivation,17 and are required for BRCA1 function, whose alteration leads to breast and ovarian cancer.18 GCN5 is also recruited to c-Myc by the cofactor TRRAP19 and is part of the PGC-1α protein complex.7 To date, the crystal structures of several HAT domains have been solved: yeast and Tetrahymena thermophila GCN5,20-22 yeast Hpa2,22 yeast Esa1,23 yeast HAT1,24 human P/CAF,25 and human MYST1 (Plotnikov, manuscript in preparation). Ternary complex structures have been solved for the HAT domain of Tetrahymena thermophila GCN5, either bound to coenzyme A (CoA) and histone H3 or H4 peptides,21, 26 complexed with a bisubstrate inhibitor,27 and bound to CoA and p53 peptide,28 providing important insights into substrate binding and specificity. We aim at resolving the structures of all human histone acetyltransferases for which structural information is not yet available and to compare their structural and functional properties. Here we report the three-dimensional structure of the HAT domain of human GCN5 in complex with AcCoA, determined to 1.74 Å resolution. DNA encoding amino acids 497–662 of human GCN5 was amplified by PCR from the Mammalian Gene Collection clone (accession code gi:10835101) and subcloned into a modified pET28a-LIC vector (for details see http://www.sgc.utoronto.ca/SGC-WebPages/toronto-vectors.php). The recombinant protein was overexpressed as an N-terminal His6-tagged protein in E. coli BL21 (DE3) Codon plus RIL (Stratagene). Cells were grown in Terrific Broth in presence of 50 μg/mL kanamycin at 37°C to an OD600 of about 0.8, induced with 0.5 mM isopropyl-1-thio-D-galactopyranoside, and incubated overnight at 15°C. For purification cell paste was resuspended in lysis buffer (50 mM sodium/potassium phosphate buffer, pH 7.5, 500 mM NaCl, 5% glycerol, and 0.1 mM phenylmethyl sulfonyl fluoride). Cells were lysed by passing through Microfluidizer (Microfluidics Corporation) at 18,000 psi. The lysate was clarified by centrifugation and the supernatant was applied to a 5 mL HiTrap Chelating column (Amersham Biosciences), charged with Ni2+. The column was washed with 50 mL of 20 mM HEPES-NaOH, pH 7.5 buffer, containing 500 mM NaCl and 50 mM imidazole. The protein was eluted with 20 mM HEPES-NaOH, pH 7.5, 500 mM NaCl, 5% glycerol, and 250 mM imidazole. To cleave the His-tag, thrombin (Sigma) was added to GCN5 containing fractions and the sample was incubated overnight at 4°C while dialyzing against 20 mM HEPES-NaOH, pH 7.5, and 150 mM NaCl. The protein was further purified to homogeneity using cation exchange chromatography on Source 30 S column (10 × 10) (Amersham Biosciences), equilibrated with 20 mM HEPES-NaOH, pH 7.5. GCN5 was eluted with a linear gradient of up to 500 mM NaCl (in 30 column volumes). Purification yield was 20 mg of protein per 1 L of culture. Purified HAT domain of human GCN5 was complexed with AcCoA at 1:5 molar ratio of protein to coenzyme, and crystallized using the sitting drop vapor diffusion method at 20°C by mixing 1 μL of 8.4 mg/mL protein solution with 1 μL of the reservoir solution containing 15% (v/v) ethanol and 100 mM Tris-HCl, pH 7.0. Before flash-freezing the crystal in liquid nitrogen, the crystal was soaked in a cryoprotectant consisting of 17% (v/v) ethanol, 100 mM Tris-HCl, pH 7.0, 9% sucrose, 4% glucose, 8% ethylene glycol, and 8% glycerol. Diffraction data were collected at 100 K on a FR-E/R axis IV++ diffraction system (Rigaku/MSC) using Cu radiation. The data were measured at wavelength 1.54 Å and processed using the program HKL2000.29 Data collection statistics are shown in Table I. The structure of the HAT domain of GCN5 was solved by molecular replacement using the program PHASER.30 The model of the HAT domain of yeast GCN5 served as a template (1YGH).20 Refinement of the model was performed with CNS.31 To monitor progress 4% of the data was used for calculation of Rfree.32 Manual rebuilding of the model was performed using the program O33 based on σ-weighted 2Fo − Fc and Fo − Fc electron density maps. Statistics on structure refinement are summarized in Table I. Structure comparisons were carried out using the program DALI34 and PYMOL.35 Figures were created with PYMOL.35 Sequence alignments were performed with MULTALIN (http://prodes.toulouse.inra.fr/multalin). The X-ray data and the atomic coordinates have been deposited into the PDB as 1Z4R. The crystal structure of the HAT domain of human GCN5 contains 1 molecule per asymmetric unit. The final model includes amino acid residues 496–658, 1 AcCoA molecule and 131 water molecules. The protein structure has a mixed α/β topology and contains seven α-helices and seven β-strands. The overall structure complexed with AcCoA is presented in Figure 1(A). The HAT domain of human GCN5 folds around a central core that contains a mixed β-sheet built up of antiparallel β-strands. Most of the AcCoA contacts are mediated by the central core domain of the enzyme. The coenzyme molecule is bound in a cleft on the surface of the protein via hydrogen bonds between the oxygen atoms of the pyrophosphate moiety of AcCoA and the backbone amide nitrogens of Val587, Gly589, Gly591, and Thr592 (Fig. 2). The carbonyl oxygen of the AcCoA thioester is hydrogen bonded to the backbone amide nitrogen of Cys579. Additional interactions were found between the pantothenic acid moiety of AcCoA and the backbone amide of Val581 and the side chain of Gln586. The adenine ring of the coenzyme forms a single direct hydrogen bond with the main chain carbonyl group of Tyr617 (Fig. 2). The fold of the central core domain, harboring the conserved motif A of the GNAT proteins (Fig. 3), and thus the mode of coenzyme binding are similar to that of other previously reported histone acetyltransferase structures.9, 37, 38 The structures of the HAT domain of human GCN5 and human P/CAF,25 for example, superimpose with a root-mean-square deviation (r.m.s.d.) of 0.61 Å for 157 Cα atoms. The r.m.s.d. values of the superimposition of the HAT domain of human GCN5 with the HAT domain of GCN5 from yeast20 and Tetrahymena thermophila,21 however, were determined with 1.27 Å for 153 Cα atoms and 1.0 Å for 152 Cα atoms, respectively. Figure 1(B) illustrates the superimposition of the crystal structures of those HAT proteins. Whereas the conserved core domains, that have a prominent role in coenzyme binding and catalysis, superimpose very well, structural differences are observed in two loop regions, located N- and C-terminal of the core domain. In human GCN5 Loop 1 connects helix α1 and strand β1 and Loop 2 is situated between helix α7 and strand β7 (Fig. 3). Comparisons of this loop regions revealed, that their B-factors are somewhat higher than average, indicating flexibility. The Loop 2 has been associated with substrate binding and specificity.26, 28, 39 Loop 1 is situated distant from the active site on the surface of the protein [Fig. 1(C)]. Since HATs function as multiprotein complexes in the cell, HAT activity and specificity might be regulated also via interaction with other proteins. Thus, Loop 1 could represent a potential interacting region. The structural basis for multiprotein complexes, however, still remains elusive. (A) Overall fold of human GCN5. The protein is shown in a rainbow color gradient from blue (N-terminus, N) to red (C-terminus, C). The secondary structure elements are labeled as defined in Figure 3. The AcCoA molecule is represented as a ball-and-stick model colored as per atom type: carbon in gray, oxygen in red, nitrogen in blue, sulfur in yellow, and phosphorus in orange. (B) Superimposition of selected HAT domain crystal structures. The divergent loop regions N- and C-terminal of the core domains (Loop 1 and 2) are highlighted in red (human GCN5), blue (human P/CAF, 1CM0), yellow (yeast GCN5, 1YGH), and green (GCN5 from Tetrahymena thermophila, 1QSR). The AcCoA molecule bound to human GCN5 is shown as a ball-and-stick model. (C) Superposition of the HAT domains of human GCN5 (red) bound to AcCoA and Tetrahymena thermophila GCN5 (green, 1QSN) bound to CoA (not included) and H3 peptide (brown). The AcCoA molecule and the glutamate residue E575 of human GCN5, functioning as a general base in catalysis, are shown as ball-and-stick models. Protein residues involved in AcCoA binding. Hydrogen bonds are shown as dashed lines. Sequence alignment of selected HATs. Amino acid sequences used in the alignment correspond to the HAT domains of human GCN5 (hGCN5), human P/CAF (hPCAF), yeast GCN5 (yGCN5), and Tetrahymena GCN5 (tGCN5). The secondary structure elements shown above the alignment correspond to hGCN5. Residues involved in AcCoA binding are highlighted with ▾. The four underlined regions, referred to as motif A–D, correspond to sequence motifs within the GNATs,36 The glutamate residue marked with ▿functions as a general base in catalysis. The divergent loop regions N- and C-terminal (Loop 1 and Loop 2) of the core domain are highlighted in gray. The current believed catalytic mechanism for GCN5-related HATs is sequential ordered and proceeds through a ternary complex, whereby the acetyl moiety of AcCoA is transferred directly from the cofactor to the side chain nitrogen of the substrate lysine residue.40 Prior to the acetylation reaction, a deprotonation of the side chain amino group of the lysine residue within the histone substrate is required. A glutamate residue located within the structurally conserved core domain is suggested to assist in this step by functioning as a general base for catalysis.41 Structure comparisons of the HAT domain of human GCN5 with the HAT domain of human P/CAF and GCN5 from Tetrahymena thermophila suggest the glutamate residue E575 of human GCN5 to serve as the general base deprotonating the substrate [Fig. 1(C)]. The crystal structure of the HAT domain of human GCN5 reported here provides information on the mechanism of the enzyme and can be used together with the known structures of HAT domain containing proteins for structure-based drug design resulting in new therapeutics for the treatment of HAT-related diseases.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.