Modulating Acetyl-CoA Binding in the GCN5 Family of Histone Acetyltransferases

Michael R Langer,Christopher J Fry,Craig L Peterson,John M Denu

doi:10.1074/jbc.m203251200

Abstract

The histone acetyltransferase (HAT) GCN5 is the founding member for a family of chromatin remodeling enzymes. GCN5 is the catalytic subunit of a large multi-subunit complex that functions in the regulation of gene activation via acetylation of lysine residues within the N-terminal tails of core histone proteins. Using acetyl-CoA as a co-substrate, the high affinity binding of acetyl-CoA is a critical first step in the reaction. Here, we examine the biochemical and biological importance of a conserved hydroxyl-bearing residue in signature motif A. Interestingly, one major exception is the Saccharomyces cerevisiae GCN5, where an alanine (Ala(190)) is located in the corresponding position. In related GCN5 family structures, a hydroxyl-containing side chain residue is hydrogen-bonded to the alpha-phosphate oxygen of CoA. We demonstrate that this key hydrogen bond contributes approximately 10-fold to the binding affinity of GCN5 HATs for acetyl-CoA. Human p300/CBP-associating factor, human GCN5, and tetrahymena GCN5 displayed dissociation constants (K(d)) for acetyl-CoA of 0.64 +/- 0.12, 0.56 +/- 0.15, and 0.62 +/- 0.17 microm, respectively. In contrast, S. cerevisiae GCN5 displayed a K(d) of 8.5 microm. When Ala(190) was replaced with threonine, the A190T derivative yielded a K(d) value of 0.56 +/- 0.1 microm for acetyl-CoA, completely restoring the higher affinity binding seen with the GCN5 homologs that naturally harbor a threonine at this position. Detailed kinetic analyses revealed that the A190T derivative was otherwise catalytically indistinguishable from wild type GCN5. We also demonstrate that the A190T allele rescued the slow growth phenotype and the defect in HO transcription caused by a deletion of GCN5. Furthermore, the A190T allele supported wild type levels of transcriptionally targeted and global histone H3 acetylation. In each case, the A190T derivative behaved similarly to wild type GCN5, suggesting that the efficacy of HAT activity by GCN5 is not limited by the availability of nuclear acetyl-CoA pools.

Highlights

Histone acetyltransferases (HATs)1 catalyze the transfer of an acetyl group from acetyl-CoA to the acceptor ⑀-amino group
Mutational Analysis of yGCN5—Previous kinetic analyses of the catalytic domains of hP/CAF and wild type yGCN5 [27,28,29, 40] indicated a completely ordered kinetic mechanism, where upon AcCoA binding to free enzyme, there is an enzyme isomerization that permits the efficient binding of polypeptide (Scheme 1)
The limiting rate in catalytic turnover is the chemistry step (k7; Scheme 1). These initial analyses demonstrated that hP/CAF and yGCN5 have nearly identical rates of catalysis when both substrates are present at saturating levels

Summary

EXPERIMENTAL PROCEDURES

Materials—P81 phosphocellulose disks were purchased from Invitrogen. Dispo-Equilibrium dialyzers were obtained from Amika Corporation. Yeast Strains and Media—The yeast strain CY569 (gcn5D, H0-lacZ) was used for the analysis of gcn5⌬ phenotypes in vivo and is described in Ref. 7 This strain was transformed with expression plasmids encoding wild type Gcn5p (CP649), a mutant gcn5p containing an A190T mutation (CP914), or no protein (empty vector; BA110). The generated mutation and integrity of the full-length GCN5 gene was verified by DNA sequencing using the PrismTM Ready Reaction DyeDeoxyTM terminator cycle sequencing kit (PerkinElmer Life Sciences). Enzymatic Assays for HATs—HAT activity was measured with [3H]acetyl-CoA and histone H3 peptide as substrates, using a radioactive P81 filter binding assay as described previously [26]. Chromatin Immunoprecipitations—Chromatin immunoprecipitations were performed in triplicate (three independent transformants) using cells grown to mid-log phase and an antibody against diacetylated histone H3 (Upstate Biotechnology, Inc.) as described previously [21], with the following modifications. Nals were quantified using a PhoshorImager and ImageQuant v4.2 (Molecular Dynamics)

RESULTS AND DISCUSSION

HAT enzyme

Miller units