Metabolic Regulation of Histone Acetyltransferases by Endogenous Acyl-CoA Cofactors

Abstract

The finding that chromatin modifications are sensitive to changes in cellular cofactor levels potentially links altered tumor cell metabolism and gene expression. However, the specific enzymes and metabolites that connect these two processes remain obscure. Characterizing these metabolic-epigenetic axes is critical to understanding how metabolism supports signaling in cancer, and developing therapeutic strategies to disrupt this process. Here, we describe a chemical approach to define the metabolic regulation of lysine acetyltransferase (KAT) enzymes. Using a novel chemoproteomic probe, we identify a previously unreported interaction between palmitoyl coenzyme A (palmitoyl-CoA) and KAT enzymes. Further analysis reveals that palmitoyl-CoA is a potent inhibitor of KAT activity and that fatty acyl-CoA precursors reduce cellular histone acetylation levels. These studies implicate fatty acyl-CoAs as endogenous regulators of histone acetylation, and suggest novel strategies for the investigation and metabolic modulation of epigenetic signaling.

Figure 1

Metabolic regulation of histone lysine acetyltransferase (KAT) activity by endogenous acyl-CoAs. By antagonizing acetyl-CoA binding, metabolic acyl-CoAs may inhibit KAT activity, thereby transducing information about the metabolic state of the cell to changes in histone acetylation.

Figure 2

A chemical proteomic strategy for profiling cellular KAT complexes. (a) Schematic of KAT chemical proteomic probe (H3K14-CoA-biotin, 1) used in this study, with ribbon representing H3 peptide. Full structure of probe and competitor are provided in the Supporting Information. (b) Spectral counts for KATs and complex members enriched by 1. (c) Immunoaffinity profiling of Gcn5, pCAF, and Mof demonstrate active site dependent enrichment of each KAT. Input represents the KAT immunoblot signal from non-enriched control, and competitor refers to the active site directed ligand H3K14-CoA.

Figure 3

Competitive chemical proteomic profiling of acyl-CoA/KAT interactions. (a) Schematic depiction of strategy. (b) Representative immunoblots depicting enrichment of Gcn5, Mof, and pCAF by probe 1 (10 μM) in the absence or presence of metabolic acyl-CoA competitors (30 μM). (c-e) Quantitative rank ordering of acyl-CoA interactions with Gcn5, Mof, and pCAF, as analyzed via gel densitometry. Values represent the average ± SD of three replicates, with percent competition calculated by comparing KAT pulldown in the presence and absence of each acyl-CoA.

Figure 4

Biochemical characterization of fatty acyl-CoA/KAT interactions. (a) Homology model of human Gcn5 bound to bisubstrate inhibitor H3K14-CoA and palmitoyl-CoA. (b) Analysis of metabolic inhibition of Gcn5 via coupled-enzyme assay Lineweaver-Burk plot depicting acetyl-CoA competitive inhibition of Gcn5 by palmitoyl-CoA. (c) Analysis of metabolic inhibition of Gcn5 via separation-based assay. Palmitoyl-CoA and related fatty acyl-CoAs inhibit Gcn5 biochemical activity more potently than CoA. Structures of fatty acyl-CoAs are provided in the Supplementary Information (Figure S1).

Figure 5

Cellular effects of palmitoyl-CoA precursors and palmitoyl-CoA synthetases on histone acetylation. (a) Long-chain acyl-CoA synthetase (ACSL)-dependent formation of palmitoyl-CoA from palmitate. (b) Top: Effect of palmitate-treatment and ACSL3 overexpression on histone acetylation in HEK-293 cells. Values under each immunoblot represent the densitometry quantified band intensity relative to the control (lane 1). All quantified values were normalized based on Ponceau staining to account for differences in loading between different lanes (Figure S5a). Bottom: Cytosolic/nuclear loading controls and verification of FLAG-ACLS3 overexpression.