Acetyl-CoA synthetase activity is enzymatically regulated by lysine acetylation using acetyl-CoA or acetyl-phosphate as donor molecule

Abstract

The AMP-forming acetyl-CoA synthetase is regulated by lysine acetylation both in bacteria and eukaryotes. However, the underlying mechanism is poorly understood. The Bacillus subtilis acetyltransferase AcuA and the AMP-forming acetyl-CoA synthetase AcsA form an AcuA•AcsA complex, dissociating upon lysine acetylation of AcsA by AcuA. Crystal structures of AcsA from Chloroflexota bacterium in the apo form and in complex with acetyl-adenosine-5'-monophosphate (acetyl-AMP) support the flexible C-terminal domain adopting different conformations. AlphaFold2 predictions suggest binding of AcuA stabilizes AcsA in an undescribed conformation. We show the AcuA•AcsA complex dissociates upon acetyl-coenzyme A (acetyl-CoA) dependent acetylation of AcsA by AcuA. We discover an intrinsic phosphotransacetylase activity enabling AcuA•AcsA generating acetyl-CoA from acetyl-phosphate (AcP) and coenzyme A (CoA) used by AcuA to acetylate and inactivate AcsA. Here, we provide mechanistic insights into the regulation of AMP-forming acetyl-CoA synthetases by lysine acetylation and discover an intrinsic phosphotransacetylase allowing modulation of its activity based on AcP and CoA levels.

Fig. 1. Bacillus subtilis AcsA and AcuA form a stable complex.

a The B. subtilis acu-operon encodes for the lysine acetyltransferase AcuA, the classical Zn²⁺-dependent KDAC AcuC and AcuB of unknown function. The acetyl-CoA synthetase AcsA is reversely transcribed upstream of the acu-operon. Acetylation of K549 in the AcsA C-terminal domain (dark green) inhibits AcsA activity, deacetylation by AcuC restores its activity. AcsA catalyses the generation of acetyl-CoA in two half-reactions. In the first half-reaction, acetate is converted to acetyl-AMP under consumption of ATP and release of pyrophosphate (PP_i), making the reaction irreversible (adenylation reaction) and in the second half-reaction the mixed anhydride acetyl-AMP is converted to the thioester acetyl-CoA (thioester-forming reaction). b AcuA forms a monomer and AcsA a dimer in solution as shown by analytical size exclusion chromatography (SEC) experiments. Source data are provided as Source Data file. c AlphaFold2 structure prediction of AcsA. AcsA is in the conformation to catalyse the first half-reaction (adenylation reaction). Superimposing the AlphaFold2 AcsA structure and AcsA•AMP-propyl•CoA of S. enterica (PDB: 1PG3) shows this conformation being incompatible with CoA binding (lower closeup). K549 of AcsA is in direct interaction distance to the ribose of AMP-propyl using the same binding site as ATP (upper closeup). d AcsA and AcuA form a stable complex in solution. Analytical SEC experiments suggest that AcsA and AcuA form an apparent heterotetramer composed of two AcsA•AcuA heterodimers. The lane labelled with + shows the acetylated B. subtilis AcsA loaded as technical control for the immunoblot, and the lane labelled with M represents the protein molecular weight marker. The experiment was repeated independently three times with similar results. Source data are provided as Source Data file. e AcuA binding stabilizes AcsA in a conformation not capable of catalysing the first or second half-reaction. Upon AcuA binding a conformational change of the flexible AcsA C-terminal domain is observed. In this conformation AcsA K549 is placed directly into the active site of AcuA, coordinated by several glutamic/aspartic acid side chains. AcsA is inactive in this conformation as the CoA binding site is occupied by the C-terminal α-helix of AcuA.

Fig. 2. Binding of AcuA to AcsA inactivates AcsA activity and acetylation of AcsA at K549 by AcuA results in dissociation of AcuA from AcsA.

a Binding of AcuA to AcsA inactivates AcsA-activity. The preformed AcsA•AcuA complex (pre) or AcsA alone (post) was incubated with/without CoA (C)/acetyl-CoA (aC), ATP/AMP and acetate, as indicated. Afterwards, AcuA was also added to the post samples and all samples were incubated. The acetyl-CoA generated by acetyl-CoA synthetase (AcsA) activity was detected indirectly by immunoblotting with an anti-acetyl-lysine antibody (IB: AcK) assessing AcsA K549-acetylation. The result was confirmed in at least two independent experiments. Lane M represents the protein molecular weight marker. Source data are provided as Source Data file. b The AcsA•AcuA complex dissociates upon acetylation of AcsA by AcuA in the presence of acetyl-CoA. SEC runs were performed with the AcsA•AcuA complex pretreated with acetyl-CoA. AcsA elutes as homodimer (13.63 ml), AcuA as monomer (17.48 ml). The peak at 19.89 ml corresponds to CoA/acetyl-CoA. Immunoblotting (IB: AcK) shows AcsA lysine acetylation and Ponceau S-red staining (PoS) shows AcsA•AcuA dissociation. Lane + indicates the acetylated B. subtilis AcsA used as technical control, the M represents the protein molecular weight marker. The experiment was repeated independently three times with similar results. Source data are provided as Source Data file. c Acetylation of AcsA K549 is performed by AcuA and can be reversed by AcuC. AcsA or AcsA K549R (10 µM) was incubated with AcuA (2 µM) in the presence/absence of acetyl-CoA (0.5 mM) or the deacetylase AcuC (20 µM), as indicated. Samples lacking AcuC contained deacetylase inhibitor SAHA (50 µM). The lane labelled with M represents the protein molecular weight marker. The result was confirmed in at least three independent experiments. Source data are provided as Source Data file. d AcsA K549Q alters the conformation of the AcsA•AcuA complex. AlphaFold2 structure predictions suggest conformational changes in the AcsA•AcuA complex upon mutation of AcsA K549Q. The AcsA K549Q C-terminal domain moves towards AcuA. Q549 is not oriented towards the AcuA active site.

Fig. 3. Molecular dynamics simulations of the catalytic mechanism exerted by AcuA to acetylate AcsA.

a Main clusters for binding of acetyl-CoA, CoA and desulfo-CoA AcuA to the AcsA•AcuA complex obtained by MD simulations R2, R3, and R4, respectively (Supplementary Table 2). Binding pockets are coloured by the fraction of time with contacts between the respective CoA derivative and AcsA•AcuA and drawn as a surface for residues >20%. Acetyl-CoA binds in an extended, productive conformation and is tightly locked, while CoA and desulfo-CoA bind in an unproductive conformation with higher flexibility. Desulfo-CoA binds similar to CoA. The CoA thiol group is distant from K549 (balls) of AcsA. R108 and D141 of AcuA form a salt bridge and keep acetyl-CoA locked in, but not in the case of CoA or desulfo-CoA. Source data are provided as Source Data file. b Catalytic mechanism proposed for AcT activity of AcuA on K549 of AcsA based on the major cluster found in MD simulations (R2). Acetyl-CoA (sticks) binding to AcsA•AcuA. Acetyl-CoA binds in a productive conformation to AcuA in the AcsA•AcuA complex. Side chains in AcuA E102, K549, and W140 (ball and stick) and their interaction distances are highlighted (charts RS^A-RS^D; RS: reaction step). MD simulations suggest E102 of AcuA acting as a general base/acid during catalysis (RS^A). The acetyl-group is positioned deeply towards K549 of AcsA enabling acetyl-group transfer (RS^B). W140 is important for acetyl-CoA binding. Acetyl-CoA is bound by W140 of AcuA forming stacking interactions of the aromatic side chain and the 4-phosphopantotheine moiety of acetyl-CoA (RS^C). A tetrahedral intermediate is formed and stabilized by the main chain amide of W140 forming an oxyanion hole (RS^D). c MD simulations revealed that K549 of AcsA forms an electrostatic network with several negatively charged residues in AcuA (D82, E85, E97, E102 and E135). The graph denotes the average time fraction of K549 being occupied by either residue via hydrogen bonding in the presence of either acetyl-CoA (R1) or CoA (R3). Source data are provided as Source Data file.

Fig. 4. Mutational analyses to confirm the catalytic acetyltransferase mechanism.

a Mutation of the totally conserved residues E102 and W104 of AcuA abolish AcuA AcT activity. AcuA (5 µM) was incubated with acetyl-CoA (200 µM) and AcsA (20 µM) for 3 min, 10 min at 20 °C and 3 h at 37 °C, as indicated. E102A completely abolishes AcT activity of AcuA. W140F shows a reduced AcT activity, while W140A completely abolishes the AcT activity of AcuA towards AcsA. All other mutations do not impair AcT activity. Acetylation of AcsA was assessed by immunoblotting using an anti-AcK-AB (IB: AcK). Loading control (LC) was performed by Ponceau S-red staining (LC: PoS). The lane labelled with M represents the protein molecular weight marker. The result was confirmed in at least three independent experiments. Source data are provided as Source Data file. b Catalytic mechanism proposed for AcT activity of AcuA on K549 of AcsA. MD simulations and mutational analyses suggest that E102 of AcuA acts as a general base/acid during catalysis. W140 is important for acetyl-CoA binding. E102 of AcuA abstracts a proton from K549 of AcsA. This enhances the nucleophilicity of K549 to attack the electrophilic carbon of the acetyl-group of acetyl-CoA. A tetrahedral intermediate is formed and stabilized by the main chain amide of W140 forming an oxyanion hole. The protonated E102 acts as catalytic acid to protonate the CoA thiolate leaving group resulting in the collapse of the intermediate to finally result in acetylated K549 and CoA.

Fig. 5. Regulation of enzymatic and non-enzymatic acetylation of AcsA by AMP and ATP.

a AMP stimulates enzymatic acetylation of AcsA by AcuA. AcsA (10 µM) was incubated for 5 min at 20 °C with AcuA (1 µM) in the presence of acetyl-CoA (1 mM) and AMP, ATP, NAD⁺, CoA or AcP (each at 2 mM). Immunoblotting using an anti-AcK-AB (IB: AcK) shows that the presence of AMP and to a lesser extent ATP stimulates enzymatic acetylation of AcsA, whereas CoA impairs it. -acetyl-CoA: control lacking acetyl-CoA. The lane labelled with M represents the protein molecular weight marker. Ponceau S-red staining represents the loading control (LC). The result was confirmed in at least three independent experiments. Source data are provided as Source Data file. b ATP stimulates non-enzymatic acetylation of AcsA at K549, which is reversed by the deacetylase AcuC. AcsA (10 µM) was incubated for 24 h at 37 °C with acetyl-CoA (0.5 mM) and AMP/ATP (2 mM) in the presence of SAHA (50 µM)/AcuC (20 µM) as indicated. Immunoblotting using an anti-AcK-AB (IB: AcK) shows that AcsA is acetylated non-enzymatically in the presence of ATP. K549 is the major acetyl-group acceptor site for non-enzymatic acetylation in the presence of ATP as K549R shows no signal. In the presence of the deacetylase AcuC K549-acetylation is reversed. The lane labelled with M represents the protein molecular weight marker. The loading control was done by Ponceau S-red staining (LC: PoS). The result was confirmed in at least three independent experiments. Source data are provided as Source Data file.

Fig. 6. AcuA acetylation of AcsA K549 in the presence of AcP and CoA results in AcsA•AcuA dissociation.

a AcuA (5 µM) acetylates AcsA (25 µM) in the presence of CoA and AcP independent of AcuB (25 µM). Without AcuA no acetylation of AcsA is detectable neither in the presence of AcP (2 mM) alone nor in the presence of AcP and CoA (2 mM). As a control, acetylation of AcsA is observed in the presence of acetyl-CoA (1 mM). All samples were incubated for 2 h at 37 °C. AcsA acetylation was detected by immunoblotting (IB: AcK). Lane M: molecular weight marker. Ponceau S-red staining: loading control (LC: PoS). The result was confirmed in at least three independent experiments. Source data are provided as Source Data file. b Acetylation of AcsA by AcuA in the presence of AcP and CoA leads to AcsA•AcuA dissociation. Treatment of AcsA•AcuA with CoA and AcP results in acetylation of AcsA and dissociation of AcsA•AcuA (elution volume: from 13.50 ml to 17.41 ml (AcsA)) as shown by Ponceau S-red staining (PoS). Immunoblotting (IB: AcK) of SEC elution fractions shows acetylation of AcsA. Lane M: molecular weight marker. The experiment was repeated independently three times with similar results. Source data are provided as Source Data file. c Desulfo-CoA cannot replace CoA in AcP-dependent acetylation of AcsA catalysed by AcuA. In the phosphotransacetylase (Pta)-assay, AcuA (10 µM) and AcsA (10 µM) were incubated with/without CoA/desulfo-CoA (dsCoA; 1 mM) and/or AcP (1 mM) for 2 h or 5 h at 37 °C. Acetylation of AcsA is only observed in samples of AcsA•AcuA containing CoA and AcP. AcsA cannot produce acetyl-CoA using desulfo-CoA (Acs-assay). AcsA (10 µM) was incubated with acetic acid (0.5 mM), ATP (0.5 mM) and CoA/dsCoA (250 µM) for 20 min at 37 °C. Afterwards, AcuA (2 µM) was added (10 min, 20 °C) to detect the generated acetyl-CoA. Detection was done by immunoblotting (IB: AcK). Ponceau S-red staining: loading control (LC: PoS). Lane M: molecular weight marker. The result was confirmed in at least two independent experiments. Source data are provided as Source Data file.

Fig. 7. H139 is important for the Pta activity of AcsA•AcuA.

a Main cluster of AcP within the active site of AcuA in the AcsA•AcuA complex from MD simulation (R3). R108 binds the phosphate group of AcP (RS^A; RS: reaction step) and the acetyl-group is stabilized in a hydrophobic pocket. The thiol group of CoA is close to the acetyl-group C-atom of AcP (RS^B), priming it for transfer of the acetyl-group. MD simulations reveal H139 of AcuA coming in close contact with the thiol of CoA, allowing to abstract a proton to activate CoA (RS^C). Source data are provided as Source Data file. b Mutational studies reveal that H139 is essential for Pta activity. AcsA (20 µM) was incubated (2 h or 5 h; 37 °C) with AcuA/AcuA mutants (5 µM), AcP (1 mM), CoA (2 mM). EDTA (10 mM) was used to unravel a potential contribution of a metal ion on catalysis. The result was confirmed in at least three independent experiments. Source data are provided as Source Data file. c In the presence of AcP and CoA K549 of AcsA is acetylated within the AcsA•AcuA complex. Acetylation is increased upon mutation of E135 in AcuA and is reversed by the classical deacetylase AcuC. AcuA (1 µM) was mixed with AcsA (10 µM) in the presence of 1 mM AcP and 2 mM CoA, incubated for 3 h at 37 °C, and then, AcuC was added for 30 min at 37 °C. Acetylation of AcsA was analysed by immunoblotting (IB: AcK). Ponceau S-red staining: loading control (LC: PoS). The result was confirmed in at least three independent experiments. Source data are provided as Source Data file. d Mechanism of Pta activity via H139 acting as a general base. The results suggest H139 acting as a general base to deprotonate and thereby activate the CoA thiol group for nucleophilic attack on the electrophilic C-atom of the acetyl-group of AcP. This is enhanced by R108 of AcuA neutralizing negative charges at the AcP phosphate. A tetrahedral intermediate forms being resolved by H139 acting as general acid protonating the phosphate leaving group finally resulting in the formation of acetyl-CoA and inorganic phosphate.

Fig. 8. Model for regulation of AcsA activity via acetylation by AcT and Pta activity.

AMP-forming acetyl-CoA synthetase AcsA forms a complex with the Gcn5-related acetyltransferase AcuA, inhibiting AcsA activity by stabilizing AcsA in a conformation incompetent to catalyse the first or second half-reaction of acetyl-CoA synthesis. In the presence of acetyl-CoA, AcuA acetylates AcsA at K549 in the C-terminal domain (I: AcT activity). The acetyltransferase reaction proceeds with a general base-mechanism, i.e. E102 of AcuA. K549-acetylation abolishes electrostatic interactions of K549 with several aspartic acid/glutamic acid side chains of AcuA, i.e. D82, E85, E97, E102, E135. In turn, this lowers the affinity of AcuA to AcsA enabling the AcsA C-terminal domain to displace AcuA adopting the adenylation-conformation of the first half-reaction in the apo form. K549-acetylated AcsA is inactive as it interferes with ATP/acyl-AMP binding and it is not capable of orienting the substrate for nucleophilic attack on the α-phosphate of ATP. In the presence of AcP and CoA AcuA•AcsA catalyses a phosphotransacetylase reaction (II: Pta activity) involving H139 of AcuA as general base to deprotonate the thiol of CoA attacking the AcP generating acetyl-CoA, which can be used by AcuA in a subsequent acetyltransferase reaction to acetylate K549 of AcsA. This enables to switch off acetyl-CoA synthetase activity under conditions of high levels of CoA and AcP (acetate dissimilation). Furthermore, it dissociates the AcsA•AcuA complex under conditions of a low ratio of acetyl-CoA/CoA, under which AcuA activity is low, resulting in replenishing a pool of noncomplexed, readily activatable K549-acetylated AcsA. Re-activation of K549-acetylated AcsA is accomplished by deacetylation by the classical Zn²⁺-dependent deacetylase AcuC or a sirtuin deacetylase SrtN. Sections in light grey are shown to contextualize the regulation of AcsA activity into the metabolic pathways.