Protein acetylation has emerged as a prominent post-translational modification that can occur on a wide variety of proteins. The metabolite acetyl-CoA is a key intermediate in energy metabolism that also serves as the acetyl group donor in protein acetylation modifications. Therefore such acetylation modifications might be coupled to the intracellular availability of acetyl-CoA. In the present article, we summarize recent evidence suggesting that the particular protein acetylation modifications enable the regulation of protein function in tune with acetyl-CoA availability and thus the metabolic state of the cell.
Fundamental aspects of the regulation of cell growth and proliferation remain incompletely understood. In investigations of the mechanisms of cell growth control, researchers have tended to focus on biological macromolecules, such as DNA, RNA and proteins, without much regard for small molecule metabolites. In recent years, a renewed interest in cellular metabolism and its influence on cell growth regulation has emerged. Cellular metabolite pools may not always be static, but may instead fluctuate in a manner dependent on cues from the environment or the nutritional state of a cell or organism. Since many of these metabolites are also key substrates or cofactors for enzymes that catalyse post-translational modifications, it follows that changes in their levels could significantly affect the activity of such enzymes and thus biological regulation [1–3]. We have observed that acetyl-CoA levels oscillate as a function of the growth and metabolic state of yeast cells. In the present article, we briefly review emerging evidence suggesting that these acetyl-CoA fluctuations are sensed via distinctive protein acetylation modifications that consequently affect the regulation of cell growth and metabolism.
Sources and fates of acetyl-CoA
Acetyl-CoA is a central metabolite in cellular energy metabolism. Intracellularly, acetyl-CoA is considered to be compartmentalized between mitochondrial and nucleocytosolic pools . Typically, glucose is a major source of acetyl-CoA. Glucose is converted into pyruvate through glycolysis in the cytoplasm. Pyruvate then is transported into mitochondria and converted into acetyl-CoA by the pyruvate dehydrogenase complex. Alternatively, mitochondrial pools of acetyl-CoA can be generated from fatty acid oxidation, CoA transferases , or mitochondrial acetyl-CoA synthetase enzymes [6,7]. Cells can also produce acetyl-CoA from available amino acids, such as glutamine and leucine . Cytosolic acetyl-CoA can be produced from mitochondrial-derived citrate via the enzyme ATP citrate lyase, or from acetate by the cytosolic acetyl-CoA synthetase enzymes [9,10]. Once generated, acetyl-CoA has multiple fates. Acetyl-CoA can be coupled to oxaloacetate for entry into the TCA (tricarboxylic acid) cycle as citrate, which then through a series of transformations is used for synthesis of ATP or particular amino acids and cofactors. Acetyl-CoA in the cytosol can be carboxylated to form malonyl-CoA, the initial step in fatty acid synthesis. Acetyl-CoA is also used as a substrate in the biosynthesis of numerous other metabolites. In addition, as discussed below, acetyl-CoA is the acetyl group donor for numerous acetylation modifications that occur on proteins [4,11–14].
Protein acetylation and the regulation of cell growth
Beyond its role in energy metabolism and biosynthesis, acetyl-CoA also functions as the acetyl group donor for protein acetylation modifications. Recently, we showed that intracellular acetyl-CoA levels oscillate as a function of the growth status of yeast cells and can in turn influence the abundance of certain protein lysine acetylation modifications [11,12]. That is, particular acetylation modifications could in fact be regulated by the availability of this acetyl donor.
During continuous growth under nutrient-limited conditions, yeast cells can become highly synchronized. The cell population undergoes robust oscillations in oxygen consumption and alternates between growth and quiescent-like phases in these ‘YMCs (yeast metabolic cycles)’ . Metabolite profiling studies revealed that acetyl-CoA levels oscillate in a very periodic manner as a function of the YMC . Upon exit of the reductive charging quiescent-like phase and entry into the oxidative growth phase, intracellular trehalose and glycogen reserves rapidly decrease which is accompanied by an increase in glucose . Shortly after this rise in glucose, acetyl-CoA levels sharply increase, precisely correlating with the temporal window of growth gene induction [15,16]. Acetyl-CoA levels decrease as cells enter the reductive charging phase, whereupon stress and survival genes are induced. Intracellular acetyl-CoA in batch culture-grown cells also increases substantially upon nutrient repletion . However, before these observations, a role for acetyl-CoA in growth regulation was perhaps not considered because of its abundance and ease of synthesis in cells grown in the glucose-rich medium.
Thus we hypothesized acetyl-CoA may function as a metabolic cue to induce cell growth upon carbon source repletion. To investigate this possibility, we added various carbon sources, including glucose, ethanol and acetate, to cells in the reductive charging phase. Each of these carbon sources was able to induce growth phase entry in the YMC. Following addition of 13C-labelled acetate, [13C]acetyl-CoA levels increased rapidly within a minute, and further induced synthesis of unlabelled 12C-acetyl-CoA, probably from other carbon sources to further promote growth . We then surveyed acetylated proteins following addition of [13C]acetate. Besides histones, Spt7p, Sgf73p and Ada3p were identified to be acetylated precisely upon growth phase entry, in tune with the increase in intracellular acetyl-CoA. These proteins are components of the transcriptional co-activator complex SAGA, which harbours histone acetyltransferase activity [18,19]. We further observed that numerous sites on histone H3 (Lys9, Lys14, Lys18, Lys23 and Lys27) and H4 (Lys5, Lys8 and Lys12) linked to gene activation were also acetylated precisely in tune with the spike in acetyl-CoA levels in oxidative phase. Subsequent ChIP-Seq analysis revealed that the acetylated histones were present primarily at a distinctive set of more than 1000 genes important for cell growth that are induced during oxidative phase. Thus the increase in acetyl-CoA induces SAGA to acetylate histones at each of these genes, in turn activating their transcription and enabling cells to commit to growth (Figure 1). Acetylation of histones at these genes is dependent on Gcn5p, the acetyltransferase enzyme within SAGA, and mutation of GCN5 or these components of SAGA disrupt the YMC .
Acetyl-CoA increase stimulates transcription of growth genes
Moreover, we observed that CLN3 belongs to this group of oxidative phase growth genes. CLN3 is the earliest expressed G1-phase cyclin, which promotes the G1–S transition in the cell cycle [20,21]. Although CLN3 transcription is almost immediate upon glucose repletion, it is not regulated by canonical glucose signalling pathways despite a requirement for glycolysis [22,23]. Thus it remained unclear how glucose stimulates CLN3 transcription to promote entry into the cell division cycle. In the YMC, CLN3 mRNA peaks shortly after the peak in acetyl-CoA levels [15,16]. In investigating the possibility that acetyl-CoA might also induce CLN3, we found that acetate alone was able to in-duce CLN3 transcription to a similar extent to either glucose or rich medium . These observations suggest that acetyl-CoA is the common downstream metabolite of acetate and glucose, which functions as a direct inducer of CLN3 transcription. Furthermore, CLN3 transcription is also dependent on SAGA-catalysed histone acetylation just like the other growth genes. Thus CLN3, which gates entry into START of the cell division cycle, is induced together with ribosome biogenesis and other growth-related genes by acetyl-CoA.
These studies using the YMC have revealed that acetyl-CoA plays a key role in regulating particular acetylation events that are important for cell growth. They suggest that the acetyltransferase Gcn5p present within SAGA has the distinctive ability to acetylate its substrates in concert with surges in acetyl-CoA. Consistent with this idea, additional Gcn5p substrates, including the transcriptional co-activator of ribosomal subunit genes Ifh1p and the subunit of the SWI/SNF chromatin remodelling complex Snf2p [24,25], were also dynamically acetylated in tune with acetyl-CoA. Interestingly, proteins that are acetylated but which are not Gcn5p substrates, were not dynamically acetylated in tune with the acetyl-CoA fluctuations that occur in the YMC . Taken together, these data suggest that in budding yeast, the acetylation of Gcn5p substrates is driven by intracellular acetyl-CoA.
Protein acetylation and the regulation of metabolism
Protein acetylation is well-known for its role in chromatin-associated processes within the nucleus, such as transcription, replication and DNA damage repair. In recent years, the improvement of mass spectrometry technologies subsequently enabled protein acetylation to be profiled on a global level in precise detail [13,14,26–28]. In these proteomic surveys, a huge number of proteins were newly identified to be acetylated, not only tremendously expanding the number of known acetylated proteins, but also establishing protein acetylation as one of the major post-translational modifications along with phosphorylation.
Besides nuclear proteins, a large number of cytosolic and mitochondrial proteins are also observed to be acetylated [26,28–30]. Non-nuclear protein acetylation has been found to occur vastly at very low stoichiometry levels , which might explain how such substrates have been overlooked until recently. Although the non-nuclear acetylated proteins span multiple functional categories including cell cycle, cytoskeleton, chaperone, and signal transduction, a large proportion of them are involved in energy production and intermediary metabolism [13,14,32]. Acetyl-CoA synthetase was one of the first reported metabolic enzymes regulated by acetylation [33,34]. The acetylation of Lys609 on Salmonella acetyl-CoA synthetase blocks adenylate intermediate synthesis, thus disrupting the enzyme's ability to synthesize acetyl-CoA from acetate. This acetylation modification is removed by sirtuins [33,34]. Subsequently, acetylation was detected on almost every enzyme in intermediary metabolism including glycolysis/gluconeogenesis, TCA cycle, amino acid metabolism and urea cycle [13,14,32]. Perhaps corresponding to different metabolic demands, acetylation of each enzyme may differentially affect enzymatic activity or protein stability (Figure 2). The p300 acetyltransferase has been reported to catalyse many of these acetylation events, and in most instances acetylation of these enzymes appears to have a negative inhibitory effect on activity . Interestingly, acetylated lysine residues were significantly more enriched in structured regions of the protein [26,27]. This is in stark contrast with phosphorylation that tends to occur more frequently in unstructured regions of proteins .
Many metabolic enzymes are acetylated
Which protein acetylation modifications are driven by acetyl-CoA?
The reports of thousands of newly acetylated proteins raise the interesting question of which might be linked to acetyl-CoA availability? Although the number of acetylation sites might be comparable with the number of phosphorylation sites [13,14,26–28], acetyltransferase enzymes are far fewer in number compared with the kinases. Anabolic and catabolic enzymes of a particular metabolic pathway also appear to be acetylated, sometimes concurrently. In yeast, Gcn5p has the ability to acetylate its substrates in tune with acetyl-CoA . However, although the p300 acetyltransferase in other species may be activated by autoacetylation in tune with acetyl-CoA , whether p300 or other acetyltransferases catalyse acetylation of substrates in tune with acetyl-CoA in vivo is less clear.
To explain how such a large number and variety of proteins could be acetylated by a small number of acetyltransferases, it has been suggested that acetyl-CoA might non-enzymatically acetylate proteins under certain conditions [32,37]. Acetyl-CoA was shown to non-enzymatically acetylate histones and synthetic polylysine residues at concentrations of 0.5 mM, which could be accelerated at higher pH . In mitochondria, the acetyl-CoA concentration has been estimated at 0.5–1.0 mM . Furthermore, due to the active proton transporter on the mitochondrial membrane, the pH of the mitochondrial matrix is higher (7.9–8.0) than that in the cytosol (7.2), which could support increased kinetics of the non-enzymatic acetylation reaction. A number of studies recently showed that in vitro mitochondrial protein extract became hyperacetylated if acetyl-CoA was provided [31,39]. Furthermore, in addition to lysine acetylation, other lysine acylation modifications were also observed to occur in physiological conditions including succinylation, butyrylation and propionylation [39–41]. Thus conditions that favour high intracellular concentration of these acyl-CoA metabolites might promote increased non-enzymatic acylation modification of proteins. Some of these could be regulatory, whereas others could be spurious and detrimental to protein function. In the case of the acetyl-CoA synthetases, the acetylation of a lysine residue in the active site inactiva-tes the enzyme as a means of feedback regulation [33,34]. Although no acetyltransferase enzyme has been reported to catalyse this modification, the sirtuins catalyse its removal to activate the enzyme [33,34]. Thus for these enzymes, high acetyl-CoA may promote non-enzymatic acetylation as a means of regulation.
However, despite the conditions in the mitochondria that might favour non-enzymatic protein acylation, a recent report identified GCN5L1 as a putative mitochondrial-localized acetyltransferase that could influence the mitochondrial acetylome . GCN5L1 was further shown to regulate mitochondrial protein homoeostasis . Future studies will reveal the extent through which such modifications are either catalysed or non-enzymatic. In other studies that implicate acetylation modifications in cellular metabolism and possibly acetyl-CoA, high-fat diet and chronic ethanol consumption were found to result in global hepatic lysine hyperacetylation [44,45]. Acetylation of metabolic enzymes, including acetyl-CoA synthetase, was observed to change over the circadian cycle [46,47]. Calorie restriction was also found to increase protein acetylation, particularly in mitochondria and could reflect an increase in mitochondrial acetyl-CoA derived from fatty acid oxidation [28,30,48]. The logic and extent through which intracellular or compartment-specific acetyl-CoA pools regulate these acetylation modifications under these various settings awaits further detailed investigation.
In summary, recent studies have now begun to reveal the underappreciated impact of metabolism and small molecule metabolites on various cellular processes. In the present article, we have briefly reviewed just one such key metabolite, i.e. acetyl-CoA, and its role in regulating protein function through acetylation modifications. Acetyl-CoA not only lies at the intersection of many key metabolic pathways, but also serves as a substrate for acetyltransferase enzymes. Certain acetyltransferases might sense the acetyl-CoA concentration through their intrinsic acetyl-CoA binding affinity and respond by acetylating downstream target proteins, such as histones. Acetyl-CoA may also acetylate proteins non-enzymatically to regulate particular metabolic enzymes, such as the acetyl-CoA synthetases. Thus acetyl-CoA levels, which are influenced by nutrient and carbon source availability, can be likened to a rheostat that reflects the metabolic state of cells. The importance of acetyl-CoA in cellular and organismal function is only beginning to be established, and future investigations will probably reveal additional intersections between acetyl-CoA and numerous biological regulatory pathways.
Coenzyme A and Its Derivatives in Cellular Metabolism and Disease: A Biochemical Society Focused Meeting held at Charles Darwin House, London, U.K., 20–21 March 2014. Organized and Edited by Ivan Gout (University College London, U.K.), Suzanne Jackowski (St. Jude Children's Research Hospital, U.S.A.) and Ody Sibon (University of Groningen, The Netherlands).
We apologize to those whose work could not be cited owing to word and reference limitations on this mini-review.
We acknowledge funding support from the National Institutes of Health [grant number R01GM094314].
Benjamin Tu is a consultant and shareholder of Peloton Therapeutics, Inc.