Anda di halaman 1dari 3

Cell Metabolism


The Logic Linking Protein Acetylation and Metabolism

Leonard Guarente1,*
1Paul F. Glenn Laboratory, Department of Biology, Massachusetts Institute of Technology, Cambridge MA 02139, USA

DOI 10.1016/j.cmet.2011.07.007

Protein acetylation now rivals phosphorylation in frequency of occurrence but is incompletely understood.
A picture is presented in which protein acetylation is linked to available energy via the NAD-dependent
deacetylases. This model suggests that protein acetylation regulates metabolic strategy and also helps store
energy in cells.

It is now apparent that a large number of the sirtuins are NAD-dependent deacety- whether it—as well as other acetylated
cellular proteins are acetylated. For lases, a high flux through glycolysis is metabolic enzymes in mammalian
example, recent studies have uncovered expected to lower the activity of SIRT1, cells—is a SIRT1 substrate, and whether
acetylation of a high fraction of mitochon- as well as the other sirtuins in the nucleus acetylation will generally be activating for
drial proteins (Kim et al., 2006), as well as and cytoplasm (SIRT2, SIRT6, and glycolytic enzymes.
most metabolic enzymes in the cytosolic/ SIRT7). Consistent with this idea, many The increase in protein acetylation in
nuclear pool (Wang et al., 2010; Zhao studies show that SIRT1 activity is low in times of plenty is analogous to the
et al., 2010). Acetylation in this latter growth conditions of glucose excess and increased phosphorylation of proteins by
cellular compartment is carried out by high during energy limitation (Finkel CDKs to drive the cell cycle under these
histone acetyl transferases (HATs) and et al., 2009; Imai and Guarente, 2010; conditions. More generally, phosphoryla-
deacetylation by class I and II histone Haigis and Sinclair, 2010). Low SIRT1 tion-based signaling pathways may
deacetylases (HDACs), as well as the activity should increase the acetylation impinge on sirtuins to exert an additional
class III deacetylases termed sirtuins. status of many metabolic enzymes in the layer of control over protein acetylation.
Among the sirtuins, SIRT1 appears to cytosolic/nuclear compartment, although How will the flow of carbon provide
have the broadest range of substrates any physiological effects on the activities acetyl-CoA for acetylation of cytosolic/
and affect the greatest number of physio- of HATs and HDACs may also be nuclear proteins under glycolytic con-
logical pathways (Finkel et al., 2009; Imai important. ditions? The pyruvate produced by
and Guarente, 2010; Haigis and Sinclair, Indeed in Salmonella, the acetylation glycolysis has two possible fates
2010). In the mitochondria, the source of level of many glycolytic enzymes was (Figure 1A). First, it can be reduced to
protein acetylation is not known, but the shown to be higher in glucose-grown cells fermentation products (bacteria), ethanol
most prominent deacetylase appears than in cells grown on a nonfermentable (yeast), or lactate (mammals), thereby
to be the sirtuin SIRT3 (Lombard et al., carbon source (Wang et al., 2010). This converting some of the NADH produced
2007). difference was at least partly mediated in glycolysis back to NAD. But second, it
Despite our growing understanding of by the bacterial sirtuin cobB, since acety- can be used to synthesize fatty acids as
the scope of protein acetylation, many lation of these enzymes appeared to be a primary mechanism of cell growth and
questions regarding this posttranslational higher in cobB mutants under all growth energy storage. In this case, the pyruvate
modification remain unanswered. How is conditions. One enzyme regulated in this is converted into acetyl-CoA by pyruvate
protein acetylation globally regulated? way is glyceraldehyde 3-phosphate dehydrogenase (PDH) in mitochondria,
And is there an underlying logic that links dehydrogenase (GAPDH), which converts and then rederived back into cytoplasmic
diet, metabolism, and protein acetyla- glyceraldehyde 3-phosphate to 1,3-bi- acetyl-CoA (Figure 1A). This latter step
tion? In this Essay, I attempt to integrate sphosphoglycerate in glycolysis, and occurs by transport of citrate from
recent findings from numerous labs to catalyzes the reverse reaction in gluco- mitochondria to the cytosol, where
suggest how the diet may dictate the level neogenesis. Remarkably, acetylated ATP-citrate lyase (ACL) converts it to
of protein acetylation, which in turn molds GAPDH drives the forward reaction used acetyl-CoA. ACL was recently shown to
the metabolic strategy of cells. in glycolysis, while deacetylated GAPDH be required for normal histone acetylation
Cells have two basic ways to produce is more effective at the reverse reaction in mammalian cells (Wellen et al., 2009),
ATP: glycolysis, which takes place in the used in gluconeogenesis. suggesting the primary pathway of stor-
cytosolic/nuclear location, and oxidative These simple findings suggest a feed- ing energy as fat may also provide the
phosphorylation, which occurs in the forward mechanism, in which diets rich carbon for protein acetylation in the cy-
mitochondria. When energy is abundant in carbohydrate energy drive glycolysis, tosolic/nuclear pool. By this view, the
in the diet, growth factors will drive convert NAD to NADH, inactivate sirtuins, high degree of protein acetylation in the
glucose uptake leading to activation of and increase acetylation and activity of cytosolic/nuclear pool during energy ex-
glycolysis, generation of ATP, and glycolytic enzymes (Figure 1A). Intrigu- cess can be seen as another mechanism,
conversion of NAD to NADH in the cyto- ingly, mammalian GAPDH is also acety- along with fat synthesis, of storing carbon
solic/nuclear pool (Figure 1A). Because lated, and it will be important to determine energy.

Cell Metabolism 14, August 3, 2011 ª2011 Elsevier Inc. 151

Cell Metabolism


(e.g., p53) or repressing (e.g., PGC-1a),

and this would allow flexibility in
regulating the many affected pathways.
The storage aspect of acetylation may
be important in transitioning from energy
excess to energy limitation (Figure 1B).
Under these conditions acetate generated
by SIRT1 deacetylation of many proteins
would be a substrate for acetyl-CoA
synthetase (ACS). Along with the oxidation
of fatty acids, this would drive the TCA
cycle and oxidative phosphorylation to
yield ATP and CO2. Reinforcing this model,
Figure 1. SIRT1 Links Protein Acetylation and Metabolic Strategy of Cells SIRT3, which is upregulated during this
(A) Glycolytic strategy for ATP production and flow of carbon (black text and arrows) in energy storage energy transition, deacetylates and acti-
tissues when energy is in excess. Mitochondria (Mito) and the TCA cycle are indicated and the remainder vates the mitochondrial ACS and long-
of the cell represents the cytosolic and nuclear compartments combined. Enzymes in blue italics are glyc-
eraldehyde-3-phosphate dehydrogenase (GAPDH), histone acetyl transferases (HATs), pyruvate dehy- chain acyl dehydrogenase (LCAD) in the
drogenase (PDH), and ATP-citrate lyase (ACL). Note a feed-forward loop of high glycolysis and NADH fatty acid oxidation pathway, both of which
production, low SIRT1 activity, and high acetylation and activity of glycolytic enzymes (e.g., GAPDH). generate acetyl-CoA (Verdin et al., 2010).
Carbon flows as follows. Glucose 6-phosphate (Glu-6P) is converted to pyruvate, which can be reduced
to lactate or converted to acetyl-CoA (Ac-CoA) by PDH. In this case, citrate from the TCA cycle is trans-
It is intriguing that SIRT1 cleaves NAD to
ported to the cytosol and converted to Ac-CoA for fatty-acid synthesis by ACL. Depletion of citrate from nicotinamide and O-acetyl ADP-ribose
the TCA cycle necessitates anapleurotic conversion of amino acids (AA) to a-ketogluterate to replenish the each reaction cycle, even though protein
deacetylation is energetically favorable.
(B) Oxidative strategy for ATP production and flow of carbon (black text and arrows) in energy producing
tissues upon energy deprivation. Glycolytic feed-forward loop in (A) is reversed, thereby activating SIRT1 NAD must subsequently be regenerated
and deacetylating and inhibiting glycolytic enzymes. SIRT1 generates O-acetyl ADP ribose and thus by salvage pathways (Haigis and Sinclair,
acetate (Ac) by deacetylating glycolytic proteins—and induces mitochondrial biogenesis and b-oxidation 2010; Imai and Guarente, 2010). How the
of fatty acids by deacetylating transcription factors (TFs) like PGC-1a. SIRT3 is also induced under energy
limitation. Note that deacetylation of acetyl-CoA synthetase (ACS) and long-chain acyl dehydrogenase cleavage and recycling of NAD make sense
(LCAD) by SIRT3 in the mitochondria is known to activate those enzymes and generate acetyl-CoA. biologically is an important topic for future
Conversely, diets poor in energy would the one hand, oxidative metabolism in In the above framework, protein acetyla-
limit glycolysis, activate SIRT1, and trigger mitochondria, for example fatty acid tion governs how cells choose glycolytic
protein deacetylation in the cytosolic/ oxidation, may be expected to increase versus oxidative metabolism as a function
nuclear pool (Figure 1B). These conditions acetyl-CoA levels in the organelle and of available energy and helps determine
would fit energetically with one of the main thereby favor protein acetylation the storage or utilization of carbon energy.
outputs of SIRT1 activity, the deacetyla- (Figure 1B). On the other, one of the mito- Because the substrates for protein acety-
tion of PGC-1a to promote mitochondrial chondrial proteins activated under these lation are NAD and acetyl-CoA, it would
biogenesis and oxidative metabolism conditions is the SIRT3 deacetylase. make sense that SIRT1, an NAD-depen-
(Rodgers et al., 2005). Oxidative phos- What does seem clear is that deacetyla- dent deacetylase, be a central metabolic
phorylation produces much more ATP per tion by SIRT3 activates mitochondrial sensor and mediator of this process. The
input glucose than fermentation, because metabolic enzymes for oxidative metabo- protein acetylation cycle depicted in the
the fuel is oxidized all the way to CO2. lism and detoxification of reactive oxygen figure may interact with other biological
Since fatty acids are oxidized in mitochon- species (Verdin et al., 2010). processes to determine the metabolic
dria under these conditions, acetyl-CoA Evolution has evidently selected for strategy of cells. For example PARP-1 inhi-
(and FADH) will be generated in this com- protein acetylation as a regulatory mecha- bition drives SIRT1 activity and oxidative
partment. In contrast, the generation of nism, as well as an energy-storage mech- metabolism by raising NAD levels (Bai
acetyl-CoA in the cytoplasm by ATP- anism, when energy is in excess. The et al., 2011). Finally, cancer cells often
citrate lyase will be limited. Both the high regulatory aspects minimally include display a preference for glycolytic/fermen-
SIRT1 activity and low cytosolic acetyl- histone acetylation to regulate gene tran- tative metabolism over oxidative metabo-
CoA levels would favor maintenance of scription, acetylation of metabolic en- lism, termed the Warburg effect. It will be
cytosolic/nuclear proteins in the deacety- zymes to favor glycolysis for ATP produc- important to determine whether manipula-
lated state during energy-poor diets. tion, and acetylation of many transcription tion of NAD or sirtuin levels is a general
Changes in the acetylation of mitochon- factors, such as p53, FOXO, PGC-1a, strategy to tip the balance of cancer cells
drial proteins with diet appear to vary from nuclear receptors, etc. to adjust expes- toward oxidative metabolism and inhibit
tissue to tissue and may be conceptually sion levels of pathways for oxidative their potential for tumorigenesis.
distinct from the above framework for versus glycolytic metabolism (Finkel
proteins in the nuclear/cytoplasmic pool et al., 2009; Imai and Guarente, 2010; Hai-
(Schwer et al., 2009). In fact, there appear gis and Sinclair, 2010). In the case of tran-
to be competing processes affecting scription factors, the acetylation of SIRT1- I thank the NIH and the Paul F. Glenn Medical
protein acetylation in this organelle. On sensitive lysines can either be activating Foundation for support.

152 Cell Metabolism 14, August 3, 2011 ª2011 Elsevier Inc.

Cell Metabolism

REFERENCES Kim, S.C., Sprung, R., Chen, Y., Xu, Y., Ball, H., Pei, Alt, F.W., and Lombard, D. (2009). Aging Cell 5,
J., Cheng, T., Kho, Y., Xiao, H., Xiao, L., et al. 604–606.
(2006). Mol. Cell 4, 607–618.
Bai, P., Cantó, C., Oudart, H., Brunyánszki, A.,
Verdin, E., Hirshey, M.D., Finley, L.W., and Haigis,
Cen, Y., Thomas, C., Yamamoto, H., Huber, A.,
Lombard, D.B., Alt, F.W., Cheng, H.L., Buck- M.C. (2010). Trends Biochem. Sci. 35, 669–675.
Kiss, B., Houtkooper, R.H., et al. (2011). Cell
Metab. 4, 461–468. enborg, J., Streeper, R.S., Mostoslavsky, R.,
Kim, J., Yancopoulos, G., Valenzuaela, D., Wang, Q., Zhang, Y., Yang, C., Xiong, H., Lin, Y.,
Murphy, A., et al. (2007). Mol. Cell. Biol. 24, Yao, J., Li, H., Xie, L., Zhao, W., Yao, Y., et al.
Finkel, T., Deng, C.X., and Mostoslavsky, R. (2009). 8807–8814. (2010). Science 327, 1004–1007.
Nature 460, 587–591.
Wellen, K.E., Hatzivassiliou, G., Sachdeva, U.M.,
Rodgers, J.T., Lerin, C., Haas, W., Gygi, S.P., Spie- Bui, T.V., Cross, J.R., and Thompson, C.B.
Haigis, M.C., and Sinclair, D.A. (2010). Annu. Rev. gelman, B.M., and Puigserver, P. (2005). Nature (2009). Science 324, 1076–1080.
Pathol. 5, 253–295. 434, 113–118.
Zhao, S., Xu, W., Jiang, W., Yu, W., Lin, Y., Zhang,
Imai, S., and Guarente, L. (2010). Trends Pharma- Schwer, B., Eckersdorff, M., Li, Y., Silva, J.C., Fer- T., Yao, J., Zhou, L., Zeng, Y., Li, H., et al. (2010).
col. Sci. 31, 212–220. min, D., Kurtev, M.V., Giallourakis, C., Comb, M.J., Science 327, 1000–1004.

Cell Metabolism 14, August 3, 2011 ª2011 Elsevier Inc. 153