|Home | About | Journals | Submit | Contact Us | Français|
Sirtuins are highly conserved NAD+-dependent protein deacetylases and/or ADP-ribosyltransferases that can extend the lifespan of several lower model organisms including yeast, worms and flies. The seven mammalian sirtuins, SIRT1 to SIRT7, have emerged as key metabolic sensors that directly link environmental signals to mammalian metabolic homeostasis and stress response. Recent studies have shed light on the critical roles of sirtuins in mammalian energy metabolism in response to nutrient signals. This review focuses on the involvement of two nuclear sirtuins, SIRT1 and SIRT6, and three mitochondrial sirtuins, SIRT3, SIRT4, and SIRT5, in regulation of diverse metabolic processes.
Yeast silent information regulator 2 (Sir2) protein and its homologues in other prokaryotes and eukaryotes, also known as sirtuins, are highly conserved NAD+-dependant protein deacetylases and/or ADP-ribosyltransferases 1-3. In yeast, Sir2 is of great importance in the maintenance of the silent chromatin at the mating-type loci, telomeres, and rRNA-encoding DNA repeats (reviewed in 4). However, sirtuins in other model organisms have been increasingly recognized as crucial regulators for a variety of cellular processes, ranging from energy metabolism and stress response to tumorigenesis and aging 5.
About a decade ago, Sir2 was first implicated as a limiting component of yeast longevity 6, 7. Since then, many members of this family have been identified as key longevity regulators in species ranging from yeast to fly (reviewed in 8, 9). For example, in S. cerevisiae, deletion of the Sir2 gene dramatically shortens lifespan, whereas an extra copy of the Sir2 extends lifespan by 40% 7. In C. elegans, overexpression of Sir2.1, a Sir2 homologue, extends lifespan by up to 50% 10. Similar results are seen in Drosophila 11. The mammalian genome encodes seven sirtuins, named SIRT1 to SIRT7 12. Currently, the role of mammalian sirtuins in the regulation of mammalian aging is an area of intense research.
The seven mammalian sirtuin proteins share a highly conserved NAD+-binding and catalytic core domain, but have distinct flanking N- and C-terminal extensions 12 (Figure (Figure1).1). Many mammalian sirtuins bears a NAD+-dependant protein deacetylase activity. A variety of acetyl-proteins have been identified as their substrates at different physiological and pathological conditions (Table (Table1),1), although whether p53 is a true substrate for SIRT7 is still not clear. The only reported activity of SIRT4, on the other hand, is the NAD+-dependant ADP-ribosyltransferase activity 13, 14. The divergent N and C-termini of sirtuins are responsible for their variation binding partners, substrates, and subcellular localization (9, Figure Figure1).1). Three of the mammalian sirtuins, SIRT1, SIRT6, and SIRT7, are nuclear proteins (Figure (Figure1).1). SIRT1, while predominately a nuclear enzyme, can also shuttle between cytosol and nucleoplasm in various tissues in response to different environmental signals 15. SIRT6 is a nuclear, chromatin-bound protein 16, whereas SIRT7 is highly enriched in nucleolus 17. Three additional mammalian sirtuins, SIRT3, SIRT4, and SIRT5, are found in mitochondria. They participate in a number of metabolic and survival processes associated with the mitochondrial activity (reviewed in 18). The other mammalian sirtuin, SIRT2, is primarily cytosolic.
Despite diverse subcellular localizations and a broad range of substrate specificities, the activity of all sirtuins is directly controlled by cellular NAD+ levels, which is an indicator of cellular metabolic status. The activity of these enzymes is also inhibited by their common enzymatic product, nicotinamide 19, and possibly by NADH 20. It is therefore not surprising that the activity of sirtuins changes in response to environmental cues that impact cellular metabolic state. A prominent example of this is Caloric Restriction (CR), a 20-40% reduction in calories consumed below ad libitum intake without malnutrition. CR has been shown to extend the median and maximum life span of numerous organisms including yeast, flies, worms, fish, as well as rodents and other higher mammals 21. In mammals, CR has been seen to ameliorate many of the pathologies associated with obesity and metabolic syndrome 22, 23.
CR is thought to promote lifespan extension through multiple signaling pathways. On the one hand, CR decreases the activity of pro-aging pathways such as insulin and growth hormone signaling and oxidative stress. On the other hand, CR stimulates the activity of cellular stress-resistance pathways including DNA repair and autophagy 24, 25, promoting cell survival in response to environmental stress. It has been shown that in yeast, CR reduces cellular NADH concentrations, thereby increasing the NAD+/NADH ratio and promoting Sir2 activity 20. In a number of lower model organisms, sirtuins are required for the lifespan extension provided by CR 11, 26. Whether sirtuins mediate the life-extending effects of CR in mammals is currently an area of great interest. Recent studies have shown that in mice, SIRT1 protein levels are elevated during CR in the brain, white adipose tissue (WAT), muscles, liver, and kidney 27, 28. CR also induces the expression of SIRT3, particularly in brown adipose tissue (BAT) 29, 30. However, during CR in mice, cellular NAD+ levels and the NAD+/NADH ratio fluctuate significantly depending on tissue type 31, suggesting that sirtuins have tissue-specific responses to CR. Nevertheless, it appears that SIRT1 regulates energy metabolism and physical responses to CR 32-34, while SIRT3 is able to mediate CR-associated reduction of oxidative damage, preventing age-associated hearing loss 35, 36. Moreover, transgenic mice overexpressing SIRT1 display multiple phenotypes resembling those of CR mice, including lower body weight, greater metabolic activity, and reduced serum levels of cholesterol, adipokines, insulin, and glucose 37-39. Taken together, these studies highlight the importance of sirtuins in CR-mediated prevention of age-associated functional decline, suggesting that sirtuins may be important therapeutic targets for a number of age-related diseases.
As the best-studied mammalian sirtuin, SIRT1 has been implicated in a variety of metabolic processes including hepatic lipid metabolism and gluconeogenesis, pancreatic insulin secretion, fat cell accumulation and maturation, central nutrient sensing, as well as circadian regulation of metabolism. Further elucidation of the role SIRT1 plays in energy metabolism will likely provide key insights into developing treatments for obesity-induced metabolic disease.
The liver is a central metabolic organ controlling key aspects of lipid and glucose metabolism in response to nutritional and hormonal signals 40. For example, in fasting conditions, the liver converts lipid and glycogen stores into available energy through the process of fatty acid β-oxidation and glycogenolysis/gluconeogenesis. In the fed condition, metabolic programs in the liver are switched on to store energy in the form of glycogen and lipid droplets.
Recent reports have shown that SIRT1 is an important regulator of hepatic glucose and lipid metabolism (Figure (Figure2).2). For instance, during short-term fasting phase, SIRT1 inhibits TORC2, a key mediator of early phase gluconeogenesis, leading to decreased gluconeogenesis 41. Prolonged fasting leads to increased SIRT1 deacetylation and activation of PGC-1α, an essential co-activator for a number of transcription factors, resulting in increased fatty acid oxidation and improved glucose homeostasis 42-44. SIRT1 also deacetylates and activates transcriptional factor Foxo1, resulting in increased gluconeogenesis 45. Consistently, adenoviral knockdown of SIRT1 reduces expression of fatty acid β-oxidation genes in the liver of fasted mice 46. Specific deletion of the exon 4 of the hepatic mouse SIRT1 gene, which resulted in a truncated non-functional SIRT1 protein, impairs fatty acid β-oxidation, thereby increasing the susceptibility of mice to high-fat diet induced dyslipidemia, hepatic steatosis, inflammation and endoplasmic reticulum (ER) stress 43. A complete deletion of hepatic SIRT1 by floxing exons 5 and 6 leads to the development of liver steatosis even under normal chow diet 47. Conversely, hepatic overexpression of SIRT1 mediated by adenovirus attenuates hepatic steatosis and ER stress, and restores glucose homeostasis in mice 48, confirming the essential role of SIRT1 in maintaining hepatic metabolic homeostasis.
SIRT1 also regulates hepatic cholesterol and bile acid homeostasis through direct modulation of the liver X receptor (LXR), farnesoid X receptor (FXR), and the sterol regulatory element binding protein (SREBP) family of transcription factors 49-52. LXR and FXR are nuclear receptors that function as important cholesterol and bile acid sensors 53. We have previously shown that SIRT1 can directly deacetylate LXRs, resulting in increased LXR turnover and target gene expression 49. Systemic deletion of SIRT1 in mice results in decreased serum HDL levels 49. Recently, Kemper et al. showed that the bile sensor FXR is also a target of hepatic SIRT1 in metabolic regulation 50. Acetylation of FXR inhibits its activity and downregulation of hepatic SIRT1 increases FXR acetylation, causing deleterious metabolic outcomes such as liver steatosis and decreased bile output. Walker et al. and Ponugoti et al. further showed that SIRT1 may also regulate hepatic lipid metabolism through deacetylation of SREBPs 51, 52, critical regulators of lipid metabolism that promote expression of lipogenic and cholesterolgenic genes involved in lipid storage 54. These two reports revealed that SIRT1 can directly deacetylate SREBP, and that SIRT1 activity is important in the fasting-dependent attenuation of SREBP 51, 52. In addition, chemical activators of SIRT1 inhibit SREBP target gene expression in vitro and in vivo, correlating with attenuated liver steatosis in diet-induced and genetically obese mice. In summary, these findings imply that hepatic SIRT1 plays a critical role in metabolic regulation and activation of SIRT1 in the liver may prove beneficial in treating obesity-associated diseases.
Adipose tissue functions both to store fat and as a conduit for hormone signaling. For example, WAT-derived hormones such as leptin and adiponectin control energy balance, glucose regulation, and fatty acid catabolism. Among numerous factors involved in adipose tissue differentiation, the nuclear receptor PPARγ plays an essential role in modulating fatty acid storage and glucose metabolism 55. SIRT1 has been shown to repress PPARγ, thereby suppressing the expression of genes such as the mouse aP2 gene and fat storage (27, Figure Figure2).2). Consistently, treatment of mice on a high fat diet with resveratrol, a polyphenol that activates SIRT1, has been shown to protect against high-fat induced obesity and metabolic derangements 56-58. Although the mechanisms of resveratrol's apparent protective effect on metabolic disorders are not fully understood, questions still remain whether or not resveratrol directly activates SIRT1 or functions through multiple signaling pathways 59, 60, these findings demonstrate that SIRT1 acts in concert with lipid regulating transcription factors to adapt gene transcription to changes in nutrient levels.
As an important non-shivering thermogenesis organ, brown adipose tissue (BAT) plays an essential role in survival of non-shivering animals and neonates. However, the role of SIRT1 in BAT is still not clear. On the one hand, SIRT1 in BAT may directly promote BAT differentiation through repression of the MyoD-mediated myogenic gene expression signature and stimulation of PGC-1α-mediated mitochondrial gene expression 61. On the other hand, SIRT1 also appears to regulate BAT differentiation and functions through non-cell autonomous mechanisms. It has been shown that SIRT1 in propiomelanocortin (POMC) expressing neurons selectively controls perigonadal WAT-to-BAT-like remodeling to increase energy expenditure in female mice 62. Consistent with these observations, activation of SIRT1 by a specific SIRT1 activator, SRT1720, enhances oxidative metabolism in skeletal muscle, liver, and BAT 63. Together, these data indicate that SIRT1 modulates BAT through both cell autonomous and non-cell autonomous mechanisms.
Pancreatic β cells are systemic metabolic sensors that release insulin in response to blood glucose levels. Dysfunction of these cells is the leading cause of type 1 diabetes mellitus, and partially contributes to the pathogenesis of type 2 diabetes. SIRT1 has been shown to be a major positive regulator for pancreatic insulin secretion, which in turn triggers glucose uptake and utilization. For example, it has been reported that increased dosage of SIRT1 in pancreatic β cells improves glucose tolerance and enhances insulin secretion in response to glucose in mice 64, whereas deletion of SIRT1 impairs glucose-stimulated insulin secretion 65. In both studies, SIRT1 has been shown to promote insulin secretion by repressing transcription of the uncoupling protein 2 (UCP2) (Figure (Figure2).2). Furthermore, it has been shown that treatment with resveratrol potentiates glucose-stimulated insulin secretion in β cells 66. In line with these observations, activation of SIRT1 by its activators in animals protects against high-fat induced obesity and insulin resistance 56-58, and modest overexpression of SIRT1 has a protective effect against high-fat induced glucose intolerance 38, 39. These observations suggest that modulation of SIRT1 activity may regulate whole-body glucose metabolism at both systemically and locally.
The brain is essential in controlling whole-body metabolism through both neurologic and endocrine functions. The hypothalamus/pituitary axis is the primary brain structure that interprets adiposity and nutrient-related inputs to regulate energy homeostasis. It has been shown that both CR and fasting enhance SIRT1 expression and activity in the hypothalamus 67, 68. Mice lacking SIRT1 in the brain show specific anterior pituitary cell defects and fail to mediate changes in pituitary signaling and physical activity in response to CR 34, while brain-specific SIRT1 transgenic mice display enhanced neural activity in the hypothalamus 68. These findings suggest that SIRT1 in the brain may function as a link between the hypothalamus/pituitary hormones and CR longevity pathways in mammals.
However, the role of SIRT1 in the central control of whole body energy homeostasis is complex (Figure (Figure2).2). In the hypothalamus, the anorexigenic POMC expressing neurons and the orexigenic agouti-related protein (AgRP) expressing neurons are the major regulators of feeding and energy expenditure 69. The POMC neurons produce satiety peptides thereby inhibiting food intake after feeding, while the AgRP neurons promote feeding in response to fasting and CR. It appears that SIRT1 plays distinct roles in these two neuron populations. A recent study has shown that inhibition of hypothalamic SIRT1 activity reverses the fasting induced decrease of FOXO1 acetylation, resulting in increased POMC and decreased AgRP expressions, thereby decreasing food intake and body weight gain 67. This hypothalamic activity of SIRT1, which is in contrast to its anti-obesity functions in the peripheral tissues, appears to act through AgRP neurons, as AgRP neuron-specific deletion of SIRT1 decreases AgRP neuronal activity, thereby alleviating the inhibitory tone on the POMC neurons, resulting in decreased food intake and body weight 70. Specific deletion of SIRT1 in POMC neurons in mice, on the other hand, causes a blunted response to leptin signaling and reduced energy expenditure, leading to hypersensitivity to diet-induced obesity 62. Nevertheless, these studies confirm that SIRT1 is an essential element in the periphery-central feedback circuits that mediate normal responses to nutrient deprivation. Consistent with these notions, central administration of small molecule SIRT1 activator has shown promise in controlling of diet-induced obesity. For example, long-term intracerebroventricular infusion of resveratrol normalizes hyperglycemia and greatly improves hyperinsulinemia in mice with diet-induced obesity and diabetes 71, although resveratrol may elicit its functions through both SIRT1 dependent and independent pathways. In summary, SIRT1 activity appears to be an important player in the central regulation of nutrient sensing. As relatively little is known at this stage, this area of sirtuin biology may yield important discoveries in coming years.
All living entities on Earth have endogenously driven 24-hour cycles in many biological processes. This circadian rhythm depends on internal clocks that work in part through chromatin modification and epigenetic control of gene expression 72. In mammals, the circadian clock is largely controlled by negative-feedback loops mediated by the heterodimeric transcription factors CLOCK-BMAL1 and their transcriptional targets, including the PER and CRY proteins that in turn directly repress CLOCK-BMAL1 activity, as well as REV-ERB and ROR nuclear receptors that control BMAL1 expression 73. The CLOCK protein itself is a transcriptional activator that functions as a histone acetyltransferase (HAT) 74, highlighting the importance of chromatin modification in the regulation of circadian gene expression.
SIRT1 has recently been linked to the regulation of the circadian rhythm (Figure (Figure2).2). Two studies have shown that SIRT1 interacts with CLOCK-BMAL1 to directly regulate the amplitude of circadian clock-controlled gene expression through deacetylation of PER2 and/or BMAL1 75, 76. However, whether it is the amount of SIRT1 or its activity that is cyclically regulated by the circadian clock remains unclear. Two recent studies appear to favor the circadian regulation of SIRT1 activity 77, 78. The expression of NAMPT, an enzyme that controls a rate-limiting step in NAD+ biosynthesis, is directly controlled by CLOCK-BMAL1. It has been proposed that this regulation leads to circadian oscillation of cellular NAD+ levels, resulting in a cyclical regulation of SIRT1 activity 77, 78. Together, these studies add a new feedback loop in the circadian clock that involves CLOCK-BMAL1, NAMPT, NAD+, and SIRT1, and provide an important link between the circadian clock and cellular metabolism.
It remains to be determined whether the observed oscillation of cellular NAD+ levels 77, 78 is indeed responsible for the oscillation of SIRT1 activity observed in earlier studies 76. An earlier study had shown that the activity of immuno-purified endogenous SIRT1 proteins display a circadian oscillating pattern when measured in vitro with fixed amount of exogenous NAD+ (Figure (Figure1D1D in 76), suggesting that post-translational modifications, or protein-protein interactions also play a role in the circadian regulation of SIRT1 activity. SIRT1 has several interacting partners that can directly inhibit or activate its activity 79-82. Its activity can also be modulated by phosphorylation 83-85, particularly by DYRK1A 86, an essential clock component that governs the rhythmic phosphorylation and degradation of CRY2 protein 87. Therefore, exploring the possible role of these factors in the circadian regulation of SIRT1 activity may provide novel insights into its function in circadian rhythm.
SIRT6, another nuclear sirtuin, plays an essential role in animal development. Previous studies have shown that among the seven mammalian sirtuins, SIRT6 deficiency causes the most striking phenotypes. The SIRT6 null mice appear to have normal embryonic development and intrauterine growth and were born with no obvious abnormalities. However, they suffer a severe metabolic imbalance, acute onset of hypoglycemia, postnatal growth retardation, and premature death at one month of age 16. Unlike the 110 kDa SIRT1 protein, SIRT6 is relatively small (around 36 kDa) with short N- and C-terminal extensions 12. Although it is a nuclear protein, it is predominately chromatin-bound and is enriched in telomeric chromatin in S-phase 16, 89. Moreover, despite possessing a robust NAD+-dependent auto-ADP-ribosyltransferase activity 88, SIRT6 is a highly specific histone 3 deacetylase that targets acetyl-H3K9 and acetyl-H3K56, playing an important role in DNA repair, telomere function, genomic stability, and cellular senescence 89-91. Human SIRT6 also deacetylates C-terminal binding protein interacting protein (CtIP) and promotes DNA end resection 92.
Recent studies using mouse models have revealed essential roles of SIRT6 in metabolic homeostasis and inflammation, two heavily inter-locked biological processes (Figure (Figure3).3). Kawahara et al. showed that SIRT6 functions at chromatin to attenuate NF-κB signaling 93. SIRT6 interacts with the RelA subunit of NF-κB and deacetylates H3K9 at target promoters. Deficiency of SIRT6 in mice results in increased NF-κB-driven gene expression programs. However, SIRT6 also appears to positively regulate Tumor necrosis factor (TNF) production at a post-transcriptional step in response to an increase in intracellular NAD+ concentrations 94, raising the possibility that SIRT6 may play discrete roles in acute and chronic inflammatory responses. SIRT6 is also a master regulator of systemic glucose metabolism. Using a whole-body knockout mouse model, Zhong et al. reported that SIRT6 functions as a corepressor of the transcription factor Hif1α to suppress glucose uptake and glycolysis 95. SIRT6 knockout mice are hypoglycemic and SIRT6-deficient cells exhibit increased Hif1α activity with augmented glucose uptake, along with upregulation of glycolysis and diminished mitochondrial respiration. Xiao et al. also reported that systemic deletion of SIRT6 in mice results in severe hypoglycemia, but it appears that SIRT6-deficiency enhances both basal and insulin-stimulated glucose uptake through increased insulin signaling and activation of AKT 96. Further, studies using tissue-specific SIRT6 knockout mouse models have indicated that SIRT6 regulates metabolic homeostasis both systemically and locally. For instance, using a liver-specific SIRT6 knockout mouse model, Kim et al. demonstrated that liver-specific deletion of SIRT6 leads to increased glycolysis and triglyceride synthesis, as well as reduced fatty acid oxidation, resulting in liver steatosis 97. Neural-specific deletion of SIRT6 does not lead to hypoglycemia 98. Instead, neural SIRT6-null mice exhibit postnatal growth retardation due to somatotropic attenuation through low growth hormone (GH) and insulin-like growth factor 1 (IGF1) levels. They gradually reach normal size over time and ultimately become obese. These observations indicate that SIRT6 regulates animal body growth through central control, while modulating glucose and lipid metabolism at both systemic and local levels (Figure (Figure33).
In line with the notion that sirtuins are important cellular metabolic sensors coupling energy status to cellular functions, three of the mammalian sirtuins, SIRT3, SIRT4, and SIRT5, are directly localized to mitochondria. These double membrane-bound organelles are the “cellular power plants”, consuming 85-95% of the oxygen used by cells to generate ATP, and account for over 90% of reactive oxygen species (ROS) generated in cells 99. Mitochondria are also at the crossroads between numerous intermediary metabolic pathways that function in nutrient adaptation in response to environmental nutrient cues. Therefore it is not surprising that defects in mitochondrial functions are associated with degenerative diseases, cancer, and aging 100.
Despite the important localization of SIRT3-5, mice with deletions of these mitochondrial sirtuins lack obvious phenotypes 101, explaining why the biochemical and biological functions of these proteins have just started to be elucidated (Figure (Figure4).4). SIRT3, the most studied mitochondrial sirtuin, is highly expressed in metabolically active tissues such as brown adipose, muscle, liver, kidney, heart, and brain 102, 103. Consistent with this expression pattern and its mitochondrial localization, deletion of SIRT3 in mice leads to striking mitochondrial protein hyperacetylation 101. These hyperacetylated proteins include subunits of oxidative phosphorylation complexes 104, metabolic enzymes such as acetyl-CoA synthetase 2 105, 106, long-chain acyl CoA dehydrogenase (LCAD) 107 and 3-hydroxy-3-methylglutaryl CoA (HMG-CoA) synthase 2 108, as well as oxidative stress reducing enzymes isocitrate dehydrogenase 2 (IDH2) 36, superoxide dismutase 2 (SOD2) 35, and MnSOD 109. A number of recent studies have reported that hyperacetylation of these mitochondrial proteins in SIRT3 deficient mice results in a variety of metabolic abnormalities including reduced ATP production, decreased rates of fatty acid oxidation and ketone body production, and fatty liver 104, 107, 108, 110. In addition, SIRT3 also regulates mitochondrial translation 111, a process that may indirectly impact all mitochondrial pathways.
Hyperacetylation of mitochondrial proteins may have biological consequences beyond abnormal metabolic homeostasis. Using SIRT3-deficient and transgenic mouse models, two studies reported that SIRT3 blocks age-related cardiac hypertrophic response 112, 113. Kim et al. showed that SIRT3 is a mitochondria-localized tumor suppressor in response to stress 114. SIRT3 also protects in vitro fertilized mouse preimplantation embryos against oxidative stress 115. Furthermore, two recent studies showed that SIRT3 deficient mice fail to suppress oxidative stress and hearing loss in response to CR 35, 36. These observations demonstrate that SIRT3 can delay the onset of a number of oxidative stress-associated pathologies in multiple tissues and suggest that SIRT3 may be a novel target for these age-associated diseases as well as aging itself. However, since SIRT3 null mice display minimal phenotypes under normal feeding conditions, particularly when they are young 101, much work still needs to be done to determine whether the effects of SIRT3 deletion on these processes are physiologically relevant, and whether they are direct consequences of particular hyperacetylated mitochondrial proteins. Furthermore, it remains to be answered whether SIRT3 is required for CR-mediated extension of lifespan or it is only a part of CR- elicited anti-aging circuits.
Unlike SIRT3, very little is known about the functions of SIRT4 and SIRT5. As a sirtuin that only displays detectable ADP-ribosyltransferase activity, SIRT4 is unique among the mammalian sirtuins 13. For example, although SIRT4 is a mitochondrial enzyme, it is not enriched in mitochondria-dense tissues such as heart and muscle 13. The only known substrate of SIRT4 is glutamate dehydrogenase (GDH), and SIRT4 interacts with GDH and suppresses its activity via ADP-ribosylation 13. It appears that loss of SIRT4 in mice results in increased insulin secretion in response to glucose and amino acids 13, while overexpression of SIRT4 in cultured cells suppresses insulin secretion 14. SIRT4 also plays a role in the regulation of fatty acid oxidation and mitochondrial gene expression in liver and muscle cells, although the mechanisms underlying these phenotypes are still unclear 116. SIRT5 has recently been reported as a specific deacetylase for carbamoyl phosphate synthetase (CPS1), the rate-limiting first step of the urea cycle in mitochondria 117, 118. Deletion of SIRT5 in mice leads to increased acetylation of CPS1 and elevated levels of ammonia after prolonged fasting, whereas SIRT5 transgenic mice display increased CPS1 activity 117. In summary, while it appears likely that SIRT3 is a major mitochondrial deacetylase that helps to prevent non-specific hyperacetylation of mitochondrial proteins by acetyl-CoA over-flood during the process of active energy production, SIRT4 and SIRT5 appear to have specific targets and functions in this organelle. Further work is needed to identify these SIRT4 and SIRT5 specific targets to elucidate any additional roles these sirtuins possess in the regulation of energy metabolism.
As one of the most important cellular metabolic sensors in cells, it has been speculated that the seven sirtuins coordinate with each other in various cellular compartments to actively monitor diverse environmental signals, modulating cellular metabolic activity, gene transcription, and genome stability, ultimately affecting aging. Indeed, perturbation of the activity of one sirtuin has been shown to impact the activities of other sirtuin members. For example, we reported that in mouse macrophages, deletion of SIRT1 leads to increased levels of chromatin-associated SIRT6 near the NF-κB binding sites, resulting in reduction of local acetyl-H3K9 levels, compensating for the hyperactivation of NF-κB induced by SIRT1 deficiency 119. Interestingly, this compensatory effect between SIRT1 and SIRT6 does not appear to exist in hepatocytes. Instead, deletion of SIRT1 causes 50% reduction of both SIRT6 mRNA and protein levels in the liver 97. This is a result of SIRT1 binding to Foxo3a and nuclear respiratory factor 1 (NRF-1) on the promoter of SIRT6 in hepatocytes, directly promoting the expression of SIRT6 under both basal and fasting conditions 97. Additional studies will be necessary to address whether the varying coordination patterns between these two nuclear sirtuins are related to the different metabolic profiles of macrophages and hepatocytes.
Intensive crosstalk between nuclear and mitochondrial sirtuins have also been implied in the literature. A recent study reported that SIRT3 is a transcriptional target of PGC-1α via an estrogen-related receptor binding element (ERRE) on its promoter 120. As PGC-1α is a direct deacetylation substrate of SIRT1 42, this observation suggests that SIRT1 could indirectly modulate the expression of SIRT3 through PGC-1α deacetylation. Additionally, Nasrin et al. recently showed that SIRT4 knockdown in hepatocytes induces an increase in fatty acid oxidation through SIRT1 116. Although molecular mechanisms underlying these connections remain to be defined, these findings suggest the existence of a sirtuin-network that may be pivotal in the maintenance of systemic metabolic homeostasis.
The discovery that overexpression of sirtuins extends lifespan in lower model organisms has evoked a flood of research on their roles in the mammalian aging process. The disruption of energy metabolism, genome stability, and stress response in sirtuin-deficient mouse models demonstrates that mammalian sirtuins are major contributors to the delicate balance between metabolism and aging. Consistent with this notion, small molecule activators of SIRT1, such as the polyphenol resveratrol and closely related derivatives, have shown promise as therapeutic agents for the treatment of obesity and related metabolic diseases 58, 121, 122, although there were debate on whether these small-molecule drugs are direct activators of SIRT1 that function to prevent obesity and diabetes 59, 60. Further research is needed on the systemic and tissue-specific functions of sirtuins in other aspects of the aging process, such as stem cell renewal and protein quality control, as well as on the identification of novel targets and modulators of these proteins. Ultimately, these studies will enhance our knowledge of sirtuin biology, aiding the development novel small molecule modulators of sirtuins to combat age-associated diseases as well as aging itself.
We thank Lawrence J. Forsberg from the UNC-Chapel Hill and members of the Li laboratory for critical reading of the manuscript. The work related to this article was supported by the Intramural Research Program of the NIH, National Institute of Environmental Health Sciences to X.L. (Z01 ES102205).