Sirtuin Regulation in CR - Part 2

[Guest guest]

The below is part 2 of a paper that is pdf-availed. Sirtuin Regulation in Calorie Restriction.Qiu X, Brown KV, Moran Y, Chen D.Biochim Biophys Acta. 2010 Aug;1804(8):1576-83. Epub 2009 Sep 24.PMID: 19782772 The beneficial effects of calorie restriction diet in extending lifespan and preventing diseases have long been recognized. Recent genetic and molecular studies in model organisms began to uncover the molecular regulation of calorie restriction response, with the gene SIR2 playing an essential role. This article summarizes the latest development on how mammalian SIR2 homologs coordinately regulate the calorie restriction response. ... CR animals have decreased levels of blood insulin in response to limited foodintake. A low insulin level suppresses glycolysis and maintains blood glucoselevels, allowing a switch to fatty acids as an energy source. In the

pancreatic ß-cells, SIRT1 suppresses UCP2, a mitochondrial uncoupler, and permits efficientATP production [44, 45]. As ATP is required to trigger insulin secretion from theß-cells, SIRT1 is a positive regulator of this biological process. Low insulin levelsin CR mice suggest that the activity of SIRT1 is likely reduced in the pancreaticß cells. Although the NAD levels in the pancreatic ß cells during CR havenot been reported, they are decreased during fasting, when insulin secretion isalso suppressed [44]. The liver synthesizes fat for storage when there is excessive food intake, andproduces glucose from non-carbohydrate carbon substrates in response tofasting. PGC-1-alpha is also a key regulator of gluconeogenesis [46]. Bydeacetylating and activating PGC-1-alpha, SIRT1 increases the expression ofenzymes catalyzing the rate-limiting steps of gluconeogenesis. It would belogical, then, if

SIRT1 activity is also increased in the livers of CR mice to activatePGC-1-alpha and promote glucose production. However, decreased SIRT1expression and increased NADH levels have been observed in the livers of CRmice, suggesting that the activity of this deacetylase is actually decreased [43].Adding support to this observation, liver-specific SIRT1 knockout mice respond toCR normally and induce gluconeogenesis to the same degree as their wild typecounterparts [43]. How do we make sense of the decreased SIRT1 activity in the liver of CR mice?Active gluconeogenesis is an acute response to the fasting signal glucagonimmediately following food withdrawal. Hepatic glucose production decreasesafter fatty acid oxidation is activated and the ketone bodies produced by the liversupply compensatory fuel for glucose-dependent tissues, such as the brain [47].The expression of PGC-1-alpha and another

transcription factor regulatinggluconeogeneis, TORC2, are both increased upon acute fasting to promoteglucose production [47, 48]. During prolonged fasting, TORC2 is quicklydeacetylated and inactivated by SIRT1, which is not active during acute fastingbut is activated during prolonged fasting. Moreover, the transcription of PGC-1-alpha is also decreased to suppress gluconeogensis when glucose is no longer in highdemand. At this stage, only FOXO1, whose gluconeogenic activity is activated bySIRT1, is active to maintain the basal level of glucose. The expression of PGC-1-alpha is greatly induced by CR, which is essential to inducemitochondrial biogenesis [40]. Increased mitochondrial biogenesis is a hallmarkof CR and is not observed under other energy limiting conditions, such as fasting.This high level of PGC-1-alpha is likely to be the dominant factor regulatinggluconeogenesis in CR liver and

triggers more glucose production than what isneeded by the CR animals. After long-term CR, animals become adapted toefficient fatty acid usage and effectively rely on ketone bodies. In addition, afterthe initial adaptation stage of the CR diet, the animals do not lose weight anymore. This suggests that there is no further break down of muscle to use theiramino acids as supplies for glucose production. Thus, CR animals do not haveas high of a demand for glucose as animals on acute fasting, and thedownregulation of SIRT1 in the liver may be necessary to suppress glucoseproduction induced by high level of PGC-1-alpha. In this case, TORC2 may stayactive to support basal levels of glucose production. Inactivation of SIRT1 during CR in the liver is further supported by increasedapoptosis [49, 50], as SIRT1 suppresses p53 [51, 52], FOXO [53, 54], Ku70 [55],E2F [56] and promotes cell survival. This can

also explain the puzzlingobservation that while the SIRT1 activator resveratrol mimics many aspects ofCR response, resveratrol and CR regulate cell cycle checkpoints and apoptosisin opposite directions in the liver [57]. Thus, as discussed earlier, increasedSIRT1 in the muscle and the WAT promotes cell survival and increases stressresistance, which are crucial for the lifespan of postmitotic cells. However,hepatocytes are mitotic cells that retain tremendous proliferation potential.Slowing the proliferation by cell cycle arrest may result in increased lifespan. How is SIRT1 regulated in a tissue-dependent manner during CR? CR increasesmetabolism, and CR animals have increased energy demand. This pushesNADH to enter the electron transport chain to produce ATP and be convertedinto NAD, resulting in increased NAD level observed in the CR muscle and WAT.However, the liver differs from other metabolic

tissues in that under fedconditions, it is engaged in fatty acid synthesis, which increases the cellularredox state. During CR, fatty acid synthesis is shut down and the cellular redoxstate is decreased, consistent with the increased NADH level observed in the CRliver. Interestingly, the changes in the redox state of the CR tissues correlate with thechanges in the expression level of SIRT1 [43]. In muscle and WAT, theNAD/NADH ratio is increased and the expression of SIRT1 is also upregulated.In the liver, both the NAD/NADH ratio and the SIRT1 expression are reduced.Since the same changes in the NAD/NADH ratio occur in the SIRT1 KO mice onCR diet, this correlation suggests that the expression of SIRT1 is regulated bythe cellular redox state. Indeed, the transcription factor HIC1 represses SIRT1expression when coupled with the redox sensing co-repressor CtBP [58]. Whenthe cellular redox state is

increased, CtBP is dissociated from HIC1, resulting in areduction in transcriptional repression of SIRT1 mediated by HIC1. Thus, NADand NADH not only regulate the enzymatic activity of SIRT1 but also itsexpression at the transcriptional level, allowing a tight control of its deacetyaseactivity by the cellular redox state. These studies are consistent with a model whereby SIRT1 is a master regulatorof CR in many tissues. It is worth noting that SIRT1 whole body knockout micehave pleiotropic metabolic phenotype. In this case, tissue-specific SIRT1knockout mice will be valuable tools to dissect the role of SIRT1 in mediating CRresponse in vivo. Calorie restriction: insights from SIRT1 activators The observations that SIRT1 increases mitochondrial biogenesis and fatty acidoxidation suggest that activation of SIRT1 may mimic many aspects of CRresponse and improve metabolic

homeostasis. Consistent with this idea, SIRT1transgenic mice display some phenotypes similar to CR mice. One transgenicline has decreased body weight, increased metabolic rate, and improved glucosetolerance. The other two lines have less overt metabolic phenotype on a chowdiet but are protected against high fat diet feeding despite the same weight gain[59-61]. It is worth noting that three existing SIRT1 transgenic lines are driven bydifferent promoters and have different overexpression patterns/levels of SIRT1.Despite differences in certain phenotypes, they all have improved metabolichomeostasis. Thus, small molecule SIRT1 activators may be the best candidatesfor the coveted CR mimetics to delay age-related diseases. Interestingly,although resveratrol does not increase lifespan of mice fed a chow diet, itsignificantly improves the physiology and survival of mice fed a high fat diet [57,62].

Resveratrol-treated mice have increased numbers of mitochondria,decreased levels of blood glucose and insulin, increased glucose tolerance andinsulin sensitivity, and improved motor activity, reminiscent of CR effects. In addition to its interaction with SIRT1, resveratrol is known to interact with manyother proteins and pathways involved in energy balance, such as mitochondrialATP synthase, complex III, fatty acid synthase, and AMP kinase [63]. Animportant question is whether the CR-mimicking effects of resveratrol aremediated by SIRT1 in vivo. Newly identified SIRT1 activators structurallyunrelated to resveratrol, such as SIR1720, have similar effects in obese rodents,such as improved insulin sensitivity and mitochondrial capacity [64, 65]. Thesespecific SIRT1 activators are 1000-fold more potent that resveratrol, suggestingthat SIRT1 activation alone is sufficient to induce the beneficial metabolic

effects. Among the growing list of SIRT1 substrates, PGC-1-alpha seems to be one of themost critical downstream effectors of resveratrol [62, 64]. Resveratrol enhancesmuscle oxidative capacity and endurance due to its ability to engender a switchfrom fast-twitch/glycolytic fibers to slow-twitch/oxidative muscle fibers. Thisswitch is coupled with an increased expression of mitochondrial genes andgenes involved in oxidative phosphorylation, consistent with the known functionof PGC-1-alpha in regulating muscle fiber types and mitochondrial biogenesis [66, 67].Resveratrol increases PGC-1-alpha activity by enhancing its SIRT1-mediateddeacetylation. The effects of resveratrol on the expression of mitochondrialgenes are abolished in cells with SIRT1 knocked down or with PGC-1-alpha containing mutated acetylation sites [62]. Diet-induced obesity is a major risk factor in the development of

insulinresistance, characterized by reduced glucose clearance in the peripheral tissues.High doses of SIRT1 activators protect mice from weight gain when fed a high fatdiet, along with improving their insulin resistance [62, 64]. However, it is worthnoting that low doses of resveratrol improve insulin resistance even when there isno change in body weight and body fat [57, 64, 65]. Thus, increasedmitochondrial capacity may be the key to combat insulin resistance. AMPK, a stress and energy sensor that is activated during low energy states tomaintain glucose, fatty acids, and energy homeostasis, is also upregulated inmice treated with SIRT1 activators [57, 64]. The activation of AMPK by SIR1720is an indirect effect, since SIR1720 does not activate AMPK in tissue culture cellsor in mice via acute treatment. Rather, SIRT1 activation leads to increasedenergy expenditure, and the resulting energy-deficit

state, reflected by decreasedATP and ADP levels, leads to AMPK activation. Activated AMPK can furtheramplify the increase in fatty acid usage and mitochondrial capacity. In addition tothis SIRT1-dependent mechanism, AMPK may also be directly activated byresveratrol. These mechanistic insights on the actions of SIRT1 activators shedlight on how the interplay of CR and SIRT1 regulates energy balance at themolecular level (Figure 2). How can SIRT1 activators mimic the CR response if SIRT1 is inactivated incertain CR tissues such as the liver? It is likely that muscle is the dominanteffector mediating the metabolic changes when small molecular SIRT1 activatorsare administered. For example, increased mitochondrial biogenesis is observedin muscles and brown adipose tissues, but not in the livers of resveratrol treatedmice [62]. Moreover, the insulin-stimulated glucose uptake is improved inmuscles

but not the livers of SRT1720 treated mice [64]. Following oraladministration, SIRT1 activators are likely to be removed in the liver. Mitochondria: the centers of CR response Mitochondria are the hubs of metabolism. At least 20% of mitochondrial proteinsare acetylated [68]. This observation underscores the importance ofmitochondrial deacetylases in metabolic regulation. Many mitochondrial proteinsare hyperacetylated in mice deficient in SIRT3, a mitochondrial sirtuin,suggesting that SIRT3 may have broad impacts in metabolic control [69]. Onesubstrate of SIRT3 is acetyl-CoA synthetase 2 (AceCS2), which converts acetateinto acetyl-CoA [70, 71]. Under ketogenic conditions such as CR, the liverreleases large amounts of acetate into the blood. Tissues such as the muscleand the heart express AceCS2 and effectively utilize acetate as an energysupply. Since deacetylation of AceCS2 activates its

enzymatic activity, SIRT3 islikely to be activated during CR [72]. SIRT3 also deacetylates many subunits ofcomplex I and increases its respiration rate [73]. This allows efficient entry ofNADH into the electron transport chain to generate ATP. Thus, activation ofSIRT3 during CR may also contribute to increased respiration. Other SIRT3substrates and their roles in CR response remain to be elucidated. Under CR conditions, amino acid catabolism is activated and the urea cycle hasto be upregulated to remove excess amounts of ammonia. SIRT5, anothermitochondrial sirtuin, deacetylates and activates carbamoyl phosphatesynthetase 1 (CPS1), an enzyme catalyzing the first step of the urea cycle [74].The expression level of SIRT5 is unchanged but the level of NAD is increased by50% in the mitochondria of the CR liver. This increased SIRT5 activity mayupregulate the urea cycle during CR by deacetylating and

activating CPS1.SIRT4 is also localized to the mitochondria. It has no detectable deacetylaseactivity but possesses NAD-dependent mono-ADP-ribosyltransferase activity.One known substrate of SIRT4 is glutamate dehydrogenase (GDH), whichconverts glutamate to alpha-ketoglutarate and allows it to enter the TCA cycle [75].By ADP-ribosylating GDH, SIRT4 suppresses its activity and prevents theutilization of amino acids as an energy source. During CR, although the NADlevel is increased in the mitochondria, the expression of SIRT4 is decreased. Thedecreased SIRT4 activity results in the activation of GDH and the utilization ofglutamate and glutamine to generate ATP. This is particularly important for thepancreatic ß-cells of CR mice. Due to the low blood glucose level, glucosestimulatedinsulin secretion is suppressed. However, the activation of GDHallows amino acid-stimulated insulin secretion to maintain the

basal level ofinsulin. GDH can also be deacetylated by SIRT3, although the functional linkbetween acetylation and ADP-ribosylation of this enzyme is not established [69]. The center of CR response is mitochondrial activation, which allows metabolicadaptation to chronic energy deficit (Figure 3). One beneficial effect ofmitochondrial activation is increased respiration and more efficient ATPproduction. In addition, it also permits a fuel usage switch from glucose to noncarbohydratessince glycolysis and fatty acid synthesis occur in the cytosol whilefatty acid oxidation, amino acid metabolism and ketogenesis are confined to themitochondria. SIR2 has evolved to have diverse paralogs in mammals; at leastfour of them are directly or indirectly involved in mitochondrial activation,emphasizing the importance of the mitochondria and sirtuins in mediating CRresponse. These four sirtuins, SIRT1, 3, 4, and

5, are found in various cellularcompartments, have different enzymatic activities, and coordinate with eachother to regulate mitochondrial activity. It is interesting to speculate whymammals evolved to have three mitochondrial sirtuins and how their substratespecificity is achieved. Figure 3. The central role of mitochondria in CR response. CR increasesmitochondrial biogenesis, fatty acid usage, and respiration via the nuclear sirtuinSIRT1. Mitochondrial sirtuins, SIRT3, 4, and 5, also mediate the CR-inducedmetabolic shift toward acetate and amino acid usage.Regulation of Sirtuin Activities by CR It is striking that the activities of all sirtuins are subject to the regulation of NAD.Yet, sirtuin activities are differentially regulated during CR. Activation of SIRT3and SIRT5 is in contrast to the inactivation of SIRT1 in the CR liver. Themitochondria and the cytoplasm/nucleus have

separate NAD pools, sinceNAD(H) cannot pass through the mitochondrial membranes. For instance,following genotoxic stress, nuclear and cytoplasmic NAD is depleted, butmitochondrial NAD still remains at the physiological level [76]. This allows fordifferential activation of sirtuins based on their subcellular localization. Asdiscussed above, the redox state of the CR liver is reduced by fatty acidsynthesis, which is confined to the cytoplasm. The increased NAD level in themitochondria results from the upregulation of the NAD biosynthetic enzymeNampt [74, 76]. It still remains to be clarified whether it is the mitochondrial or thecytoplasmic Nampt that is activated. The second model proposes that increasedcytoplasmic Nampt elevates the level of nicotinamide mononucleotide (NMN), theprecursor of NAD. The mitochondrial membrane is permeable to NMN, and thetranslocation of NMN into the mitochondria increases the

NAD level. Furthermore, different sirtuins have different Km's for NAD [77-79]. This isrelevant because the same redox state may have different effects on differentsirtuins. This is particularly important for sirtuins reacting to the same NAD pool,such as SIRT3, 4, and 5. Other than HIC1, the transcription of SIRT1 is alsoregulated by p53 and Foxo3a in response to nutrient conditions [80]. In additionto NAD, NADH, and regulation at the transcriptional level, other regulatorymechanisms are also worth exploring. SIRT1 was found to have 13phosphorylation sites and one sumoylation site, and in vitro deacetylation activityassays showed that dephosphorylated or desumoylated SIRT1 has decreasedactivity [81, 82] . In addition, an endogenous activator (AROS) [83]and inhibitor(DBC1) [73, 84]of SIRT1 have been identified. It would be interesting to seewhether posttranscriptional modifications or endogenous

regulators of sirtuinsalso contribute to sirtuin regulation during CR. Cancer CR is known to prevent the incidence of cancer. The observation that SIRT1suppresses the apoptosis pathway mediated by p53, FOXO, and Ku70 andincreases stress resistance raises a tantalizing question: does SIRT1 activationpromote cancer? If so, how can we reconcile CR-induced SIRT1 activation andcancer prevention by CR? Many lines of evidence do support the cancerpromotingpotential of SIRT1. First, Deleted in Breast Cancer 1 (DBC1), likely atumor suppressor, promotes p53-mediated apoptosis through specific inhibitionof SIRT1 [73, 84]. Second, newly identified small molecule activators of p53 withthe potential to decrease tumor growth act through inhibiting SIRT1 and SIRT2[3]. Third, the loss of tumor suppressor HIC1 promotes tumorigenesis via SIRT1activation and the subsequent attenuation of p53 [85].

Finally, SIRT1 knockoutmice show increased p53 hyperacetylation and radiation-induced apoptosis [86]. However, in vivo evidence suggests that SIRT1 is a tumor suppressor. In ap53±background, mice with deficient SIRT1 develop cancers in multiple tissues[87]. BRCA-1 mutant mice develop mammary tumors due to the suppression ofSIRT1 [88]. In addition, SIRT1 overexpression reduces colon cancer formation inthe APCmin/+ mouse model [89]. How can SIRT1 prevent cancer when itsuppresses p53-induced apoptosis? First, SIRT1 modifies histones, ensuresproper chromosome condensation and segregation during mitosis, and preventsgenetic instability [87]. SIRT1 reverses acetylation of histone H3 on lysine 56(H3K56), which is critical in packaging DNA into chromatin following DNAreplication and repair [90]. In addition, SIRT1 maintains genomic stability bysilencing repetitive DNA [91]. SIRT1 also promotes DNA damage

repair throughNBS1 [92]. Increased genomic stability, together with enhanced DNA damagerepair, puts less demand on the cellular surveillance system, such as p53-dependent apoptosis. Second, despite its general anti-apoptotic activity, SIRT1may promote cancer-specific cell death by suppressing Survivin, an inhibitor ofapoptosis that is highly expressed in most human tumors but is completelyabsent in terminally differentiated cells [88]. Finally, SIRT1 also prevents cancerby suppressing oncogenes, such as ß-catenin [89]. Thus, SIRT1 activationduring CR may increase stress resistance by suppressing apoptosis and at thesame time, prevent cancer by increasing genomic stability, promoting cancer-specific cell death and suppressing oncogenes (Figure 4). Figure 4. SIRT1 increases stress resistance yet prevents cancer. SIRT1increases stress resistance by suppressing apoptosis. However, SIRT1

preventscancer by increasing genomic stability, promoting cancer-specific apoptosis, andsuppressing certain oncogenes. Inflammation Inflamm-aging, a term first coined by Claudio Franceschi [93, 94], refers to thephenomenon that aging is accompanied by a low-grade chronic, systemic up-regulation of the pro-inflammatory response. Inflammaging is caused byoveractive innate immune response and repressed adaptive immunity. Genesinvolved in inflammation are upregulated with age [95-98], and these age-relatedchanges can be reverted by CR regimen [97-99]. Evidence is emerging for a roleof sirtuins in dampening inflammation, as both SIRT1 and SIRT6 inhibit NF-kappa-B, akey factor for the inflammatory response [100, 101]. This is further supported bythe observation that resveratrol attenuates inflammation and SIRT1 KO micedevelop inflammatory lesions [102-105]. However, more in vivo evidence

isneeded to support this idea. Future directions The groundbreaking studies on SIRT1 activators in metabolic disease mousemodels are the first proof of principle that it is possible to design small moleculeCR mimetics to combat aging. Despite the unequivocal effects of SIRT1activators on metabolic regulation and aging-related deterioration, it does notextend the lifespan of mice fed a regular diet [106]. Tissue-specific regulation ofSIRT1 and coordination between multiple sirtuins to regulate CR also suggestthat a more complete understanding of sirtuin regulation by CR is necessarybefore we can fully take advantage of the beneficial effects of this dietaryregimen. In addition, the central role of mitochondria in CR response raises thepossibility of targeting mitochondria for preventing metabolic diseases. Finally,despite the intense studies on sirtuins and metabolic regulation, the

physiologicalrole of sirtuins in regulating biological pathways related to other diseases of agingare still awaiting further validation. These studies will serve as the basis for usingsirtuin activators or inhibitors to prevent or treat other diseases.

-- Aalt Pater