The Role of Sirtuin-1 in the Vasculature: Focus on Aortic Aneurysm

Sirtuin-1 (SirT1) is a nicotinamide adenine dinucleotide-dependent deacetylase and the best characterized member of the sirtuins family in mammalians. Sirtuin-1 shuttles between the cytoplasm and the nucleus, where it deacetylates histones and non-histone proteins involved in a plethora of cellular processes, including survival, growth, metabolism, senescence, and stress resistance. In this brief review, we summarize the current knowledge on the anti-oxidant, anti-inflammatory, anti-apoptotic, and anti-senescence effects of SirT1 with an emphasis on vascular diseases. Specifically, we describe recent research advances on SirT1-mediated molecular mechanisms in aortic aneurysm (AA), and how these processes relate to oxidant stress and the heme-oxygenase (HO) system. HO-1 and HO-2 catalyze the rate-limiting step of cellular heme degradation and, similar to SirT1, HO-1 exerts beneficial effects in the vasculature through the activation of anti-oxidant, anti-inflammatory, anti-apoptotic, and anti-proliferative signaling pathways. SirT1 and HO-1 are part of an integrated system for cellular stress tolerance, and may positively interact to regulate vascular function. We further discuss sex differences in HO-1 and SirT1 activity or expression, and the potential interactions between the two proteins, in relation to the progression and severity of AA, as well as the ongoing efforts for translational applications of SirT1 activation and HO-1 induction in the treatment of cardiovascular diseases including AA.

SirT1 is essential for VSM structural and functional homeostasis (Li et al., 2011;Gorenne et al., 2013;Fry et al., 2015;Chen et al., 2016). Physiological cyclic stretch promotes VSM contractile properties via SirT1/FOXO3a, thus maintaining vasoconstriction (Huang et al., 2015). Interestingly, lack of SirT1 in VSM-specific SirT1-ablated mice does not induce any vascular functional impairment per se (Fry et al., 2015). However, in response to pro-inflammatory and pro-oxidant stressors, such as angiotensin II (angII) infusion, arterial ligation, hyperlipidemic apoE −/− genetic background, aging and a diet rich in fat and sucrose, VSM SirT1 is absolutely essential to prevent maladaptive arterial remodeling that leads to vascular disease, such as atherosclerosis (Gorenne et al., 2013), aortic dissection (Fry et al., 2015), and arterial stiffness (Fry et al., 2016). Differential effects of SirT1 expression or activity levels in the vasculature are uniquely dependent on disease type, stage and interacting factors and thoroughly reviewed elsewhere (D'Onofrio et al., 2018;Man et al., 2019). In this review, we will focus on yet another role of SirT1 in vascular disease, namely aortic aneurysm, which only recently started to be appreciated.

SIRT1 IN AORTIC ANEURYSM
Aortic aneurysms are abnormal aortic enlargements which can develop in the thoracic (TAA) or abdominal (AAA) regions. Although the pathogenesis behind different forms of TAA and AAA may differ greatly, a combination of genetic predisposition and hypertension, particularly in Marfan's and related syndromes, generally contribute to the development of TAA prone to dissections, while smoking, male gender and diabetes are the major risk factors for AAA prone to ruptures. Overall, the most troublesome aspect of AA (TAA and AAA), particularly in non-syndromic and idiopathic forms, is that they often remain clinically undiagnosed until the aortic wall dissects or ruptures, potentially causing sudden death (Davies, 1998;Mody et al., 2014). Currently, treatment options against these potentially lethal vascular conditions are limited to blood pressure control and surgical repair, for which the risk of mortality remains high at 30-50%, depending on the repair method employed (Virani et al., 2020); therefore, novel therapeutic targets are urgently needed. We have recently shown that mice with VSM-specific deletion of SirT1 have a drastically increased mortality (70%) in response to angII infusion due to aortic wall dissection, particularly in the thoracic region, which resulted from excess oxidant production and oxidantstimulated matrix metalloproteinases (MMPs) activation (Fry et al., 2015). Consistent with these findings, VSM-specific overexpression of SirT1 protects aged mice on apoE-deficient genetic background, against angII-induced AAA and rupture by opposing vascular senescence and inflammation . Moreover, calorie restriction, which is known to activate SirT1 (Guarente, 2013), prevents angII-induced AAA through SirT1-dependent deacetylation of H3 at lysine 9 on the MMP2 gene promoter, which downregulates MMP2 and subsequent elastin fragmentation in the aortic wall .
Furthermore, endothelial SirT1 has been shown to counteract the deleterious effects of angII on AA formation, believed to be driven mainly by endothelial, but not VSM, angII type 1α receptors (ATR1α) (Rateri et al., 2011). In a model of Marfan's syndrome (fibrillin-1 mutant mice, Fbn1 mgR/mgR ), deletion of endothelial ATR1α decreased the incidence of TAA (Galatioto et al., 2018). Since angII administration is known to downregulate SirT1 expression and activity in aortic endothelial cells (Marampon et al., 2013), these findings suggest that angII could accelerate the development of TAA in renin-dependent hypertensive Marfan individuals by suppressing SirT1/eNOS and impairing endothelial function.
It is worth mentioning that SirT1 in hematopoietic cells has recently emerged as an important mediator of AA formation. Macrophage-specific deficiency of SirT1 increased the incidence and exacerbated disease severity in a mouse model of angIIinduced AAA, by increasing the pro-inflammatory inducible nitric oxide synthase, a marker of M1 macrophages, while decreasing arginase and mannose receptor, two markers of M2 macrophages. On the contrary, SirT1 overexpression in macrophages, achieved by adenoviral transfection, had an opposite effect (Zhang et al., 2018). Overall, multiple studies indicate that SirT1 in variety of vascular cells is essential for the maintenance of vascular wall integrity and to prevent AA.

SIRT1 AND OXIDATIVE STRESS IN AORTIC ANEURYSM
An imbalance between the production of reactive oxygen species (ROS) and the cellular anti-oxidant systems, defined as oxidant stress, is a hallmark of vascular diseases, including atherosclerosis and diabetic endothelial dysfunction (Brown and Griendling, 2015;Quintana and Taylor, 2019). A role of oxidant stress in the development and progression of AA has been described recently (Quintana and Taylor, 2019). In human and animal studies, NADPH oxidase (Nox) has been identified as a major source of ROS contributing to the development of AA (Miller et al., 2002;Yang et al., 2010;Jimenez-Altayo et al., 2018;van der Pluijm et al., 2018). Ablation of Nox1, Nox2, Nox4 isoforms or of the noncatalytic Nox subunit p47 in mice prevented the development of AAA (Lu et al., 2016;Siu et al., 2017). This is explained by the fact that Nox-derived ROS are important second messengers in the vasculature, however, excessive ROS lead to increase MMPs activation as well as inflammatory and apoptotic factors, which contribute to the pathogenesis of AAA (Finkel, 2011;Miki and Funato, 2012;Holmstrom and Finkel, 2014;Brown and Griendling, 2015;Forrester et al., 2018;Quintana and Taylor, 2019). Similarly, the expression of anti-oxidant enzymes such as superoxide dismutase, glutathione peroxidase, glutathione reductase, and glutathione S-transferase decreases in aortae of AAA and TAA patients (Dubick et al., 1999;Liao et al., 2008;Zuniga-Munoz et al., 2017).
Endothelial and VSM SirT1 are known to regulate the cellular redox state in the vascular wall by multiple mechanisms, including direct deacetylation of FOXOs, NF-kB, Nrf2, mitochondrial superoxide dismutase and Nox, which overall decrease ROS production (Fry et al., 2015;Huang et al., 2015;Zhang et al., 2017). Moreover, SirT1 inhibits the mitochondrial adaptor protein p66 Shc , a critical modulator of intracellular redox state and a major contributor of oxidative damage-induced endothelial dysfunction in the settings of diabetes (Zhou et al., 2011;Paneni et al., 2013), by directly deacetylating its lysine 81 (Kumar et al., 2017). Taken together, these well-established anti-oxidant effects of SirT1 are consistent with the findings that VSM-specific SirT1 deletions in mice, and decreased expression of SirT1 in human aorta, are associated with aortic dissection or aneurysms (Fry et al., 2015;Chen et al., 2016).
Interestingly, SirT1 itself is a redox-sensitive enzyme. We have shown that SirT1 oxidative post-translational modifications, namely S-glutathionylation, at cysteine residues of its catalytic domain, profoundly affects its enzymatic activity (Zee et al., 2010;Shao et al., 2014). Reversible oxidative post-translational modifications, such as S-glutathionylation, result in activation or inactivation of proteins, thereby regulating signaling cascades and preventing irreversible oxidation of protein thiols in the settings of oxidative stress Byrne et al., 2020). Consistent with our findings, redox factor-1 and apurinic/apyrimidinic endonuclease-1, two cellular reducing agents, are able to restore NO bioavailability and endotheliumdependent vasorelaxation, through the reduction of conserved cysteine sulfhydryls in the SirT1 deacetylase domain (Jung et al., 2013). The loss of this fine-tuned mechanism in human endothelial cells exposed to oxidant stress, such as after exposure to cigarette smoke or hydrogen peroxide, have been associated with endothelial dysfunction . More recently, we found that H 2 O 2 , as well as TGF-β1, a pro-fibrotic cytokine known to be activated in Marfan's syndrome and to upregulate Nox4 in the vasculature (Lu et al., 2016;Siu et al., 2017;Zuniga-Munoz et al., 2017;Jimenez-Altayo et al., 2018), increases the reversible oxidation of SirT1 in aortic VSM cells (E. Budbazar and F. Seta, unpublished results). Therefore, an impairment of SirT1 activity by oxidative post-translational modifications may be causally linked to the development of AA, at least in individuals with Marfan's syndrome, possibly by exacerbating the deleterious effects of oxidant stress, whereas preventing or boosting SirT1 activity in the aortic wall may prevent AA.

SIRT1 AND HO-1 IN AORTIC ANEURYSM
Heme oxygenase-1 (HO-1) and 2 (HO-2) catalyze the ratelimiting step of the cellular degradation of heme. In the presence of oxygen and cofactors, HO-1 and HO-2 metabolize heme into carbon monoxide (CO), free iron and biliverdin, subsequently converted to bilirubin (BR; Abraham and Kappas, 2005). Both HO-1 and HO-2 isoforms are catalytically active in the vasculature (Thorup et al., 1999;Zhang et al., 2001), as demonstrated by HO-dependent release of CO in rat aorta, renal and cerebral arteries, and gracilis muscle arterioles Zhang et al., 2001).
HO-1 and HO-2 regulate vascular tone and hemodynamic function, mainly through CO production (Kaide et al., , 2004Zhang et al., 2001;Rodriguez et al., 2003;Arregui et al., 2004). We and others have shown that HO inhibition causes vasoconstriction (Kaide et al., , 2004Zhang et al., 2001;Rodriguez et al., 2004) while CO generally promotes vasodilation (Kaide et al., , 2004Arregui et al., 2004;Rodriguez et al., 2011). The vasoconstrictor effects associated with HO inhibitors are greater after NO synthesis inhibition Rodriguez et al., 2003), linked in part, to amplification of prevailing neurohormonal constrictor mechanisms (Rodriguez et al., 2003). Importantly, NO synthesis inhibition elevates CO formation both in vivo and in vitro (Rodriguez et al., 2004). Therefore, the significance of the HO-CO system might be particularly relevant in the settings of reduced NO bioavailability, as it is the case for conditions associated with increased oxidant stress and reduced NO levels, including numerous vascular and renal diseases (Salom et al., 2007;Rodriguez et al., 2011;Bonacasa et al., 2013).
Overall, HO-1 and SirT1 are part of the integrated system that modulates the cellular response to stress and might positively interact to regulate cardiovascular function. Treatment with cobalt protoporphyrin, a HO-1 chemical inducer (Sodhi et al., 2015), or HO-1 overexpression in macrophages (Nakamura et al., 2018), consistently enhanced SirT1 expression. Conversely, SirT1 activity can modulate HO-1 biological effects in hepatic cells (Sodhi et al., 2015;Nakamura et al., 2018), adypocites (Lakhani et al., 2019), or myocardial tissue (Yu et al., 2019). Moreover, in a human model of endogenous AT1R antagonism, SirT1 directly associates with improved NO-dependent vasodilation via HO-1 (Davis et al., 2013). These interactions suggest that positive feedback mechanisms between SirT1 and HO-1 might be at play in the vasculature, and they may affect HO-1mediated cytoprotection in cardiovascular diseases. To this end, pharmacological activation of miR-181b/SirT1/HO-1 signaling axis reduced the development of ang II-induced AAA in apoE-/-mice (Hou et al., 2019). Although the exact molecular mechanisms accounting for the interaction between SirT-1 and HO-1, and its functional significance in vascular pathology, remain to be fully elucidated, the development of specific targeted therapies that modulate the SirT1/HO-1 axis could represent a new therapeutic strategy for the management of AA disease, as well as other vascular diseases.

SEX DIFFERENCES, OXIDATIVE STRESS, AND SUSCEPTIBILITY TO AA
Aortic aneurysm is more likely to occur in men than in women (Singh et al., 2001) but women have a greater risk of AA rupture and poorer outcome than men. Several factors may contribute to sex-specific susceptibility of male to AA namely differences in mechanical properties of the aorta, levels of aortic wall MMPs, renin angiotensin system activity, or inflammatory and immune responses, which might be influenced by sex hormones, as thoroughly reviewed elsewhere (Boese et al., 2018;Robinet et al., 2018). Interestingly, to dissociate the effects of sex hormones from sex chromosome, Alsiraj et al. (2017) used the four core genotype mouse model to create XX and XY female offspring, showing that XY phenotypic females had increased AA incidence and severity compared with XX females, through increased inflammation and oxidative stress. Moreover, castration prevented the progression of disease in a model of angII-induced AA . Since estrogen and testosterone also differently influence redox balance (Boese et al., 2018), the higher susceptibility of male to oxidative stress and AA in animal studies, might be related to either the effects of sex hormones or sex chromosomes genes on oxidant pathways.
Increased activation of Nox and ROS production are known to be higher (Dantas et al., 2004), while the antioxidant capacity is decreased (Rodriguez et al., 2010;Bonacasa et al., 2013), in males compared to females in preclinical and clinical studies (Sartori-Valinotti et al., 2007). Estrogen increases HO-1 expression in cultured endothelial cells (Baruscotti et al., 2010;Marcantoni et al., 2012) and the uterus of ovariectomized animals (Cella et al., 2006). In contrast, HO-1 expression is higher in females compared to males in several experimental models of cardiovascular disease (Zampino et al., 2006;Bonacasa et al., 2013). Noteworthy, our own work and others' , showed that increased HO-1 expression and/or activity in female rats was associated with decreased nitrosative stress and glomerular damage in the diabetic kidney (Bonacasa et al., 2013) and a lower susceptibility to cardiac ischemia compared to males (Posa et al., 2013). In line with this notion, increased HO-1 expression and activity in the female offspring of diabetic mothers, was associated with lower nitrotyrosine levels, and improved cardiovascular function in females compared to males (F. Rodriguez and J. M. Sanchez, unpublished results). Therefore, antioxidant HO-1 activity could improve vascular function partly by compensating for the loss of NO bioavailability, in vascular diseases, particularly in females.
Similarly to HO-1, SirT1 activity may be influenced by sex or sex hormones. Estrogen induces SirT1 and SirT1/AMPK/histone H3 pathway, which relates to the cardiovascular protective effects of estrogen therapy (Bendale et al., 2013). Moreover, SirT-1 downregulation in the female, but not male, hearts directly correlated with a decline in mitochondrial antioxidative defense and a pro-inflammatory shift (Barcena de Arellano et al., 2019). Consistently, sex-specific downregulation of the SirT1/AMPK/FOXO3a/PGC-1α regulatory network was observed in male, but not female offspring, in response to maternal high-fat feeding (Nguyen et al., 2017). Overall, the possibility of a sex-specific regulation of SirT1 and HO-1 in the context of oxidative stress, such as in AA, is intriguing. Further studies addressing the functional significance of sex specific changes in SirT1 and HO-1 in AA, and the correlation with clinical features are warranted.

TRANSLATIONAL THERAPEUTIC APPLICATIONS OF SIRT1 OR HO ACTIVATION FOR AA
Substantial R&D has been invested in the quest of compounds that could boost SirT1 activity for the treatment of cardiovascular diseases. Thus far, numerous plant-derived molecules have shown promising results, although proving their pharmacological specificity remains a challenge. Liquorice-derived licochalcone attenuates angII-induced AAA by modulating the miR-181b/SirT1/HO-1 signaling axis (Hou et al., 2019). Resveratrol, a grape-derived non-toxic polyphenol known to activate SirT1, decreases the incidence of AAA in angII-infused apoE −/− mice fed a high fat diet by increasing ACE2 and downregulating NF-kB, Akt, ERK1/2 and ATR1 in human aortic VSM (Moran et al., 2017). Similarly, resveratrol, but not another putative SirT1 agonist SRT1720, prevented the development of TAA in Fbn1 C1041G/+ mice, a model of Marfan's syndrome, by stimulating an endothelial cell-derived factor, which downregulated miR-29b and upregulated Bcl-2 in VSM cells, inhibiting their apoptosis (Hibender et al., 2016). However, the beneficial effects of resveratrol and other putative SirT1-specific agonists, such as SRT1720 and SRT2104, are attributed to both SirT1-dependent and -independent mechanisms (Park et al., 2012;Zordoky et al., 2015;Bonnefont-Rousselot, 2016;Hibender et al., 2016;Moran et al., 2017;van Andel et al., 2019). It is well established that resveratrol can activate AMPK, inhibit cyclooxygenases and affect a variety of other enzymes (Bonnefont-Rousselot, 2016;Wicinski et al., 2018). Likewise, novel SirT1 activators, which were effective in decreasing arterial stiffness and improving the lipid profile of healthy cigarette smokers and elderly volunteers in clinical trials (Libri et al., 2012;Venkatasubramanian et al., 2013Venkatasubramanian et al., , 2016, function as partial agonists of SirT1 or exhibit offtarget effects (Pacholec et al., 2010). Therefore, attempts to develop alternative approaches to activate SirT1, rather than direct agonists, are an active field of research. Manipulating the metabolism of NAD + , the essential co-factor for SirT1 deacetylase activity, has proven very promising, as in some cases, it recapitulates the effects of resveratrol (Baur, 2010). Both niacin and nicotinamide supplementation increases NAD + levels and NAD + -dependent SirT1 activity in aortas and prevents the development of AAA in mice (Horimatsu et al., 2019). Similarly, nicotinamide phosphoribosyltransferase (NAMPT), also known as visfatin, a methyltransferase crucial for the synthesis of the NAD + precursor nicotinamide mononucleotide (Galli et al., 2013), has been shown to maintain aortic VSM integrity through NAD/SirT1 pathway (Watson et al., 2017). Additionally, we propose that understanding the role of oxidative post-translational modifications, such as S-glutathionylation of SirT1 cysteine residues, may lead to novel therapeutic strategies (Figure 1). We previously shown that a redoxresistant mutant SirT1 and overexpression of glutaredoxin-1, a thioltransferase that removes S-glutathionylation on SirT1, protects against liver metabolic disease (Shao et al., 2014(Shao et al., , 2017. Whether a similar approach could be implemented in the clinic to treat AA and other cardiovascular diseases warrants further research. Lipid-lowering agents widely used in clinical settings, such as probucol or statins, showed anti-proliferative effects on VSM through HO-1 induction (Deng et al., 2004;Lee et al., 2004), and decreased the development of experimental AA by inducing HO-1 gene expression (Azuma et al., 2016;Piechota-Polanczyk et al., 2018;Chen et al., 2020). Likewise, simvastatin-treated FIGURE 1 | A graphical abstract of oxidative modifications of cysteine thiols and pharmacological agents that increase SirT1 and HO-1 activity. The reactive thiolate anions of SirT1 exposed to ROS can react to form reversible (red and yellow, temporary inactivation) or irreversible (black, permanent inactivation). Glutathione (GSH) adducts added by S-glutathionylation inactivate SirT1. SirT1-GSH adducts may be reversed by glutaredoxin-1 (Glrx1), forming oxidized GSH (GSSG) and active SirT1 (green). HO-1 may interact with SirT1 through positive feedback mechanisms and metabolize heme into carbon monoxide (CO), free iron (Fe 2+ ) and biliverdin, subsequently converted to bilirubin (BR). NAD + , nicotinamide adenine dinucleotide; NAMPT, nicotinamide phosphoribosyltransferase; ROS, reactive oxygen species.
Frontiers in Physiology | www.frontiersin.org patients with AA, showed higher levels of anti-oxidant HO-1 in vascular tissue (Piechota-Polanczyk et al., 2018). Taken together, these findings strongly support a key protective role of HO-1 in limiting AA progression, and that statins and probucol may limit AA progression through mechanisms involving HO-1, probably independently of their lipid-lowering effects.
In contrast, it is important to consider that deleterious effects of HO-1 (Jais et al., 2014) and SirT1 overexpression (Kawashima et al., 2011) have been reported. In the case of HO-1 induction, these deleterious effects relate to increased reactive free iron (Suttner and Dennery, 1999) or reduced heme availability (Duvigneau et al., 2019). Interestingly, it has been suggested that there is a beneficial threshold of HO-1 overexpression (Suttner and Dennery, 1999;Namba et al., 2014) operating in specific subcellular localizations (Namba et al., 2014). Therefore, attempts at targeting HO-1 enzymatic activity to treat AA, as well as other vascular diseases, should aim at generating sufficient amounts of HO-1 activity, in specific cell compartments, maintaining suitable levels of intracellular heme and heme derived products formed by HO, avoiding any undesirable effects. Similarly, SirT1 activation or overexpression should be titrated to avoid unwanted effects.
Overall, HO-1 and SirT1 are part of gene complexes termed vitagenes (Calabrese et al., 2014), which provide an integrated response to control oxidant stress-induced tissue injury. Vitagenes products (i.e., heat shock proteins, BR, CO, gluthathione) mediate anti-oxidant, anti-apoptotic, anti-proliferative and anti-inflammatory actions (Abraham and Kappas, 2005;Calabrese et al., 2014). Dietary antioxidants (Abraham and Kappas, 2005;Moran et al., 2017;Hou et al., 2019) and clinically widely used compounds, such us statins, have shown favorable pleiotropic effects partly mediated by HO-1 induction (Lee et al., 2004;Piechota-Polanczyk et al., 2018) or SirT1 expression (Strycharz et al., 2018), providing a strong rationale for their therapeutic benefits in cardiovascular diseases, including AA (Hou et al., 2019). Therefore, the importance of developing pharmacological agents that activate an integrated response against oxidant stress in AA seems evident. Still, there is no definitive evidence from large scale clinical studies with anti-oxidant supplementation (Egea et al., 2017), requiring a better understanding of how each compound would affect the sources of ROS in specific cell compartments, modulate NAD + levels (Horimatsu et al., 2019), or affect ROS-induced reversible modifications, such as protein S-glutathionylation (Shao et al., 2014(Shao et al., , 2017, for the suitable translation into clinical settings.

AUTHOR CONTRIBUTIONS
EB wrote and reviewed the manuscript, and prepared the figure. FR and FS wrote and reviewed the manuscript. JS reviewed the manuscript. All authors contributed to the article and approved the submitted version.