{"paper_id":"cb630df7-34e3-45a8-a0c8-6585913b6134","body_text":"Sirtuins are NAD + -dependent deacylases. They link cellular energy status to epigenetic, metabolic and stress responses. The founding member, yeast Sir2, silences transcription and extends replicative lifespan under caloric restriction. Its mammalian homologues, SIRT1–SIRT7, are key regulators of ageing across eukaryotes [ 1 ].\nSir2 was subsequently identified as an NAD + -dependent histone deacetylase, directly linking cellular metabolism to epigenetic control [ 2 ,  3 ]. Sir2 overexpression extends replicative lifespan in yeast and promotes longevity in Caenorhabditis elegans and Drosophila melanogaster, effects closely associated with caloric restriction and NAD +  availability [ 4 – 6 ].\nIn mammals, seven homologues (SIRT1–SIRT7) were identified in the late 1990 s and early 2000 s (Fig.  1 ). Although they share a conserved catalytic core, they differ in localization and regulatory function. SIRT1, the closest orthologue of yeast Sir2, initially attracted considerable attention due to its roles in metabolic adaptation, stress resistance and genome maintenance through deacetylation of transcriptional regulators such as PGC-1α and FOXO family members [ 7 – 9 ]. However, it soon became clear that sirtuin biology extends beyond SIRT1 and beyond deacetylation alone. SIRT5 catalyses desuccinylation and demalonylation [ 10 ,  11 ], SIRT6 mediates long-chain deacylation [ 12 ], and select isoforms exhibit mono-ADP-ribosylation activity [ 13 ]. These diverse enzymatic activities underpin isoform-specific functions in mitochondrial homeostasis, chromatin organization, immune regulation and cellular stress responses [ 14 – 17 ]. These findings have reframed sirtuins as versatile metabolic regulators not merely chromatin-associated deacetylases. Fig. 1 Discovery, diversification and compartmentalization of the sirtuin family.  a  Discovery and catalytic diversification. Yeast Sir2 was identified in 1989 as a transcriptional silencer and was subsequently shown to be an NAD + -dependent histone deacetylase; Sir2 overexpression extends replicative lifespan in yeast. Mammalian SIRT1, the closest orthologue of Sir2, was identified in the late 1990 s and regulates metabolic adaptation, stress resistance and genome maintenance through deacetylation of PGC-1α and FOXO. SIRT2–7 were identified in the early 2000 s, revealing catalytic diversity beyond deacetylation: SIRT5 catalyses desuccinylation and demalonylation, SIRT6 mediates long-chain deacylation, and select isoforms exhibit mono-ADP-ribosylation.  b  Subcellular compartmentalization. SIRT1, SIRT6 and SIRT7 are nuclear (SIRT7 is enriched in the nucleolus); SIRT3, SIRT4 and SIRT5 are mitochondrial; SIRT2 is predominantly cytosolic.  c  Functional division of labour by compartment. Nuclear sirtuins regulate chromatin remodelling, DNA repair and metabolic gene regulation; mitochondrial sirtuins control oxidative phosphorylation, fatty-acid oxidation, amino-acid metabolism and the urea cycle; cytosolic SIRT2 modulates cytoskeletal dynamics and cell-cycle progression. Catalytic diversity and compartmentalization jointly confer functional specificity, forming a distributed NAD +  metabolic sensing network\nDiscovery, diversification and compartmentalization of the sirtuin family.  a  Discovery and catalytic diversification. Yeast Sir2 was identified in 1989 as a transcriptional silencer and was subsequently shown to be an NAD + -dependent histone deacetylase; Sir2 overexpression extends replicative lifespan in yeast. Mammalian SIRT1, the closest orthologue of Sir2, was identified in the late 1990 s and regulates metabolic adaptation, stress resistance and genome maintenance through deacetylation of PGC-1α and FOXO. SIRT2–7 were identified in the early 2000 s, revealing catalytic diversity beyond deacetylation: SIRT5 catalyses desuccinylation and demalonylation, SIRT6 mediates long-chain deacylation, and select isoforms exhibit mono-ADP-ribosylation.  b  Subcellular compartmentalization. SIRT1, SIRT6 and SIRT7 are nuclear (SIRT7 is enriched in the nucleolus); SIRT3, SIRT4 and SIRT5 are mitochondrial; SIRT2 is predominantly cytosolic.  c  Functional division of labour by compartment. Nuclear sirtuins regulate chromatin remodelling, DNA repair and metabolic gene regulation; mitochondrial sirtuins control oxidative phosphorylation, fatty-acid oxidation, amino-acid metabolism and the urea cycle; cytosolic SIRT2 modulates cytoskeletal dynamics and cell-cycle progression. Catalytic diversity and compartmentalization jointly confer functional specificity, forming a distributed NAD +  metabolic sensing network\nA defining feature of mammalian sirtuins is their distinct subcellular localization, which provides a structural basis for functional specialization. Despite sharing an NAD + -dependent catalytic mechanism, individual sirtuins occupy discrete compartments: SIRT1, SIRT6 and SIRT7 are primarily nuclear; SIRT3, SIRT4 and SIRT5 are localized to mitochondria; and SIRT2 is predominantly cytoplasmic [ 18 ]. This compartmentalization restricts substrate accessibility and aligns each isoform with distinct metabolic environments. Increasing evidence suggests that nuclear, cytosolic and mitochondrial NAD +  pools are dynamically regulated and partially insulated rather than freely interchangeable, rendering sirtuin activity highly sensitive to compartment-specific NAD +  availability [ 19 ]. Direct measurement of compartment-specific NAD +  dynamics in vivo, however, remains technically challenging. Within the nucleus, SIRT1, SIRT6 and SIRT7 integrate NAD +  levels with transcriptional and chromatin regulation: SIRT1 modulates stress and metabolic pathways via deacetylation of p53, FOXOs and PGC-1α; SIRT6 contributes to DNA repair, glucose metabolism and inflammatory control; and SIRT7 regulates ribosomal biogenesis and RNA polymerase I activity [ 20 – 22 ]. In mitochondria, SIRT3, SIRT4 and SIRT5 regulate oxidative phosphorylation, fatty acid oxidation, amino acid metabolism and the urea cycle through NAD + -dependent deacylation and ADP-ribosylation reactions [ 23 – 25 ]. SIRT2, positioned in the cytoplasm, modulates cytoskeletal organization, cell-cycle progression and stress responses through deacetylation of non-histone substrates [ 26 ]. This spatial organization establishes a functional division of labour in which NAD +  acts as a shared metabolic currency, while subcellular localization confers regulatory specificity.\nDespite extensive investigation of individual isoforms, current understanding of sirtuin biology is fragmented by isoform-centric and disease-specific perspectives. This reductionist approach has generated seemingly contradictory findings, where the same sirtuin may exert protective or deleterious effects depending on tissue, metabolic state or timing, which points to the need for a unifying conceptual framework. This review conceptualizes the sirtuin family as a distributed metabolic sensing network that connects NAD +  availability with transcriptional, epigenetic and post-translational regulation across subcellular compartments [ 27 ]. This view rests on the idea that spatially segregated NAD +  pools and compartment-specific substrate landscapes confer functional specificity despite a shared catalytic dependency [ 18 ,  28 ,  29 ]. From this perspective, dysregulation of sirtuin-mediated metabolic sensing contributes to disease initiation and progression differently depending on tissue metabolic state, disease stage and stress duration, shaping whether cells undergo adaptive repair, maladaptive remodeling or irreversible decline [ 30 ]. This review integrates these principles to show how compartmentalized sirtuin activities collectively shape metabolic and stress responses across diseases. However, the relative contribution of each isoform to specific metabolic states is poorly quantified, and whether functional redundancy among compartmentalized sirtuins exists is largely unexplored. The available evidence does not yet permit definitive conclusions on either point, and we acknowledge this as an explicit limitation of the present synthesis. Compartmentalization itself provides a principled basis for approaching these gaps: if sirtuin function is spatially defined, then isoform-specific modulation, rather than global NAD +  elevation, offers a rational route to resolving relative contributions and redundancy experimentally. By framing sirtuins as context-sensitive integrators of metabolism and cell fate, this review aims to provide a coherent conceptual foundation for interpreting the expanding literature and to identify principles that may guide the development of more selective sirtuin-targeted interventions. A central thesis is that the limited and often heterogeneous clinical efficacy of global NAD + -boosting strategies is not incidental, but a predictable consequence of compartmentalization. Because global NAD +  elevation is blind to subcellular compartment, it necessarily co-activates sirtuin isoforms with opposing, compartment-specific actions. Isoform-selective, compartment-aware modulation therefore represents the more rational route to therapeutic translation.\n\nAs core components of a distributed NAD + -sensing network, individual sirtuin isoforms are distinguished first by their catalytic preferences and subcellular compartmentalisation. Although sirtuins were initially defined as NAD + -dependent deacetylases, it is now recognized that the mammalian sirtuin family catalyzes a broad spectrum of lysine deacylation reactions beyond canonical deacetylation [ 31 ]. Individual isoforms show clear substrate preferences, acting on a variety of acyl-lysine modifications. This catalytic diversity partitions sirtuin function within the family, helping to account for why specific isoforms preferentially regulate chromatin structure, metabolic enzymes or stress-response proteins [ 32 – 34 ].\nHowever, catalytic preference alone does not fully explain sirtuin functional specificity. Nuclear sirtuins (SIRT1, SIRT6 and SIRT7) mainly regulate acetylated histones and transcriptional regulators, shaping gene expression programs, chromatin accessibility and transcriptional memory [ 35 ]. By contrast, mitochondrial sirtuins (SIRT3-5) regulate oxidative metabolism, amino acid catabolism and redox balance through deacylation of metabolic enzymes. SIRT2, which is largely cytosolic but stress-responsive, targets both metabolic enzymes and structural proteins involved in cell-cycle progression and stress adaptation [ 36 ]. Moreover, even when both catalytic preference and subcellular localization are considered, the biological output of sirtuins is not solely determined by their enzymatic properties. Because many lysine acylations arise from metabolic flux rather than dedicated enzymatic writing, sirtuin activity is highly sensitive to substrate availability [ 25 ].rather than total [ 37 ].\nA defining feature of the sirtuin family is a largely conserved subcellular compartmentalisation, although this segregation is not absolute as noted above. Despite a conserved catalytic core and obligate dependence on NAD + , sirtuins are predominantly segregated into the nucleus (SIRT1, SIRT6 and SIRT7), mitochondria (SIRT3-5) and cytosol (SIRT2). This spatial separation constrains substrate access and couples each isoform to a distinct metabolic and redox microenvironment [ 38 ].\nNAD +  is not uniformly distributed within the cell. Nuclear, cytosolic and mitochondrial NAD +  pools are dynamically regulated and only partially equilibrated [ 39 ]. Accordingly, sirtuin activity reflects local NAD +  availability within specific compartments not total cellular NAD +  abundance. Global elevations in NAD +  may preferentially activate specific sirtuins, depending on which compartmental pools and which isoforms are expressed or poised for activation [ 40 ].\nThis spatial coupling helps explain why identical mechanistic perturbations can produce divergent sirtuin-dependent responses across tissues. For example, viral infection may preferentially suppress SIRT3–5 activity by depleting mitochondrial NAD + , while leaving nuclear sirtuins largely intact [ 41 ]. In contrast, chromatin stress or DNA damage can selectively engage nuclear sirtuins even when mitochondrial metabolism is preserved. Thus, NAD +  availability is interpreted locally, allowing sirtuin activity to be tuned by compartment-specific metabolic cues instead of a global switch.\nSirtuin activity is regulated through post-translational modifications, protein–protein interactions, subcellular relocalisation and substrate availability. Accordingly, sirtuin protein abundance correlates poorly with functional output, reflecting the importance of regulatory mechanisms that operate at the level of enzyme activation and substrate engagement [ 42 ].\nSeveral sirtuins undergo context-dependent stabilisation or degradation through phosphorylation- or ubiquitination-dependent pathways that couple protein turnover to stress signalling. Similarly, interactions with scaffolding proteins or transcriptional complexes can restrict sirtuins to defined substrate pools, conferring selectivity that is not evident from in vitro enzymatic assays [ 43 ]; these regulatory layers are particularly prominent for sirtuins acting on non-histone substrates, where transient interactions and local concentration effects dominate.\nAn important methodological implication of this complexity is that genetic ablation and pharmacological inhibition of sirtuins are not mechanistically equivalent [ 44 ]. Genetic ablation abolishes both catalytic and non-catalytic functions and permits compensatory rewiring over developmental or chronic timescales, whereas small-molecule modulators typically impose acute, substrate-selective perturbations that may preferentially affect specific deacylation reactions. Failure to distinguish between these modes of perturbation has contributed to apparent inconsistencies in the literature. This distinction is not merely technical. It reflects the underlying principle that sirtuin output is shaped by regulatory context beyond catalytic activity alone. Integrating genetic and pharmacological evidence within a unified framework that accounts for substrate availability, subcellular localisation and compensatory adaptation is therefore essential for resolving contradictions and mapping the true regulatory logic of the sirtuin family. However, the relative contribution of each regulatory layer (transcriptional abundance, post-translational modification, subcellular redistribution, and local NAD +  availability) to net sirtuin output is poorly defined for most isoforms, and whether these layers operate hierarchically or redundantly under different physiological conditions has not been established.\nThe biochemical and organizational features of the sirtuin family indicate that functional specialization emerges from the interplay of catalytic repertoire and subcellular localization (which jointly determine substrate access and regulatory context) rinstead of isoform identity alone. Sirtuin activity is in turn licensed by compartment-specific NAD +  pools and by the metabolic processes that generate acylated substrates, linking enzymatic output to local metabolic state. Superimposed on these constraints, regulatory inputs at the level of protein stability, complex formation and substrate access introduce additional layers of control that can override changes in gene expression.\nThese principles help explain the functional divergence observed across sirtuin biology. Conflicting outcomes attributed to the same sirtuin more often reflect differences in tissue type, metabolic state, stress duration or the subcellular compartment engaged, not genuine functional contradiction [ 45 ]. Acute versus chronic perturbations, for example, can engage distinct regulatory layers; similarly, the same sirtuin may fulfil fundamentally different roles in immune, stromal and parenchymal cells. From a translational perspective, this family-level logic has important implications. Interventions that globally elevate NAD +  are likely to activate multiple sirtuins in a compartment- and tissue-biased manner, whereas isoform-selective modulators must contend with substrate selectivity and cell-type- and state-dependent dual functions [ 46 ]. Accordingly, predicting therapeutic efficacy and avoiding unintended consequences requires an integrated understanding of catalytic specificity, subcellular localization and metabolic state.\nThese biochemical and regulatory principles frame the subsequent discussion of sirtuin function in metabolism, inflammation and disease. Instead of being considered in isolation, sirtuins can be viewed as a distributed network of NAD + -licensed regulators whose collective behaviour emerges from spatial organisation and metabolic signalling.\n\nBuilding on the compartmentalized architecture outlined above, we next consider how individual sirtuin isoforms (operating within a distributed NAD + -sensing network) coordinate glycolysis, fatty acid oxidation, and mitochondrial quality control across subcellular compartments. Across mammalian cells, metabolic adaptation to hypoxia, inflammation and oncogenic stress is frequently initiated at the level of glycolysis. Enhanced glycolytic flux supports rapid ATP production under oxygen limitation and generates metabolic intermediates and redox signals that influence downstream biosynthetic and oxidative pathways, including lipid metabolism. As such, glycolysis is an entry point at which cellular stress responses intersect with broader programs of metabolic reprogramming. Sirtuins, as NAD + -dependent deacylases distributed across the nucleus, cytosol and mitochondria, are well positioned to modulate this early metabolic response. Through regulation of hypoxia-responsive transcriptional programs and post-translational control of key glycolytic enzymes, individual sirtuins influence the magnitude and routing of glycolytic throughput. These early sirtuin-dependent effects on glycolysis shape both immediate energy production and the redox state and carbon availability that condition subsequent engagement of lipid anabolic and catabolic pathways.\nHIF-1α functions as a master regulator of oxygen homeostasis that couples hypoxia to metabolic reprogramming, initiating glycolytic activation while shaping downstream carbon flux and redox balance [ 47 ]. In the nucleus, SIRT1 limits acetylation-dependent stabilization of HIF-1α, such that reduced SIRT1 activity is associated with prolonged HIF-1α persistence and enhanced pro-glycolytic transcriptional programs, whereas restoration of SIRT1 restrains glycolytic flux in disease-relevant contexts [ 47 ].\nSIRT6 provides a parallel chromatin-based layer of control by repressing HIF-1α target genes through promoter-associated histone deacetylation, thereby constraining the amplitude of hypoxia-induced glycolytic programs under normoxic or low-stress conditions [ 48 – 50 ]. Consistent with this model, loss of SIRT6 releases this transcriptional restraint under hypoxic or inflammatory pressure and is linked to exacerbated glycolytic activation and worsened inflammatory pathology in ulcerative colitis [ 51 ].\nBeyond transcriptional regulation, sirtuins also gate HIF-1α proteostasis. In the cytosol, SIRT2-mediated deacetylation of HIF-1α enhances its coupling to the oxygen-sensing machinery by promoting PHD2-dependent hydroxylation and subsequent VHL-mediated degradation, thereby limiting excessive HIF-1α accumulation in tumour contexts [ 52 ]. Mitochondrial sirtuins provide an additional layer of regulation by modulating redox and metabolic signals that influence HIF-1α stability. Through regulation of mitochondrial respiration and reactive oxygen species, SIRT3 indirectly tunes HIF-1α activity during hypoxic stress and the associated glycolytic and angiogenic responses [ 53 ]. In metabolic disease models, loss of SIRT3 has been linked to enhanced HIF-1α–LIPIN1 signalling, suppression of PPARα-dependent oxidative pathways and increased lipid uptake, collectively contributing to fatty liver progression [ 54 ]. Similarly, SIRT5 contributes to hypoxic adaptation by desuccinylation upstream regulators that reduce HIF-1α ubiquitination and degradation, thereby enhancing HIF-1α stability and promoting glycolytic rewiring in colorectal cancer [ 55 ]. These findings support a multi-compartment model in which nuclear transcriptional control, cytosolic regulation of protein stability and mitochondrial redox signalling jointly determine the magnitude and duration of HIF-1α-driven glycolysis. Recent work further indicates that SIRT4 can modulate HIF-1α stability under pharmacologically induced mitochondrial stress, but not under basal conditions [ 56 ] (Fig.  2 ). Fig. 2 Multi-compartmental sirtuin control of HIF-1α-dependent glycolysis. Sirtuins across nuclear, cytosolic and mitochondrial compartments converge on HIF-1α to regulate glycolytic flux through coordinated transcriptional, post-translational and redox-dependent mechanisms. Distinct sirtuin isoforms modulate HIF-1α stability, activity and chromatin output via NAD + -dependent deacetylation, deacylation, hydroxylation and ubiquitination, collectively shaping hypoxic adaptation and metabolic homeostasis\nMulti-compartmental sirtuin control of HIF-1α-dependent glycolysis. Sirtuins across nuclear, cytosolic and mitochondrial compartments converge on HIF-1α to regulate glycolytic flux through coordinated transcriptional, post-translational and redox-dependent mechanisms. Distinct sirtuin isoforms modulate HIF-1α stability, activity and chromatin output via NAD + -dependent deacetylation, deacylation, hydroxylation and ubiquitination, collectively shaping hypoxic adaptation and metabolic homeostasis\nAerobic glycolysis is increasingly understood as a process governed by discrete, rate-limiting enzymatic checkpoints, not as a uniformly regulated metabolic programme. While hypoxia-inducible factor 1α (HIF-1α)-driven transcription establishes a permissive framework for glycolytic activation, accumulating evidence indicates that the magnitude and routing of glycolytic flux are ultimately determined at key enzymatic nodes. In this checkpoint-centric model, sirtuins operate in parallel with transcriptional programs as NAD + -sensitive modulators of glycolytic throughput. By sensing changes in redox state and lysine acylation, sirtuins impose node-specific control over glycolytic enzymes that determine pathway commitment, redox balance and downstream metabolic signalling.\nA clinically relevant illustration of this principle is provided by oxaliplatin-resistant colorectal cancer, in which PARP-driven NAD +  depletion suppresses SIRT1 activity and is associated with enhanced glycolytic throughput, whereas restoration of SIRT1 attenuates glycolytic output and mitigates chemoresistance [ 57 ]. This metabolic shift cannot be explained by changes in glycolytic enzyme abundance alone, but instead reflects altered NAD + -dependent sirtuin activity. These observations demonstrate that modulation of sirtuin activity can reprogramme glycolytic flux independently of transcriptional induction, setting the stage for node-specific control at key enzymatic checkpoints.\nAt the entry point of glycolysis, hexokinase 2 (HK2) functions as a key commitment checkpoint whose activity determines whether glucose is channelled into sustained glycolytic flux. Multiple sirtuins target HK2, although the direction and mechanism of regulation are highly stress-type and tissue specific. Environmental stressors provide instructive examples of this context specificity. Low-dose arsenic exposure promotes aerobic glycolysis through engagement of a SIRT1–p53–HK2 axis, in which SIRT1 suppresses p53 expression and deacetylates p53 at lysine 382, thereby relieving p53-mediated repression of HK2 and augmenting glycolytic output [ 58 ]. In contrast, chronic hydroquinone exposure is associated with the selection of a pro-glycolytic SIRT1–HK2 programme that supports malignant progression; in this context, genetic or pharmacological inhibition of SIRT1 suppresses HK2 expression, reduces glycolytic throughput and attenuates tumorigenicity [ 57 ]. These examples illustrate how SIRT1 modulates early glycolytic commitment in a stress-dependent manner, not enforcing a fixed metabolic outcome. Regulation of HK2, however, extends beyond SIRT1, indicating that multiple sirtuin pathways converge on this entry checkpoint.\nSIRT6 illustrates the stress-dependent variability of sirtuin control at glycolytic entry points. While often characterised as a transcriptional brake on glycolysis, SIRT6 can, under defined therapeutic and signalling conditions, facilitate glycolytic commitment. In EGFR-mutant non–small-cell lung cancer, sustained SIRT6 activity promotes a HIF-1α–HK2 axis that maintains glycolytic flux under EGFR inhibition, thereby contributing to erlotinib resistance. Disruption of this axis suppresses glycolysis and restores drug responsiveness, which indicates that SIRT6 does not impose a fixed metabolic outcome but instead modulates glycolytic entry in a chromatin- and stress-dependent manner [ 59 ]. By contrast, in inflammatory glioblastoma contexts, SIRT6 acts as a chromatin-level constraint on glycolytic entry. Through cooperation with the transcription factor MZF1, SIRT6 reinforces nucleosomal occupancy at the HK2 promoter, thereby limiting HK2 inducibility and restraining glycolytic activation [ 60 ]. These divergent observations likely reflect the capacity of SIRT6 to integrate transcription factor context, chromatin accessibility and stress signalling to shape HK2 output, rather than enforcing a fixed regulatory direction. Accordingly, HK2 functions as a tunable commitment step that couples NAD + /acetylation state with tissue- and insult-specific regulatory control.\nBeyond the entry checkpoint at HK2, control of glycolysis is further refined at its distal steps. Lactate dehydrogenase A (LDHA) and pyruvate kinase M2 (PKM2) act as key regulatory nodes where glycolytic flux intersects with redox balance and signalling. LDHA supports lactate production by regenerating NAD +  from NADH, thereby influencing the NAD + /NADH ratio that governs sirtuin activity. This relationship is reciprocal, as sirtuin-dependent metabolic state also feeds back to shape LDHA function. Beyond direct transcriptional control of LDHA, sirtuins also regulate LDHA-centred pathways through chromatin-mediated feedback. For example, loss of SIRT6 increases histone H3K9 acetylation at the Got1 promoter via TIP60, activating a Got1–LDHA axis that supports elevated glycolytic output in disease settings [ 61 ]. In addition to transcriptional and chromatin-based regulation, LDHA activity is also directly tuned by sirtuin-dependent post-translational modifications. In aggressive prostate cancer, SIRT5 functions as an LDHA desuccinylase, at lysine 118, linking LDHA succinylation to enzymatic activity, lactate production and metastatic potential [ 62 ]. Thus, LDHA emerges as a terminal redox node directly coupled to NAD + -dependent sirtuin activity.\nPKM2 offers a parallel case at the distal end of glycolysis: it controls pyruvate formation and can also enter the nucleus to support pro-growth transcriptional programs. In this setting, SIRT6 is a key regulator of PKM2 at multiple levels. One well-defined route is post-transcriptional: SIRT1 and SIRT6 deacetylate the splicing factor hnRNP A1, shifting splicing from PKM2 toward PKM1 and thereby limiting the PKM2-linked signalling state [ 63 ]. Similarly, SIRT6 can directly deacetylate nuclear PKM2 at lysine 433, promoting exportin-4–dependent cytoplasmic relocation and further restricting PKM2’s transcriptional co-activator functions [ 50 ]. However, PKM2 is not an exclusive target of SIRT6. In distinct cellular and pathological contexts, other sirtuins can engage PKM2 through alternative acetylation sites to achieve different metabolic outcomes. For example, in models of Parkinson’s disease, SIRT1 directly deacetylates PKM2 at lysines 135 and 206, reducing PKM2 enzymatic activity and lactate production, thereby constraining glial activation and neuroinflammatory pathology [ 64 ]. Beyond SIRT6 and SIRT1, SIRT2 further extends PKM2 regulation in a tissue- and stress-dependent manner. Under hepatic metabolic stress, SIRT2 deacetylation couples PKM2 acetylation status to ubiquitination and protein stability, thereby shaping glycolytic flux and energy balance [ 65 ]. In inflammatory immunity, the same modification restrains PKM2’s non-canonical nuclear kinase activity linked to STAT3 signalling and Th17-driven pathology [ 66 ]. These findings identify PKM2 as a recurrent sirtuin-sensitive node where metabolic flux control and signal transduction converge.\nThese studies support a checkpoint-centric view in which sirtuins modulate glycolytic flux through discrete, high-control nodes rather than acting as uniform activators or repressors of glycolysis. The metabolic outcome of sirtuin action is therefore jointly determined by tissue identity, chromatin state and stress signalling. This framework provides a coherent basis for understanding the opposing roles of individual sirtuins across disease settings (Fig.  3 ). Fig. 3 Sirtuin-mediated control of glycolytic flux at rate-limiting enzymatic checkpoints. Sirtuins regulate glycolytic output by imposing node-specific control over key rate-limiting enzymes, including HK2 at pathway entry, PKM2 at the flux and signalling interface, and LDHA at the terminal redox step. Through NAD + -dependent deacetylation and deacylation, distinct sirtuin isoforms tune enzyme activity, stability, localization and transcriptional coupling, enabling stress-responsive rewiring of glycolysis in response to metabolic stress and disease states\nSirtuin-mediated control of glycolytic flux at rate-limiting enzymatic checkpoints. Sirtuins regulate glycolytic output by imposing node-specific control over key rate-limiting enzymes, including HK2 at pathway entry, PKM2 at the flux and signalling interface, and LDHA at the terminal redox step. Through NAD + -dependent deacetylation and deacylation, distinct sirtuin isoforms tune enzyme activity, stability, localization and transcriptional coupling, enabling stress-responsive rewiring of glycolysis in response to metabolic stress and disease states\nInstead of acting as redundant regulators, sirtuins form a compartment-stratified network whose influence shifts between mitochondrial redox control and nuclear transcriptional constraint depending on stress duration. This dynamic organisation helps explain the stress-duration-dependent effects of sirtuin isoforms on glycolytic flux. Such flux tuning also determines whether glycolytic carbon is retained for lactate production or diverted toward anabolic pathways, most notably lipid synthesis via the citrate–ACLY–acetyl-CoA axis.\nBeyond shaping glycolytic throughput, checkpoint control determines how carbon flux is redistributed toward biosynthetic programs. One major route for such redistribution is mitochondrial citrate export and its conversion to nuclear acetyl-CoA by ATP-citrate lyase (ACLY), thereby linking glucose-derived carbon to chromatin acetylation and transcriptional control of lipogenic capacity. Within this interface, nuclear sirtuins translate acetyl-CoA availability into transcriptional control of lipid synthesis. SIRT6 constrains de novo lipogenesis by repressing SREBP1/2-driven lipogenic gene programs through promoter-associated histone deacetylation, thereby coupling glycolytic carbon routing to chromatin-level restraint of lipid anabolic output [ 67 ,  68 ].\nWhile nuclear sirtuins primarily constrain lipogenic transcription at the chromatin level, mitochondrial sirtuins can influence lipid-related cell fate decisions through metabolic priming. Among them, SIRT4 has been implicated in early adipogenic commitment. In experimental models, SIRT4 promotes adipocyte differentiation by enhancing branched-chain amino acid (BCAA) catabolism, thereby increasing leucine oxidation and facilitating PPARγ-driven transcriptional programs [ 69 ]. Reduced SIRT4 expression in multiple diabetic mouse models further suggests a potential link between SIRT4-mediated BCAA metabolism and adipose tissue dysfunction, although causal relationships remain to be fully elucidated [ 69 ]. This pro-adipogenic role of SIRT4 contrasts with the predominantly antilipogenic and chromatin-restrictive actions reported for nuclear sirtuins such as SIRT1 and SIRT6, which reflects an isoform-specific division of labour in control of adipogenesis within the sirtuin family.\nExtending this chromatin-based framework, SIRT6 further integrates circadian and nutrient signals to fine-tune SREBP-dependent lipogenesis. SREBP1 orchestrates transcription of FASN, ACC1, SCD1 and other lipogenic genes, and its rhythmic recruitment to chromatin is coordinated by SIRT6 together with the core clock components CLOCK and BMAL1, thereby generating daily oscillations in hepatic fatty acid and cholesterol synthesis [ 70 ]. Loss of SIRT6 disrupts these lipid rhythms and is associated with fatty acid metabolic defects, indicating that SIRT6 does not merely repress lipogenesis but helps encode circadian timing into the chromatin accessibility of SREBP programmes, thereby stabilizing lipid homeostasis across feeding-fasting cycles.\nBeyond temporal control, SIRT6 reinforces lipogenic restraint through coordinated regulation of the SREBP1 signalling axis. It attenuates SREBP1 expression and processing, promotes AMPK-dependent inhibitory phosphorylation of SREBP1, and is itself repressed by intragenic miR-33b/a embedded within SREBP loci, establishing a reciprocal feedback circuit that buffers lipogenic output across nutrient states [ 71 ]. In the liver, this restraint is further implemented through direct deacetylation of key lipogenic transcriptional regulators, including SREBP1, ChREBP and LXRα, and hepatocyte-specific Sirt6 deletion markedly exacerbates diet-induced steatosis [ 72 – 74 ]. Consistent with a signalling-chromatin coupling mechanism, SIRT6 also activates AMPKα and inhibits mTORC1 to repress SREBP1c and improve hepatic and pancreatic coordination of glucose and lipid metabolism during high-fat feeding or starvation. Nutritional and environmental cues converge on this axis: black tea theaflavins upregulate SIRT6, enhance AMPK activity and downregulate SREBP1 and its targets to correct dyslipidaemia in obese mice [ 75 ], whereas arsenic-induced lncRNA PU.1 AS relieves EZH2-mediated repression of Sirt6, increases SIRT6 expression and limits hepatocellular triglyceride accumulation [ 76 ].\nSIRT6 couples glycolytic carbon flux to chromatin-mediated lipogenesis through ATP citrate lyase (ACLY), an enzyme that links cellular metabolism to histone acetylation. ACLY generates cytosolic and nuclear acetyl-CoA from citrate, and nuclear ACLY-derived acetyl-CoA fuels histone-acetyltransferase-dependent histone acetylation at growth-promoting loci [ 77 ,  78 ]. Thus, citrate export from glycolysis is translated into chromatin accessibility and activation of lipogenic gene programmes. By restricting nuclear ACLY accumulation, SIRT6 is proposed to lower nuclear acetyl-CoA availability and reduce histone acetylation at pro-invasive loci, consistent with suppression of adhesion and migration programmes and reduced invasiveness. Thus, SIRT6 does not merely repress downstream enzymes, but modulates the epigenetic substrate supply that determines the transcriptional permissiveness of lipogenic pathways. In parallel with this chromatin-level regulation, SIRT6 represses FASN and other lipogenic enzymes, and SIRT6-AMPK signalling downregulates SCD1 and improves hepatic lipid handling [ 77 – 80 ]. These downstream effects constitute the execution layer of SIRT6-dependent lipogenic restraint. Consistent with this model, dietary polyphenols and herbal extracts such as EGCG and Rhubarb tea activate a SIRT6-AMPK-SREBP1-FASN cascade, increase hepatic SIRT6 expression and reduce triglyceride and diglyceride accumulation in models of diet-induced metabolic dysfunction [ 79 ,  80 ]. These studies identify SIRT6 as a chromatin-based brake on SREBP- and ACLY-driven lipogenesis, linking glycolysis-derived acetyl-CoA supply to epigenetic control of lipid metabolic output.\nBeyond fatty acid synthesis, SIRT6 also extends its chromatin-based control to cholesterol biosynthesis. It is recruited by FoxO3 to the Srebp2 promoter, where it deacetylates H3K9 and H3K56, compacts chromatin and represses Srebp2 transcription [ 81 ,  82 ]. Consistent with this mechanism, liver-specific Sirt6 deletion elevates cholesterol levels, whereas SIRT6 overexpression corrects diet-induced and genetically driven hypercholesterolaemia, supporting a SIRT6-FoxO3-SREBP2 axis as a key checkpoint in cholesterol homeostasis. In addition, SIRT6 deacetylates and stabilises FoxO3a, enhancing its nuclear activity in pathways that coordinate lipogenesis, FAO and oxidative stress. Under exercise conditions, this axis contributes to suppression of PCSK9 through epoxyeicosatrienoic acid (EET) signalling [ 83 ]. Through these FoxO-dependent circuits, SIRT6 integrates fatty acid and cholesterol regulation within a unified chromatin-based framework of lipid metabolic control.\nWhile mitochondrial sirtuins such as SIRT4 influence adipogenic commitment, nuclear sirtuins, particularly SIRT6, impose epigenetic restraint on both fatty acid and cholesterol biosynthesis through coordinated regulation of ACLY, SREBP1/2 and FoxO-dependent circuits.\nBy limiting de novo lipogenesis and buffering transcriptional responses to nutrient excess, this framework shifts cellular metabolism toward balanced lipid utilisation, coordinating lipogenic suppression with fatty acid oxidation (FAO) and mitochondrial energy metabolism.\nIn addition to lipogenic control, metabolic homeostasis requires efficient fatty acid β-oxidation (FAO). FAO provides sustained ATP production during fasting, hypoxia, inflammatory stress, and its activation represents a coordinated, metabolic shift not a simple change in fuel preference. In this process, the sirtuin family acts as a compartmentalised NAD + -sensitive regulatory system. Nuclear sirtuins promote transcription of FAO-related genes, cytosolic sirtuins modulate lipid utilization, and mitochondrial sirtuins directly regulate β-oxidation enzymes and redox balance. Through these coordinated actions, sirtuins adjust FAO flux to match cellular energy demand and environmental stress.\nAt the nuclear level, SIRT1 acts as an NAD + -sensitive transcriptional regulator that promotes metabolic flexibility by shifting energy production away from glycolysis and toward fatty acid β-oxidation under stress. In colorectal carcinoma, nutrient deprivation and elevated ROS induce SIRT1 upregulation, triggering deacetylation and nuclear export of β-catenin, suppression of glycolytic throughput and a compensatory increase in FAO-dependent mitochondrial ATP production [ 84 ]. Conversely, smoking-associated metabolic stress reduces SIRT1 expression, impairs PPARα-CPT1a-driven FAO and leads to pathological lipid accumulation [ 85 ]. Beyond these stress-induced changes in expression, SIRT1 activity can be further amplified through NSD2-mediated methylation of the SIRT1 activator AROS, enhancing FAO capacity and contributing to tumour metabolic reprogramming [ 86 ]. These observations support SIRT1 as a nuclear switch that links environmental stress cues to transcriptional programmes sustaining FAO competence.\nSIRT6 complements SIRT1 by establishing the chromatin environment that supports FAO activation. Through direct association with PPARα and its target promoters, SIRT6 deacetylates the co-activator NCOA2 at Lys780, thereby enhancing PPARα-driven transcription of FAO genes while simultaneously repressing SREBP-dependent lipogenesis [ 87 ]. This SIRT6-PPARα axis is supported by multiple in vivo models in which pharmacological or natural compounds increase SIRT6 expression, enhance hepatic FAO and improve metabolic homeostasis in a SIRT6-dependent manner [ 70 ,  88 – 91 ]. However, small-molecule SIRT6 activators such as MDL-800 paradoxically suppress PPARα and FAO through oxidative stress-driven JNK activation, independent of SIRT6 deacetylase activity, which reflects the paradoxical and pathway-specific complexity of targeting this pathway [ 92 ]. Together, SIRT1 and SIRT6 form a nuclear regulatory layer that shifts cellular metabolism away from glycolytic and lipogenic programmes toward sustained fatty acid oxidation.\nBeyond its nuclear transcriptional roles, SIRT2 also regulates FAO in the cytosol through reversible deacetylation of key FAO enzymes, thereby expanding the functional scope of the sirtuin family. In diabetic cardiomyopathy, a stress-induced mitochondrial pool of SIRT2 emerges. In these settings, SIRT2 directly deacetylates CPT2, a key enzyme, and modulates β-oxidation capacity in a tissue- and stress-specific manner [ 93 ]. This mitochondrial relocalization represents a non-canonical regulatory mechanism in which SIRT2 extends beyond its classical cytosolic functions to directly influence mitochondrial fatty acid metabolism.\nAt the mitochondrial level, SIRT3 acts as a central regulator that coordinates β-oxidation capacity with structural and redox adaptations necessary for sustained oxidative metabolism. In human induced pluripotent stem cell-derived cardiomyocytes, SIRT3 preserves an adult-like oxidative phenotype by deacetylating the mitochondrial fusion OPA1. This action maintains mitochondrial network integrity, enhances respiratory capacity and promotes a metabolic shift toward fatty acid oxidation while limiting glycolytic dependence [ 94 ]. Beyond cardiomyocytes, adipocyte-specific overexpression of SIRT3 increases L-carnitine levels and reduces long-chain acylcarnitine accumulation. These changes facilitate efficient long-chain FAO and improve lipid metabolism, thereby enhancing insulin sensitivity [ 95 ].\nSIRT5 adds an additional enzymatic layer to mitochondrial FAO regulation through its desuccinylase activity. In diabetic cardiomyopathy, SIRT5 deficiency results in hyper-succinylation of CPT2 at Lys424, impairing FAO efficiency and leading to fatty acylcarnitine accumulation and cardiac lipotoxicity [ 96 ]. In steroid-insensitive asthma, SIRT5 upregulation enhances FAO but aggravates mitochondrial oxidative stress. Conversely, pharmacological inhibition with the selective SIRT5 inhibitor MC3482 increases succinylation of FAO enzymes, reduces FAO flux and alleviates mitochondrial dysfunction and airway inflammation [ 97 ]. These findings indicate that SIRT5 fine-tunes FAO activity through reversible lysine succinylation, but its metabolic impact is strongly shaped by tissue type and substrate availability.\nBeyond its transcriptional cooperation with PPARα, SIRT6 acts as a nuclear integrator linking FAO to cellular stress adaptation across metabolic tissues. In the liver, SIRT6 preserves mitochondrial integrity and limits lipotoxic injury, thereby restraining progression from steatosis to hepatocellular carcinoma [ 98 – 100 ]. Mechanistically, SIRT6 activation enhances hepatic FAO capacity and stabilises NRF2-dependent antioxidant responses, improving lipid homeostasis in fatty liver models [ 101 – 103 ]. In addition, SIRT6 modulates endoplasmic reticulum stress through regulation of XBP1s, further coordinating metabolic and stress signaling pathways [ 104 ,  105 ].\nThe protective metabolic functions of SIRT6 extend beyond the liver to the kidney and vasculature. In diabetic nephropathy and acute kidney injury, SIRT6 preserves mitochondrial integrity and limits lipid peroxidation, thereby restraining ferroptotic injury via modulation of Nrf2/GPX4 signalling and ACMSD suppression [ 106 ,  107 ]. SIRT6 also maintains podocyte and endothelial function, mitigating vascular injury [ 108 ,  109 ]. In macrophages, SIRT6 promotes lipophagy and cholesterol efflux, reducing foam cell formation [ 107 ,  110 ]. Moreover, myeloid-specific Sirt6 deletion implicates SIRT6 in systemic energy expenditure and adipose–immune crosstalk [ 111 ].\nFAO is thus governed by a spatially organized regulatory hierarchy spanning three subcellular compartments. In the nucleus, SIRT1 and SIRT6 promote the metabolic shift toward FAO by repressing glycolytic and lipogenic programmes while activating PPAR-dependent oxidative transcription. SIRT2 introduces a stress-responsive layer, extending sirtuin control to mitochondrial FAO enzymes through stress-induced cytoplasmic-to-mitochondrial relocalization. Within mitochondria, SIRT3 and SIRT5 fine-tune β-oxidation capacity by regulating mitochondrial structure, acylation status and redox balance. This compartmentalized sirtuin network positions FAO under integrated transcriptional and post-translational control, linking thermogenesis and energetic demands (Fig.  4 ). Fig. 4 Compartmentalized sirtuin regulation of lipid metabolic homeostasis.  a  Glycolysis–lipid interface: SIRT6 couples glycolytic flux to lipogenesis by controlling nuclear acetyl-CoA availability and deacetylating histones and SREBP1/2 to restrain lipogenic gene expression.  b  Fatty acid β-oxidation: Nuclear SIRT1 and SIRT6 promote FAO gene transcription through PGC-1α/PPARα and NCOA2 deacetylation, whereas mitochondrial SIRT3, SIRT5 and SIRT2 regulate β-oxidation capacity via OPA1 deacetylation, CPT2 desuccinylation and stress-induced mitochondrial relocalization.  c  Lipid mobilization, thermogenesis and adipogenic commitment: SIRT6, SIRT1 and SIRT4 modulate lipolysis (ATGL/HSL), thermogenesis (PGC-1α–UCP1) and adipocyte differentiation.  d  Lipotoxic stress and metabolic homeostasis: Sirtuins restrain ROS, ER stress and ferroptosis to protect against lipotoxic stress, with context-dependent outcomes\nCompartmentalized sirtuin regulation of lipid metabolic homeostasis.  a  Glycolysis–lipid interface: SIRT6 couples glycolytic flux to lipogenesis by controlling nuclear acetyl-CoA availability and deacetylating histones and SREBP1/2 to restrain lipogenic gene expression.  b  Fatty acid β-oxidation: Nuclear SIRT1 and SIRT6 promote FAO gene transcription through PGC-1α/PPARα and NCOA2 deacetylation, whereas mitochondrial SIRT3, SIRT5 and SIRT2 regulate β-oxidation capacity via OPA1 deacetylation, CPT2 desuccinylation and stress-induced mitochondrial relocalization.  c  Lipid mobilization, thermogenesis and adipogenic commitment: SIRT6, SIRT1 and SIRT4 modulate lipolysis (ATGL/HSL), thermogenesis (PGC-1α–UCP1) and adipocyte differentiation.  d  Lipotoxic stress and metabolic homeostasis: Sirtuins restrain ROS, ER stress and ferroptosis to protect against lipotoxic stress, with context-dependent outcomes\nActivation of fatty acid oxidation requires prior mobilization of stored lipids. Lipolysis and lipid-droplet remodelling release fatty acids that fuel mitochondrial β-oxidation and thermogenic programmes in brown and beige adipocytes. These processes form a coordinated axis linking lipid mobilization, oxidative metabolism and heat production.\nWithin this axis, sirtuins function as NAD + -dependent regulators that integrate nutritional and hormonal cues to adjust lipid utilization and thermogenic capacity. SIRT1 promotes lipolysis and beige adipocyte formation through activation of PGC-1α-dependent transcriptional programmes [ 112 ]. Mitochondrial sirtuins such as SIRT3 may reinforce this response by sustaining oxidative metabolism and limiting reactive oxygen species. Although other isoforms are also implicated in adipose lipid handling, their contributions appear to depend on tissue and metabolic context.\nAmong nuclear isoforms, SIRT6 provides a direct link between chromatin regulation and lipid mobilization. SIRT6 enhances lipolysis partly through deacetylation of KAT8, which restores transcription of ATGL and HSL [ 113 ]. Similarly, SIRT6 promotes thermogenic gene expression by deacetylating FoxO1 and facilitating Ucp1 transcription via a proposed mTORC2–SIRT6–FoxO1 pathway [ 114 ]. Consistent with these findings, reduced SIRT6 and ATGL expression in human adipose tissue supports the translational relevance of this regulatory axis [ 115 ]. During intermittent fasting, SIRT6 is required for optimal white-to-brown adipose conversion, as Sirt6 deficiency suppresses p38 MAPK–ATF2 signalling and UCP1 expression [ 116 ].\nThese data identify SIRT6 as a nuclear coordinator of lipolysis and thermogenic reprogramming. In the broader family context, SIRT1 and mitochondrial sirtuins complement this function by maintaining PGC-1α-driven oxidative adaptation and mitochondrial integrity, illustrating a division of labour across subcellular compartments.\nSustained fatty acid oxidation and lipid clearance are essential to prevent lipotoxic injury during chronic lipid excess. Cholesterol overload and accumulation of toxic lipid disrupt mitochondrial and endoplasmic reticulum function, promoting oxidative stress and driving inflammatory responses that contribute to the progression from steatosis to steatohepatitis.\nUnder these conditions, sirtuins regulate lipid turnover and mitochondrial oxidation to preserve tissue integrity. Terf2ip loss reduces Sirt1 transcription and suppresses AMPK signalling, impairing lipophagy and mitochondrial β-oxidation. Restoration of Terf2ip expression or pharmacological activation of SIRT1 reactivates lipid-clearing pathways and attenuates hepatic lipotoxic inflammation [ 117 ].\nSIRT6 also participates in limiting lipid-driven oxidative stress in metabolic organs. In high-fat diet-induced cardiac hypertrophy, Sirt6 deficiency increases myocardial lipid accumulation and reactive oxygen species production. Conversely, SIRT6 is associated with elevated ENDOG and SOD2 expression and reduced mitochondrial oxidative damage [ 118 ]. In the liver, SIRT6 engages a PGC-1α–ENDOG pathway linked to suppression of oxidative stress during high-fat diet-induced injury, and metformin may exert part of its hepatoprotective effect through this mechanism [ 118 ]. However, SIRT6-mediated oxidative regulation is tissue- and tumour-stage dependent. In certain tumour settings, SIRT6-driven iron mobilization and lipid-peroxidation pathways can enhance susceptibility to oxidative lipid damage. In anaplastic thyroid carcinoma, SIRT6 promotes ferroptotic vulnerability by inducing NCOA4-dependent ferritinophagy and increasing lipid peroxidation [ 119 ]. These findings reveal the disease- and tissue-specific nature of SIRT6-mediated oxidative lipid stress [ 118 – 120 ] (Fig.  5 ). Fig. 5 SIRT6 as an input-integrating epigenetic brake on lipogenesis and lipid-stress outcomes. Schematic overview of how SIRT6 integrates NAD +  availability, circadian CLOCK–BMAL1 cues, nutrient stress and oxidative/ER stress to regulate lipid metabolism across multiple tissues. Through four interconnected modules (circadian gating of SREBP1 chromatin recruitment, AMPK-mTORC1-mediated energy gating of lipogenesis, direct deacetylation of lipogenic transcription factors and chromatin marks, and regulation of nuclear acetyl-CoA supply via ACLY), SIRT6 restrains de novo lipogenesis while coordinating fatty acid oxidation, lipolysis and thermogenesis. These integrated actions promote metabolic protection in lipid-stressed tissues by limiting steatosis, oxidative injury and ferroptotic susceptibility, but can confer metabolic vulnerability in select malignancies, which highlights the context-specific outputs of SIRT6-dependent lipid regulation\nSIRT6 as an input-integrating epigenetic brake on lipogenesis and lipid-stress outcomes. Schematic overview of how SIRT6 integrates NAD +  availability, circadian CLOCK–BMAL1 cues, nutrient stress and oxidative/ER stress to regulate lipid metabolism across multiple tissues. Through four interconnected modules (circadian gating of SREBP1 chromatin recruitment, AMPK-mTORC1-mediated energy gating of lipogenesis, direct deacetylation of lipogenic transcription factors and chromatin marks, and regulation of nuclear acetyl-CoA supply via ACLY), SIRT6 restrains de novo lipogenesis while coordinating fatty acid oxidation, lipolysis and thermogenesis. These integrated actions promote metabolic protection in lipid-stressed tissues by limiting steatosis, oxidative injury and ferroptotic susceptibility, but can confer metabolic vulnerability in select malignancies, which highlights the context-specific outputs of SIRT6-dependent lipid regulation\n\nImmune activation is inseparable from metabolic reprogramming: nutrient utilization, mitochondrial function and redox balance actively instruct effector differentiation, tolerance and tissue repair, not serving as passive by-products of inflammation. Metabolism and immunity thus operate as an integrated regulatory system, not parallel processes.\nThe sirtuin network extends its NAD + -sensing logic from metabolism into immunity: the same compartmentalised isoforms that govern glycolytic flux and mitochondrial quality control also shape inflammatory signalling, cytokine production, and immune cell fate decisions. Within this framework, sirtuins (SIRT1–SIRT7) function as NAD + -dependent sensors that translate cellular energetic status into immune outcomes. Because their enzymatic activity depends on NAD +  availability, they are uniquely positioned to couple metabolic flux to transcriptional control, stress adaptation and inflammatory signaling across stromal, innate and adaptive compartments.\nTheir actions are compartmentalized. Nuclear isoforms such as SIRT1 and SIRT6 remodel chromatin landscapes and shape cytokine programs and tolerance pathways. In contrast, cytosolic and mitochondrial members (SIRT2–SIRT5) recalibrate glycolysis, fatty acid oxidation, and organelle stress responses, thereby influencing inflammatory amplification and cell-death susceptibility. SIRT7 further links nucleolar stress to lymphocyte differentiation.\nHowever, key questions remain unresolved: to what extent immunometabolic sirtuin effects are cell-type-specific versus conserved across immune lineages, and whether the seven sirtuins operate as independent checkpoints or as a coordinated network in which the composite NAD + /NADH ratio simultaneously tunes multiple isoforms to produce a unified immune output. The sirtuin family constitutes a mechanistic bridge between metabolic cues and inflammatory regulation. This conceptual integration provides the foundation in the sections that follow.\nSirtuins regulate epithelial-stromal communication and barrier stability through at least two functionally distinct mechanisms: maintaining epithelial barrier integrity and restraining stromal fibroblast activation. In endometriosis, SIRT1 is proposed to promote epithelial plasticity by enabling endometrial epithelial cells to evade damage- or oncogene-induced senescence, thereby facilitating EMT-associated transcriptional programmes and invasive behaviour. Clinical lesions show increased stromal fibroblasts with reduced epithelial components, suggesting a remodelled niche that may further reinforce SIRT1-driven EMT through microenvironmental feedback. Nevertheless, this remains a correlative observation; whether the stromal expansion is a cause or consequence of epithelial SIRT1 activity has not been established [ 121 ]. Maintenance of barrier integrity involves both SIRT1-dependent epithelial plasticity and SIRT2-regulated junctional stability. In the gut, pharmacological inhibition of SIRT2 has been reported to alleviate colitis by stabilizing surface E-cadherin and strengthening adherens junctions. Genetic deletion of SIRT2 does not recapitulate this effect, and a PROTAC degrader targeting SIRT2 likewise fails to protect. The original study attributed this discrepancy to the fact that SIRT2 inhibitors suppress only some but not all SIRT2 activities, whereas complete protein ablation removes all catalytic and non-catalytic functions, emphasizing the therapeutic value of substrate-dependent inhibitors over complete SIRT2 elimination [ 122 ]. This divergence indicates that enzymatic inhibition and genetic deletion are not mechanistically equivalent, and reveals context- and substrate-dependent functions of sirtuins in barrier tissues.\nWithin stromal compartments, SIRT6 functions as a chromatin-based repressor of fibroblast activation. By limiting senescence-associated secretory phenotypes (SASP) and fibrogenic programs, SIRT6 helps maintain stromal homeostasis in multiple tissues [ 123 – 127 ]. In breast cancer, a subset of cancer-associated fibroblasts (CAFs) exhibits a senescent, metabolically reprogrammed myofibroblastic phenotype characterized by high TSPAN8 expression. In these cells, SIRT6 levels are markedly reduced, and its stability is controlled by a TSPAN8-MAPK11-RBBP6 phosphorylation pathway [ 123 ,  124 ]. Decreased SIRT6 correlates with enhanced protumour activity and increased secretion of SASP cytokines such as IL-6 and IL-8. Combined SIRT6 activation (MDL-800) and TSPAN8 inhibition reverses CAF-induced stemness and chemoresistance [ 123 ], providing functional rescue evidence supporting a causal role for SIRT6 loss in stromal tumour promotion.\nBeyond tumour stroma, SIRT6 exerts antifibrotic and anti-inflammatory effects in multiple fibroblast-like populations. In rheumatoid arthritis, SIRT6 is downregulated in fibroblast-like synoviocytes. Restoring its expression suppresses MyD88–ERK signalling and reduces IL-6 and TNF production, thereby limiting synovial inflammation and tissue injury [ 125 ]. A similar pattern is observed in keloid fibroblasts and photoaged dermal fibroblasts. Reduced SIRT6 promotes fibroblast proliferation, collagen deposition and inflammatory cytokine release, whereas SIRT6 restoration attenuates these changes by restraining MAPK–ERK and NRF2–HO-1–NF-κB pathways [ 126 ,  127 ]. In hepatic stellate cells, SIRT6 directly deacetylates Smad3 to repress fibrogenic gene transcription, mitigating inflammation and fibrosis in non-alcoholic steatohepatitis [ 128 ]. These findings are consistent with a shared mechanism in which SIRT6 limits fibroblast activation and inflammatory signalling across tissues, thereby restraining pathological fibrosis. However, it remains unclear whether SIRT6 activity in distinct fibroblast populations is governed by shared upstream signals or tissue-specific regulatory inputs, and whether these mechanisms are causally linked across acute and chronic fibrotic settings.\nSirtuins operate at both sides of the stromal–epithelial interface: SIRT1 and SIRT2 modulate epithelial plasticity and barrier integrity, whereas SIRT6 functions as a broadly acting repressor of fibroblast and stellate cell activation. However, the functional crosstalk between these two compartments is largely unexplored; for instance, whether SIRT1-driven stromal remodelling in endometriosis feeds back to alter SIRT6-dependent fibroblast quiescence has not been tested. Furthermore, key mechanistic conclusions currently rely heavily on pharmacological or overexpression studies, with loss-of-function approaches sometimes yielding discordant results (as exemplified by SIRT2 in colitis). Resolving these discrepancies and achieving a genuine compartment-specific understanding will require genetic models that permit cell-type-selective deletion of individual sirtuins, combined with systematic profiling of the epithelial and stromal sirtuin interactomes across different tissues and disease states. Such efforts will be essential to transform the current descriptive knowledge into a predictive framework that distinguishes universal sirtuin functions from context-specific adaptations.\nSirtuins act as immunometabolic checkpoints in innate immunity by coupling NAD + -dependent sensing of metabolic and redox states to epigenetic and signalling outputs. In innate immune and barrier-associated cells, distinct SIRT isoforms exert cell-type- and stimulus-specific control over inflammatory activation, antiviral responses, and immune tolerance, yet the specific signals that dictate which sirtuin is engaged in a given context remain poorly defined. Broadly, nuclear sirtuins (SIRT1, SIRT6) regulate transcriptional and chromatin-based immune signalling, whereas cytosolic and mitochondrial sirtuins (SIRT2, SIRT3, SIRT5) shape metabolic checkpoints and organelle homeostasis; SIRT7 links nucleolar stress to immune differentiation, though its innate immune roles are just emerging.\nSIRT1 is a prototypical nuclear regulator of antiviral innate immunity: it directly deacetylates IRF3 and IRF7, thereby restraining virus-induced type I (IFN-I) interferon signalling, and its inhibition or genetic deletion enhances IFN-I responses and antiviral defence [ 129 ]. Beyond antiviral responses, SIRT1 also contributes to antibacterial immunity. In murine pneumonia model, acute  Pseudomonas  infection induces a functional decline in SIRT1 protein, which correlates with reduced neutrophil recruitment and an immature, low-MPO phenotype, ultimately impairing bacterial clearance and worsening disease severity [ 130 ]. SIRT1 protein levels decline during infection even in mice with genetic overexpression, suggesting that therapeutic strategies may need to stabilize SIRT1 protein not merely enhancing its transcription. These findings position SIRT1 as a signalling-level regulator of both antiviral and antibacterial innate responses. However, the endogenous contribution of SIRT1 to immune homeostasis under physiological conditions, as opposed to acute infection models, is largely unexplored.\nSirtuin-mediated innate immune control extends beyond nuclear signal transduction to metabolic and organelle stress checkpoints, particularly in macrophages. This shift from nuclear to cytosolic regulation is exemplified by SIRT2.\nSIRT2 directly couples glycolytic flux to phagocytic function. Under acute ethanol exposure, SIRT2 deacetylates the glycolytic enzyme PFKP at K394/395, reducing glycolytic flux and impairing Atg4B-dependent LC3 activation [ 131 ]. This suppression of LC3-associated phagocytosis compromises bacterial clearance in murine sepsis models [ 132 ]. This work illustrates how a cytosolic sirtuin can regulate innate immunity by linking glycolysis to non-canonical autophagy, bypassing classical interferon pathways. SIRT2 tunes cell-death thresholds within pathogen survival niches. During Brucella abortus infection, elevated SIRT2 in bovine placental trophoblasts promotes mitochondrial apoptosis; paradoxically, pharmacological SIRT2 inhibition further enhances apoptotic signalling while reducing intracellular Brucella survival. This paradoxical result suggests that SIRT2 restrains apoptosis at a level that favours pathogen persistence, and that selective intervention may disrupt this balance [ 131 ].\nRecent evidence identifies SIRT2 as a cytosolic checkpoint on DNA sensing. It is currently the only sirtuin reported to directly suppress the cGAS–STING pathway through cytosolic condensate regulation [ 133 ]. Mechanistically, SIRT2 deacetylates G3BP1 at K257, K276, and K376, disrupting cGAS–G3BP1 condensates required for efficient DNA binding and cGAS activation. Pharmacological inhibition of SIRT2 with AGK2 enhances IFN responses and protects mice from HSV-1 infection, which reveals an isoform-specific antiviral mechanism that is distinct from the nuclear IRF regulation by SIRT1 and the mitochondrial metabolic checkpoints controlled by SIRT3 and SIRT5 [ 133 ].\nSIRT3, a mitochondrial sirtuin, coordinates efferocytosis and inflammasome activation with organelle homeostasis in innate immune cells. In cardiac macrophages, SIRT3 promotes MerTK-dependent efferocytic clearance of apoptotic cardiomyocytes and resolution of inflammation; Sirt3 deficiency exacerbates angiotensin II-induced cardiac inflammation by impairing this clearance and disrupting mitochondrial iron homeostasis [ 134 ]. Mechanistically, SIRT3 deacetylates Frataxin (FXN) to limit mitochondrial iron accumulation and support phagocytic clearance. Independently, during Salmonella Typhimurium infection, SIRT3 deacetylates NLRC4, promoting NLRC4 activation, ASC oligomerization, and IL-1β production; Sirt3 deficiency impairs this inflammasome response [ 135 ]. These findings identify SIRT3 as a mitochondrial checkpoint that integrates iron metabolism, efferocytosis, and inflammasome signalling. However, whether the efferocytic and inflammasome functions are mediated by distinct SIRT3 substrates or converge on common downstream pathways remains to be determined.\nSIRT5 provides an additional layer of metabolic control over innate inflammatory responses under oxidative and environmental stress. In eosinophilic asthma, SIRT5 overexpression in OVA-induced mouse models and HDM-stimulated airway epithelial cells suppresses eosinophilic inflammation by limiting ROS accumulation and activating the Nrf2–HO-1 antioxidant pathway [ 136 ]. Beyond the airway, SIRT5 also supports vascular homeostasis under systemic stress. In simulated microgravity, SIRT5 downregulation increases lysine succinylation of ERO1A at K396, promoting endothelial apoptosis and vascular dysfunction, whereas SIRT5 restoring expression reverses these effects [ 137 ]. These findings identify SIRT5 as a metabolic stress‑responsive regulator that preserves tissue integrity by constraining oxidative damage and cell death. However, current evidence relies heavily on overexpression and systemic stress models; the role of endogenous SIRT5 in basal innate immune homeostasis remains unclear.\nAlthough much innate immunity research has centred on macrophages, SIRT6 plays a key role in preserving endothelial and epithelial barrier homeostasis under inflammatory and metabolic stress. In lung microvascular endothelial cells, SIRT6 overexpression or pharmacological activation (UBCS039) reduces LPS-induced VCAM1 expression and monocyte adhesion by inhibiting NF-κB signalling [ 138 ]. In diabetic foot ulcers, reduced SIRT6 expression correlates with impaired healing, restoring SIRT6 activates Nrf2 signalling, limits oxidative stress and promotes angiogenesis [ 139 ]. During dengue virus infection, SIRT6 suppresses excessive innate immune activation by interfering with NF-κB p65 DNA binding and attenuating RLR- and TLR3-driven cytokine production [ 140 ]. In the intestinal epithelium, SIRT6 maintains R-spondin-1 expression and protects against inflammation-induced apoptosis [ 141 ]. These findings position SIRT6 as a brake on barrier‑associated inflammatory activation and a promoter of tissue repair. However, most studies employed overexpression or pharmacological activation; the contribution of endogenous SIRT6 to barrier immune homeostasis under steady state remains to be rigorously tested.\nBeyond barrier regulation, SIRT6 exerts stimulus- and disease-duration-dependent immunometabolic control in macrophage-related inflammation. In atherosclerosis, deletion of Sirt6 in bone marrow-derived cells increases MSR1 expression, enhances oxidised LDL uptake and promotes foam cell formation through derepression of c-Myc, a target normally restrained by SIRT6 [ 142 ]. During acute inflammatory stimulation with LPS, SIRT6 accumulates at the perinuclear endoplasmic reticulum and promotes TNF maturation via its demyristoylase activity, whereas pharmacological inhibition limits TNF release and alleviates sepsis and obesity-associated inflammation. In contrast, under chronic inflammatory conditions, SIRT6 supports resolution and tissue protection. It has been reported to enhance efferocytosis by facilitating miR-216/217 processing and activating the DEL-1–CD36 axis, although the underlying enzymatic activity responsible and the direct target remain to be defined [ 143 ]. In the kidney, proximal tubular Sirt6 deletion aggravates macrophage infiltration and fibrosis [ 144 ]; in adipose tissue, SIRT6 promotes IL-4-driven M2 polarisation, correlating with improved metabolic profiles in obesity [ 145 ]. SIRT6 emerges as a macrophage-centred regulator that balances lipid handling, cytokine release, and inflammatory resolution. However, it remains to be clarified whether the apparently opposing acute (pro-TNF) and chronic (pro-resolution) functions are mediated by distinct enzymatic activities, subcellular localizations, or cell-type-specific interactors.\nThe sirtuin family regulates innate immunity through a hierarchical architecture that mirrors its subcellular compartmentalization. SIRT1 acts at the nuclear signalling level by directly deacetylating IRF3 and IRF7 to tune antiviral and antibacterial transcriptional responses [ 130 ,  132 ]. SIRT2, SIRT3 and SIRT5 operate as cytosolic and mitochondrial metabolic-organelle checkpoints: SIRT2 couples glycolytic flux to phagocytic clearance and restrains cGAS–STING DNA sensing through condensate regulation [ 131 ,  134 ]; SIRT3 integrates iron homeostasis with efferocytosis and inflammasome activation [ 135 ,  136 ]; SIRT5 preserves tissue integrity by constraining ROS accumulation and aberrant succinylation under oxidative stress [ 137 ,  138 ]. SIRT6, uniquely, spans multiple subcellular compartments to integrate lipid metabolism, cytokine maturation [ 143 ,  144 ,  146 ], and barrier homeostasis [ 139 ,  140 ,  142 ,  145 ] in macrophages, endothelial cells [ 141 ,  147 ]. By contrast, the innate immune functions of SIRT4 and SIRT7 remain largely uncharacterized beyond emerging evidence for SIRT7 in nucleolar stress and inflammatory gene silencing. Several gaps in the current evidence limit a fully integrated view. First, for most isoforms, endogenous loss-of-function studies under steady-state conditions are lacking; the majority of conclusions derive from overexpression, pharmacological modulation, or acute challenge models, and whether basal sirtuin activity at physiological NAD +  levels exerts tonic immune control has not been tested. Second, it has not been established whether the seven sirtuins operate as independent checkpoints or as a coordinated NAD + -sensing network in which the composite NAD + /NADH ratio simultaneously tunes multiple isoforms to produce a unified immune output. Third, systematic substrate profiling under identical immune-stimulatory conditions (necessary to distinguish genuine isoform-specific functions from overlapping activities) has yet to be performed for the full family. A cross-cutting observation is that the same sirtuin can exert cell-type-dependent functional divergence that, in some contexts, produces opposing outcomes. SIRT6 provides a clear illustration. In adipose tissue macrophages, SIRT6 promotes IL-4-dependent M2 polarization that maintains systemic insulin sensitivity [ 146 ]; similarly, in mast cells, Sirt6 protects against fibroinflammation by suppressing galectin-3-mediated M1 macrophage polarization [ 146 ]. A more direct example of opposing outcomes emerges in the kidney, where proximal tubular Sirt6 restrains fibrogenesis through β-catenin deacetylation [ 145 ], yet in renal macrophages Sirt6-driven M1-to-M2 polarization can paradoxically exacerbate fibrosis when M2 macrophages persist during the AKI-to-CKD transition [ 147 ]. This cell-type-dependent duality is consistent with the checkpoint concept outlined above: the same sirtuin isoform gates inflammatory outcomes through distinct substrates in different cellular compartments. From a translational perspective, systemic SIRT6 activation could simultaneously produce beneficial and adverse effects in different tissues, which emphasizes the need for cell-type-restricted therapeutic strategies.\nSirtuins contribute to immune tolerance at several checkpoints that map onto the same compartmentalized architecture described for innate immunity. SIRT1 maintains tissue-specific immune privilege by repressing stress ligands such as MICA/B and ULBP3. Loss of SIRT1 derepresses these ligands, which in turn activate NF-κB/STAT3 signalling, induce CXCL9/10/11 production, and promote cytotoxic CD8 +  T-cell infiltration. Restoration of SIRT1 activity suppresses these stress signals and re-establishes local immune tolerance, identifying SIRT1 as a regulator of tissue-specific immune restraint [ 148 ]. In contrast to the tolerance-supporting role of SIRT1, SIRT2 favours pathogenic Th17 polarization in systemic autoimmunity. In systemic lupus erythematosus, SIRT2 deacetylates p70S6K and c-Jun, sustaining mTORC1–HIF-1α–RORγt signalling, enhancing IL-17A production, and suppressing IL-2. Genetic or pharmacological inhibition of SIRT2 rebalances the Th17–IL-2 axis and alleviates disease manifestations in preclinical models, which highlights isoform-specific control of effector T-cell fate decisions [ 149 ].\nSIRT3 links metabolic fitness to immune tolerance in both adaptive and innate compartments. In activated CD4 + T cells, SIRT3 deacetylates and stabilises PFKFB3, sustaining glycolytic ATP supply and reducing apoptosis; accordingly, SIRT3 loss worsens antigen-induced arthritis and can be functionally rescued by restoring PFKFB3 [ 150 ]. In macrophages exposed to metabolic stress, SIRT3 limits succinate dehydrogenase hyperacetylation, thereby restraining succinate accumulation and preventing succinate-driven epigenetic silencing of the tolerogenic factor KLF4 [ 151 ]. Together, these findings place SIRT3 as a metabolic stress-responsive regulator that preserves effector viability while protecting innate tolerance programs. At mucosal surfaces, SIRT5 can shift inflammation in the opposite direction. By supporting glutaminolysis-dependent M2 polarization, SIRT5 promotes type 2-skewed inflammation in eosinophilic chronic rhinosinusitis with nasal polyps; genetic deletion reduces disease severity, which shows that individual sirtuins can either reinforce or erode tolerance depending on tissue and metabolic state [ 152 ].\nSIRT7 contributes to immune tolerance by stabilizing lineage-defining transcription programs. In early B-cell development, SIRT7 deacetylates Pax5 at Lys198, maintaining its stability and supporting orderly B-lymphopoiesis; disruption of this axis impairs precursor differentiation and increases susceptibility to malignant transformation [ 153 ]. Beyond lineage commitment, SIRT7 also modulates peripheral T-cell balance. In immune thrombocytopenia, it desuccinylates STAT3 at Lys573, restoring the Th17–Treg equilibrium and limiting pathogenic inflammation [ 154 ]. Central tolerance is likewise shaped by sirtuin activity in non-haematopoietic compartments. In medullary thymic epithelial cells, SIRT6 supports proliferation, tissue-restricted antigen expression and regulatory T-cell generation. Conditional Sirt6 ablation disrupts these processes and results in spontaneous multi-organ autoimmunity, consistent with defective central tolerance [ 155 ]. These observations indicate that sirtuin-mediated tolerance operates at three interconnected levels: lineage stabilization (SIRT7–Pax5), peripheral effector balance (SIRT7–STAT3, SIRT2-Th17, SIRT3-KLF4) and thymic antigen-driven selection (SIRT6). This three-tier architecture is consistent with the compartmentalised NAD + -sensing model proposed in above: nuclear sirtuins (SIRT6, SIRT7) govern transcriptional programs of tolerance and lineage commitment, whereas mitochondrial sirtuins (SIRT3) couple metabolic fitness to tolerogenic signalling. However, several limitations temper a fully integrated picture. First, the existing data are heavily weighted toward adaptive tolerance; the role of sirtuins in peripheral tolerance mechanisms, particularly regulatory T-cell (Treg) stability and function, is largely unexplored despite the well-established dependence of Tregs on mitochondrial fatty acid oxidation, a process directly linked to SIRT3 activity. Second, for SIRT1 and SIRT7, evidence of tolerance function derives predominantly from overexpression or disease models; whether endogenous sirtuin activity at basal NAD +  levels contributes to tonic tolerance maintenance has not been tested. Third, a striking and as-yet-unexploited parallel with above is the bidirectional nature of sirtuin tolerance control: SIRT1, SIRT6 and SIRT7 enforce tolerance, whereas SIRT2 and, in certain contexts, SIRT5 erode it. This raises the question of whether the same NAD +  threshold that triggers innate inflammatory checkpoints simultaneously shifts the tolerance–immunity balance. The direction and magnitude of these effects depend on cell type, metabolic context and tissue environment (Fig.  6 ). Fig. 6 Organelle-stratified sirtuin regulation of immunometabolic disease progression.  a  Metabolic and inflammatory cues: nutrient availability, the NAD + /NADH ratio, and cellular and inflammatory stress provide the inputs that set sirtuin activity.  b  Compartmentalized sirtuin checkpoints: nuclear SIRT1, SIRT6 and SIRT7, cytosolic SIRT2, mitochondrial SIRT3 and SIRT5, and nucleolar SIRT7 impose compartment-specific control over immune signalling, chromatin and metabolism.  c  Cell-type-specific functions: sirtuins exert distinct roles in barrier epithelia and stroma, innate immune cells and adaptive lymphocytes, with context-dependent divergent outcomes.  d  Integrated immunometabolic outcomes: these activities converge on barrier integrity, controlled inflammation and resolution, tissue repair, immune tolerance and metabolic homeostasis.  e  Net outcome depends on cell type, metabolic state and tissue environment, uniting previously disparate autoimmune pathways and nominating sirtuin modulators as precision therapeutics for restoring tolerance without broad immunosuppression\nOrganelle-stratified sirtuin regulation of immunometabolic disease progression.  a  Metabolic and inflammatory cues: nutrient availability, the NAD + /NADH ratio, and cellular and inflammatory stress provide the inputs that set sirtuin activity.  b  Compartmentalized sirtuin checkpoints: nuclear SIRT1, SIRT6 and SIRT7, cytosolic SIRT2, mitochondrial SIRT3 and SIRT5, and nucleolar SIRT7 impose compartment-specific control over immune signalling, chromatin and metabolism.  c  Cell-type-specific functions: sirtuins exert distinct roles in barrier epithelia and stroma, innate immune cells and adaptive lymphocytes, with context-dependent divergent outcomes.  d  Integrated immunometabolic outcomes: these activities converge on barrier integrity, controlled inflammation and resolution, tissue repair, immune tolerance and metabolic homeostasis.  e  Net outcome depends on cell type, metabolic state and tissue environment, uniting previously disparate autoimmune pathways and nominating sirtuin modulators as precision therapeutics for restoring tolerance without broad immunosuppression\nThe tumour immune microenvironment (TIME) is characterized by hypoxia, nutrient limitation and oxidative stress, conditions that reshape immune cell metabolism and function. Within this metabolically constrained setting, sirtuins act as NAD + -dependent sensors that couple energy status to chromatin regulation, signalling thresholds and cell fate decisions in tumour-infiltrating immune cells. This section examines how individual sirtuins modulate macrophage polarization, effector T-cell fitness, regulatory cell expansion and stromal–immune crosstalk in a compartment-specific manner.\nSIRT1 displays compartment-specific effects in the TIME. In tumour-associated macrophages, oxidative stress disrupts extracellular SIRT1, leading to intracellular SIRT1 deficiency in M2-like TAMs. This state is associated with PD-L1 upregulation, CD8 +  T-cell exhaustion and enhanced immune evasion [ 156 ]. In contrast, tumour-cell-intrinsic SIRT1 can reinforce immunosuppression through paracrine signalling. In colorectal cancer, SIRT1 increases CX3CL1 expression, promoting expansion of suppressive TNFRSF9 +  Treg cells via CX3CR1–SATB1/BTG2 axis; CX3CR1 blockade enhances the efficacy of PD-1 inhibition in preclinical models [ 157 ]. Thus, these observations indicate that SIRT1 influences immune evasion through distinct mechanisms in macrophage and tumour compartments.\nSIRT2 primarily regulates the metabolic fitness of cytotoxic lymphocytes in nutrient-limited tumours. In tumour-reactive CD8 +  T cells, SIRT2 deacetylates glycolytic and mitochondrial enzymes, suppressing both glycolysis and oxidative phosphorylation. Genetic deletion or pharmacological inhibition restores proliferation and effector function, and elevated SIRT2 expression in tumour-infiltrating lymphocytes correlates with poor immunotherapy response [ 158 ]. SIRT2 also modulates tumour immune surveillance through additional mechanisms: SIRT2 inhibition induces MLH1 deficiency and enhances CD8 +  T-cell infiltration and antitumour immunity in colorectal cancer models [ 159 ,  160 ], while its effects on NK cells are model-dependent, enhancing tumouricidal function in hepatocellular carcinoma but inhibiting NK-cell activity in melanoma [ 161 ,  162 ]. Together, these data position SIRT2 as a metabolic checkpoint that restrains adaptive and innate antitumour immunity.\nMyeloid lipid metabolism is a major sirtuin-controlled route to immunosuppression in the TME. In tumour-associated macrophages, the SENP1–SIRT3 axis increases mitochondrial acetyl-CoA and supports cholesterol biosynthesis, which reinforces M2 polarization and is accompanied by reduced CD8 +  T-cell abundance and function [ 163 ]. In bladder cancer, loss of SIRT4 is associated with impaired CD8 +  T-cell chemotaxis and cytotoxicity together with enhanced tumour-cell proliferation, and higher SIRT4 expression correlates with a more favourable immunotherapy response [ 164 ]. In hepatocellular carcinoma, SIRT5 downregulation drives peroxisomal hypersuccinylation and excess bile acid production, promoting M2-like macrophage polarization through nuclear receptor signalling; bile acid sequestration with cholestyramine reverses this circuit and limits tumour growth in SIRT5-low models [ 165 ].\nBeyond immune cell-intrinsic programmes, stromal sirtuins shape immunosuppressive niches. In breast cancer, TSPAN8-MAPK11-dependent proteasomal degradation reduces SIRT6 in a subset of myofibroblastic CAFs, activating SASP programmes with elevated IL-6 and IL-8 and increased therapy resistance; combined SIRT6 activation (MDL-800) and TSPAN8 blockade attenuates these protumour stromal features [ 123 ,  124 ,  166 ]. By contrast, in anaplastic thyroid carcinoma, tumour-cell-intrinsic SIRT6 promotes ferroptosis, which demonstrates that the impact of SIRT6 depends on the cellular compartment [ 119 ].\nIn summary, sirtuins shape the tumour immune microenvironment through compartment- and isoform-specific mechanisms. SIRT1 and SIRT5 influence macrophage and regulatory T-cell programmes via metabolic and extracellular signalling pathways, whereas SIRT2 and SIRT3 regulate the metabolic threshold of cytotoxic T and natural killer cells. SIRT4 links mitochondrial homeostasis to immune surveillance, and SIRT6 modulates stromal senescence and ferroptosis-associated immunogenicity. This compartment- and isoform-specific division of labour mirrors the architecture described for innate immunity and tolerance, reinforcing the concept that the sirtuin family operates as a distributed NAD + -sensing network whose immunological output is determined by which isoforms are engaged in which cellular compartments.\nRather than acting uniformly as pro- or anti-tumour factors, individual sirtuins adjust immune activation according to cellular context and metabolic constraints. A recurrent theme across innate immunity, tolerance and tumour immunology (as exemplified by the opposing effects of SIRT2 on NK cells in different tumour models [ 161 ,  162 ] and of SIRT6 in stromal versus tumour-cell compartments [ 120 ,  123 ]) is that the same sirtuin isoform can produce antithetical immunological outcomes depending on cell type, tissue environment and NAD +  availability. These mechanistic insights provide a rationale for selective therapeutic modulation, which will be discussed in the following section. The outcome of sirtuin targeting in the TIME is likely to depend on tumour type, immune composition and metabolic state, a consideration that applies equally to innate and adaptive immune contexts and underscores the need for context-guided rather than pan-sirtuin therapeutic strategies.\n\nWhen the distributed sirtuin regulatory network becomes dysregulated (whether through NAD +  depletion, isoform-specific loss, or aberrant acylation), its failure manifests across diverse disease states in a tissue- and context-specific manner. Metabolic syndrome and related disorders, including obesity, type 2 diabetes, metabolic dysfunction-associated steatotic liver disease and gestational diabetes, are characterized by chronic energy surplus, reduced metabolic flexibility and progressive NAD +  decline. These changes promote mitochondrial dysfunction, lipid accumulation and persistent low-grade inflammation, driving disease progression. As NAD + -dependent deacylases, sirtuins directly link cellular metabolic state to transcriptional, enzymatic and redox regulation [ 167 ,  168 ]. Within this context, different sirtuin isoforms contribute to metabolic control at distinct cellular levels. Nuclear SIRT1 and SIRT6 regulate gene expression programs that influence lipid and glucose metabolism; SIRT2 modulates cytosolic stress responses and metabolic enzyme activity; mitochondrial SIRT3–5 maintain oxidative metabolism and redox balance; and SIRT7 remains less well defined in systemic metabolic disease.\nSIRT1 and SIRT6 play important roles in systemic metabolic protection. SIRT6 suppresses excessive glycolysis and lipid synthesis, while SIRT1 helps maintain insulin sensitivity by deacetylating key metabolic and stress-related proteins. Together, these nuclear sirtuins link reduced NAD +  levels to changes in gene expression and metabolic control [ 169 ,  170 ]. Although these findings establish SIRT1 and SIRT6 as central metabolic regulators, the direction of causality in human disease (whether sirtuin decline drives metabolic decompensation or reflects compensatory adaptation to chronic overnutrition) remains unresolved in most cases.\nMitochondrial sirtuins (SIRT3-5) provide additional metabolic support. SIRT3, the major mitochondrial deacetylase in metabolic tissues, is essential for maintaining mitochondrial function under metabolic stress. Reduced SIRT3 expression is associated with impaired oxidative metabolism and weakened antioxidant defense in several disease models, including diabetic osteoporosis [ 171 ,  172 ], periodontitis [ 173 ], hypoglycemia-induced astrocyte neurotoxicity [ 174 ] and renal fibrosis [ 171 – 176 ]. SIRT4 and SIRT5 also contribute to mitochondrial regulation, particularly in stress responses related to cell survival and metabolic stability.\nSIRT2 has context-dependent and stage-dependent effects. Under hyperglycemia, it limits excessive β-cell expansion by deacetylating oxidative phosphorylation (OXPHOS) enzymes [ 177 ]. It also promotes mitochondrial biogenesis in trophoblasts through PGC-1α and suppresses TGFβ–SMAD2/3–mediated fibrosis in kidney disease. In these settings, SIRT2 appears to protect tissues by preventing excessive metabolic or stress responses.\nTogether, nuclear SIRT1/SIRT6 and mitochondrial SIRT3 represent key regulators of systemic metabolic stability, while SIRT2 and SIRT4/5 provide additional support in specific contexts. This helps explain why global NAD +  precursors such as NMN and NR show broad metabolic benefits in experimental models, whereas isoform-selective modulators may be needed for more targeted therapy.\nCardiovascular and renal diseases are closely linked to chronic metabolic stress. Long-term nutrient excess, oxidative stress and reduced NAD +  levels contribute to endothelial dysfunction, inflammation, cardiac remodelling and fibrosis. In these conditions, sirtuins regulate gene expression, protein modification and mitochondrial function. Nuclear SIRT1, SIRT6 and SIRT7 influence transcriptional control; cytosolic SIRT2 regulates protein acetylation; and mitochondrial SIRT3 maintains mitochondrial integrity [ 168 ].\nSIRT1 and SIRT6 play important protective roles in renal and vascular disease. In diabetic nephropathy, gut-derived indole-3-propionic acid stabilizes SIRT1 by preventing phosphorylation-dependent ubiquitination; this restores PGC-1α-mediated mitochondrial biogenesis and reduces ROS-mediated glomerular injury [ 178 ]. SIRT1 also protects against toxin-induced acute kidney injury by maintaining PINK1/Parkin mitophagy [ 179 ]. In lupus-prone mice, it reduces renal inflammation by limiting NLRP3 inflammasome activation through NF-κB and ROS/TRPM2-Ca 2+  signalling pathways [ 180 ]. SIRT6 also exerts complementary protective effects. It helps maintain mitochondrial dynamics, suppresses lipid accumulation and oxidative stress programs, and reduces vascular calcification and foam cell formation. These protective functions are often reduced in disease conditions.\nSIRT2 and SIRT3 provide key post-translational and mitochondrial support. In angiotensin II-driven hypertension, SIRT2 deacetylates Septin4-K174 in podocytes, reducing PARP1/caspase-3 mediated apoptosis [ 181 ]. It also inhibits TGF-β-induced SMAD2/3 activation and limits tubular fibrosis in chronic kidney disease [ 182 ]. SIRT3 improves mitochondrial function in macrophages, decreases lactate-driven NLRP3 activation, and reduces perivascular adipose fibrosis [ 183 ]. SIRT7 protects against hyperglycemic memory in diabetic nephropathy. High glucose lowers SIRT7 expression, leading to persistent ELK1-dependent DAPK3 transcription and ongoing endothelial inflammation even after glucose returns to normal. Overexpression of SIRT7 improves renal function and reduces inflammation in preclinical models [ 184 ]. Overall, SIRT1 and SIRT6 act as main transcriptional and epigenetic regulators against cardio-renal metabolic injury. Their effects are supported by SIRT2 and SIRT3, which maintain protein stability and mitochondrial function, and by SIRT7, which helps reverse metabolic memory. This layered mechanism suggests that NAD +  precursors and isoform-selective modulators may help slow the progression from metabolic disorder to organ damage. The limitations of current evidence should be acknowledged: most mechanistic conclusions derive from gain-of-function or acute knockout models, and whether basal sirtuin activity at physiological NAD +  levels exerts tonic metabolic control in humans remains untested. Neurodegenerative disorders, in which metabolic failure and neuroinflammation converge, are discussed later [ 185 ].\nSIRT1 directly regulates pancreatic β-cell function and insulin secretion. In pancreatic β cells, Sirt1 represses the mitochondrial uncoupling protein UCP2, an action that raises ATP levels and thereby couples glucose metabolism to insulin secretion; Sirt1 deficiency or inhibition impairs glucose-stimulated insulin secretion [ 186 ]. This establishes SIRT1 as a positive regulator of β-cell secretory capacity, positioning its loss as a potential contributor to β-cell dysfunction in type 2 diabetes. β-cell senescence provides a second, disease-relevant layer of sirtuin regulation. The cholesterol oxidation product 7-ketocholesterol (7-KC) accelerates pancreatic β-cell senescence by inhibiting the SIRT1/CDK4–Rb–E2F1 signalling pathway, thereby reducing insulin secretion and regenerative capacity, a mechanism linking lipid stress to β-cell aging in type 2 diabetes [ 187 ]. These findings reveal opposing inputs on the β-cell SIRT1 axis: physiological SIRT1 activity supports insulin secretion, whereas lipid-induced loss of SIRT1 signalling drives senescence, illustrating how metabolic context determines whether sirtuin activity protects or compromises β-cell function. Sirtuins are key regulators of adipose tissue remodelling in obesity, governing adipogenesis, beige adipocyte conversion and inflammation [ 188 ]. SIRT6 in particular has multiple roles in adipose tissue inflammation, where it modulates the inflammatory program of adipocytes and adipose-resident immune cells, with its loss or downregulation linked to exacerbated adipose inflammation in obesity [ 189 ]. Beyond adipocyte-intrinsic regulation, endothelial sirtuins contribute to whole-body metabolic homeostasis. Endothelial SIRT3 deficiency predisposes brown adipose tissue to whitening in diet-induced obesity: both global and brown-adipose-tissue-endothelium-specific Sirt3 knockout accelerate diet-induced weight gain, accompanied by glucose intolerance, insulin resistance and brown adipose tissue whitening [ 190 ]. This positions endothelial SIRT3 as a guardian of brown adipose tissue thermogenic capacity, linking vascular sirtuin signalling to systemic energy balance. In the liver, mitochondrial sirtuins protect against steatosis. Irisin, a skeletal-muscle-derived myokine, alleviates hepatic steatosis by activating the autophagic SIRT3 pathway, thereby improving hepatic lipid homeostasis in high-fat-diet-induced metabolic dysfunction-associated steatotic liver disease (MASLD) [ 191 ]. Together with the established roles of SIRT1/SIRT6 in hepatic lipid metabolism, this reinforces a model in which mitochondrial sirtuins (SIRT3 in particular) mediate autophagy-dependent lipid clearance that opposes steatosis progression. Across β cells, adipose tissue and liver, sirtuins exert tissue-specific and context-dependent effects: SIRT1 supports β-cell insulin secretion yet its loss also permits senescence, while SIRT3 protects both brown adipose tissue thermogenesis and hepatic lipid clearance. A notable gap is that the relative contribution of individual sirtuins to human metabolic disease (versus their well-documented roles in rodent models) is poorly quantified, and whether sirtuin-directed strategies can achieve tissue-selective metabolic benefit without systemic NAD +  modulation remains an open translational question.\nCancer cells use sirtuin-dependent metabolic and epigenetic regulation to survive stress and therapy. The roles of the seven mammalian sirtuins are highly tumour-type- and treatment-dependent. SIRT6, SIRT3, and SIRT4 often act as tumour suppressors, whereas SIRT7 and frequently SIRT5 promote tumour progression. SIRT1 and SIRT2 display dual roles depending on the setting.\nSIRT6 is one of the most consistent tumour-suppressive sirtuins in many cancer types. It inhibits glycolysis and acetyl-CoA-driven anabolic pathway, maintains genome stability, and limits stromal senescence. Loss of SIRT6 weakens these protective effects, while SIRT6 activators (MDL-800/811) can restore them in several models. SIRT1 shows strong tissue-type dependence. It promotes ferroptosis in glioma [ 192 ] and reduces pro-glycolytic H3K18 lactylation in gastric cancer [ 193 ], however, it can also contribute to therapy resistance, including taxol resistance through polyploid giant cancer cells [ 194 ], 5-FU resistance via p53/miR-101 [ 195 ] and escape from doxorubicin-induced senescence by restoring autophagy [ 196 ].\nSIRT2 similarly displays stage- and cell-type-dependent polarity. It can be oncogenic in early neoplasia through effects on mitotic fidelity and EMT programmes (SMC1A mitotic fidelity [ 197 ]; Snail/EMT [ 198 ]), yet it can acquire tumour-suppressive roles in other settings through DNA repair or pathway decoupling (BER via OGG1 [ 199 ]; AKR1C1-STAT3 decoupling [ 200 ]). This bidirectionality can be interpreted as reprogramming of sirtuin dependencies across tumour evolution rather than as contradiction.\nMitochondrial SIRT3 and SIRT4 generally limit tumorigenesis by maintaining oxidative metabolism and imposing metabolic checkpoints. By contrast, SIRT5 often acts as an oncogenic vulnerability in solid tumours and leukaemia, where it sustains respiratory capacity and redox-supporting pathways, promotes resistance to ferroptosis, and facilitates immune evasion. SIRT7 promotes PD-L1 expression and is linked to resistance to sorafenib, gemcitabine and radiotherapy. It also accelerates PTEN degradation, supports Fascin-mediated metastasis, and dampens endoplasmic reticulum stress across multiple tumor types, including HCC, PDAC, endometrial cancer, ESCC, and melanoma [ 16 ].\nThe sirtuin family contributes a compartmentalized regulatory network in cancer, with distinct isoforms playing opposing roles. Nuclear SIRT6 and SIRT1 function as a metabolic-epigenetic brake, mitochondrial SIRT3 and SIRT4 act as tumour-suppressive checkpoints, SIRT5 and SIRT7 are frequently co-opted as oncogenic drivers, and SIRT2 displays bifunctional stage-dependent switching. This isoform specificity clarifies why non-selective NAD +  boosters yield inconsistent outcomes, whereas isoform-selective SIRT2/5/7 inhibitors, SIRT6 activators and SIRT3/4-preserving approaches are advancing as precision combination therapies. During cancer metabolic rewiring, sirtuins enforce a functional stratification: SIRT1 and SIRT6 restrain glucose-driven anabolic metabolism, SIRT3 and SIRT4 promote oxidative restraint, SIRT5 sustains respiration-linked redox networks, and SIRT2 can amplify acetyl-CoA- and proline-dependent biosynthesis. SIRT6 stands out as a consistent suppressor of tumour metabolic reprogramming, broadly curbing glycolytic and acetyl-CoA-linked anabolic drift. In contrast, SIRT2 can serve as an oncogenic accelerator in specific settings. In oesophageal squamous cell carcinoma, SIRT2 activates ATP-citrate lyase to expand cytosolic acetyl-CoA pools and promote lipogenesis; SIRT2 inhibition suppresses tumour growth [ 201 ]. In colorectal cancer, SIRT2 deacetylates IDH1 at K224 to enhance α-ketoglutarate-dependent redox homeostasis, with IDH1 hyperacetylation correlating with metastasis and poor survival [ 202 ]. In hepatocellular carcinoma, SIRT2 deacetylates mitochondrial P5CS to fuel proline-linked respiration and tumor expansion [ 203 ]. SIRT5 sustains multiple oncogenic circuits by desuccinylating and stabilising MTHFD2 in breast cancer, boosting one-carbon metabolism-derived NADPH and facilitating chemoresistance [ 204 ].\nSirtuins establish a functionally stratified network in cancer metabolism. While SIRT6, and contextually SIRT1, SIRT3, and SIRT4, typically restrain anabolic flux and enforce tumor-suppressive checkpoints, SIRT2 and SIRT5 are often co-opted to drive oncogenic pathways that enhance acetyl-CoA availability, proline metabolism, and redox support. This functional divergence clarifies the inconsistent outcomes of non-selective NAD +  augmentation and rationalizes the development of isoform-selective strategies, such as SIRT2/5 inhibition or SIRT6 activation, to target specific metabolic vulnerabilities. The role of any individual sirtuin is ultimately defined by tumor type, stage, and metabolic dependencies, which reflects the necessity for biomarker-guided therapeutic application. However, for most of these tumour-type-specific observations, whether sirtuin dysregulation is a driver of malignant progression or a consequence of tumour-driven metabolic reprogramming remains unresolved.\nNeurodegenerative disorders and CNS are characterized by the convergence of metabolic dysfunction, loss of protein homeostasis, and chronic inflammation, which collectively undermine neuronal and glial resilience. Within this pathological landscape, sirtuins function as NAD + -dependent regulators that integrate genomic, metabolic, and inflammatory responses. A compartmentalized division of labour emerges: nuclear sirtuins (SIRT1, SIRT6) govern transcriptional programs critical for long-term cell fate, while cytosolic SIRT2 and mitochondrial SIRT3 regulate axonal integrity, bioenergetics, and redox balance. Disease manifestations often reflect which cellular compartment is compromised first: nuclear instability perpetuates neuroinflammation, whereas mitochondrial failure limits the energy required for synaptic repair.\nWithin this framework, SIRT6 serves as a central protective factor in CNS homeostasis, with SIRT1 and SIRT3 providing contextual support via neuroimmune and mitochondrial mechanisms, respectively. For instance, in models of optic nerve injury, microglial SIRT1 deacetylates TSC2 to restrain mTORC1 activity, limiting excessive synaptic pruning and preserving retinal ganglion cell function [ 185 ]. In neuropathic pain, the ketone body β-hydroxybutyrate exerts anti-inflammatory effects by engaging SIRT3 to restore mitochondrial membrane potential and suppress microglial ROS in the spinal cord [ 205 ].\nBuilding on this compartmentalized division of labour, the specific mechanisms through which sirtuins influence neurodegeneration can be organized along three routes: tau pathology, mitochondrial quality control and neuronal DNA repair, with NAD +  availability as the common thread. SIRT1 is the most consistently implicated sirtuin in tau biology, and it regulates tau through multiple distinct mechanisms. A mitophagy-based pathway links the autophagy-initiating kinase ULK1 to inhibition of pathogenic tau acetylation through the ULK1–NAD + /SIRT1 axis, suggesting that mitophagy restricts early tau pathology via SIRT1-dependent deacetylation [ 206 ]. SIRT1 also controls tau O-GlcNAcylation indirectly: it deacetylates CREB, inhibiting CREB phosphorylation and thereby regulating OGT-dependent tau modification, a pathway relevant to the reduced tau O-GlcNAcylation observed in Alzheimer's disease brain [ 207 ]. Through a third route, Sirt1 enhances tau exon 10 inclusion and improves spatial memory in Htau mice, linking SIRT1 activity to tau isoform balance and cognitive function [ 208 ]. In the mitochondrial compartment, SIRT3 and SIRT6 preserve neuronal bioenergetics. SIRT3-mediated deacetylation of succinate dehydrogenase subunit A (SDHA) rescues mitochondrial bioenergetics and contributes to neuroprotection in rotenone-induced Parkinson's disease models [ 209 ]. SIRT6, which has been implicated in DNA repair and has critical roles in brain development and neurodegeneration, is a key regulator of mitochondrial function in the brain, as shown by integrated transcriptomic and metabolomic analyses [ 210 ]. The NAD + –sirtuin axis also has direct therapeutic relevance in neurodegeneration. NAD +  supplementation normalizes key Alzheimer's features and DNA damage responses in an AD mouse model carrying a DNA-repair deficiency, providing in vivo evidence that restoring NAD +  availability counteracts neuronal DNA damage [ 211 ]. In Parkinson's disease, sirtuins are increasingly recognized as regulators of mitochondrial dysfunction, oxidative stress and α-synuclein pathology that drive dopaminergic neurodegeneration [ 212 ]. Sirtuins thus form a convergent node linking NAD +  metabolism, tau post-translational modification and mitochondrial quality control in neurodegeneration. The field remains limited by the dominance of mouse models over human data, and whether the protective effects observed for SIRT1/SIRT3/SIRT6 can be achieved with isoform-selective or NAD + -based interventions without the mixed outcomes seen with global NAD +  elevation remains unresolved.\nSirtuins protect the cardiovascular system through compartment-specific actions in the endothelium, the myocardium and the cardiac mitochondrial network. In the endothelium, SIRT1 maintains vascular function against metabolic stress. Chronic SIRT1 supplementation in diabetic mice improves endothelial function by suppressing oxidative stress, positioning SIRT1 as a protective factor against diabetic vascular injury [ 213 ]. This endothelial role aligns with the broader function of SIRT1 as a metabolic-stress sensor that couples NAD +  availability to vascular homeostasis [ 214 ].\nIn the myocardium, SIRT6 opposes aging-related cardiac decline. Aging is associated with reduced Sirt6 activity together with cardiac hypertrophy and fibrosis; aged hearts show increased mitochondrial DNA damage, shortened telomeres and accumulation of 8-oxo-dG adducts, and Sirt6 mitigates aging-induced cardiomyocyte senescence and cardiac hypertrophy [ 215 ]. SIRT7, the nucleolar sirtuin that regulates RNA polymerase I transcription, contributes to myocardial tissue repair by maintaining transforming growth factor-β signalling [ 216 ]. In hypertensive hearts, SIRT7 expression falls while cuproptosis rises (decreased FDX1, increased HSP70), and restoring SIRT7 ameliorates myocardial remodelling and cardiac dysfunction through modulation of YAP–ATP7A signalling [ 217 ,  218 ].\nIn the cardiac mitochondria, SIRT3 is the central regulator. The SIRT3 activator 2-APQC alleviates myocardial hypertrophy and fibrosis by regulating mitochondrial homeostasis, supporting SIRT3 as a therapeutic target in cardiac remodelling [ 219 ]. This builds on a long-standing literature establishing mitochondrial SIRT3 as a determinant of cardiac energy metabolism and heart disease [ 220 ,  221 ]. Sirtuins span the full cardiovascular axis: SIRT1 at the endothelium, SIRT6 and SIRT7 in the myocardium, and SIRT3 in the mitochondrial compartment. Sirtuins also sit at the centre of cardiac ageing and fibrosis, where their protective roles make them candidate targets for age-related cardiac remodelling [ 222 ,  223 ]. Whether isoform-specific modulation can achieve cardiovascular benefit without the systemic effects of global NAD +  manipulation remains an open translational question.\nIn myocardial ischemia–reperfusion injury, a complication of reperfusion therapy for acute myocardial infarction, the sirtuin family provides protection against the reactive-oxygen burst, mitochondrial dysfunction and programmed cell death that drive cardiomyocyte loss [ 224 ,  225 ].\nSirtuins occupy a central position at the interface between metabolic state, stem cell fate and cellular senescence. Because stem cell exhaustion and cellular senescence are two hallmarks of aging [ 226 ,  227 ], the sirtuin family provides a direct mechanistic link between nutrient sensing and the maintenance of regenerative capacity across tissues. In embryonic stem cells (ESCs), SIRT1 preserves pluripotency by restraining p53 activity. Silencing of the pluripotency factor Oct4 in human ESCs leads to activation of the tumour suppressor p53 and subsequent differentiation, an effect that depends on SIRT1: Oct4 depletion reduces SIRT1 expression, which in turn increases acetylation of p53 at lysine 120 and 164 and derepresses p53-dependent differentiation programs. Ectopic SIRT1 expression in Oct4-silenced cells prevents p53 activation and maintains the undifferentiated state [ 228 ]. SIRT1 thus functions as a downstream effector of Oct4, coupling the pluripotency network to the p53 stress axis.\nIn adult somatic stem cells, SIRT6 maintains stem cell homeostasis through epigenetic restraint of Wnt signalling. In haematopoietic stem cells (HSCs), SIRT6 interacts with LEF1 and deacetylates histone H3 at lysine 56 to suppress transcription of Wnt target genes, a mechanism that preserves HSC quiescence and self-renewal capacity; deletion of Sirt6 impairs HSC quiescence and drives premature exhaustion [ 229 ]. A parallel protective role has been documented in human bone marrow mesenchymal stem cells (hBM-MSCs), where knockdown of SIRT6 enables senescence, as evidenced by increased senescence-associated β-galactosidase activity, elevated p16 expression, and reduced proliferative and migratory capacity [ 230 ]. These findings position SIRT6 as a general anti-aging and stemness-preserving factor across multiple stem cell compartments.\nThe nucleolar sirtuin SIRT7 guards genome integrity and opposes cellular senescence by maintaining ribosomal DNA (rDNA) stability. SIRT7 deficiency compromises rDNA integrity and permits senescence induced by rDNA instability [ 231 ]. Consistent with this, SIRT7 has been implicated in promoting ribosomal RNA expression, aiding DNA damage repair and regulating chromatin compaction, functions that collectively link nucleolar homeostasis to the aging process [ 232 ]. At the family level, sirtuins integrate metabolic cues with the stemness–differentiation balance, with distinct isoforms governing pluripotency, lineage specification and adult stem cell maintenance in a compartment-specific manner [ 233 ]. Together, these observations support a model in which nuclear sirtuins (SIRT1, SIRT6, SIRT7) function as guardians of stem cell function and barriers against senescence: SIRT1 through the p53 axis, SIRT6 through epigenetic restraint of Wnt signalling, and SIRT7 through rDNA stability. However, several questions remain unresolved: whether the opposing effects of individual sirtuins on self-renewal versus differentiation reflect distinct substrate specificity or divergent metabolic contexts has not been systematically addressed, and the role of mitochondrial sirtuins (SIRT3–SIRT5) in stem cell metabolism and aging is comparatively understudied.\nThe sirtuin family was discovered through genetic studies linking Sir2 to replicative lifespan extension in yeast under caloric restriction, and aging has remained central to sirtuin biology ever since. In mammals, sirtuins regulate lifespan-associated processes through their shared requirement for NAD + , a coenzyme that declines with age and thereby couples metabolic state to the activity of the entire sirtuin network. NAD +  is an essential cofactor for sirtuins and poly(ADP-ribose) polymerases (PARPs), functioning both as a redox coenzyme and as a signalling molecule that controls NAD + -sensing enzymes [ 234 ]. NAD +  levels decline in multiple tissues during aging, and this decline is now understood to be driven in part by the NADase CD38, whose expression increases with age. CD38 metabolizes the NAD +  precursor NMN, and its age-related upregulation dictates NAD +  decline and mitochondrial dysfunction through a SIRT3-dependent mechanism; NAD +  levels and mitochondrial function are preserved in old CD38-knockout mice [ 235 ].\nThe NAD + –sirtuin axis is a central mediator of caloric restriction, the most robust longevity intervention in model organisms. Because sirtuins are NAD + -dependent, caloric restriction is proposed to activate sirtuins through elevated NAD +  availability, and pharmacological strategies to raise NAD + , including the precursors nicotinamide mononucleotide (NMN) and nicotinamide riboside (NR), have shown therapeutic potential in vivo across metabolic and age-related contexts [ 236 ]. The NAD + –sirtuin connection therefore links nutrient availability to organismal aging through a conserved, druggable pathway [ 237 ]. Sirtuins also mediate the pancreatic effects of caloric restriction, regulating β-cell function, metabolism and longevity [ 238 ].\nAmong the seven mammalian sirtuins, SIRT1 and SIRT6 are the most consistently implicated in aging-related diseases, functioning in metabolic adaptation, genome maintenance and inflammation, and their modulation has been proposed as a therapeutic strategy across cardiovascular, neurodegenerative and metabolic disorders of aging [ 239 ]. Sex-specific effects are emerging: hepatocyte-specific Sirt6 knockout accelerates female-specific aging through SULT1E1-mediated estrogen depletion [ 240 ]. A recent advance has identified an endogenous mechanism of sirtuin activation. Lithocholic acid (LCA), a bile acid that accumulates in mammals during caloric restriction, binds the protein TULP3 to activate sirtuins and AMPK, thereby slowing aging in animal models [ 241 ]. This finding extends the NAD + –sirtuin–longevity framework by identifying a caloric-restriction-derived metabolite that directly engages the sirtuin network, suggesting that the anti-aging effects of caloric restriction are mediated by both NAD +  availability and specific endogenous activators.\nThese findings establish the sirtuin family as a conserved nexus linking NAD +  metabolism, caloric restriction and the aging process. However, several questions remain unresolved: whether NAD +  decline is a cause or a consequence of aging has not been definitively established, and the extent to which NAD + -boosting interventions translate from rodent models to human aging remains uncertain given heterogeneous clinical outcomes. The identification of endogenous activators such as LCA raises the further question of whether physiological sirtuin activation, rather than pharmacological NAD +  elevation, represents a more tractable route to modulating human healthspan.\n\nThe distributed, compartmentalized architecture of the sirtuin network presents both opportunities and obstacles for therapeutic intervention: isoform selectivity, compartment-specific drug exposure, and the risk of disrupting network-level coordination must all be navigated. This section evaluates current strategies against this biological complexity.\nDirect pharmacological modulation refers to small molecules that bind sirtuin proteins and alter catalytic activity through allosteric or competitive mechanisms. In contrast to indirect NAD + -based approaches, these strategies aim to achieve isoform selectivity at the level of enzyme engagement. However, selectivity is multidimensional, involving isoform preference, substrate bias, and compartment-specific exposure. SIRT1 has been among the most intensively explored sirtuin isoforms and serves as a useful paradigm for both the promise and limitations of direct sirtuin targeting. Early enthusiasm was driven by the discovery that resveratrol [ 242 ], a naturally occurring polyphenol, functions as an allosteric activator of SIRT1 [ 243 ]. Resveratrol enhances SIRT1-mediated deacetylation and engages the SIRT1-AMPK-PGC-1α axis, improving mitochondrial function, oxidative stress resistance and metabolic homeostasis across numerous preclinical models. However, clinical translation has been challenging. In a Phase I trial of micronized resveratrol (SRT501) in patients with hepatic metastases, 5 g/day produced dose-limiting gastrointestinal intolerance (nausea, vomiting, diarrhea) [ 244 ]; at ≤ 1 g/day, resveratrol is generally well tolerated across multiple disease settings. Although modest improvements in inflammatory and metabolic biomarkers have been reported in some trials, disease-modifying effects have been weak or absent. Concerns about renal toxicity derive from preclinical rodent studies at suprapharmacological doses [ 245 ] and have not been prominent in human data. The Sirtris Pharmaceuticals program, which advanced SRT501 and second-generation SIRT1 activators (SRT1720, SRT2104) into Phase II trials, was ultimately discontinued after SRT2104 failed to meet primary efficacy endpoints in chronic obstructive pulmonary disease and inflammatory disease [ 246 ].\nA conceptual advance has recently emerged with the development of NAD + -independent SIRT1 activators based on an (E)−2’-demethylsulfonic acid scaffold [ 247 ]. Unlike classical STACs, these compounds directly participate as acetyl acceptors in the deacetylation reaction, enabling SIRT1 activation under NAD + -depleted conditions characteristic of ageing and metabolic disease. Although highly selective and efficacious in preclinical models, their pharmacokinetic optimization and translational relevance remain to be established. Two decades of SIRT1-directed drug discovery illustrate an important lesson: chemical tractability alone is insufficient to overcome the biological complexity of sirtuin regulation in vivo.\nBeyond SIRT1, increasing efforts have focused on targeting other sirtuin family members, revealing pronounced isoform-specific opportunities and limitations. SIRT2 inhibitors, including AGK2, AK-7, display biochemical selectivity and central nervous system penetration [ 248 ], enabling extensive preclinical exploration in neurodegenerative disease, cancer and inflammatory models. However, no SIRT2-directed agents have yet progressed to clinical testing, reflecting ongoing uncertainties regarding therapeutic windows and tissue-specific effects. Recent identification of S-ketamine as a stereospecific, NAD + -independent SIRT2 modulator highlights alternative regulatory strategies but also underscores the early stage of translational development [ 249 ].\nSIRT3, the primary mitochondrial deacetylase, is a promising therapeutic node among sirtuins. Natural activators such as honokiol, polydatin [ 250 ] and melatonin [ 251 ] enhance mitochondrial antioxidant capacity and bioenergetic efficiency, while structure-guided efforts have yielded synthetic allosteric activators with improved potency and pharmacokinetics. A notable example is SKLB-11A [ 252 ], which activates SIRT3 through a distinct conformational mechanism and confers robust cardioprotection in preclinical models. Despite these advances, SIRT3-directed strategies remain confined to preclinical stages.\nSIRT5 stands out as the isoform with relatively mature chemical toolbox. Highly selective substrate-competitive inhibitors, including ε-N-thioglutaryl-lysine derivatives and pyrazolone-based compounds, have enabled detailed dissection of SIRT5-dependent metabolic pathways [ 253 ]. However, the lack of clearly defined clinical indications and limited in vivo validation have thus far constrained translational progress. By contrast, SIRT6 is a compelling nuclear target, with allosteric activators such as MDL-800 [ 254 ], MDL-801 [ 254 ] and MDL-811 [ 255 ] showing strong efficacy in models of cardiometabolic, inflammatory and ischemic disease. An orally available SIRT6-modulating compound (IMU-856) has entered early-phase clinical testing, marking a potential step forward for isoform-selective sirtuin therapeutics [ 256 ].\nBy contrast, SIRT7 inhibition is supported by more explicit chemical evidence, although still at an early translational stage. Structure-based modelling and virtual screening have yielded candidate inhibitors with preferential affinity for SIRT7, proposed as starting points for liver cancer therapy. Complementary approaches using cyclic tripeptide inhibitors demonstrate low-micromolar inhibition of human SIRT7, establishing proof-of-principle pharmacophores but with inherent peptide-associated limitations [ 257 ]. In disease models, SIRT7 inhibition has also been achieved using bioactive natural products, such as Oroxylin A, to reprogramme hepatic stellate cell fate in liver fibrosis [ 258 ]. Nonetheless, the field remains constrained by limited isoform selectivity profiling, insufficient target-engagement validation, and a lack of standardized pharmacodynamic biomarkers, which underscores the early translational status of SIRT7-directed therapeutics. Together, these efforts highlight that progress in isoform-selective sirtuin pharmacology is constrained less by chemical feasibility than by the need for validated biomarkers, tissue-specific target engagement and disease-contextualized indication selection. Direct pharmacological and chemical tools targeting individual sirtuins are summarized in Table  1 .\n Table 1 Direct modulators and chemical tools targeting members of the sirtuin family SIRT family member Regulatory factor Modality/mechanism Effect on SIRT Clinical status SIRT1 Resveratrol [ 242 – 245 ] Allosteric activator; links SIRT1 to AMPK–PGC-1α signalling ↑ Phase I–II clinical trials ( NCT01354977 ); heterogeneous results; well tolerated at ≤ 1 g/d, higher doses (~5 g/d) associated with adverse effects including renal injury, contributing to trial discontinuation in some settings SIRT1 SRT1720; SRT2104 [ 246 ] Second-generation SIRT1-activating compounds (STACs; binds SIRT1 N‑terminal regulatory domain ↑ SRT1720: Preclinical (metabolic and neurodegenerative models) SRT2104: Phase II completed (T2DM, psoriasis;  NCT01154101 ); no late‑stage development SIRT1 ((E)−2′-demethylsulfonic acid scaffold) [ 247 ] NAD + ‑independent activator; acts as acetyl acceptor ↑ Preclinical; validated under NAD + ‑depleted conditions SIRT2 AGK2; AK-7 [ 248 ,  259 ,  260 ] Selective SIRT2 inhibitors with central nervous system penetrance ↓ Preclinical (neurodegenerative, cancer, inflammatory models) SIRT2 S-ketamine [ 249 ] Stereospecific, NAD + ‑independent SIRT2 modulator; enhances SIRT2–p65 interaction ↑ Preclinical; identified via repurposing for depression SIRT3 Honokiol; Polydatin; Melatonin [ 250 ,  251 ,  261 ] Natural product SIRT3 activators; enhance mitochondrial antioxidant capacity ↑ Preclinical models SIRT3 SKLB-11A [ 252 ] Synthetic allosteric SIRT3 activator with a distinct conformational mechanism ↑ Preclinical models SIRT5 ε-N-thioglutaryl-lysine derivatives; pyrazolone-based compounds [ 253 ] Substrate-competitive SIRT5 inhibitors ↓ Preclinical; chemical probes for SIRT5 SIRT6 MDL-800; MDL-801; MDL-811 [ 254 ,  255 ] Allosteric SIRT6 activators ↑ Preclinical (cardiometabolic, inflammatory, ischemic models) SIRT6 IMU-856 [ 256 ] Orally available SIRT6-modulating compound ↑ Phase I completed (coeliac disease; ACTRN12620000901909) SIRT7 Virtual-screening-derived candidate inhibitors [ 262 ] Structure-based SIRT7 inhibitors ↓ Preclinical; proposed for liver cancer SIRT7 Cyclic tripeptide inhibitors [ 257 ] Low-micromolar SIRT7-selective inhibitors ↓ Preclinical; proof-of-principle pharmacophores SIRT7 Oroxylin A [ 258 ] Natural SIRT7 inhibitor; induces HSC senescence ↓ Preclinical (anti-fibrotic in liver fibrosis) SIRT2 SirReal-based PROTACs [ 263 – 265 ] Cereblon-recruiting degraders enabling selective SIRT2 protein degradation ↓ (protein degradation) Most mature examples of sirtuin-directed PROTAC strategies SIRT6 SIRT6 degraders (PROTAC technology) [ 266 ] Potent degraders suppress hepatocellular carcinoma growth ↓ (protein degradation) Hepatocellular carcinoma models Symbol footnotes: ↑, increased activity:↓, decreased activity\nDirect modulators and chemical tools targeting members of the sirtuin family\nSRT1720: Preclinical (metabolic and neurodegenerative models)\nSRT2104: Phase II completed (T2DM, psoriasis;  NCT01154101 ); no late‑stage development\nSymbol footnotes: ↑, increased activity:↓, decreased activity\nIn contrast to direct small-molecule modulators that bind sirtuins at the catalytic or allosteric level, indirect strategies alter sirtuin activity by reshaping the metabolic or signalling environments in which these enzymes operate. Given that all seven mammalian sirtuins are obligate NAD + -dependent deacylases, interventions that influence NAD +  availability, redox balance, nutrient sensing or transcriptional control can modulate sirtuin function across a wide range of tissues and disease states. Such approaches often lack isoform selectivity but may exert broader physiological impact by engaging the family-level regulatory logic described in earlier sections of this review.\nRestoration of intracellular NAD +  levels represents the most extensively explored indirect approach. Precursors such as nicotinamide riboside (NR) and nicotinamide mononucleotide (NMN) increase cellular NAD +  pools and have demonstrated metabolic and cardiovascular benefits in experimental models. However, NAD +  elevation does not activate a single sirtuin in isolation. Instead, it enhances the activity of multiple compartmentalized sirtuins in a manner determined by tissue distribution, subcellular NAD +  pools and metabolic demand. As discussed above, nuclear, cytosolic and mitochondrial NAD +  pools are only partially equilibrated; therefore, global NAD +  boosting may preferentially engage certain isoforms depending on disease context.\nThis principle is illustrated in studies often described as “SIRT7 activation.” In most cases, SIRT7 is not directly targeted by a selective agonist. Rather, NAD +  precursor supplementation or metabolic reprogramming increases SIRT7 activity as a downstream consequence of improved NAD +  availability or altered transcriptional regulation. For example, nicotinamide riboside supplementation phenocopies Sirt7 overexpression in pulmonary hypertension models, while hypoxia-induced NAD +  production via PYCR1 enhances Sirt7 deacetylase activity within ELK4-dependent programmes [ 267 ]. In other contexts, increased SIRT7 expression, such as during ketogenic diet exposure or GLSP-mediated signalling, has been interpreted as functional activation [ 268 ,  269 ]. These findings indicate that most in vivo “SIRT7 activation” reflects metabolic licensing rather than direct catalytic stimulation.\nBeyond NAD +  precursors, pharmacological agents that alter cellular energy balance can indirectly regulate sirtuin activity. Metformin is a prototypical example. By inhibiting mitochondrial complex I, metformin lowers ATP levels, activates AMPK and suppresses mTORC1 signalling. This shift in energy state enhances the AMPK–SIRT1 axis and modulates downstream pathways including PGC-1α, lipid metabolism and inflammatory signalling [ 270 ,  271 ]. The clinical benefits of metformin cannot be attributed solely to SIRT1 activation; rather, SIRT1 participates as one component of a broader metabolic network.\nThis systems-level engagement is conceptually consistent with the compartmentalized regulatory framework established above. Energy-sensing interventions alter glycolytic flux, fatty acid oxidation and mitochondrial redox state, thereby reshaping the acylation landscape on which sirtuins act. In this context, sirtuins function less as primary drug targets and more as interpreters of metabolic reconfiguration. Indirect and systems-level regulatory mechanisms are outlined in Table  2 .\n Table 2 Indirect and systems-level mechanisms regulating sirtuin activity in physiological and disease contexts SIRT family member Regulatory factor Regulation type/mechanism Effect on SIRT Disease or experimental context SIRT1 Metformin [ 270 ,  271 ] Indirect engagement of the AMPK–SIRT1 axis through cellular energy balance; suppresses of mTORC1–SREBP signalling ↑ (indirect) Cardiometabolic benefits supported by large clinical datasets SIRT7 Nicotinamide riboside (NR) [ 272 ] NAD +  precursor; described as activating Sirt7 in vivo via increased NAD +  availability ↑ (indirect) Pulmonary hypertension models SIRT7 Hypoxia-induced NAD +  production via PYCR1 [ 267 ] Enhanced NAD +  availability increases Sirt7 deacetylase activity within an ELK4-dependent transcriptional programme ↑ (indirect) Hypoxia-associated tumour growth SIRT7 β-hydroxybutyrate-mediated promoter acetylation [ 268 ] Promotes SIRT7 promoter acetylation → increased SIRT7 expression ↑ (expression upregulation) Ketogenic diet exposure SIRT7 GLSP-driven enhancement of Sirt7–Keap1–Nrf2 signalling [ 269 ] Transcriptional upregulation of SIRT7; enhances SIRT7–Keap1–Nrf2 signalling ↑ (signalling enhancement) Vascular ageing models Symbol footnotes: ↑, increased activity: ↓, decreased activity\nIndirect and systems-level mechanisms regulating sirtuin activity in physiological and disease contexts\nSymbol footnotes: ↑, increased activity: ↓, decreased activity\nProteolysis-targeting chimeras (PROTACs) are a versatile modality to modulate sirtuin function by inducing selective protein degradation rather than enzymatic inhibition, offering a complementary strategy to conventional small-molecule inhibitors. To date, the most mature examples are centred on SIRT2, where sirtuin-rearranging ligands (SirReals) have been repurposed into PROTAC architectures capable of driving efficient and isoform-selective SIRT2 degradation via cereblon recruitment [ 263 ,  264 ]. These degraders demonstrated that removal of the SIRT2 protein can yield biological outcomes distinct from catalytic inhibition, including differential effects on microtubule acetylation and cellular morphology [ 265 ]. Subsequent studies further extended this concept by developing next-generation SIRT2-targeted PROTACs with enhanced antiproliferative activity in cancer models, supporting the therapeutic promise of sirtuin degradation strategies [ 122 ,  273 ]. Beyond SIRT2, SIRT6 has also been targeted using PROTAC technology. One such degrader, SZU-B6, induced near-complete SIRT6 degradation, impaired DNA damage repair and enhanced radiosensitization in hepatocellular carcinoma cell lines, and exhibited synergistic antitumor activity with sorafenib or irradiation in xenograft models [ 266 ]. These findings highlight the feasibility of extending protein-degradation approaches to other nuclear sirtuins implicated in cancer and metabolism. More broadly, comprehensive reviews of HDAC- and sirtuin-directed PROTACs underscore that epigenetic enzymes may be well suited for degradation-based pharmacology, owing to their scaffolded protein–protein interactions and scaffold-dependent non-enzymatic functions [ 274 ,  275 ].\nWhile direct PROTACs against SIRT7 have not yet been reported, several mechanistic studies emphasize that selective elimination of individual sirtuins can produce phenotypes not achievable by inhibition alone, providing a conceptual framework for future SIRT7 degrader development. The growing chemical biology toolbox for sirtuins (including activity-based probes to assess functional engagement and degradation outcomes [ 276 ]) is expected to facilitate this transition. Current evidence positions PROTAC-mediated sirtuin degradation as an advancing strategy, with SIRT2 and SIRT6 serving as precedents and SIRT7 representing a plausible next direction for targeted protein degradation approaches.\nCurrent therapeutic efforts targeting sirtuins illustrate both the tractability and the limitations of this enzyme family as drug targets. While small-molecule modulators, NAD +  precursors and emerging degradation strategies have demonstrated preclinical efficacy and acceptable safety profiles, their clinical performance has been heterogeneous and frequently dependent on patient selection and disease stage. These outcomes point to the need for improved isoform selectivity, clearer pharmacodynamic biomarkers and a more nuanced understanding of how disease-specific metabolic states shape sirtuin function in vivo. Rather than representing a failure of the sirtuin concept, these challenges highlight the transitional nature of the field and set the stage for a broader, family-wide reassessment of sirtuin biology and therapeutic strategy. Key challenges for PROTAC-based sirtuin modulation include achieving tissue-selective degradation, minimizing off-target proteostasis perturbation and establishing pharmacodynamic biomarkers that distinguish degradation from inhibition.\nA recent synthesis of context-dependent sirtuin pharmacology in cancer highlights metabolic-epigenetic switches and biomarker stratification as the basis for precision therapeutic targeting [ 277 ].\nA central principle emerging from this analysis is that therapeutic selectivity must be defined at the level of subcellular compartment and isoform, not merely at the level of the sirtuin family. Sirtuin function is compartmentalized, so interventions that raise NAD +  globally (such as NAD +  precursors) necessarily engage all seven isoforms across every compartment. This broad engagement likely underlies their heterogeneous clinical outcomes. Direct, isoform-selective modulators and PROTAC-mediated degradation restrict the intervention to a single isoform, matching the therapeutic strategy to the compartment-specific disease mechanism. A second principle is that the direction of intervention, whether activation, inhibition, or degradation, should be matched to disease context. The same sirtuin can be protective in one tissue and pathogenic in another, so a fixed agonist or inhibitor is unlikely to be broadly useful. Degradation, by contrast, removes the target entirely and can recapitulate phenotypes that inhibition alone cannot achieve, as illustrated by SIRT2-directed PROTACs. These principles carry three implications for developing more selective sirtuin-targeted interventions. First, prioritize isoform-selective over family-wide modulation; second, match the modality (activation, inhibition, degradation) to the tissue- and disease-specific role of the target isoform; and third, incorporate pharmacodynamic biomarkers that report isoform-specific target engagement rather than global NAD +  status.\n\nSirtuins have evolved as a family of NAD + -dependent deacylases that collectively function as a distributed metabolic-sensing network, rather than as isolated enzymatic regulators. Across subcellular compartments, SIRT1–SIRT7 translate fluctuations in nutrient availability, redox balance and cellular stress into coordinated transcriptional, epigenetic and post-translational responses. This compartmentalized organization enables sirtuins to regulate core processes such as glycolysis, fatty acid oxidation, mitochondrial quality control, inflammation and immune tolerance with a high degree of context sensitivity. Viewing sirtuins through a family-wide lens therefore helps reconcile their apparent pleiotropy and provides a conceptual framework in which metabolic state, spatial localization and substrate accessibility jointly shape biological outcomes.\nA recurring theme emerging from this review is that many seemingly contradictory roles attributed to individual sirtuins reflect differences in metabolic context, tissue identity and temporal dynamics rather than intrinsic functional inconsistency. The same sirtuin can exert protective or deleterious effects depending on NAD +  pool availability, disease stage and cellular lineage. This tissue- and disease-stage dependence is particularly evident in cancer, inflammation and metabolic disease, where sirtuin-mediated regulation can either restrain maladaptive metabolic reprogramming or be co-opted to support pathological states. Recognizing sirtuins as context-sensitive modulators, rather than fixed oncoproteins or tumour suppressors, is therefore important for accurate interpretation of experimental data and for rational therapeutic targeting.\nDespite substantial progress, important gaps remain in our understanding of sirtuin biology. First, the extent to which different sirtuin isoforms functionally cooperate or compensate for one another across subcellular compartments is poorly defined. Second, the mechanisms governing compartment-specific NAD +  pool dynamics and their selective impact on individual sirtuins are still incompletely understood. Third, non-canonical functions of sirtuins, including scaffold roles and non-enzymatic activities revealed by protein degradation approaches, may be underappreciated. Finally, several family members, most notably SIRT4 and SIRT7, remain relatively understudied in immune regulation and human disease, representing promising opportunities for future investigation.\nFrom a translational perspective, this review highlights why early attempts to therapeutically target sirtuins have yielded mixed clinical outcomes. Strategies focused on single isoform activation or inhibition often fail to account for the interconnected, family-level regulatory logic of sirtuin signalling. Emerging approaches offer new opportunities to overcome these limitations, including isoform-selective modulators, NAD + -independent activators, PROTAC-mediated degradation and combination therapies integrating metabolic and immune interventions. Successful clinical translation is likely to require both improved chemical selectivity and robust pharmacodynamic biomarkers and patient stratification based on metabolic and disease context.\nIn conclusion, the sirtuin family represents a key regulatory axis at the intersection of metabolism, epigenetics and immune homeostasis. By framing sirtuins as a coordinated, compartmentalised network rather than as individual drug targets, this review provides a unifying conceptual foundation for interpreting the expanding literature and for guiding future therapeutic development. Advances in understanding sirtuin cross-talk, NAD +  compartmentalization and isoform-specific vulnerabilities have the potential to transform sirtuin biology from a field characterized by complexity and controversy into one driven by greater mechanistic clarity and translational precision. Realizing this potential will depend on integrating mechanistic insights with rigorous in vivo validation and carefully designed clinical studies.","source_license":"CC-BY-4.0","license_restricted":false}