Xiangyu Wu
Tingye He1,2†- 1Department of Anesthesiology, The Affiliated Hospital, Southwest Medical University, Luzhou, Sichuan, China
- 2Anesthesiology and Critical Care Medicine Key Laboratory of Luzhou, Southwest Medical University, Luzhou, Sichuan, China
Postoperative cognitive dysfunction (POCD) remains a significant challenge in perioperative medicine, especially among older adults. Despite its prevalence, existing models centered on transient neuroinflammation fail to explain why cognitive deficits often persist long after systemic immune responses resolve. This review proposes a new framework: POCD is driven not by ongoing inflammation, but by a stable shift in microglial identity. We describe a closed-loop “inflammatory memory circuit” in which mitochondrial dysfunction, chromatin remodeling, and persistent polarization co-evolve to lock microglia into a hypersensitive, neurotoxic state. Recent studies suggest that surgical trauma triggers mitochondrial damage and mtDNA release, initiating innate immune activation via the cGAS–STING and NLRP3 pathways. These events engage epigenetic machinery—including HDAC3, DNMT3a, and long non-coding RNAs like MEG3—which reinforce transcriptional programs that lower activation thresholds and amplify cytokine output. Sustained M1-like polarization further propagates this loop, driving neuronal injury even in the absence of continued systemic cues. We outline experimental strategies to validate this model, including time-resolved single-cell transcriptomics and chromatin accessibility profiling. Therapeutically, we highlight HDAC inhibitors, SIRT1 agonists, and lncRNA-targeted interventions as potential strategies to disrupt the circuit before state-locking occurs. By reframing POCD as a glial fate transition rather than a transient immune reaction, this model offers mechanistic clarity and opens a path toward time-sensitive, precision interventions.
1 Introduction: rethinking POCD through the lens of inflammatory memory
Postoperative cognitive dysfunction continues to pose a significant neurocognitive challenge in the aging surgical population. Despite improvements in anesthetic safety and perioperative care, many patients—particularly the elderly—experience a constellation of cognitive disturbances, including episodic memory lapses, executive dysfunction, and impaired attention that may persist for weeks or even months following otherwise uneventful surgeries (Park et al., 2024; Zeng et al., 2023). Recent high-quality studies estimate that early POCD occurs in approximately 20%–40% of elderly patients after non-cardiac surgery (Miller et al., 2018; Punjasawadwong et al., 2018). Alarmingly, POCD is increasingly recognized as a risk factor for premature mortality and accelerated neurodegenerative trajectories, yet these lasting effects often follow a brief and seemingly reversible physiological insult—transient exposure to anesthesia and surgical trauma (Steinmetz et al., 2009; Yang X. et al., 2022; Zhao et al., 2024).
Traditional mechanistic accounts have focused on acute neuroinflammatory responses, transient blood–brain barrier disruption, and peripheral immune activation. While these mechanisms are undoubtedly involved in the perioperative window, they typically resolve within days (Alam et al., 2018; Zhang Z. et al., 2024). This temporal mismatch between a short-lived systemic insult and a protracted period of cognitive dysfunction suggests the involvement of a more persistent, CNS-intrinsic process (Tan et al., 2021; Wang T. et al., 2025). Indeed, several studies have reported the normalization of peripheral inflammatory markers despite ongoing neurocognitive symptoms, raising the possibility that the insult has induced a more enduring transformation within the central nervous system—particularly in its immune-competent glial population (Femenía et al., 2018; Zhang and Yin, 2023).
This review proposes a conceptual shift: rather than interpreting POCD as a failure of inflammation to resolve, we argue that it reflects a maladaptive form of neuroimmune memory. We hypothesize that microglia, upon exposure to perioperative stress, enter a primed state characterized by chromatin imprinting, altered mitochondrial homeostasis, and a persistent bias toward pro-inflammatory functional states (Huang et al., 2023; Martins-Ferreira et al., 2021; Neher and Cunningham, 2019). This locked inflammatory configuration is not merely a residual of injury but represents an epigenetically stabilized phenotype that continues to influence neural circuits long after the primary trigger has receded (Wu et al., 2023). Mounting evidence supports this notion. Studies in aging and neurodegeneration have shown that microglia retain molecular signatures of past insults, including persistent shifts in mitochondrial reactive oxygen species (mtROS) handling, histone acetylation patterns, and non-coding RNA expression (Cheray and Joseph, 2018; Haley et al., 2019; Nakanishi et al., 2011). These features not only render microglia more reactive to secondary stimuli but constitute an epigenetically maintained “neuroimmune memory” state analogous to trained immunity in peripheral macrophages (Zhang X. et al., 2022). In the context of POCD, this memory may manifest as a sustained distortion of the glial environment—one that perpetuates synaptic dysfunction and impairs recovery, even in the absence of ongoing inflammation (Jung et al., 2023; Qi et al., 2023; Tan et al., 2024).
To better articulate this concept, we introduce the “three-axis inflammatory memory model” which describes a recursive interaction between mitochondrial stress, epigenetic remodeling, and microglial functional state transition. In this framework, mitochondrial perturbation serves as the initial trigger, epigenetic reprogramming functions as the stabilizing layer, and pro-inflammatory microglial behavior acts as the amplifying output (Suo et al., 2025; Ying et al., 2024; Zhang L. et al., 2025). These three axes collectively sustain a pathological feedback loop that resists spontaneous resolution and provides a mechanistic basis for the persistent neurocognitive sequelae observed in POCD. We further propose that this model offers multiple experimentally testable avenues. With the aid of single-cell RNA sequencing, ATAC-seq, and histone modification profiling at distinct postoperative intervals, it becomes possible to chart the epigenetic trajectory of microglial states from injury through recovery (Suo et al., 2025). Additionally, we explore potential therapeutic entry points—including SIRT1–PGC-1α modulators, histone deacetylase inhibitors, and long non-coding RNA silencers—that may help dismantle this locked phenotype (Chen et al., 2020; Yang X. et al., 2021; Ye et al., 2023). Finally, we highlight promising candidate biomarkers—such as circulating mtDNA, histone deacetylase (HDAC) activity levels, and MEG3 expression—that could be leveraged to predict susceptibility and inform personalized intervention strategies (Chen et al., 2020; Ye et al., 2023; Zhang L. et al., 2025).
By reframing POCD as a disease of maladaptive neuroimmune imprinting rather than transient inflammation, we aim to redirect mechanistic inquiry and therapeutic innovation toward the underlying cellular programs that sustain cognitive vulnerability. This shift in perspective may yield not only new scientific insights but also practical approaches to identifying and reversing the memory-encoded pathologies at the core of postoperative brain dysfunction.
2 Postoperative inflammation inertia: decoupling stimulus and persistence
2.1 Persistence of microglial activation beyond cytokine resolution in preclinical models
Rodent models of surgery-induced neuroinflammation consistently demonstrate that microglial activation persists well beyond the transient wave of peripheral cytokines typically triggered by surgical stress. For instance, studies employing tibial fracture or laparotomy under general anesthesia reveal sustained morphological changes in hippocampal and cortical microglia, accompanied by upregulation of Iba1, CD68, and MHC-II expression for up to 7 days postoperatively (Feng et al., 2017; Qiu et al., 2016; Zhu et al., 2016). By contrast, levels of IL-6, TNF-α, and HMGB1 in serum and cerebrospinal fluid return to baseline within 48–72 h, indicating that the acute inflammatory surge is both sharp and short-lived (Zou et al., 2025). This discrepancy suggests that microglia do not simply mirror the temporal profile of systemic inflammation, but instead undergo a more durable shift in functional state (Qiu et al., 2016). Importantly, behavioral studies in the same models show memory deficits and reduced fear conditioning performance that align temporally with glial activation rather than with cytokine elevation (Meng et al., 2019). These findings challenge linear models of inflammation resolution and raise the possibility that microglia, once activated, may adopt a new operational identity that perpetuates dysfunction in the absence of continued external provocation.
Several candidate mechanisms have been proposed. Mitochondrial damage–associated molecular patterns (mtDAMPs), including extracellular mtDNA and cardiolipin, may persist long after tissue healing and act as latent triggers of microglial reactivity (Lin et al., 2022; Liu Y. et al., 2024). In parallel, metabolic reprogramming—marked by enhanced glycolysis and impaired oxidative phosphorylation—resembles a Warburg-like shift, stabilizing microglia in an energetically pro-inflammatory state (Baik et al., 2019; Yang S. et al., 2021). Transcriptomic analyses further support this view, revealing sustained upregulation of chromatin remodelers and pattern recognition receptors, suggesting that once microglia cross an activation threshold, they remain in a functionally reprogrammed state uncoupled from external stimuli (Miao et al., 2023).
Such evidence supports a paradigm in which microglial activation becomes intrinsically sustained—not by persistent signaling, but by a transition in cellular identity. This “phenotypic inertia” likely reflects chromatin imprinting and metabolic stabilization, establishing a new homeostatic set point for microglial behavior. In this light, the neuroimmune system may serve not merely as a reactive interface, but as a durable memory substrate for prior inflammatory episodes.
2.2 Human evidence for central inflammatory inertia following surgery
Although direct human brain sampling in the postoperative context is constrained by ethical and technical limitations, converging lines of evidence suggest that central inflammation may outlast systemic recovery. Positron emission tomography (PET) imaging using radiolabeled translocator protein (TSPO) ligandshas been used in studies in autoimmune encephalitis, neurodegeneration, and tuberous sclerosis complex have revealed prolonged central microglial activation that correlates with cognitive decline, even after peripheral inflammatory markers normalize (Kagitani-Shimono et al., 2023; Rossano et al., 2024; Wang J. et al., 2025). These elevations occur independently of systemic cytokine normalization, pointing to a central mechanism that continues to evolve after peripheral markers subside.
In rodent models simulating clinical surgery, sustained elevations of IL-1β and IFN-γ have been detected in the hippocampus up to 7 days postoperatively, accompanied by microglial accumulation in both the hippocampus and prefrontal cortex (Qiu et al., 2016). Moreover, transcriptomic analysis of hippocampal tissue following IFN-γ-induced microglial priming reveals sustained upregulation of chromatin remodeling enzymes, mitochondrial stress-response genes, and MHC-II antigen presentation pathways, reflecting a persistently activated microglial transcriptional state (Zhang J. et al., 2020).
We propose the term inflammatory inertia to describe this phenomenon: a sustained, stimulus-decoupled microglial activation state that represents a qualitative shift rather than a quantitative extension of inflammation. This is not simply prolonged inflammation in the classical sense; rather, it is a form of neuroimmune state-locking, akin to epigenetic reprogramming in cancer or immune cell exhaustion in chronic viral infections. The inflammatory burden may dissipate, but the cell’s functional identity does not revert (Huang et al., 2023; Neher and Cunningham, 2019; Zhang X. et al., 2022). This perspective alters both our understanding and our therapeutic priorities. The critical determinant of POCD may not be the peak intensity or duration of inflammation, but the point at which microglia cross a regulatory threshold and enter a stable, maladaptive phenotype. Unlike acute inflammation, which is inherently self-limiting, this inertia resists resolution and perpetuates subtle but persistent circuit dysfunction (Jung et al., 2023). In recognizing this phenomenon as a neuroimmune state transition, we shift the therapeutic focus from suppressing cytokine production to intercepting—or reversing—the cellular reprogramming that maintains microglial pathogenicity. This reorientation also informs biomarker discovery: rather than tracking inflammatory surges, future diagnostics should aim to detect epigenetic or metabolic indicators of glial state locking.
Taken together, existing preclinical and mechanistically relevant clinical findings support a revised model of postoperative neuroinflammation—not as a fading echo of peripheral insult, but as a form of innate immune memory embedded within the brain’s microglial population. This framework reflects a shift from stimulus–response coupling to a persistent, self-sustaining inflammatory configuration—an epigenetically encoded microglial phenotype that resembles trained immunity in peripheral macrophages and persists even after resolution of the initial insult (Huang et al., 2023; Neher and Cunningham, 2019; Zhang X. et al., 2022). Peripheral trained immunity has been proposed to contribute to sustained neuroimmune activation. Surgical or systemic inflammatory stimuli may induce epigenetic reprogramming of circulating monocytes, a phenomenon termed “trained immunity,” characterized by persistent pro-inflammatory responses including IL-1β, IL-6, and TNF-α release (Netea et al., 2020). Although direct clinical evidence remains limited, these systemic cytokines may modulate central immune homeostasis, potentially reinforcing microglial priming via soluble mediators rather than direct cell infiltration. This paradigm emphasizes targeting persistent immune cell states that underlie neuroimmune dysregulation, beyond merely suppressing acute inflammation.
3 A three-axis mechanism for inflammatory memory state
The persistence of microglial activation in POCD cannot be adequately explained by transient stimuli alone. Instead, it appears to originate from a feedback-coupled biological network—an interdependent triad of mitochondrial stress, chromatin imprinting, and phenotypic persistence that mutually reinforce one another (Zhang X. et al., 2022; Zhang L. et al., 2025). This “three-axis” model of inflammatory memory does not describe a sequence of independent processes but a co-evolving system that traps microglia in a pathologically sensitized state. Each axis contributes uniquely: mitochondrial stress acts as the initial signal integrator, epigenetic reprogramming stabilizes a maladaptive identity, and sustained polarization amplifies neurotoxic output. The following section focuses on the initiating role of mitochondrial stress.
3.1 Mitochondrial stress as the fast trigger
Among the earliest events after surgical insult is the emergence of mitochondrial dysfunction in central nervous system immune cells, particularly microglia. Surgery presents multiple metabolic challenges—hypoxia, anesthetic toxicity, and systemic inflammation—that converge upon mitochondria, leading to rapid alterations in microglial morphology and bioenergetics. Within hours post-injury, murine studies have reported swelling of mitochondrial cristae, elevated mtROS, and activation of redox-sensitive inflammatory signaling (He et al., 2022; Ye et al., 2016).
These stress responses are not incidental. mtROS and mitochondrial DNA (mtDNA), two potent damage-associated molecular patterns (DAMPs), act as immunogenic signals. mtROS facilitates lipid peroxidation and inflammasome activation via NLRP3, promoting caspase-1–dependent maturation of IL-1β and IL-18 (Antón and Traba, 2022; Han et al., 2021; Wu et al., 2019). Concurrently, mtDNA—hypomethylated and structurally akin to bacterial genomes—activates the cGAS–STING pathway once released into the cytosol (Yang et al., 2024). This induces type I interferons and NF-κB signaling cascades, contributing to an antiviral-like transcriptional profile within hippocampal microglia (Liu Y. et al., 2024). Notably, inhibition of either NLRP3 or STING alleviates postoperative cognitive deficits in rodent models, supporting a causal link (Yang et al., 2024; Zhang L. et al., 2025).
Microglial detection of these mitochondrial DAMPs is mediated through pattern recognition receptors such as NOD-like receptors (Zhang L. et al., 2025). But this detection is not a transient event; it leads to persistent transcriptional reprogramming involving IRF7 and chromatin-modifying complexes (Bowen et al., 2024; Saeki et al., 2024; Tanaka et al., 2015). These molecular signatures render microglia hyperresponsive to subsequent perturbations, a property known as trained immunity in peripheral innate cells (Neher and Cunningham, 2019; Zhang X. et al., 2022). Additionally, the impairment of mitophagy, particularly through PINK1–Parkin dysregulation, results in accumulation of dysfunctional mitochondria and continued DAMP release (Wu et al., 2021; Yao et al., 2024). This failure of mitochondrial quality control sets the stage for a feedback loop in which oxidative stress and immune activation sustain one another (Lv et al., 2024).
Cumulatively, the mitochondrial axis initiates and perpetuates inflammatory activation in microglia (van Horssen et al., 2019). Far from a transient trigger, mitochondrial dysfunction transforms microglia into long-term effectors through metabolic and redox instability (Sangineto et al., 2023; Wang et al., 2024). This process acts as the first pivot in the inflammatory memory circuit, preparing the cellular landscape for more durable epigenetic reprogramming.
3.2 Epigenetic reprogramming as the slow engraver
Where mitochondrial dysfunction primes the acute inflammatory reactivity of microglia, it is the epigenetic machinery that inscribes this state into a persistent transcriptional identity. Chromatin modifications, DNA methylation, and long non-coding RNA (lncRNA) coordination collectively reshape the regulatory landscape of microglia, transforming them into hypersensitive effectors of neuroinflammation (Cai et al., 2020; Yang X. et al., 2021). In this axis—the “slow engraver”—epigenetic alterations serve not as echoes of inflammation but as its lasting codex.
DNA methylation provides a parallel and complementary layer of transcriptional control. DNMT3a, a de novo DNA methyltransferase, is upregulated in neuroinflammatory and neurodegenerative conditions and targets promoter regions involved in mitochondrial metabolism and synaptic regulation (McGregor et al., 2025). These modifications repress oxidative phosphorylation genes and sustain glycolytic, pro-inflammatory phenotypes. Inhibiting DNMT3a reverses these changes, restoring metabolic flexibility and attenuating inflammatory output, indicating its role in locking microglia into a maladaptive state. LncRNAs add further complexity to this regulatory architecture. MEG3 and MALAT1, two well-studied lncRNAs, are upregulated in microglia under neuroinflammatory conditions (Cai et al., 2020; Li et al., 2020). MALAT1, similarly, binds EZH2 to mediate epigenetic silencing of anti-inflammatory regulators (Cai et al., 2020; Xu et al., 2022).
The net result is a transcriptionally poised but epigenetically locked microglial state. Single-cell chromatin accessibility profiling (scATAC-seq) demonstrates persistent rapidly induced upon re-challenge (Saeki et al., 2024; Zhang X. et al., 2022). This suggests that epigenetic reprogramming creates a landscape primed for pathological recall. Critically, these modifications are maintained independent of ongoing inflammatory stimuli, supporting the concept of microglial memory. Transcription factors such as PU.1, IRF8, and C/EBPβ orchestrate this process, linking early mitochondrial signals with durable chromatin remodeling (Saeki et al., 2024). Their activation patterns form regulatory loops that reinforce accessibility at inflammatory gene loci and silence neuroprotective programs (Unlu et al., 2007; Zhou et al., 2019). Thus, epigenetic reprogramming in POCD is not simply an outcome of prior activation—it is the mechanism that sustains it.
This second axis of inflammatory memory ensures that microglia retain a biased, hypersensitive state long after the inciting event has passed. As a result, even minimal perturbations can provoke exaggerated responses, sustaining neuronal injury and cognitive decline. In the context of POCD, this axis transforms inflammation from an event into a condition—a locked transcriptional program embedded in chromatin and reinforced by lncRNA scaffolds.
3.3 Microglial polarization as the executor and amplifier
With microglial inflammatory memory set in motion by mitochondrial danger signaling and epigenetic imprinting, the third axis—phenotypic polarization—emerges as the effector limb of persistent neuroinflammation. This axis is not a passive endpoint but an active amplifier that embeds and perpetuates the pathological program. At the center of this mechanism lies the sustained M1-like polarization phenotype: a pro-inflammatory state marked by upregulation of iNOS, CD86, and MHC-II and downregulation of markers such as CD206 and arginase-1 (Du et al., 2024; Zhao et al., 2019).
Under normal conditions, microglia exhibit plasticity between inflammatory and reparative states. In the postoperative brain, however, this flexibility is compromised. Prolonged M1 polarization has been observed for over two weeks in rodent POCD models, correlating with hippocampal synapse loss, impaired long-term potentiation, and memory deficits (Li et al., 2022; Qi et al., 2023). These M1-polarized microglia act as autonomous effectors, releasing TNF-α, IL-1β, and IFN-γ, which not only impair neurons directly but also fuel oxidative stress and mitochondrial dysfunction in glia and neurons alike (Qiu et al., 2016; Wang et al., 2015). Importantly, this phenotype feeds back into the mitochondrial and epigenetic axes (Ji et al., 2024). TNF-α has been shown to induce mitochondrial permeability and mtDNA release in neurons, reinforcing cGAS–STING activation in microglia (Liu Y. et al., 2024; Willemsen et al., 2021). Meanwhile, M1 signaling maintains NF-κB activation via sustained IκB kinase activity and transcriptional reinforcement at enhancer regions (Yang et al., 2024). Key transcription factors—including STAT1, and lncRNAs like MEG3—further consolidate this state by enhancing inflammatory gene expression (Ma et al., 2024; Przanowski et al., 2014). This phenotype is stabilized, in part, by suppression of anti-inflammatory programs involving STAT6 and PPARγ (Liu X. et al., 2024; Yao et al., 2023). Concurrently, the anti-inflammatory axis is suppressed. SIRT1, a regulator of mitochondrial homeostasis and NF-κB restraint, is downregulated after surgery, particularly in aged animals (Yan et al., 2019; Zhang J. et al., 2022). This loss disables metabolic resilience and accelerates polarization inertia. Microglia, in this context, cease to function as sentinels and instead become long-term amplifiers of neurotoxicity (Liu et al., 2025). Emerging single-cell transcriptomic studies reveal that microglia are far from homogeneous, instead comprising multiple phenotypic subsets beyond the classical M1/M2 dichotomy (Masuda et al., 2020). Among these, disease-associated microglia (DAM) represent a transcriptionally distinct subtype characterized by upregulation of TREM2 and APOE, expanding in contexts of chronic neuroinflammation and neurodegeneration (Keren-Shaul et al., 2017). These findings underscore that discrete microglial populations engage specific transcriptional programs depending on the nature and chronicity of the insult. Aging itself is known to increase microglial heterogeneity, as demonstrated by single-cell transcriptomic studies identifying pro-inflammatory and interferon-responsive clusters in aged murine brains, suggesting that the aged microglial landscape may predispose to exaggerated neuroimmune responses following surgical stress (Hammond et al., 2019).
This third axis thus closes the loop: what begins as a stress response becomes an encoded inflammatory phenotype, executed through persistent polarization and sustained by mitochondrial dysfunction and epigenetic reprogramming. The resulting system is no longer dependent on the initial insult—it has become its own driver of pathology. Collectively, these three axes—mitochondrial stress propagation, chromatin remodeling and epigenetic lock-in, and sustained polarization—do not operate in isolation. Instead, they reinforce one another in a self-perpetuating manner. To integrate these findings, we outline a unified mechanistic model: a circular “inflammatory memory circuit” that may underlie the persistence of postoperative cognitive dysfunction (Figure 1).
Figure 1. Inflammatory memory circuit in postoperative cognitive dysfunction (POCD). A closed-loop model linking mitochondrial stress, epigenetic lock-in, and sustained microglial polarization in postoperative cognitive dysfunction. Surgical stress induces mtROS and mtDNA release, which activate innate immune sensors (NLRP3 inflammasome and cGAS–STING). This response initiates chromatin remodeling via HDAC3 suppression and MEG3–EZH2–H3K27me3 axis engagement. The resulting epigenetic landscape enforces a fate-locked M1-like phenotype, characterized by TNF-α overproduction and loss of plasticity. These outputs further amplify upstream mitochondrial perturbation, completing a self-sustaining pathogenic loop.
4 A closed-loop model of inflammatory memory in POCD
The persistence of cognitive dysfunction following transient surgical insult necessitates a shift from linear models of postoperative neuroinflammation to a more dynamic framework (Hovens et al., 2012; Liu et al., 2023). Rather than fading with the resolution of systemic signals, inflammation in POCD appears to be sustained by a recursive network of cellular reprogramming. We propose that this durability arises from a closed-loop “inflammatory memory circuit”—a self-reinforcing system in which mitochondrial dysfunction, chromatin remodeling, and phenotypic polarization are not merely parallel processes, but mutually catalytic nodes in a circular pathophysiological sequence. Once microglia cross a reactivity threshold, they enter a metastable state in which each axis of the circuit perpetuates the others, transforming a short-term perturbation into a long-term disorder.
4.1 Initiation: mitochondrial danger signaling
The loop is initiated by acute metabolic stress imposed by surgery and anesthesia, which elicits mitochondrial dysfunction in microglia (Yang et al., 2024). mtROS and mitochondrial DNA (mtDNA) are released as DAMPs, activating canonical innate immune pathways including the NLRP3 inflammasome and the cGAS–STING axis (Ji et al., 2024; Liu Y. et al., 2024). These pathways catalyze the production of IL-1β, IL-18, and type I interferons, establishing a rapid pro-inflammatory program that primes the immune transcriptome (Yang et al., 2024; Zhang L. et al., 2025).
4.2 Stabilization: epigenetic engraving of an inflammatory identity
Rather than resolving with the attenuation of DAMP signaling, this initial activation leads to chromatin reconfiguration. DNMT3a has been shown to enhance innate immune activation by upregulating HDAC9, which deacetylates TBK1 and facilitates downstream antiviral signaling (Li et al., 2016). The long non-coding RNA MEG3 interacts with chromatin remodelers such as EZH2 to modulate epigenetic repression of pro-inflammatory genes in postoperative cognitive dysfunction (Ma et al., 2024). This results in an “epigenetic lock-in,” rendering microglia hyper-responsive and transcriptionally biased even in the absence of external stimuli (Zhou X. et al., 2024).
4.3 Execution and amplification: polarization-driven feedback
These epigenetically biased microglia adopt a persistent M1-like polarization phenotype, marked by sustained NF-κB and STAT1 activity, high levels of TNF-α and IL-1β, and suppression of anti-inflammatory mediators like PPARγ and SIRT1 (Liu Y. et al., 2024; Yan et al., 2019; Zhang Q. et al., 2024). Crucially, the outputs of this phenotype feed directly into upstream axes: cytokines such as TNF-α exacerbate neuronal mitochondrial damage, triggering further mtDNA release and reinforcing cGAS–STING signaling (Ji et al., 2024; Yang et al., 2024). Thus, polarization functions as both the executor and amplifier of the circuit, transforming microglia into autonomous drivers of neuroinflammation (Kong et al., 2022).
4.4 Pathophysiological consequence: a self-reinforcing neuroimmune state
The convergence of these three axes creates a system in which the original insult is no longer required to maintain pathology. Mitochondrial dysfunction reinforces epigenetic priming; chromatin remodeling sustains polarization; and M1 polarization recycles mitochondrial stress. This circuitous architecture recasts POCD not as an extended response to injury, but as a state transition—one in which microglia have acquired a memory-like identity that persists beyond resolution and resists reversal. Aging also exacerbates key regulatory failures in microglia. In older brains, PINK1/Parkin-mediated mitophagy is impaired, leading to accumulation of damaged mitochondria that release mtDNA and trigger innate immune sensors (e.g., cGAS–STING and NLRP3), thereby locking microglia in a pro-inflammatory state (Gulen et al., 2023; Lv et al., 2024). Concurrently, aging is associated with pronounced epigenetic alterations in microglia that compromise their ability to regulate inflammatory gene expression. For example, several HDAC isoforms—including HDAC1 and HDAC3—are upregulated in senescent microglia and aged hippocampal tissue, correlating with enhanced microglial and senescence markers (Auzmendi-Iriarte et al., 2022). These epigenetic vulnerabilities likely render aged microglia more susceptible to adopting maladaptive, persistent activation states following surgical stress (Fonken et al., 2018).
By conceptualizing POCD as an emergent property of this inflammatory memory circuit, we provide a framework that explains both the delayed onset and the prolonged duration of symptoms. More importantly, this model identifies multiple nodal points—metabolic, epigenetic, and phenotypic—at which intervention may disrupt the loop and restore neuroimmune equilibrium.
5 Hypothesis testing and experimental strategies
To empirically validate the inflammatory memory circuit proposed in POCD, we outline a two-pronged strategy: temporal mapping of microglial state transitions, and targeted pharmacologic interventions aimed at disrupting key axes of the circuit. This approach is further supported by biomarker development to enable translational application and real-time clinical stratification.
5.1 Temporal mapping of state transitions
A defining feature of the inflammatory memory model is the transition from transient activation to a metastable, memory-locked microglial state (Huang et al., 2023; Martins-Ferreira et al., 2021; Zhang X. et al., 2022). To resolve this trajectory, we propose longitudinal sampling at six critical postoperative timepoints—1, 6 h, 1, 3, 7, and 14 days—encompassing both the acute phase and the putative chromatin-locking window. These stage-specific changes are visualized in a trajectory schematic that links cellular states, molecular regulators, and detection strategies across the postoperative timeline (Figure 2).
Figure 2. Microglial state trajectory and epigenetic transitions following surgery. This schematic summarizes postoperative microglial state transitions across defined timepoints (1 h–7 days), integrating phenotypic markers with regulatory mechanisms. The progression from resting to activated and fate-locked states corresponds with HDAC3 downregulation, MEG3/EZH2 upregulation, and increased H3K27me3 levels. Relevant detection strategies, including scRNA-seq, ATAC-seq, and immunohistochemistry, are aligned to each timepoint to support stage-specific analysis and potential therapeutic targeting.
In rodent models of postoperative cognitive dysfunction, microglia may be isolated from the hippocampus using fluorescence-activated cell sorting (FACS) based on canonical markers such as CD11b (Rayaprolu et al., 2020). Ingle-cell RNA sequencing (scRNA-seq) and ATAC-seq allow parallel profiling of transcriptome and chromatin accessibility in microglia. Focus should include mitochondrial regulators (e.g., TFAM, NDUFS3), polarization markers (iNOS, Arg1, CD86), and chromatin modifiers including HDAC3, DNMT3a, and EZH2. Accessibility changes at enhancer regions targeted by PU.1, IRF8, and NF-κB serve as epigenetic indicators of durable transcriptional reprogramming (Martins-Ferreira et al., 2021; Zhang X. et al., 2022). In parallel, mitochondrial stress should be quantified by cytosolic mtDNA abundance via qPCR and mtROS accumulation using MitoSOX fluorescence, which are validated indicators of mitochondrial damage in POCD-related microglial activation (Ji et al., 2024; Zhang L. et al., 2025). Immunohistochemistry should track co-localization of Iba1? microglia with MHC-II, cGAS–STING, and H3K27me3 to distinguish activation status and epigenetic locking, as evidenced in POCD and neurodegenerative models (Yang et al., 2024; Zhou X. et al., 2024). This multimodal approach enables differentiation of transient activation from irreversible microglial reprogramming and supports the hypothesis of bifurcation into an inflammatory memory state.
5.2 Pharmacologic interruption and biomarker development
A second axis of validation assesses whether early therapeutic intervention can interrupt or reverse microglial reprogramming. Here, agents targeting specific circuit components are administered at key timepoints before state-locking occurs.
For mitochondrial reprogramming, PGC-1α agonists such as ZLN005 may prevent mtDAMP release and subsequent inflammasome activation. In POCD models, ZLN005 administration significantly upregulated PGC-1α signaling, suppressed NLRP3 activation, and restored mitochondrial function (Wu et al., 2024; Zhang W. et al., 2024). Behavioral improvements in novel object recognition and fear conditioning tests were observed following treatment at early postoperative intervals (1–6 h), with parallel reductions in IL-1β, IL-6, and restoration of mitochondrial biogenesis markers (Wu et al., 2024). To simultaneously modulate mitochondrial and epigenetic axes, SIRT1 activators such as SRT1720 offer a dual mechanism: deacetylation of NF-κB and restoration of PGC-1α-driven mitochondrial biogenesis (Pang et al., 2024; Sung et al., 2024). These interventions are especially relevant in aged animals where SIRT1 expression is reduced and mitochondrial dysfunction prominent (Mitchell et al., 2014).
Age-stratified analyses are essential to define the temporal therapeutic window. In aged POCD models, epigenetic checkpoints such as Mef2C are rapidly lost, predisposing microglia to primed, inflammation-permissive states (Wu et al., 2023). If chromatin imprinting and loss of immune regulation occur within the first few hours after surgery, reversal strategies may become ineffective once microglia cross a transcriptional “point of no return” (Zhou X. et al., 2024). These findings support the model’s prediction of irreversible inflammatory memory transitions when early interventions are missed (Zhang X. et al., 2025). In parallel, the circuit model enables rational biomarker development.
Although direct evidence in POCD is still emerging, mitochondrial DNA (mtDNA) levels in blood have been shown to correlate with cognitive impairment in non-POCD human cohorts and may reflect systemic mitochondrial dysfunction (Tian et al., 2025). Plasma mtDNA has also been associated with neuroinflammatory phenotypes in geriatric populations and HIV-related cognitive decline (Johnston et al., 2022). Peripheral HDAC3 activity, as measured by fluorometric assays, has been implicated in epigenetic regulation relevant to cognitive decline, and its elevation is observed in both aged animals and systemic inflammatory models affecting cognition (Sathishkumar et al., 2016; Yang L. et al., 2022). Longitudinal tracking of these markers at 1, 3, and 7 days could identify patients at highest risk for POCD—even before behavioral symptoms manifest.
In aggregate, these strategies provide a mechanistic and translational pipeline: dissecting temporal dynamics, intervening pharmacologically, and predicting trajectory via liquid biopsy. The inflammatory memory circuit thus becomes not only a conceptual model, but an experimentally tractable and clinically actionable framework. Its success will hinge on identifying the precise point at which microglia shift from reversible responders to autonomous propagators of dysfunction—and on learning how to stop them before the circuit completes.
6 From inflammatory trigger to cellular state transition: a conceptual shift
6.1 Reframing POCD as a microglial state transition
The prevailing model of POCD frames it as a temporary response to acute insult—surgical trauma and anesthesia trigger peripheral inflammation, which transiently breaches the blood–brain barrier, activates glial cells, and disrupts neuronal communication (Qiu et al., 2023; Riedel et al., 2014; Terrando and Akassoglou, 2022). While conceptually intuitive, this linear framework fails to resolve a key paradox: how can a fleeting peripheral event produce cognitive decline that persists for weeks or months, especially in aged patients without prior cognitive impairment?
To reconcile this discrepancy, we propose a redefinition of the underlying pathology: POCD is not merely the downstream consequence of unresolved inflammation—it is the consequence of a fate transition within microglia. Following the initial insult, a subset of microglia fails to return to homeostasis and instead enters a metastable, maladaptive state. This transition is not defined by continued exposure to stimuli, but by an intrinsic transformation stabilized by mitochondrial dysfunction, chromatin remodeling, and sustained polarization (Suo et al., 2025; Ying et al., 2024). The change is not incremental but categorical—marked by a lowered activation threshold, rewired gene expression programs, and a loss of plasticity (Tan et al., 2023; Zhang X. et al., 2025). This concept reflects a broader biological principle. Microglia, like neurons, are highly responsive to contextual cues. Their identity is plastic, shaped by ongoing input from the tissue microenvironment, metabolic status, and epigenomic configuration (Cheray and Joseph, 2018; Lauro and Limatola, 2020). When these signals converge beyond a critical threshold—particularly through mtROS accumulation and cytosolic mtDNA release—they engage transcriptional circuits that rapidly shift toward inflammatory dominance (Lauro and Limatola, 2020; Nakanishi et al., 2011). Over hours to days, this response is consolidated: anti-inflammatory repressors are silenced, inflammatory enhancers become more accessible, and long non-coding RNAs such as MEG3 further entrench the pro-inflammatory identity (Tan et al., 2023; Ye et al., 2023).
What emerges is not prolonged inflammation, but the formation of a new cellular baseline. Microglia no longer fluctuate in response to external stimuli; instead, they adopt a fixed identity that continues to shape brain function long after the surgical insult resolves. Although direct evidence of microglial epigenetic memory in human POCD is currently lacking, clinical research has begun to explore translational biomarkers of neuroinflammation. For example, translocator protein (TSPO) PET imaging has been employed to detect central neuroinflammatory activity associated with postoperative cognitive decline in elderly patients (Forsberg et al., 2017). Additionally, microglia-associated biomarkers such as soluble TREM2 in cerebrospinal fluid and exosomal non-coding RNAs in plasma are under investigation as indicators of perioperative neuroimmune activation (Henjum et al., 2016). Future integration of single-cell transcriptomic and epigenomic tools in postmortem brain tissue may clarify whether surgery induces long-lasting microglial reprogramming in humans. POCD, from this perspective, is the clinical manifestation of glial reprogramming—a state transition in which immune sentinels become autonomous effectors of neurotoxicity.
6.2 Comparative paradigms in neuronal and glial reprogramming
The state-transition model proposed for POCD is not without precedent. Across multiple biological domains, transient perturbations have been shown to induce durable changes in cellular identity—a principle evident in neuronal plasticity, trained innate immunity, and astroglial fixation (Li et al., 2021; Muscat and Barrientos, 2021; Tercan et al., 2021).
In the nervous system, experience-dependent transcriptional plasticity is a cornerstone of memory formation. Neurons exposed to high-frequency stimuli undergo persistent alterations in chromatin structure at genes such as Arc, Egr1, and BDNF, modifying their responsiveness to future inputs (Ibarra et al., 2022; Penke et al., 2011). These transcriptional changes are stabilized by activity-dependent enhancers and histone modifications, facilitating memory encoding—but also contributing to pathologies such as epilepsy and chronic pain when dysregulated (Ding et al., 2017; Zhang X. et al., 2022). Similarly, microglia exposed to surgery-related stressors appear to “memorize” the insult, not through synaptic connectivity, but through stable shifts in inflammatory gene regulation (Neher and Cunningham, 2019; Wendeln et al., 2018).
The innate immune system provides a parallel in the concept of trained immunity. Contrary to long-standing dogma, monocytes, NK cells, and macrophages can develop memory-like properties after infection or inflammation (Netea et al., 2016; Netea et al., 2020). This training involves metabolic reprogramming (e.g., increased glycolysis) and epigenetic remodeling of inflammatory loci, producing a heightened response upon re-exposure to even unrelated antigens (Cheng et al., 2014; Saeed et al., 2014). Microglia in POCD appear to mirror this behavior: brief exposure to mitochondrial DAMPs generates a transcriptionally biased state, in which subsequent perturbations evoke exaggerated inflammatory responses (Schindler et al., 2018; Zhang L. et al., 2025). Astrocytes also exhibit a state-transition phenomenon in neurodegenerative diseases. In Alzheimer’s disease, for instance, astrocytes shift into the neurotoxic “A1” phenotype, characterized by synapse elimination, pro-inflammatory cytokine secretion, and persistent transcriptomic alteration (Hulshof et al., 2022). Importantly, this phenotype is not easily reversible and is maintained independent of ongoing external stimuli (Lopes et al., 2021). The parallels with M1-polarized microglia in POCD are striking: both glial cell types, once pushed past a functional tipping point, become active contributors to neural injury rather than protectors of homeostasis.
These examples underscore a unifying principle: cells across physiological systems can undergo stable, stimulus-independent reprogramming in response to transient perturbation. Whether in the form of memory consolidation, immune training, or gliosis, these transitions rely on an initial stressor interpreted through the lens of mitochondrial metabolism and epigenetic regulation (Liotti et al., 2022; Walters and Zovkic, 2015). The inflammatory memory circuit described here fits squarely within this framework, positioning POCD as one manifestation of a more general biological motif. By shifting focus from inflammation as a transient response to inflammation as a reprogrammed identity, this model realigns therapeutic objectives. It suggests that recovery from POCD may require more than anti-inflammatory suppression—it may necessitate interventions that reset or rewire the epigenetic and metabolic programs underlying microglial fate. To consolidate these mechanistic insights into actionable strategies, we compiled representative targets, candidate interventions, and measurable readouts across the inflammatory memory circuit. These entries reflect preclinical feasibility and potential translational applicability (Table 1).
Table 1. Therapeutic targets, intervention strategies, and biomarker readouts within the inflammatory memory circuit.
7 Translational implications: intervening in the state transition loop
Reframing POCD as a state-locked, microglia-mediated condition necessitates a corresponding shift in therapeutic approach. Traditional strategies that aim to suppress inflammation may prove inadequate once microglia have entered a maladaptive fate. If the persistence of POCD reflects not unresolved cytokine signaling but a durable reprogramming of immune identity, then interventions must act at or before the point of cellular transition. This perspective opens a new translational landscape: one centered on interrupting the trajectory toward pathological imprinting rather than extinguishing its aftermath.
7.1 Pharmacologic interruption of the inflammatory memory circuit
Three mechanistic targets emerge as promising candidates for therapeutic modulation: HDAC inhibition, lncRNA interference, and SIRT1 activation. Each strategy addresses a distinct axis within the inflammatory memory circuit and offers an entry point for preventing or reversing glial reprogramming.
Histone deacetylase inhibitors, particularly those selective for HDAC3, can restore acetylation balance at key regulatory loci and suppress the formation of transcriptional memory (Sun et al., 2022). Compounds such as RGFP966 and vorinostat have demonstrated efficacy in reversing microglial hyperreactivity and improving cognitive outcomes in preclinical neuroinflammation models (Zhang M. et al., 2020). In the context of POCD, their window of effectiveness is narrow. Administration within six hours of surgical insult is likely necessary to preserve epigenetic flexibility and prevent entrenchment of the inflammatory phenotype (Lin et al., 2024). lncRNA-targeted approaches offer a complementary axis of control. MEG3, shown to scaffold chromatin repressors and modulate splicing of inflammatory transcripts, represents an actionable node for antisense oligonucleotide (ASO) or CRISPRi-based suppression (Hasenson et al., 2022; Li et al., 2024). In rodent models, early inhibition of MEG3 dampens microglial reactivity and cytokine production (Li et al., 2020; Meng et al., 2021). Future delivery platforms such as CNS-penetrant exosomes or lipid nanoparticles could enable rapid, region-specific suppression of these transcriptional amplifiers in the perioperative setting. SIRT1 agonists, including SRT1720 and resveratrol derivatives, offer dual benefits: suppression of NF-κB transcriptional activity and restoration of mitochondrial biogenesis via PGC-1α signaling (Sung et al., 2024; Wan et al., 2016). In aged animals, where SIRT1 activity is intrinsically lower and mitochondrial resilience is impaired, such agents may prevent both the metabolic trigger and the epigenetic lock-in that underlies the inflammatory memory loop (Gano et al., 2014; Zeng et al., 2021). While bioavailability limitations remain a concern for some compounds, newer formulations are in development that may allow clinical application.
These therapeutics, if deployed early, hold potential to redirect microglial fate away from chronic reactivity. Success hinges not merely on their anti-inflammatory potency, but on their alignment with the timing of the state transition itself.
7.2 Molecular profiling for risk stratification
Given the narrow therapeutic window, early identification of high-risk individuals becomes critical. The inflammatory memory framework supports a multi-marker profiling strategy based on mitochondrial, epigenetic, and transcriptomic features.
Mitochondrial DNA released into circulation after tissue injury or surgery may act as a DAMP, promoting innate immune activation and systemic inflammation (Zhang et al., 2010). This can be quantified via qPCR in blood samples collected intra- or immediately postoperatively. Peripheral HDAC3 activity, measurable in blood mononuclear cells via fluorometric assays, may reflect central epigenetic shifts, especially when analyzed in tandem with systemic inflammatory profiles (Monisha et al., 2023; Sathishkumar et al., 2016). Serum exosomal MEG3 expression offers a glial-specific readout of transcriptional reprogramming and has been potentia in neurodegenerative and ischemic contexts (Li et al., 2020; Zhou T. et al., 2024). Its early upregulation could function as both a diagnostic biomarker and a pharmacodynamic target for lncRNA-directed interventions.
Combining these molecular indicators into a composite risk index, alongside clinical factors such as age, cognitive reserve, and surgical complexity, would enable stratification of patients into risk tiers. High-risk individuals could be prioritized for early intervention, improving both therapeutic efficiency and resource allocation.
7.3 The six-hours window: from insight to intervention
A consistent signal emerges across mitochondrial, epigenetic, and transcriptional domains: the transition toward a locked microglial state occurs rapidly—often within the first 6–12 h postoperatively (Ji et al., 2024; Zhang L. et al., 2025). This defines a narrow but actionable window during which microglial fate remains modifiable. Intervention after this point likely encounters cells that have already undergone chromatin remodeling and metabolic fixation. In such cases, treatment must shift from prevention to reversal—a far more complex and less reliable endeavor. Analogous to thrombolysis in stroke or steroid administration in spinal trauma, timing is everything. The success of anti-POCD strategies will depend not only on drug choice, but on the speed with which diagnosis and treatment can be initiated.
In sum, the closed-loop model of POCD not only redefines pathogenesis—it recalibrates intervention. Therapeutics targeting the early reprogramming phase, guided by dynamic molecular profiling, may offer a path to preserving cognitive function in the face of surgical insult. By breaking the feedback loop before it becomes self-sustaining, we shift the clinical paradigm from late-stage damage control to early-stage fate correction.
However, a strict “6 h window” for intervention is clinically impractical, so alternative neuroimmune-modulating strategies should be highlighted. One promising approach is transient microglial suppression and replacement via CSF1R inhibitors: by pharmacologically depleting resident microglia and allowing repopulation, one can reset the inflammatory state beyond the acute window (Zhang X. et al., 2025). Such microglial repopulation has been shown to attenuate postoperative neuroinflammation and cognitive deficits in aged models even when initiated outside of narrow time constraints. Dynamic Biomarker Monitoring: Similarly, delays in detecting peripheral biomarkers limit their clinical utility. To bridge this gap, intraoperative or real-time monitoring techniques could be employed – for example, rapid point-of-care assays or real-time CSF sampling during surgery – to track surges in neuroinflammatory markers and guide timely interventions. Central Validation of MEG3: Finally, while we propose lncRNA MEG3 in circulating exosomes as a microglia-specific biomarker, it lacks direct CNS validation. We should discuss correlating MEG3-containing exosomes with established neuroinflammation measures such as TSPO-targeted PET imaging or microglial markers in CSF to ensure this peripheral readout truly reflects central microglial activation. Linking peripheral MEG3 signals with TSPO-PET (which visualizes glial activation, albeit not microglia-exclusive) or CSF inflammatory profiles would strengthen its translational relevance.
8 Conclusion and future outlook
Postoperative cognitive dysfunction has long eluded mechanistic clarity, often attributed to transient neuroinflammation without a consistent explanatory model for its persistence (Muscat et al., 2021; Vacas et al., 2013). The inadequacy of acute-response paradigms becomes particularly evident in patients who exhibit cognitive decline long after systemic markers of inflammation have normalized (Peng et al., 2023). This review proposes a shift in perspective: POCD is not merely a delayed consequence of resolved inflammation but reflects a durable, maladaptive state transition in microglia—one that embeds the memory of injury into the brain’s immune architecture.
At the center of this framework lies a self-reinforcing “inflammatory memory circuit” composed of three tightly coupled biological axes. Mitochondrial stress initiates the cascade through the release of mtROS and mtDNA, triggering NLRP3 and cGAS-STING pathways. Epigenetic mechanisms—including dysregulated histone deacetylation, DNA methylation, and lncRNA-mediated repression—subsequently engrave a transcriptional identity biased toward hyperresponsiveness. Finally, sustained M1-like polarization operationalizes this memory into chronic effector activity, further damaging neuronal mitochondria and feeding back into the cycle. The result is a metastable microglial phenotype with lowered activation thresholds and impaired capacity for resolution—one that no longer requires ongoing inflammatory input to sustain its pathogenic state.
This model explains several otherwise puzzling clinical observations. It accounts for delayed-onset or non-resolving POCD despite apparent systemic recovery (Muscat et al., 2021). It clarifies the limited efficacy of late anti-inflammatory interventions, which fail to reverse fate-locked glial states (Terrando et al., 2010). And it predicts the heightened vulnerability of aging microglia, whose compromised mitophagy and epigenetic rigidity render them more susceptible to transition into a memory-locked phenotype (Ji et al., 2024; Zhang L. et al., 2025).
More importantly, the model establishes clear inflection points for therapeutic and diagnostic intervention. Because the inflammatory memory state is defined by a transition—rather than a gradient—its trajectory is theoretically measurable and interceptable. Markers such as circulating mtDNA, elevated HDAC3 activity, and serum MEG3 levels could provide early signals of fate specification, enabling pre-symptomatic risk stratification (Pinti et al., 2021; Yang X. et al., 2022; Ye et al., 2023). If administered during this early window—within approximately 6 h post-insult—agents such as HDAC inhibitors, SIRT1 agonists, or lncRNA-targeting oligonucleotides may prevent the closure of the pathological circuit altogether (Ma et al., 2024; Shi et al., 2020; Yang et al., 2020).
Future research must now move toward validating this trajectory-based model in clinical and translational settings. Longitudinal single-cell studies, real-time chromatin profiling, and non-invasive biomarker development will be key to identifying patients before cognitive symptoms emerge. Equally critical is the development of scalable, fast-acting therapeutic platforms that match the narrow timing requirements of intervention.
Ultimately, POCD should no longer be conceptualized as a lingering trace of surgical stress. It is a circuit-encoded disorder of microglial fate, in which temporary perturbation becomes enduring dysfunction. Yet this model also opens space for therapeutic optimism: because state transitions are dynamic, they are not necessarily irreversible. If caught early enough, the trajectory toward cognitive decline may be diverted—restoring the brain’s innate immune balance and preserving cognitive integrity after surgery.
Author contributions
XW: Conceptualization, Formal Analysis, Methodology, Visualization, Writing – original draft. TH: Conceptualization, Formal Analysis, Methodology, Visualization, Writing – original draft. FH: Conceptualization, Formal Analysis, Methodology, Visualization, Writing – original draft. LL: Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review and editing.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. This work was supported by the Sichuan Science and Technology Program, grant number 2022YFS0632.
Acknowledgments
We appreciate Figdraw (www.figdraw.com) for its assistance in schematic diagram creation.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The authors declare that no Generative AI was used in the creation of this manuscript.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
Alam, A., Hana, Z., Jin, Z., Suen, K., and Ma, D. (2018). Surgery, neuroinflammation and cognitive impairment. EBioMedicine 37, 547–556. doi: 10.1016/j.ebiom.2018.10.021
Antón, O., and Traba, J. (2022). Measurement of cytosolic mitochondrial DNA after NLRP3 inflammasome activation. Methods Mol. Biol. 2459, 117–129. doi: 10.1007/978-1-0716-2144-8_12
Auzmendi-Iriarte, J., Moreno-Cugnon, L., Saenz-Antoñanzas, A., Grassi, D., de Pancorbo, M., Arevalo, M., et al. (2022). High levels of HDAC expression correlate with microglial aging. Expert Opin. Ther. Targets 26, 911–922. doi: 10.1080/14728222.2022.2158081
Baik, S., Kang, S., Lee, W., Choi, H., Chung, S., Kim, J., et al. (2019). A breakdown in metabolic reprogramming causes microglia dysfunction in Alzheimer’s disease. Cell. Metab. 30, 493–507.e6. doi: 10.1016/j.cmet.2019.06.005.
Bowen, C., Nguyen, H., Lin, Y., Bagchi, P., Natu, A., Espinosa-Garcia, C., et al. (2024). Proximity labeling proteomics reveals Kv1.3 potassium channel immune interactors in microglia. Mol. Cell. Proteomics 23:100809. doi: 10.1016/j.mcpro.2024.100809
Cai, L., Tu, L., Huang, X., Huang, J., Qiu, N., Xie, G., et al. (2020). LncRNA MALAT1 facilitates inflammasome activation via epigenetic suppression of Nrf2 in Parkinson’s disease. Mol. Brain 13:130. doi: 10.1186/s13041-020-00656-8
Chen, Y., Zhang, Y., Ye, G., Sheng, C., Kong, L., and Yuan, L. (2020). Knockdown of lncRNA PCAI protects against cognitive decline induced by hippocampal neuroinflammation via regulating SUZ12. Life Sci. 253:117626. doi: 10.1016/j.lfs.2020.117626
Cheng, S., Quintin, J., Cramer, R., Shepardson, K., Saeed, S., Kumar, V., et al. (2014). mTOR- and HIF-1α-mediated aerobic glycolysis as metabolic basis for trained immunity. Science 345:1250684. doi: 10.1126/science.1250684
Cheray, M., and Joseph, B. (2018). Epigenetics control microglia plasticity. Front. Cell. Neurosci. 12:243. doi: 10.3389/fncel.2018.00243
Ding, X., Liu, S., Tian, M., Zhang, W., Zhu, T., Li, D., et al. (2017). Activity-induced histone modifications govern Neurexin-1 mRNA splicing and memory preservation. Nat. Neurosci. 20, 690–699. doi: 10.1038/nn.4536
Du, J., Chen, C., and Chen, J. (2024). Cirsilineol improves anesthesia/surgery-induced postoperative cognitive dysfunction through attenuating oxidative stress and modulating microglia M1/M2 polarization. PeerJ 12:e18507. doi: 10.7717/peerj.18507
Femenía, T., Giménez-Cassina, A., Codeluppi, S., Fernández-Zafra, T., Katsu-Jiménez, Y., Terrando, N., et al. (2018). Disrupted neuroglial metabolic coupling after peripheral surgery. J. Neurosci. 38, 452–464. doi: 10.1523/JNEUROSCI.1797-17.2017
Feng, X., Valdearcos, M., Uchida, Y., Lutrin, D., Maze, M., and Koliwad, S. (2017). Microglia mediate postoperative hippocampal inflammation and cognitive decline in mice. JCI Insight 2:e91229. doi: 10.1172/jci.insight.91229
Fonken, L., Frank, M., Gaudet, A., D’Angelo, H., Daut, R., Hampson, E., et al. (2018). Neuroinflammatory priming to stress is differentially regulated in male and female rats. Brain Behav. Immun. 70, 257–267. doi: 10.1016/j.bbi.2018.03.005
Forsberg, A., Cervenka, S., Jonsson Fagerlund, M., Rasmussen, L., Zetterberg, H., Erlandsson Harris, H., et al. (2017). The immune response of the human brain to abdominal surgery. Ann. Neurol. 81, 572–582. doi: 10.1002/ana.24909
Gano, L., Donato, A., Pasha, H., Hearon, C., Sindler, A., and Seals, D. (2014). The SIRT1 activator SRT1720 reverses vascular endothelial dysfunction, excessive superoxide production, and inflammation with aging in mice. Am. J. Physiol. Heart Circ. Physiol. 307, H1754–H1763. doi: 10.1152/ajpheart.00377.2014
Gulen, M., Samson, N., Keller, A., Schwabenland, M., Liu, C., Glück, S., et al. (2023). cGAS-STING drives ageing-related inflammation and neurodegeneration. Nature 620, 374–380. doi: 10.1038/s41586-023-06373-1
Haley, M., Brough, D., Quintin, J., and Allan, S. (2019). Microglial priming as trained immunity in the brain. Neuroscience 405, 47–54. doi: 10.1016/j.neuroscience.2017.12.039
Hammond, T., Marsh, S., and Stevens, B. (2019). Immune signaling in neurodegeneration. Immunity 50, 955–974. doi: 10.1016/j.immuni.2019.03.016
Han, X., Xu, T., Fang, Q., Zhang, H., Yue, L., Hu, G., et al. (2021). Quercetin hinders microglial activation to alleviate neurotoxicity via the interplay between NLRP3 inflammasome and mitophagy. Redox Biol. 44:102010. doi: 10.1016/j.redox.2021.102010
Hasenson, S., Alkalay, E., Atrash, M., Boocholez, A., Gershbaum, J., Hochberg-Laufer, H., et al. (2022). The association of MEG3 lncRNA with nuclear speckles in living. Cells 11:1942. doi: 10.3390/cells11121942
He, B., Yu, H., Liu, S., Wan, H., Fu, S., Liu, S., et al. (2022). Mitochondrial cristae architecture protects against mtDNA release and inflammation. Cell Rep. 41:111774. doi: 10.1016/j.celrep.2022.111774
Henjum, K., Almdahl, I., Årskog, V., Minthon, L., Hansson, O., Fladby, T., et al. (2016). Cerebrospinal fluid soluble TREM2 in aging and Alzheimer’s disease. Alzheimers Res. Ther. 8:17. doi: 10.1186/s13195-016-0182-1
Hovens, I., Schoemaker, R., van der Zee, E., Heineman, E., Izaks, G., and van Leeuwen, B. (2012). Thinking through postoperative cognitive dysfunction: How to bridge the gap between clinical and pre-clinical perspectives. Brain Behav. Immun. 26, 1169–1179. doi: 10.1016/j.bbi.2012.06.004
Huang, M., Malovic, E., Ealy, A., Jin, H., Anantharam, V., Kanthasamy, A., et al. (2023). Microglial immune regulation by epigenetic reprogramming through histone H3K27 acetylation in neuroinflammation. Front. Immunol. 14:1052925. doi: 10.3389/fimmu.2023.1052925
Hulshof, L., van Nuijs, D., Hol, E., and Middeldorp, J. (2022). The role of astrocytes in synapse loss in Alzheimer’s disease: A systematic review. Front. Cell. Neurosci. 16:899251. doi: 10.3389/fncel.2022.899251
Ibarra, I., Ratnu, V., Gordillo, L., Hwang, I., Mariani, L., Weinand, K., et al. (2022). Comparative chromatin accessibility upon BDNF stimulation delineates neuronal regulatory elements. Mol. Syst. Biol. 18:e10473. doi: 10.15252/msb.202110473
Ji, Y., Ma, Y., Ma, Y., Wang, Y., Zhao, X., Jin, D., et al. (2024). SS-31 inhibits mtDNA-cGAS-STING signaling to improve POCD by activating mitophagy in aged mice. Inflamm. Res. 73, 641–654. doi: 10.1007/s00011-024-01860-1
Johnston, C., Siegler, E., Rice, M., Derry-Vick, H., Hootman, K., Zhu, Y., et al. (2022). Plasma cell-free mitochondrial DNA as a marker of geriatric syndromes in older adults with HIV. J. Acquir. Immune Defic. Syndr. 90, 456–462. doi: 10.1097/QAI.0000000000002993
Jung, H., Lee, D., You, H., Lee, M., Kim, H., Cheong, E., et al. (2023). LPS induces microglial activation and GABAergic synaptic deficits in the hippocampus accompanied by prolonged cognitive impairment. Sci. Rep. 13:6547. doi: 10.1038/s41598-023-32798-9
Kagitani-Shimono, K., Kato, H., Soeda, F., Iwatani, Y., Mukai, M., Ogawa, K., et al. (2023). Extension of microglial activation is associated with epilepsy and cognitive dysfunction in Tuberous sclerosis complex: A TSPO-PET study. Neuroimage Clin. 37:103288. doi: 10.1016/j.nicl.2022.103288
Keren-Shaul, H., Spinrad, A., Weiner, A., Matcovitch-Natan, O., Dvir-Szternfeld, R., Ulland, T., et al. (2017). A Unique microglia type associated with restricting development of Alzheimer’s disease. Cell 169, 1276–1290.e17. doi: 10.1016/j.cell.2017.05.018
Kong, L., Li, W., Chang, E., Wang, W., Shen, N., Xu, X., et al. (2022). mtDNA-STING axis mediates microglial polarization via IRF3/NF-κB signaling after ischemic stroke. Front. Immunol. 13:860977. doi: 10.3389/fimmu.2022.860977
Lauro, C., and Limatola, C. (2020). Metabolic reprograming of microglia in the regulation of the innate inflammatory response. Front. Immunol. 11:493. doi: 10.3389/fimmu.2020.00493
Li, C., Li, Q., Liu, S., Li, J., Yu, W., Li, Y., et al. (2022). sVCAM1 in the hippocampus contributes to postoperative cognitive dysfunction in mice by inducing microglial activation through the VLA-4 receptor. Mol. Neurobiol. 59, 5485–5503. doi: 10.1007/s12035-022-02924-1
Li, L., Acioglu, C., Heary, R., and Elkabes, S. (2021). Role of astroglial toll-like receptors (TLRs) in central nervous system infections, injury and neurodegenerative diseases. Brain Behav. Immun. 91, 740–755. doi: 10.1016/j.bbi.2020.10.007
Li, T., Luo, Y., Zhang, P., Guo, S., Sun, H., Yan, D., et al. (2020). LncRNA MEG3 regulates microglial polarization through KLF4 to affect cerebral ischemia-reperfusion injury. J. Appl. Physiol. 129, 1460–1467. doi: 10.1152/japplphysiol.00433.2020
Li, X., Zhang, Q., Ding, Y., Liu, Y., Zhao, D., Zhao, K., et al. (2016). Methyltransferase Dnmt3a upregulates HDAC9 to deacetylate the kinase TBK1 for activation of antiviral innate immunity. Nat. Immunol. 17, 806–815. doi: 10.1038/ni.3464
Li, Y., Fan, H., Han, X., Sun, J., Ni, M., Hou, X., et al. (2024). Long Non-coding RNA MALAT1 protects against spinal cord injury via suppressing microRNA-125b-5p mediated microglial M1 polarization, neuroinflammation, and neural apoptosis. Mol. Neurobiol. 61, 2136–2150. doi: 10.1007/s12035-023-03664-6
Lin, M., Liu, N., Qin, Z., and Wang, Y. (2022). Mitochondrial-derived damage-associated molecular patterns amplify neuroinflammation in neurodegenerative diseases. Acta Pharmacol. Sin. 43, 2439–2447. doi: 10.1038/s41401-022-00879-6
Lin, Q., Wang, J., Wang, X., Pan, C., Jin, S., Char, S., et al. (2024). Hippocampal HDAC6 promotes POCD by regulating NLRP3-induced microglia pyroptosis via HSP90/HSP70 in aged mice. Biochim. Biophys. Acta Mol. Basis Dis. 1870:167137. doi: 10.1016/j.bbadis.2024.167137
Liotti, A., Ferrara, A., Loffredo, S., Galdiero, M., Varricchi, G., Di Rella, F., et al. (2022). Epigenetics: An opportunity to shape innate and adaptive immune responses. Immunology 167, 451–470. doi: 10.1111/imm.13571
Liu, J., Wang, Y., Sun, H., Lei, D., Liu, J., Fei, Y., et al. (2025). Resveratrol ameliorates postoperative cognitive dysfunction in aged mice by regulating microglial polarization through CX3CL1/CX3CR1 signaling axis. Neurosci. Lett. 847:138089. doi: 10.1016/j.neulet.2024.138089
Liu, X., Wang, J., Jin, J., Hu, Q., Zhao, T., Wang, J., et al. (2024). S100A9 deletion in microglia/macrophages ameliorates brain injury through the STAT6/PPARγ pathway in ischemic stroke. CNS Neurosci. Ther. 30:e14881. doi: 10.1111/cns.14881
Liu, Y., Yang, W., Xue, J., Chen, J., Liu, S., Zhang, S., et al. (2023). Neuroinflammation: The central enabler of postoperative cognitive dysfunction. Biomed. Pharmacother. 167:115582. doi: 10.1016/j.biopha.2023.115582
Liu, Y., Zhang, B., Duan, R., and Liu, Y. (2024). Mitochondrial DNA leakage and cGas/STING pathway in microglia: Crosstalk between neuroinflammation and neurodegeneration. Neuroscience 548, 1–8. doi: 10.1016/j.neuroscience.2024.04.009
Lopes, C., Cunha, R., and Agostinho, P. (2021). Astrocytes and adenosine A2 A receptors: Active players in Alzheimer’s disease. Front. Neurosci. 15:666710. doi: 10.3389/fnins.2021.666710
Lv, H., Huang, J., Zhang, X., He, Z., Zhang, J., and Chen, W. (2024). Xenon ameliorates chronic post-surgical pain by regulating mitophagy in microglia and rats mediated by PINK1/Parkin pathway. PeerJ 12:e16855. doi: 10.7717/peerj.16855
Ma, Z., Niu, H., Qi, H., and Li, Y. (2024). Functional role of lncRNA MEG3 on pyroptosis through interacting with EZH2 and YTHDC1 in postoperative cognitive dysfunction. Brain Res. Bull. 217:111060. doi: 10.1016/j.brainresbull.2024.111060
Martins-Ferreira, R., Leal, B., Costa, P., and Ballestar, E. (2021). Microglial innate memory and epigenetic reprogramming in neurological disorders. Prog. Neurobiol. 200:101971. doi: 10.1016/j.pneurobio.2020.101971
Masuda, T., Sankowski, R., Staszewski, O., and Prinz, M. (2020). Microglia heterogeneity in the single-cell era. Cell. Rep. 30, 1271–1281. doi: 10.1016/j.celrep.2020.01.010
McGregor, B., Raihan, M., Brishti, A., Hur, J., and Porter, J. (2025). Deciphering motor dysfunction and microglial activation in mThy1- α-synuclein mice: A comprehensive study of behavioral, gene expression, and methylation changes. Front. Mol. Neurosci. 18:1544971. doi: 10.3389/fnmol.2025.1544971
Meng, B., Li, X., Lu, B., Liu, R., Yuan, H., Zhai, X., et al. (2019). The investigation of hippocampus-dependent cognitive decline induced by anesthesia/surgery in mice through integrated behavioral Z-scoring. Front. Behav. Neurosci. 13:282. doi: 10.3389/fnbeh.2019.00282
Meng, J., Ding, T., Chen, Y., Long, T., Xu, Q., Lian, W., et al. (2021). LncRNA-Meg3 promotes Nlrp3-mediated microglial inflammation by targeting miR-7a-5p. Int. Immunopharmacol. 90:107141. doi: 10.1016/j.intimp.2020.107141
Miao, J., Chen, L., Pan, X., Li, L., Zhao, B., and Lan, J. (2023). Microglial metabolic reprogramming: Emerging insights and therapeutic strategies in neurodegenerative diseases. Cell. Mol. Neurobiol. 43, 3191–3210. doi: 10.1007/s10571-023-01376-y
Miller, D., Lewis, S., Pritchard, M., Schofield-Robinson, O., Shelton, C., Alderson, P., et al. (2018). Intravenous versus inhalational maintenance of anaesthesia for postoperative cognitive outcomes in elderly people undergoing non-cardiac surgery. Cochrane Database Syst. Rev. 8:CD012317. doi: 10.1002/14651858.CD012317.pub2
Mitchell, S., Martin-Montalvo, A., Mercken, E., Palacios, H., Ward, T., Abulwerdi, G., et al. (2014). The SIRT1 activator SRT1720 extends lifespan and improves health of mice fed a standard diet. Cell. Rep. 6, 836–843. doi: 10.1016/j.celrep.2014.01.031
Monisha, K., Mahema, S., Chokkalingam, M., Ahmad, S., Emran, T., Prabu, P., et al. (2023). Elucidating the histone deacetylase gene expression signatures in peripheral blood mononuclear cells that correlate essential cardiac function and aid in classifying coronary artery disease through a logistic regression model. Biomedicines 11:2952. doi: 10.3390/biomedicines11112952
Muscat, S., and Barrientos, R. (2021). The perfect cytokine storm: How peripheral immune challenges impact memory function in aging. Brain Plast. 7, 47–60. doi: 10.3233/BPL-210127
Muscat, S., Deems, N., D’Angelo, H., Kitt, M., Grace, P., Andersen, N., et al. (2021). Postoperative cognitive dysfunction is made persistent with morphine treatment in aged rats. Neurobiol. Aging 98, 214–224. doi: 10.1016/j.neurobiolaging.2020.11.008
Nakanishi, H., Hayashi, Y., and Wu, Z. (2011). The role of microglial mtDNA damage in age-dependent prolonged LPS-induced sickness behavior. Neuron Glia Biol. 7, 17–23. doi: 10.1017/S1740925X1100010X
Neher, J., and Cunningham, C. (2019). Priming microglia for innate immune memory in the brain. Trends Immunol. 40, 358–374. doi: 10.1016/j.it.2019.02.001
Netea, M., Domínguez-Andrés, J., Barreiro, L., Chavakis, T., Divangahi, M., Fuchs, E., et al. (2020). Defining trained immunity and its role in health and disease. Nat. Rev. Immunol. 20, 375–388. doi: 10.1038/s41577-020-0285-6
Netea, M., Joosten, L., Latz, E., Mills, K., Natoli, G., Stunnenberg, H., et al. (2016). Trained immunity: A program of innate immune memory in health and disease. Science 352:aaf1098. doi: 10.1126/science.aaf1098
Pang, J., Yin, L., Jiang, W., Wang, H., Cheng, Q., Jiang, Z., et al. (2024). Sirt1-mediated deacetylation of PGC-1α alleviated hepatic steatosis in type 2 diabetes mellitus via improving mitochondrial fatty acid oxidation. Cell. Signal. 124:111478. doi: 10.1016/j.cellsig.2024.111478
Park, S., Kim, J., Ha, Y., Kim, K., Yi, S., and Koo, B. (2024). Preoperative mild cognitive impairment as a risk factor of postoperative cognitive dysfunction in elderly patients undergoing spine surgery. Front. Aging Neurosci. 16:1292942. doi: 10.3389/fnagi.2024.1292942
Peng, W., Lu, W., Jiang, X., Xiong, C., Chai, H., Cai, L., et al. (2023). Current progress on neuroinflammation-mediated postoperative cognitive dysfunction: An update. Curr. Mol. Med. 23, 1077–1086. doi: 10.2174/1566524023666221118140523
Penke, Z., Chagneau, C., and Laroche, S. (2011). Contribution of Egr1/zif268 to activity-dependent Arc/Arg3.1 transcription in the dentate gyrus and area CA1 of the hippocampus. Front. Behav. Neurosci. 5:48. doi: 10.3389/fnbeh.2011.00048
Pinti, M., Ferraro, D., and Nasi, M. (2021). Microglia activation: A role for mitochondrial DNA? Neural Regen. Res. 16, 2393–2394. doi: 10.4103/1673-5374.313034
Przanowski, P., Dabrowski, M., Ellert-Miklaszewska, A., Kloss, M., Mieczkowski, J., Kaza, B., et al. (2014). The signal transducers Stat1 and Stat3 and their novel target Jmjd3 drive the expression of inflammatory genes in microglia. J. Mol. Med. 92, 239–254. doi: 10.1007/s00109-013-1090-5
Punjasawadwong, Y., Chau-In, W., Laopaiboon, M., Punjasawadwong, S., and Pin-On, P. (2018). Processed electroencephalogram and evoked potential techniques for amelioration of postoperative delirium and cognitive dysfunction following non-cardiac and non-neurosurgical procedures in adults. Cochrane Database Syst. Rev. 5:CD011283. doi: 10.1002/14651858.CD011283.pub2
Qi, Z., Yu, Y., Su, Y., Cao, B., Shao, H., and Yang, J. (2023). M1-type microglia-derived extracellular vesicles overexpressing IL-1R1 promote postoperative cognitive dysfunction by regulating neuronal. Inflammation 46, 2254–2269. doi: 10.1007/s10753-023-01875-6
Qiu, L., Ji, M., Zhang, H., Yang, J., Sun, X., Tang, H., et al. (2016). NADPH oxidase 2-derived reactive oxygen species in the hippocampus might contribute to microglial activation in postoperative cognitive dysfunction in aged mice. Brain Behav. Immun. 51, 109–118. doi: 10.1016/j.bbi.2015.08.002
Qiu, Y., Mo, C., Xu, S., Chen, L., Ye, W., Kang, Y., et al. (2023). Research progress on perioperative blood-brain barrier damage and its potential mechanism. Front. Cell. Dev. Biol. 11:1174043. doi: 10.3389/fcell.2023.1174043
Rayaprolu, S., Gao, T., Xiao, H., Ramesha, S., Weinstock, L., Shah, J., et al. (2020). Flow-cytometric microglial sorting coupled with quantitative proteomics identifies moesin as a highly-abundant microglial protein with relevance to Alzheimer’s disease. Mol. Neurodegener. 15:28. doi: 10.1186/s13024-020-00377-5
Riedel, B., Browne, K., and Silbert, B. (2014). Cerebral protection: Inflammation, endothelial dysfunction, and postoperative cognitive dysfunction. Curr. Opin. Anaesthesiol. 27, 89–97. doi: 10.1097/ACO.0000000000000032
Rossano, S., Johnson, A., Smith, A., Ziaggi, G., Roetman, A., Guzman, D., et al. (2024). Microglia measured by TSPO PET are associated with Alzheimer’s disease pathology and mediate key steps in a disease progression model. Alzheimers Dement. 20, 2397–2407. doi: 10.1002/alz.13699
Saeed, S., Quintin, J., Kerstens, H., Rao, N., Aghajanirefah, A., Matarese, F., et al. (2014). Epigenetic programming of monocyte-to-macrophage differentiation and trained innate immunity. Science 345:1251086. doi: 10.1126/science.1251086
Saeki, K., Pan, R., Lee, E., Kurotaki, D., and Ozato, K. (2024). IRF8 defines the epigenetic landscape in postnatal microglia, thereby directing their transcriptome programs. Nat. Immunol. 25, 1928–1942. doi: 10.1038/s41590-024-01962-2
Sangineto, M., Ciarnelli, M., Cassano, T., Radesco, A., Moola, A., Bukke, V., et al. (2023). Metabolic reprogramming in inflammatory microglia indicates a potential way of targeting inflammation in Alzheimer’s disease. Redox Biol. 66:102846. doi: 10.1016/j.redox.2023.102846
Sathishkumar, C., Prabu, P., Balakumar, M., Lenin, R., Prabhu, D., Anjana, R., et al. (2016). Augmentation of histone deacetylase 3 (HDAC3) epigenetic signature at the interface of proinflammation and insulin resistance in patients with type 2 diabetes. Clin. Epigenet. 8:125. doi: 10.1186/s13148-016-0293-3
Schindler, S., Frank, M., Annis, J., Maier, S., and Klegeris, A. (2018). Pattern recognition receptors mediate pro-inflammatory effects of extracellular mitochondrial transcription factor A (TFAM). Mol. Cell. Neurosci. 89, 71–79. doi: 10.1016/j.mcn.2018.04.005
Shi, J., Zou, X., Jiang, K., and Wang, F. (2020). SIRT1 mediates improvement of cardiac surgery-induced postoperative cognitive dysfunction via the TLR4/NF-κB pathway. World J. Biol. Psychiatry 21, 757–765. doi: 10.1080/15622975.2019.1656820
Steinmetz, J., Christensen, K., Lund, T., Lohse, N., and Rasmussen, L. (2009). Long-term consequences of postoperative cognitive dysfunction. Anesthesiology 110, 548–555. doi: 10.1097/ALN.0b013e318195b569
Sun, W., Zhang, N., Liu, B., Yang, J., Loers, G., Siebert, H., et al. (2022). HDAC3 Inhibitor RGFP966 ameliorated neuroinflammation in the cuprizone-induced demyelinating mouse model and LPS-stimulated BV2 Cells by downregulating the P2X7R/STAT3/NF-κB65/NLRP3 activation. ACS Chem. Neurosci. 13, 2579–2598. doi: 10.1021/acschemneuro.1c00826
Sung, J., Kim, S., Kang, Y., Park, S., and Choi, H. (2024). SIRT1-dependent PGC-1α deacetylation by SRT1720 rescues progression of atherosclerosis by enhancing mitochondrial function. Biochim. Biophys. Acta Mol. Cell. Biol. Lipids 1869:159453. doi: 10.1016/j.bbalip.2024.159453
Suo, Z., Xiao, T., Qu, Y., Zheng, Y., Xu, W., Zhou, B., et al. (2025). Aged hippocampal single-cell atlas screening unveils disrupted neuroglial system in postoperative cognitive impairment. Aging Cell. 24:e14406. doi: 10.1111/acel.14406
Tan, X., Dai, H., Yao, J., Wang, J., Dai, Y., Zhang, T., et al. (2024). Hippocampal estrogens rescued the decline of synaptic plasticity after surgery and anesthesia by inhibiting microglia overactivation. Behav. Brain Res. 459:114794. doi: 10.1016/j.bbr.2023.114794
Tan, X., Qiu, L., and Sun, J. (2021). Research progress on the role of inflammatory mechanisms in the development of postoperative cognitive dysfunction. Biomed. Res. Int. 2021:3883204. doi: 10.1155/2021/3883204
Tan, X., Wang, J., Yao, J., Yuan, J., Dai, Y., Sun, M., et al. (2023). Microglia participate in postoperative cognitive dysfunction by mediating the loss of inhibitory synapse through the complement pathway. Neurosci. Lett. 796:137049. doi: 10.1016/j.neulet.2023.137049
Tanaka, T., Murakami, K., Bando, Y., and Yoshida, S. (2015). Interferon regulatory factor 7 participates in the M1-like microglial polarization switch. Glia 63, 595–610. doi: 10.1002/glia.22770
Tercan, H., Riksen, N., Joosten, L., Netea, M., and Bekkering, S. (2021). Trained immunity: Long-term adaptation in innate immune responses. Arterioscler. Thromb. Vasc. Biol. 41, 55–61. doi: 10.1161/ATVBAHA.120.314212
Terrando, N., and Akassoglou, K. (2022). Breaking barriers in postoperative delirium. Br. J. Anaesth. 129, 147–150. doi: 10.1016/j.bja.2022.05.004
Terrando, N., Monaco, C., Ma, D., Foxwell, B., Feldmann, M., and Maze, M. (2010). Tumor necrosis factor-alpha triggers a cytokine cascade yielding postoperative cognitive decline. Proc. Natl. Acad. Sci. U S A. 107, 20518–20522. doi: 10.1073/pnas.1014557107
Tian, Q., Zweibaum, D., Qian, Y., Oppong, R., Pilling, L., Casanova, F., et al. (2025). Mitochondrial DNA copy number associated dementia risk by somatic mutations and frailty. Geroscience 47, 825–835. doi: 10.1007/s11357-024-01355-1
Unlu, S., Kumar, A., Waterman, W., Tsukada, J., Wang, K., Galson, D., et al. (2007). Phosphorylation of IRF8 in a pre-associated complex with Spi-1/PU.1 and non-phosphorylated Stat1 is critical for LPS induction of the IL1B gene. Mol. Immunol. 44, 3364–3379. doi: 10.1016/j.molimm.2007.02.016
Vacas, S., Degos, V., Feng, X., and Maze, M. (2013). The neuroinflammatory response of postoperative cognitive decline. Br. Med. Bull. 106, 161–178. doi: 10.1093/bmb/ldt006
van Horssen, J., van Schaik, P., and Witte, M. (2019). Inflammation and mitochondrial dysfunction: A vicious circle in neurodegenerative disorders? Neurosci. Lett. 710:132931. doi: 10.1016/j.neulet.2017.06.050
Walters, B., and Zovkic, I. (2015). Building up and knocking down: An emerging role for epigenetics and proteasomal degradation in systems consolidation. Neuroscience 300, 39–52. doi: 10.1016/j.neuroscience.2015.05.005
Wan, X., Wen, J., Koo, S., Liang, L., and Garg, N. (2016). SIRT1-PGC1α-NFκB pathway of oxidative and inflammatory stress during Trypanosoma cruzi infection: Benefits of SIRT1-targeted therapy in improving heart function in chagas disease. PLoS Pathog. 12:e1005954. doi: 10.1371/journal.ppat.1005954
Wang, H., Ma, R., Fang, H., Xue, Z., and Liao, Q. (2015). Impaired spatial learning memory after isoflurane anesthesia or appendectomy in aged mice is associated with microglia activation. J. Cell. Death 8, 9–19. doi: 10.4137/JCD.S30596
Wang, J., Ge, J., Deng, B., Lin, H., Yang, W., Sheng, T., et al. (2025). Microglial activation visualized by [18F]-DPA714 PET Is a potential marker of severity and prognosis for anti-LGI1 encephalitis. Eur. J. Neurol. 32:e70107. doi: 10.1111/ene.70107
Wang, T., Huang, X., Sun, S., Wang, Y., Han, L., Zhang, T., et al. (2025). Recent advances in the mechanisms of postoperative neurocognitive dysfunction: A narrative review. Biomedicines 13:115. doi: 10.3390/biomedicines13010115
Wang, X., Hu, J., Xie, S., Li, W., Zhang, H., Huang, L., et al. (2024). Hidden role of microglia during neurodegenerative disorders and neurocritical care: A mitochondrial perspective. Int. Immunopharmacol. 142:113024. doi: 10.1016/j.intimp.2024.113024
Wendeln, A., Degenhardt, K., Kaurani, L., Gertig, M., Ulas, T., Jain, G., et al. (2018). Innate immune memory in the brain shapes neurological disease hallmarks. Nature 556, 332–338. doi: 10.1038/s41586-018-0023-4
Willemsen, J., Neuhoff, M., Hoyler, T., Noir, E., Tessier, C., Sarret, S., et al. (2021). TNF leads to mtDNA release and cGAS/STING-dependent interferon responses that support inflammatory arthritis. Cell. Rep. 37:109977. doi: 10.1016/j.celrep.2021.109977
Wu, J., Guo, Y., Li, W., Zhang, Z., Li, X., Zhang, Q., et al. (2023). Microglial priming induced by loss of Mef2C contributes to postoperative cognitive dysfunction in aged mice. Exp. Neurol. 365:114385. doi: 10.1016/j.expneurol.2023.114385
Wu, T., Liang, X., He, K., Wei, T., Wang, Y., Zou, L., et al. (2019). The role of NLRP3 inflammasome activation in the neuroinflammatory responses to Ag2Se quantum dots in microglia. Nanoscale 11, 20820–20836. doi: 10.1039/c9nr06778g
Wu, X., Ding, S., Wang, G., Zhang, W., and He, K. (2024). ZLN005 reduces neuroinflammation and improves mitochondrial function in mice with perioperative neurocognitive disorders. J. Inflamm Res. 17, 8135–8146. doi: 10.2147/JIR.S482051
Wu, X., Gong, L., Xie, L., Gu, W., Wang, X., Liu, Z., et al. (2021). NLRP3 deficiency protects against intermittent hypoxia-induced neuroinflammation and mitochondrial ROS by Promoting the PINK1-parkin pathway of mitophagy in a murine model of sleep apnea. Front. Immunol. 12:628168. doi: 10.3389/fimmu.2021.628168
Xu, H., Ye, W., and Shi, B. (2022). LncRNA MALAT1 regulates USP22 expression through EZH2-mediated H3K27me3 modification to accentuate sepsis-induced myocardial dysfunction. Cardiovasc. Toxicol. 22, 813–830. doi: 10.1007/s12012-022-09758-2
Yan, J., Luo, A., Gao, J., Tang, X., Zhao, Y., Zhou, B., et al. (2019). The role of SIRT1 in neuroinflammation and cognitive dysfunction in aged rats after anesthesia and surgery. Am. J. Transl. Res. 11, 1555–1568.
Yang, C., Bao, F., Zhong, J., Zhang, L., Deng, L., Sha, Q., et al. (2020). The inhibitory effects of class I histone deacetylases on hippocampal neuroinflammatory regulation in aging mice with postoperative cognitive dysfunction. Eur. Rev. Med. Pharmacol. Sci. 24, 10194–10202. doi: 10.26355/eurrev_202010_23240
Yang, L., Hao, J., Gao, Y., Yang, X., Shen, X., Wang, H., et al. (2022). HDAC3 of dorsal hippocampus induces postoperative cognitive dysfunction in aged mice. Behav. Brain Res. 433:114002. doi: 10.1016/j.bbr.2022.114002
Yang, N., Zhong, W., Sha, H., Zhang, C., Jin, L., Duan, J., et al. (2024). mtDNA-cGAS-STING axis-dependent NLRP3 inflammasome activation contributes to postoperative cognitive dysfunction induced by sevoflurane in mice. Int. J. Biol. Sci. 20, 1927–1946. doi: 10.7150/ijbs.91543
Yang, S., Qin, C., Hu, Z., Zhou, L., Yu, H., Chen, M., et al. (2021). Microglia reprogram metabolic profiles for phenotype and function changes in central nervous system. Neurobiol. Dis. 152:105290. doi: 10.1016/j.nbd.2021.105290
Yang, X., Huang, X., Li, M., Jiang, Y., and Zhang, H. (2022). Identification of individuals at risk for postoperative cognitive dysfunction (POCD). Ther. Adv. Neurol. Disord. 15:17562864221114356. doi: 10.1177/17562864221114356
Yang, X., Zhang, Y., Chen, Y., He, X., Qian, Y., Xu, S., et al. (2021). LncRNA HOXA-AS2 regulates microglial polarization via recruitment of PRC2 and epigenetic modification of PGC-1α expression. J. Neuroinflammation 18:197. doi: 10.1186/s12974-021-02267-z
Yao, M., Wang, X., Lin, H., Shu, H., Xu, Z., Tang, L., et al. (2024). LncRNA Tug1 regulates post-stroke microglial pyroptosis via PINK1/parkin-mediated mitophagy. Inflammation doi: 10.1007/s10753-024-02219-8 Online ahead of print.
Yao, Z., Cai, L., Zhao, A., Yang, L., Chen, Z., Zhang, Y., et al. (2023). Electroacupuncture alleviates neuroinflammation by regulating microglia polarization via STAT6/PPARγ in ischemic stroke rats. Neuroscience 532, 23–36. doi: 10.1016/j.neuroscience.2023.09.007
Ye, J., Jiang, Z., Chen, X., Liu, M., Li, J., and Liu, N. (2016). Electron transport chain inhibitors induce microglia activation through enhancing mitochondrial reactive oxygen species production. Exp. Cell. Res. 340, 315–326. doi: 10.1016/j.yexcr.2015.10.026
Ye, L., Cheng, X., Shi, Y., Liu, Z., Xiong, Y., and Huang, Y. (2023). Long non-coding RNA MEG3 alleviates postoperative cognitive dysfunction by suppressing inflammatory response and oxidative stress via has-miR-106a-5p/SIRT3. Neuroreport 34, 357–367. doi: 10.1097/WNR.0000000000001901
Ying, J., Deng, X., Du, R., Ding, Q., Tian, H., Lin, Y., et al. (2024). Mitochondrial modulation treating postoperative cognitive dysfunction neuroprotection via DRP1 inhibition by Mdivi1. Sci. Rep. 14:26155. doi: 10.1038/s41598-024-75548-1
Zeng, K., Long, J., Li, Y., and Hu, J. (2023). Preventing postoperative cognitive dysfunction using anesthetic drugs in elderly patients undergoing noncardiac surgery: A systematic review and meta-analysis. Int. J. Surg. 109, 21–31. doi: 10.1097/JS9.0000000000000001
Zeng, Z., Yu, K., Hu, W., Cheng, S., Gao, C., Liu, F., et al. (2021). SRT1720 pretreatment promotes mitochondrial biogenesis of aged human mesenchymal stem cells and improves their engraftment in postinfarct nonhuman primate hearts. Stem. Cells Dev. 30, 386–398. doi: 10.1089/scd.2020.0149
Zhang, J., He, H., Qiao, Y., Zhou, T., He, H., Yi, S., et al. (2020). Priming of microglia with IFN-γ impairs adult hippocampal neurogenesis and leads to depression-like behaviors and cognitive defects. Glia 68, 2674–2692. doi: 10.1002/glia.23878
Zhang, J., Liu, Y., Li, H., Hu, Y., Yu, S., Liu, Q., et al. (2022). Stellate ganglion block improves postoperative cognitive dysfunction in aged rats by SIRT1-mediated white matter lesion repair. Neurochem. Res. 47, 3838–3853. doi: 10.1007/s11064-022-03800-z
Zhang, L., Li, J., Li, C., Wu, Y., Liu, S., Li, Q., et al. (2025). Role of microglial mitophagy in alleviating postoperative cognitive dysfunction: A mechanistic study. Mol. Neurobiol. 62, 2376–2395. doi: 10.1007/s12035-024-04405-z
Zhang, M., and Yin, Y. (2023). Dual roles of anesthetics in postoperative cognitive dysfunction: Regulation of microglial activation through inflammatory signaling pathways. Front. Immunol. 14:1102312. doi: 10.3389/fimmu.2023.1102312
Zhang, M., Zhao, Q., Xia, M., Chen, J., Chen, Y., Cao, X., et al. (2020). The HDAC3 inhibitor RGFP966 ameliorated ischemic brain damage by downregulating the AIM2 inflammasome. FASEB J. 34, 648–662. doi: 10.1096/fj.201900394RRR
Zhang, Q., Li, Y., Liu, Y., Wang, X., Yang, Y., and Shi, L. (2024). The cGAS/STING signaling pathway is involved in sevoflurane induced neuronal necroptosis via regulating microglia M1 polarization. Cell. Signal. 119:111195. doi: 10.1016/j.cellsig.2024.111195
Zhang, Q., Raoof, M., Chen, Y., Sumi, Y., Sursal, T., Junger, W., et al. (2010). Circulating mitochondrial DAMPs cause inflammatory responses to injury. Nature 464, 104–107. doi: 10.1038/nature08780
Zhang, W., Wu, C., Ge, M., Yuan, X., Han, S., Zhao, F., et al. (2024). The PGC-1α/ERRα/ULK1 pathway contributes to Perioperative neurocognitive disorders by inducing mitochondrial dysfunction and activating NLRP3 inflammasome in aged mice. Neuropharmacology 260:110119. doi: 10.1016/j.neuropharm.2024.110119
Zhang, X., Cui, X., Sun, N., Wu, X., Pan, X., Wang, R., et al. (2025). Microglial repopulation alleviates surgery-induced neuroinflammation and cognitive impairment in a ZEB1-dependent manner. FASEB J. 39:e70440. doi: 10.1096/fj.202402492R
Zhang, X., Kracht, L., Lerario, A., Dubbelaar, M., Brouwer, N., Wesseling, E., et al. (2022). Epigenetic regulation of innate immune memory in microglia. J. Neuroinflammation 19:111. doi: 10.1186/s12974-022-02463-5
Zhang, Z., Yang, W., Wang, L., Zhu, C., Cui, S., Wang, T., et al. (2024). Unraveling the role and mechanism of mitochondria in postoperative cognitive dysfunction: A narrative review. J. Neuroinflammation 21:293. doi: 10.1186/s12974-024-03285-3
Zhao, Q., Wan, H., Pan, H., and Xu, Y. (2024). Postoperative cognitive dysfunction-current research progress. Front. Behav. Neurosci. 18:1328790. doi: 10.3389/fnbeh.2024.1328790
Zhao, R., Ying, M., Gu, S., Yin, W., Li, Y., Yuan, H., et al. (2019). Cysteinyl leukotriene receptor 2 is involved in inflammation and neuronal damage by mediating microglia M1/M2 polarization through NF-κB pathway. Neuroscience 422, 99–118. doi: 10.1016/j.neuroscience.2019.10.048
Zhou, N., Liu, K., Sun, Y., Cao, Y., and Yang, J. (2019). Transcriptional mechanism of IRF8 and PU.1 governs microglial activation in neurodegenerative condition. Protein Cell. 10, 87–103. doi: 10.1007/s13238-018-0599-3
Zhou, T., Yang, C., Wang, J., Fang, L., Xia, Q., and Liu, Y. (2024). Identification of serum exosomal lncRNAs and their potential regulation of characteristic genes of fibroblast-like synoviocytes in rheumatoid arthritis. Int. Immunopharmacol. 143:113382. doi: 10.1016/j.intimp.2024.113382
Zhou, X., Lv, M., Duan, Z., Liu, W., Yan, F., Liu, J., et al. (2024). CHTOP promotes microglia-mediated inflammation by regulating cell metabolism and inflammatory gene expression. J. Immunol. 212, 677–688. doi: 10.4049/jimmunol.2300572
Zhu, Y., Peng, K., Meng, X., and Ji, F. (2016). Attenuation of neuroinflammation by dexmedetomidine is associated with activation of a cholinergic anti-inflammatory pathway in a rat tibial fracture model. Brain Res. 1644, 1–8. doi: 10.1016/j.brainres.2016.04.074
Keywords: microglial inflammatory memory, epigenetic reprogramming, mitochondrial dysfunction, postoperative cognitive dysfunction, neuroimmune plasticity
Citation: Wu X, He T, He F and Liu L (2025) Is postoperative cognitive dysfunction a disease of microglial inflammatory memory? A state-transition model from metabolic stress to epigenetic lock-in. Front. Mol. Neurosci. 18:1648161. doi: 10.3389/fnmol.2025.1648161
Received: 16 June 2025; Accepted: 24 July 2025;
Published: 06 August 2025.
Edited by:
Eva Maria Jimenez-Mateos, Trinity College Dublin, IrelandCopyright © 2025 Wu, He, He and Liu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Li Liu, bml1bml1ZG9jdG9yQHN3bXUuZWR1LmNu
†These authors have contributed equally to this work and share first authorship