Lithium as a Potential Senostatic Agent in Central Nervous System Aging and Bipolar Disorder

1. Senescence and Brain Aging

1.1. Overview of Cellular Senescence

Cellular senescence is a stress response characterized by permanent cell cycle arrest and extensive phenotypic and molecular alterations. It is induced by factors such as DNA damage, telomere shortening, oxidative stress, and oncogenic signaling. In addition to replicative senescence, stress-induced premature senescence (SIPS) can arise independently of telomere length, typically through persistent activation of the DNA damage response (DDR).

At the molecular level, senescence is primarily governed by the p53/p21 and p16/Rb pathways, which enforce stable cell-cycle arrest. Sustained DDR signaling via ATM/ATR stabilizes p53 and induces p21, while p16 inhibits CDK4/6, maintaining retinoblastoma protein (Rb) activity and repressing E2F-dependent transcription. These processes are accompanied by chromatin remodeling, including heterochromatin formation and epigenetic changes, as well as alterations in the nuclear envelope, such as loss of lamin B1. Mitochondrial dysfunction, increased reactive oxygen species (ROS), impaired mitophagy, and activation of pro-survival pathways further stabilize the senescent phenotype [1,2].

Given its heterogeneity, senescence cannot be identified by a single marker and requires integrated assessment. Common features include SA-β-Gal activity, increased expression of p16 and p21, and persistent DNA damage foci (γ-H2A.X, 53BP1). Additional supportive indicators comprise lamin B1 loss, lipofuscin accumulation, elevated lysosomal content, HMGB1 relocalization, and selected surface markers (e.g., DPP4, uPAR) [3,4].

A defining functional feature of senescent cells is the senescence-associated secretory phenotype (SASP), which includes cytokines (e.g., IL-6, IL-8), chemokines, growth factors, matrix-remodeling enzymes, and extracellular vesicles. SASP is regulated by pathways such as NF-κB, C/EBPβ, mTOR, cGAS-STING, and JAK/STAT. While transient SASP can support tissue repair and immune-mediated clearance of damaged cells, chronic SASP activation contributes to sustained inflammation, tissue dysfunction, and the propagation of senescence in neighboring cells [5]. With age, senescent cells accumulate due to cumulative damage and reduced immune surveillance, impairing tissue function and contributing to cardiovascular disease, type 2 diabetes, osteoarthritis, pulmonary fibrosis, chronic kidney disease, and neurodegeneration [6].

1.2. Cellular Senescence in Neurons and Glial Cells

Cellular senescence in the central nervous system (CNS) is a heterogeneous, cell-type–specific process that extends beyond classical proliferative arrest and reshapes tissue homeostasis. Brain aging involves region-specific vulnerability, synaptic remodeling, reduced dendritic complexity, and impaired proteostatic and metabolic resilience, accompanied by the accumulation of senescent cells that promote neuroinflammation and cognitive decline. CNS senescence reflects both post-mitotic stress adaptation in neurons and canonical senescence in proliferative glial cells [7,8].

Neurons develop a senescence-like phenotype in response to chronic genotoxic, oxidative, and metabolic stress. Although post-mitotic, aged neurons exhibit persistent DNA damage signaling and increased p21 expression [9,10], independent of cell-cycle re-entry. This state is associated with mitochondrial dysfunction, impaired autophagy, lipofuscin accumulation, elevated ROS, defective mitophagy, and altered NAD⁺ metabolism. Neurons also display chromatin remodeling, lamin B1 loss, cytoplasmic chromatin fragments, and activation of cGAS–STING signaling. Functionally, they adopt a SASP-like secretory phenotype, producing IL-6, IL-8, CCL2, and matrix remodeling enzymes, thereby contributing to local inflammation [8,10]. However, markers such as SA-β-Gal and lipofuscin should be interpreted cautiously in post-mitotic cells. Overall, neuronal senescence-like states represent maladaptive stress responses rather than true cell-cycle arrest.

In contrast, glial cells—including astrocytes and microglia—undergo canonical senescence due to their retained proliferative capacity. This is characterized by increased expression of p16INK4a and p21, persistent DDR, epigenetic remodeling, and robust SASP. Senescent astrocytes exhibit impaired glutamate uptake, disrupted metabolic support, altered lipid homeostasis, and reduced synaptogenic capacity [11,12,13]. Senescent microglia exhibit dystrophic morphology, impaired phagocytosis, and heightened inflammatory responses, including increased secretion of IL-1β, IL-6, TNF-α, and complement components. These phenotypes are accelerated by pathological stimuli, including amyloid-β, tau, α-synuclein, irradiation, oxidative stress, and toxins [13,14]. Mechanistically, NF-κB, mTOR, and cGAS–STING integrate stress signals into sustained SASP output [15].

Oligodendrocyte progenitor cells (OPCs) are also vulnerable to stress-induced senescence. Senescent OPCs adopt a pro-inflammatory phenotype and impair neuronal excitability via CCL3/CCL5–CCR5 signaling [16]. Given the importance of white matter integrity, impaired remyelination may contribute to both neurodegenerative progression and psychiatric symptoms [17]. In multiple sclerosis, senescent-like microglia and progenitor cells in demyelinating lesions are associated with reduced remyelination and functional decline [18,19]. Senescent glia also exert non-cell-autonomous toxicity, inducing synaptic dysfunction and neuronal death [12].

Accumulation of senescent glia in vivo correlates with neurodegeneration, including synaptic loss, tau pathology, and cognitive decline [20,21]. In Alzheimer’s (AD) and Parkinson’s (PD), increased p16INK4a-positive glial cells, persistent DNA damage signaling, and SASP are observed in human and animal studies [11,12,22]. Senescent microglia and astrocytes show heightened pro-inflammatory activity and impaired homeostatic functions, including reduced phagocytosis and metabolic support [12,13]. Additionally, mitochondrial dysfunction and NLRP3 inflammasome activation amplify IL-1β signaling, disrupting synaptic plasticity and memory [23,24]. Reducing senescent cell burden alleviates neuroinflammation and improves neuronal function in experimental models [25]. At the systems level, persistent SASP disrupts blood–brain barrier integrity and stem cell niches and promotes paracrine senescence [12].

Overall, CNS senescence integrates DNA damage, mitochondrial dysfunction, epigenetic instability, and inflammation across neuronal and glial populations. Neurons exhibit senescence-like stress adaptation, whereas glia undergo canonical senescence. Their accumulation links brain aging with neurodegenerative disease through chronic inflammation and impaired regeneration.

1.3. Senescence as a Systems-Level Amplifier of Neuroinflammation

Cellular senescence remodels the CNS neuroimmune architecture, shifting transient inflammation into chronic, self-sustaining circuits. In the aging brain, the accumulation of senescent neurons, glia, progenitors, and vascular cells forms a persistent inflammatory niche characterized by sustained DAMP signaling, impaired immune resolution, and extracellular matrix remodeling [8,23,24]. This is driven by persistent activation of the DNA damage response and cytoplasmic chromatin accumulation, resulting from nuclear envelope instability and lamin B1 loss, thereby activating the cGAS–STING pathway and inducing type I interferons and an NF-κB–dependent SASP [26,27]. This creates a feed-forward loop in which inflammatory signaling propagates paracrine senescence and amplifies tissue-level dysfunction. Mitochondrial dysfunction further contributes to this axis by releasing mitochondrial DNA and increasing reactive oxygen species, both of which act as endogenous DAMPs [23].

Inflammasome activation represents an additional link between senescence and neuroinflammation. Mitochondrial stress and cytosolic DNA promote NLRP3 activation in microglia, leading to caspase-1–dependent maturation of IL-1β and IL-18 [23,24,28]. Persistent inflammasome signaling drives chronic innate immune activation. Senescence also alters microglial phenotypes by disrupting receptor signaling, including TREM2-dependent pathways that affect metabolism and phagocytosis [29]. In SASP-rich environments, microglia show reduced debris clearance, heightened inflammatory responses, and impaired synaptic surveillance [12,13,18]. Complement components (C1q, C3) further contribute to aberrant synaptic elimination and progressive synaptic loss [30].

At the neurovascular interface, endothelial and pericyte senescence promotes blood-brain-barrier (BBB) dysfunction, increasing permeability to peripheral immune mediators and amplifying CNS inflammation [31]. Systemic inflammaging, marked by elevated IL-6 and TNF-α, interacts bidirectionally with CNS senescence, thereby sustaining immune activation [23,24]. Senescence also impairs the resolution of neuroinflammation. Resistance to apoptosis and reduced immune clearance prolong inflammatory signaling and disrupt regeneration^8,24. Hippocampal neurogenesis declines in pro-inflammatory environments, and oligodendrocyte progenitor differentiation is impaired in demyelinating conditions ^16,19. In neurodegenerative diseases, senescence-driven inflammatory circuits synergize with proteinopathies to accelerate neuronal vulnerability and network dysfunction [8,32].

1.4. Cellular Senescence Linking Neurodegeneration and Psychiatric Disorders

Emerging evidence indicates that cellular senescence is a convergent process linking chronic neuroinflammation with circuit-level dysfunction in neurodegenerative and neuropsychiatric disorders [7,8,32]. Rather than an isolated phenotype, senescence disrupts neuronal–glial communication, synaptic stability, and immune–brain crosstalk, influencing cognitive and behavioral outcomes [8,10]. The role of cellular senescence is most evident in age-related neurodegeneration; however, accumulating evidence suggests that similar mechanisms may also contribute to psychiatric disorders, including schizophrenia, bipolar disorder (BD), and major depression [23,24,33]. This association is partly mediated by immune senescence, which links these conditions through chronic inflammation and dysregulated immune function [24,34].

For example, older adults with major depressive disorder exhibit elevated levels of SASP factors compared to cognitively normal, never-depressed controls [35]. Moreover, an exacerbated SASP has been associated with poorer cognitive performance [36], reduced response to antidepressant treatment [37], and worse physical health outcomes [38]. Major depressive disorder is also linked to elevated inflammatory markers, microglial activation, and NLRP3 signaling [39,40]. Cytokines such as IL-1β and TNF-α impair synaptic plasticity, reduce brain-derived neurotrophic factor (BDNF) levels, and disrupt neurogenesis [39,41]. Although direct evidence of CNS senescence is limited, peripheral immune senescence—such as T-cell exhaustion—has been reported and may affect the brain via cytokine signaling and metabolic pathways [34,42].

Senescence may contribute to psychiatric disease, acting as an amplifier that integrates genomic instability, mitochondrial dysfunction, inflammasome activation, complement signaling, and immune aging into a persistent inflammatory state [8,23,32]. In neurodegeneration, this accelerates neuronal loss [12,22], whereas in psychiatric disorders, it may influence vulnerability and disease course by affecting neuroplasticity [24,34].

1.5. Senescence as a Key Mechanism of Accelerated Aging in Psychiatric Disorders

Major psychiatric disorders, such as schizophrenia, BD, and major depression, are associated with reduced life expectancy by 10-20 years [43]. This is due to increased somatic morbidity and premature mortality (also by suicide). Accelerated aging may also partly explain these phenomena; all these illnesses exhibit features of accelerated aging. For example, according to a recent review, schizophrenia, BD, and major depression present accelerated epigenetic aging based on predefined DNA methylation patterns [44].

Of these three conditions, the data on accelerated aging in BD will be presented in more detail, as lithium, the drug discussed here, is primarily therapeutic in this condition. Apart from the epigenetic alterations mentioned above, the mechanisms in BD may include, among others, telomere attrition, mitochondrial dysfunction, and inflammation. Oxidative stress is considered a major driver of telomere shortening [45], a process frequently associated with BD and a potential marker of accelerated biological aging [46]. Research on telomere length in BD has shown that BD patients tend to have significantly shorter telomeres than healthy controls [47,48,49]. Interestingly, telomere shortening has also been observed in unaffected siblings of BD patients, suggesting that impaired telomere maintenance may either represent a vulnerability factor for BD or reflect the strong heritability of telomere dynamics [50]. Importantly, telomere shortening in BD occurs alongside other markers of accelerated biological and cellular aging. Patients with BD were also found to exhibit multiple hallmarks of premature aging, including increased brain-predicted age difference (brain-PAD), elevated SASP indices, and altered mitochondrial DNA copy number [45,49,51,52,53]. Notably, elevated SASP levels in older adults with BD were associated with greater depressive symptom severity and higher somatic disease burden, further supporting the concept of enhanced senescence-associated inflammatory activity in BD.

1.6. Senotherapeutic Strategies: From Systemic Targeting to CNS Applications

Cellular senescence, as a modifiable driver of aging, has led to “senotherapy,” which targets senescent cells by eliminating them (senolytics) or by suppressing harmful SASP effects (senomorphics), thereby reducing tissue dysfunction without inducing cell death [7]. Senolytics target senescent cells’ reliance on pro-survival pathways (e.g., BCL-2, PI3K/AKT). In animal models, their clearance reduces fibrosis, improves metabolism and function, and extends healthspan. Agents such as navitoclax, dasatinib plus quercetin, and FOXO4–p53 peptides show efficacy, supporting the role of senescent cells as active drivers of dysfunction [54,55,56]. In parallel, senomorphic strategies reduce pro-inflammatory SASP components without removing cells [57]. Targeting mTOR, NF-κB, JAK/STAT, or inflammasome pathways lowers inflammation and improves tissue function in aging models [58]. These approaches may offer advantages in contexts where complete ablation of senescent cells is impractical or potentially deleterious.

Translating senotherapeutic strategies to the CNS is challenging due to differences between senescence-like states in neurons and proliferative glia, as well as BBB limitations and the need for cell-specific targeting [8,9,10]. Nevertheless, reducing senescent glial burden or SASP-driven neuroinflammation has been shown to restore synaptic function, enhance remyelination, and improve cognition in preclinical models [12].

Within the CNS, senescence interacts with mitochondrial dysfunction, DNA damage, cGAS–STING, complement signaling, and NLRP3 activation [59]. These pathways offer targets to modulate neuroinflammation without cell loss, highlighting agents that affect autophagy, GSK-3, mitochondria, and inflammatory transcription [16,24,32]. This convergence raises the question of whether neuropsychiatric drugs can modulate senescence pathways.

Lithium, used for mood stabilization and having neuroprotective properties, targets these mechanisms and may influence senescence-driven neuroinflammation in the aging brain [60]. The suggestion of lithium's anti-aging effect was first put forward by Salarda et al. [61]. Here, we perform a systematic analysis of lithium as a potential senostatic agent.

2. Lithium in Psychiatry

2.1. From Historical Background to Current Clinical Applications

The psychiatric use of lithium dates back to the late 19th century, when Carl Lange reported its beneficial effects in periodic depression, based on the concept of “uric acid diathesis” [62]. Earlier, lithium salts were introduced for the treatment of gout. A major breakthrough came in 1949, when John Cade demonstrated the efficacy of lithium carbonate in the treatment of mania [63]. This finding marked the beginning of modern lithium therapy and stimulated further research into its role in mood disorders.

In the following decades, lithium’s therapeutic profile was progressively established. Its prophylactic effect in mood disorders was confirmed, leading to the recognition of long-term lithium treatment as a cornerstone in preventing both manic and depressive recurrences [64,65]. Its antidepressant properties were also demonstrated, and in later years, lithium became widely used as an augmentation strategy in treatment-resistant depression. At the same time, a distinct subgroup of patients, termed “excellent lithium responders” (ERs), was identified. These individuals achieve complete remission under lithium monotherapy and are typically characterized by a classical episodic course of illness and a family history of bipolar disorder [66,67].

Currently, the primary indication for lithium is the prevention of manic and depressive recurrences in mood disorders, particularly BD. This recommendation has been consistently confirmed in 21st-century meta-analyses demonstrating lithium’s superiority over placebo, especially in preventing manic episodes [68,69]. Lithium is also effective in the acute treatment of mania, with efficacy comparable to other mood stabilizers, although in more severe cases it is often combined with antipsychotic drugs to achieve faster symptom control [70]. In depressive episodes, lithium monotherapy is not typically considered a first-line option; however, it plays a significant role as an adjunctive treatment, enhancing the efficacy of antidepressants in treatment-resistant depression and, in some cases, producing rapid clinical improvement [71,72].

Importantly, lithium exerts a unique anti-suicidal effect, which appears to be at least partly independent of its mood-stabilizing properties and contributes to reduced mortality in patients with mood disorders [73]. Beyond its psychiatric applications, lithium has also demonstrated antiviral activity, including inhibition of herpesvirus and potential effects against respiratory viruses. In recent years, increasing attention has been given to its neuroprotective properties, with evidence suggesting a reduced risk of dementia and possible therapeutic effects in neurodegenerative disorders such as Alzheimer’s disease [74].

2.2. The Effect of Lithium on Morbidity, Mortality, and Brain Aging in Bipolar Disorder

Long-term lithium treatment in BD is associated with reduced morbidity and mortality. It has protective effects on the skeletal system, increasing bone mineral density and lowering the risk of osteoporosis, and on the cardiovascular system, including favorable effects on cardiac structure and a reduced risk of stroke [75,76,77,78]. Beyond this, lithium use has been linked to a lower risk of asthma and certain cancers, without increasing the risk of diabetes [79,80,81].

Large population studies confirm that patients with BD have markedly increased mortality, particularly due to suicide. Among mood stabilizers, lithium treatment is consistently associated with the greatest reduction in both suicide risk and overall mortality. These findings have been supported by large cohort analyses in Taiwan, which demonstrated significantly lower mortality rates in lithium-treated patients compared to those not receiving lithium [82]. The most recent analysis of 15,384 health records of individuals with BD in Catalonia, Spain, found that lithium exposure was associated with a significantly lower mortality risk (RR=0.69) compared with no exposure [83].

Recent research has also shown that chronic lithium treatment was associated with modulatory effects on aging biomarkers, particularly those related to telomeres [84], and with a reduction in brain-PAD [53].

2.3. Lithium’s Mechanisms of Mood Stabilization

Current evidence indicates that lithium exerts pleiotropic effects at the cellular and molecular levels. Its key mechanisms include modulation of electrolyte balance and membrane transport, regulation of second messenger systems—particularly the phosphatidylinositol (PI) pathway—and inhibition of glycogen synthase kinase-3β (GSK-3β), a central pathway. Lithium also influences neurotransmission and exhibits antiviral properties [85].

Recent molecular and in vitro studies further link lithium’s actions to major pathogenic processes implicated in bipolar disorder, including low-grade inflammation, mitochondrial dysfunction, circadian rhythm disturbances, gut–brain axis alterations, and purinergic signaling. Across these mechanisms, inhibition of PI signaling and GSK-3β consistently emerges as a dominant effect and is frequently associated with clinical response, particularly in the prophylactic treatment of mood disorders [85].

2.4. Neuroprotection and Anti-Dementia Effects of Lithium

The first evidence of lithium-induced increases in brain grey matter emerged in 2000 [86]. Subsequent studies have shown that both short- and long-term lithium treatment can increase the volume of key regions, including the prefrontal cortex, anterior cingulate cortex, and hippocampus [87]. These findings support a neuroprotective role of lithium, further reinforced by population studies linking its use to reduced risk and severity of dementia [88]. In patients with BD, long-term treatment with lithium was associated with a reduced rate of dementia, in contrast to such treatment with anticonvulsants, antidepressants, and antipsychotics [89].

Clinical studies indicate that lithium may have beneficial effects in amnestic mild cognitive impairment (aMCI) and AD. In patients with aMCI, lithium treatment has been associated with reduced levels of hyperphosphorylated tau protein, stabilization of cognitive and functional performance, and a potential decrease in the rate of conversion to AD. In individuals with AD, lithium has been linked to preservation of global cognitive function and slower cognitive decline compared to untreated groups. Importantly, long-term follow-up data suggest that these benefits may persist, with improved cognitive outcomes, including better verbal fluency, observed in lithium-treated patients [90,91,92,93].

Further meta-analytic evidence in MCI and AD provides a more nuanced picture. Some studies report a slowing of cognitive decline and modest cognitive benefits associated with lithium, potentially influenced by treatment duration [94,95]. Analyses in larger patient cohorts with AD have also suggested favorable efficacy and tolerability compared with certain disease-modifying therapies [96]. However, more recent evidence indicates that, despite a satisfactory safety profile, lithium does not consistently translate into measurable cognitive or functional improvement in AD dementia [97].

Emerging evidence also points to a relationship between lithium availability and dementia risk. Reduced endogenous lithium levels have been observed in the prefrontal cortex of individuals with cognitive impairment, while epidemiological studies show an inverse association between dementia prevalence and lithium concentration in drinking water [98,99,100]. These findings support a mechanistic explanation for lithium’s potential neuroprotective effects. Central to this is the inhibition of GSK-3β, leading to decreased tau phosphorylation and amyloid deposition. Additional mechanisms include modulation of autophagy, attenuation of inflammation and oxidative stress, and possible antiviral effects, particularly against herpesviruses implicated in Alzheimer’s disease pathogenesis [101,102,103].

Figure 1. Mechanisms of lithium action in bipolar disorder and neurodegeneration.

3. Evidence for Lithium as a Potential Senotherapeutic Agent

3.1. Effect on Senescence Biomarkers

There is mounting evidence that microdoses of lithium have beneficial effects on several symptoms of brain aging in patients with Alzheimer's disease (AD) and in AD animal models. These benefits include improvements in memory maintenance, decreased senile plaque density, and reduced neuronal cell loss [91,104]. Experiments on human iPSC-derived astrocytes in vitro revealed that lithium reduced expression of p16 and p21 as well as reduced SA-β-Gal expression in astrocytes subjected to amyloid β-induced senescence [105]. Interestingly, these effects appeared independent of GSK-3β inhibition, as no significant changes in Ser9-GSK-3β phosphorylation were observed following lithium treatment. This finding suggests that the anti-senescent activity of lithium may be dose-dependent and mediated through distinct molecular pathways. Previous studies showed that high concentrations of lithium chloride (20 mM) reduced GSK-3–dependent nuclear accumulation of p53 and p21 in WI-38 fibroblasts [106], whereas findings at a much lower concentration (2.5 μM) appear to involve alternative cytoprotective mechanisms that regulate cellular senescence independently of classical GSK-3β signaling.

Additional evidence for lithium's anti-senescent properties, including potentially non-canonical mechanisms of action beyond direct GSK-3β inhibition, comes from neuronal models of SIPS. In neuronal cells exposed to hydrogen peroxide–induced oxidative stress, lithium treatment reduced several hallmarks of senescence, including SA-β-Gal activity, lipofuscin content (according to Sudan Black B staining), and the formation of senescence-associated heterochromatin foci (SAHF). Lithium also attenuated cell-cycle arrest and modulated the expression of key senescence-related proteins, including p53, p21, p16INK4a, and SIRT1. Importantly, these effects were accompanied by alterations in the senescence-associated miR-34a/Sirt1/p53 signaling axis, suggesting that lithium-mediated suppression of neuronal senescence may involve epigenetic and post-transcriptional regulatory mechanisms [107]. The potential relevance of modulating miR-34a in this context is further supported by studies identifying miR-34a as a central pro-senescent and pro-inflammatory regulator in vascular aging. In vascular smooth muscle cells (VSMCs), increased miR-34a expression was associated with enhanced SASP activation, elevated IL-6 secretion, and promotion of osteoblastic differentiation and vascular calcification. Moreover, conditioned medium derived from miR-34a-overexpressing cells accelerated both senescence and calcification in neighboring cells, indicating that miR-34a may contribute to paracrine propagation of senescence-associated inflammatory signaling [108].

Further support for the antioxidant-mediated anti-senescent effects of lithium comes from studies conducted in healthy human volunteers, in which lithium administration modulated systemic oxidative stress parameters even in the absence of psychiatric or neurodegenerative pathology. Oxidative stress is considered a major driver of cellular senescence, particularly in neuronal tissues, where excessive ROS promotes lipid peroxidation, DNA damage, mitochondrial dysfunction, and activation of pro-senescent signaling pathways [109]. In healthy subjects receiving therapeutic lithium doses for approximately 2–4 weeks, lithium significantly reduced superoxide dismutase (SOD) activity and decreased the SOD/catalase (CAT) ratio, while CAT activity itself remained unchanged. The reduction in the SOD/CAT ratio was interpreted as reflecting lower intracellular hydrogen peroxide accumulation and, consequently, reduced oxidative stress burden. Importantly, these biochemical alterations occurred without significant changes in thiobarbituric acid-reactive substances (TBARS), suggesting that lithium primarily modulates upstream ROS-processing pathways before overt lipid peroxidation becomes detectable [110]. In an animal model of amphetamine-induced mania, lithium was shown to prevent and partially reverse lisdexamfetamine- and d-amphetamine-associated hyperlocomotion together with oxidative stress–related alterations in the rat brain, including glutathione depletion and increased lipid peroxidation, particularly within the prefrontal cortex and striatum [111].

Counteraction of oxidative stress–related cellular damage by lithium may also involve stimulation of DNA repair mechanisms, particularly by enhancing non-homologous end-joining (NHEJ) pathways through upregulation of DNA ligase IV. In ischemic retinal neurocytes, lithium pretreatment reduced the number of γ-H2AX foci, indicating attenuation of DNA double-strand breaks, while simultaneously increasing ligase IV expression. Mechanistically, this effect involved cooperative transcriptional regulation by nuclear respiratory factor 1 (Nrf-1) and phosphorylated CREB1 (P-CREB1), both of which bind to the ligase IV promoter and enhance its transcription under ischemic conditions [112]. Consistent with these observations, lithium also accelerated repair of irradiation-induced DNA double-strand breaks in hippocampal neurons by enhancing DNA-dependent protein kinase (DNA-PK)-mediated NHEJ activity. Lithium treatment increased DNA-PK threonine 2609 foci formation and enhanced DNA end-joining activity, coinciding with reduced γ-H2AX accumulation and decreased neuronal apoptosis. Importantly, these neuroprotective effects were markedly attenuated following genetic or pharmacological inhibition of DNA-PK, further supporting a central role of NHEJ activation in lithium-mediated genomic protection [113]. Consistent with CREB-dependent signaling, lithium has also been shown to enhance neuronal resilience by epigenetically regulating neuroprotective pathways. In rat hippocampal neurons, lithium increased viability against glutamate-induced cytotoxicity while upregulating BDNF, Bcl2, and Bcl-XL expression, and suppressing Bax, Bad, and caspase-3. Importantly, lithium selectively increased transcription of BDNF exon IV by hypomethylating its promoter region, suggesting that epigenetic modulation may further contribute to lithium-mediated protection against stress-related neuronal damage and senescence-associated processes [114].

Notably, the direct effects of lithium on senescence may be strongly cell-type dependent. In endothelial cells, lithium down-regulates p53 and induces a senescent-like phenotype, characterized by a marked increase in MMP-1 expression, a senescence-associated marker of extracellular matrix remodeling. Interestingly, this effect was not associated with canonical GSK-3β inhibition, further supporting the notion that lithium may regulate senescence through alternative signaling pathways. Moreover, MMP-1 upregulation was shown to actively reinforce endothelial senescence, as exogenous MMP-1 enhanced the number of SA-β-Gal–positive cells, whereas MMP-1 silencing attenuated lithium-induced senescence [115].

In contrast to these findings, other studies suggest that lithium may exert context-dependent anti-senescent and regenerative effects, particularly when combined with epigenetic and pro-regenerative compounds. A recent study investigating pharmacological partial reprogramming in replicative senescent endothelial cells demonstrated that a cocktail comprising lithium carbonate, valproic acid, and tranilast improved multiple hallmarks of endothelial dysfunction and senescence. Treatment with low-dose lithium carbonate (0.3 mM) in combination with valproic acid (0.5 mM) and tranilast (30 μM) reduced the proportion of SA-β-Gal–positive cells, enhanced proliferation, and restored migratory and angiogenic capacity to levels observed in non-senescent endothelial cells. Moreover, treated cells exhibited reduced expression of key senescence-associated regulators, including p14, p16, and p53, fewer DNA double-strand breaks, and attenuated SASP, as reflected by lower levels of TNF-α, IL-6, and IL-8. Importantly, these functional improvements occurred without loss of endothelial identity or adverse effects on non-senescent endothelial cells. Mechanistically, the beneficial effects were attributed to a transient induction of Yamanaka factor–mediated partial reprogramming, suggesting that lithium may support cellular rejuvenation under specific pharmacological and temporal conditions [116]. Notably, lithium was not administered as a monotherapy in this study, making it difficult to distinguish its direct contribution from synergistic interactions with valproic acid and tranilast. However, additional studies employing lithium in combination with other compounds further indicate that lithium-containing treatment regimens may modulate mechanisms closely associated with cellular aging and senescence, particularly oxidative stress, excitotoxicity, and macromolecular damage. In primary cultured rat cerebral cortical cells, chronic treatment with lithium and valproate at therapeutically relevant concentrations significantly attenuated glutamate-induced intracellular calcium accumulation, lipid peroxidation, protein oxidation, DNA fragmentation, and cell death. Importantly, the treatment did not alter basal oxidative stress parameters, suggesting that the protective effects emerged predominantly under pathological cellular stress conditions [117].

Oxidative stress is associated with telomere shortening [45], a potential marker of accelerated biological aging in BD [46]. Chronic lithium administration has emerged as a potential modulator of this aging-related biomarker. Several studies have demonstrated that lithium treatment is associated with longer telomeres in BD patients [118,119], suggesting a protective effect against telomere erosion and cellular senescence. It was also found that the duration of lithium administration correlates with telomerase reverse transcriptase (TERT) expression, the enzyme that contributes to telomere elongation [120].

Furthermore, lithium use was also associated with reduced brain-PAD values [53], indicating a broader potential effect on biological brain aging. Given the well-established relationship between oxidative stress, genomic instability, and telomere shortening, these findings support the hypothesis that lithium may exert anti-senescent effects, at least in part, by attenuating oxidative stress and preserving telomere integrity [83].

3.3. Effects on Senesce-Associated Secretory Phenotype (SASP)

Chronic stress dysregulation increases pro-inflammatory cytokines (IL-1β, IL-6, TNF-α), oxidative stress, and neurobiological damage [121], contributing to neuronal atrophy, synaptic dysfunction, and cell death, which lithium counteracts [122]. SASP links inflammation to cellular aging, and lithium may reduce pro-inflammatory signaling, thereby supporting immunomodulatory and senostatic effects.

Clinical and experimental data show that BD is characterized by elevated levels of TNF-α, IL-1β, and IL-6 during mood episodes, whereas lithium reduces these cytokines and promotes anti-inflammatory mediators such as IL-10 [123]. These effects result from lithium’s modulation of intracellular signaling pathways regulating immune responses [124]. In line with the role of inflammation in biological aging, recent findings indicate that older adults with BD exhibit a greater burden of cellular senescence, as reflected by higher SASP index scores than non-psychiatric controls. The SASP index, based on circulating proteins associated with senescence, was significantly elevated in BD, independent of age, sex, psychopathology, and physical health status. Moreover, higher SASP scores were associated with older age, greater depressive symptom severity, and increased somatic comorbidity, suggesting a link between senescence-related processes and clinical as well as systemic disease burden in BD [33]. Inhibition of GSK-3 is the probable mechanism of lithium’s anti-SASP effects [125]. The drug also activates the Wnt/β-catenin pathway and reduces microglial activation and neuro-inflammation [124].

In vitro, lithium reduces pro-inflammatory cytokine secretion in lipopolysaccharide (LPS)-stimulated cells, promoting an anti-inflammatory phenotype. It lowers IL-2, likely by inhibiting pro-caspase-1, and reduces IL-1β without affecting IL-6. [126]. Lithium also modulates the function and survival of T-lymphocytes derived from patients with BD [127,128].

In an AD mouse model, chronic lithium administration significantly reduced elevated levels of TNF-α, TNF receptor 1, and MARCO in the brain, indicating broad suppression of pro-inflammatory signaling [129]. Similarly, lithium attenuates astrocyte-mediated inflammation by inhibiting TLR4 expression and reducing LPS-induced IL-6 production, highlighting its role in modulating innate immune signaling pathways in the CNS [130]. Moreover, lithium reduces arachidonic acid production, a key component of the inflammatory cascade, and limits microglial activation following injury or stress [126].

Further evidence from autoimmune and inflammatory models underscores the importance of GSK-3 inhibition in mediating lithium’s effects. Lithium treatment suppresses pro-inflammatory cytokine production in experimental autoimmune encephalomyelitis (EAE), reduces leukocyte infiltration and demyelination [131]. Some studies report increased IFN-γ levels and decreased IL-4 levels in lithium-treated healthy models, suggesting a potential pro-inflammatory shift under non-pathological conditions [132]. Additionally, elevated TNF-α levels have been observed in certain cohorts of lithium-treated bipolar patients, indicating variability in clinical response [123].

3.4. Modulation of Microglial Activation and Phenotype

In response to inflammation, microglia become activated, producing proinflammatory cytokines, nitric oxide (NO), and prostaglandins that exacerbate neuronal damage. They polarize into proinflammatory M1 or neuroprotective M2 phenotypes; chronic stimuli (e.g., Aβ, phosphorylated tau) favor M1, sustaining inflammation and neurodegeneration. In vitro and in vivo studies demonstrate that lithium attenuates microglial activation and reduces the production of proinflammatory cytokines, especially under LPS-induced inflammatory conditions [133,134]. Lithium suppresses innate immune response by downregulating Toll-like receptor 4 (TLR4) via the phosphoinositide 3-kinase/protein kinase B (PI3K/Akt) pathway, leading to FoxO1 inactivation and reduced inflammatory signaling [135]. Beyond suppressing inflammatory mediators, lithium modulates glial dynamics by reducing astrocyte proliferation while enhancing microglial survival and proliferation in vitro [130]]. Moreover, lithium attenuates broader inflammatory cascades, including arachidonic acid metabolism, and reduces cytokine production in response to inflammatory stimuli [133]. In an animal model of AD, lithium treatment decreases neuroinflammatory markers, including IL-6, CXCL1, and microglial surface receptors (e.g., Trem2), and reduces microglial recruitment to affected neuronal regions [129].

In Parkinson’s disease, activated microglia and elevated proinflammatory cytokine levels are consistently observed in the substantia nigra, contributing to dopaminergic neuronal loss. Key signaling molecules such as GSK-3β further amplify this process by promoting NF-κB-mediated transcription of inflammatory mediators, thereby sustaining a detrimental feedback loop between neuronal injury and microglial activation [136,137]. Lithium’s known inhibitory effects on GSK-3β suggest an additional mechanism by which it may disrupt this cycle and mitigate neuroinflammation. Moreover, lithium’s immunomodulatory effects are evident in demyelinating conditions such as EAE, where treatment reduces leukocyte infiltration, demyelination, and microglial activation, with potential implications for remyelination processes. Although lithium also influences purinergic signaling, its effects on microglial inflammatory responses in this context appear more limited, indicating cell-type-specific mechanisms of action [138].

3.5. Effects on Innate Immune Signaling

Lithium exerts context-dependent effects, showing both pro- and anti-inflammatory properties. In epithelial cells, lithium rapidly activates NF-κB, leading to increased DNA binding and transcriptional activity, as well as activation of MAPK pathways (p38, ERK), even in the absence of classical inflammatory stimuli [138]. NF-κB activation downstream of pattern recognition receptors such as TLR4, TLR7, and RIG-I drives cytokine and chemokine production, positioning this pathway as a central mediator of innate immune response. Conversely, lithium can also inhibit NF-κB signaling by upregulating IκBα, preventing NF-κB nuclear translocation, and reducing inflammatory gene expression [139]. It further suppresses TLR4-mediated signaling by promoting p105 degradation, activating ERK, and reducing NF-κB activity, thereby decreasing pro-inflammatory responses and increasing macrophage apoptosis [140]. Crosstalk among TLRs, JAK/STAT, and other signaling pathways further shapes inflammatory outcomes, and dysregulation of this network contributes to chronic inflammation [141]. In the CNS, lithium influences TLR-related neuroinflammation and modulates immune cell activity and cytokine production, although clinical effects remain variable [142,143].

4. Conclusions and Future Perspectives

Figure 2. Proposed senostatic mechanisms of lithium in the aging brain and bipolar disorder.

The question of whether lithium may be considered a senostatic agent opens an important and conceptually novel perspective on its role in psychiatry and neurobiology. The available evidence suggests that lithium influences multiple processes closely associated with cellular senescence, including oxidative stress, chronic inflammation, mitochondrial dysfunction, genomic instability, and impaired cellular resilience. Importantly, many of these effects appear to converge not on elimination of senescent cells, but rather on attenuation of senescence-associated dysfunction and inflammatory signaling. In this regard, lithium fits more closely within the conceptual framework of senostatic rather than senolytic interventions.

At the same time, the current evidence remains insufficient to definitively classify lithium as a true senostatic compound. Existing studies are highly heterogeneous in their experimental models, lithium concentrations, exposure durations, and senescence markers used for evaluation. Moreover, senescence itself is increasingly recognized as a dynamic, context-dependent biological state rather than a single, definable phenotype. This complexity is particularly relevant in the CNS, where post-mitotic neurons frequently exhibit senescence-like adaptations that differ substantially from canonical senescence observed in proliferative cells. Consequently, interpretation of lithium’s effects on “senescence” requires considerable caution.

Importantly, lithium does not uniformly suppress senescence-associated phenotypes across all experimental settings. Some studies indicate potentially pro-senescent or maladaptive effects, particularly in endothelial cells, while others demonstrate protective, regenerative, or anti-inflammatory actions. These discrepancies likely reflect cell-type specificity and the dual nature of stress-response pathways, in which partial activation may be adaptive under some conditions but detrimental under others. It also remains unclear to what extent lithium acts directly on senescence-regulating mechanisms versus indirectly through broader stabilization of cellular homeostasis.

Nevertheless, lithium remains uniquely positioned in senotherapeutic research due to its established clinical use, long-term safety data, and pleiotropic biological actions. Future studies should move beyond isolated biomarkers and focus on integrated models of brain aging that combine molecular, cellular, neuroimaging, and clinical dimensions. Determining whether lithium truly modifies senescence-driven neuroprogression—or merely mitigates some of its downstream consequences—will be essential for defining its place within emerging senotherapeutic strategies for neuropsychiatric disorders.