The Role and Therapeutic Potential of the cGAS‐STING Signaling Pathway in Alzheimer's Disease

Aging is a primary risk factor for AD, correlating strongly with a marked increase in its prevalence; the majority of AD cases are late‐onset, typically manifesting after the age of 65 (Liu 2022). In the United States, estimates indicate a prevalence of AD as high as 11% in adults aged 65 and older, rising to approximately 32% in those aged 85 and older (Dintica and Yaffe 2022). This age‐related susceptibility is closely linked to progressive structural changes in the brain, including cerebral volume reduction, decreased synaptic density, ventricular dilatation, and the accumulation of pathological hallmarks such as amyloid plaques and neurofibrillary tangles (Hou et al. 2019). Cellular senescence is a key driver of physiological aging and is implicated in numerous age‐related diseases (Kumari and Jat 2021). During cellular senescence, the cGAS‐STING pathway drives disease progression through a defined sequence: damage signal recognition, inflammation triggering, and exacerbation of AD pathology. First, in aging mouse models, cGAS recognizes cytoplasmic chromatin fragments (CCFs) released by senescent cells. STING activation subsequently induces the senescence‐associated secretory phenotype (SASP), leading to the release of pro‐inflammatory factors such as IL‐1β and TNFα (Dou et al. 2017, Passarella et al. 2024). These SASP factors activate astrocytes, impairing their phagocytic capacity for Aβ and upregulating the expression of Aβ‐generating enzymes like BACE1, which collectively accelerates Aβ deposition (Kong et al. 2020). Second, mitochondrial dysfunction in microglia—the brain's resident immune cells—leads to mitochondrial DNA (mtDNA) release, which activates the cGAS‐STING pathway and drives their transformation into a disease‐associated microglia (DAM) phenotype (Gulen et al. 2023, Tan et al. 2022). This DAM phenotype is characterized by a significant reduction in the clearance of both Aβ and pathological tau, coupled with the release of pro‐inflammatory factors that promote excessive tau phosphorylation in neurons. This leads to the formation of neurofibrillary tangles (NFTs), neuronal death, and cognitive impairment (Lin et al. 2021). Furthermore, in neurodegenerative diseases, immune responses to DNA double‐strand breaks (DSBs), mediated by NF‐κB, are a major driver of brain aging. When neuronal genomic stability is compromised, cGAS expression is upregulated, and it acts as an upstream regulator to activate the NF‐κB pathway, a mechanism validated in AD animal models (Welch et al. 2022).

Obesity, a complex metabolic disorder characterized by excessive adipose tissue accumulation, exerts pathological effects that extend beyond metabolic dysregulation. Dysfunctional adipokine signaling and impaired adipocyte physiology are central to its pathogenesis and significantly elevate the risk of neurodegenerative diseases, including AD (Uddin et al. 2020). Notably, the association between obesity and AD is age‐dependent. Midlife obesity represents a significant risk factor that can trigger AD‐related neuropathological changes years before clinical dementia symptoms manifest (Lee 2011, Moser and Pike 2016). Conversely, overweight status in later life is associated with a reduced risk of AD, mild cognitive impairment (MCI), and vascular dementia (Doruk et al. 2010).

The expansion of adipose tissue in obesity worsens AD pathology through multiple mechanisms: On the one hand, it promotes a state of chronic low‐grade inflammation that, together with oxidative stress, contributes to mitochondrial membrane disruption and genomic instability, ultimately resulting in mitochondrial impairment. The release of mtDNA from injured mitochondria can initiate cGAS‐STING signaling, creating a self‐sustaining cycle that reinforces sterile inflammation (Wu et al. 2019, Bai et al. 2017). Pro‐inflammatory cytokines such as IFN‐β, produced by activated peripheral immune cells, can traverse the blood‐brain barrier to activate the cGAS‐STING pathway in brain microglia, thereby promoting their M1 polarization (Kong et al. 2022, Preeti et al. 2024). Subsequently, IL‐6 released by these M1 microglia disrupts neuronal synaptic structures and suppresses the expression of Aβ clearance proteins such as LRP1, thereby exacerbating Aβ accumulation (Ham et al. 2019, Yang et al. 2016). Conversely, in the context of obesity, a deficiency in the adipose tissue protein DsbA‐L and excessive fatty acid accumulation activate the cGAS‐STING pathway by increasing mitochondrial ROS and promoting ceramide synthesis, respectively (Cruz et al. 2018, Huang et al. 2025). Ceramides further exacerbate mtDNA release by altering mitochondrial permeability and concurrently inhibit central insulin signaling pathways. This leads to cerebral insulin resistance, which ultimately promotes tau phosphorylation and neuronal death (Xu et al. 2020).

Cardiovascular disease (CVD), a leading global cause of disability and mortality, encompasses a spectrum of conditions, including heart failure, myocardial infarction, and coronary artery disease. Its high incidence and mortality rates impose a substantial burden on both public health and socioeconomic systems (Zhang et al. 2023). Epidemiological studies have established a strong pathological link between CVD and AD, indicating that CVD patients face a significantly elevated risk of developing AD compared to the general population (Tublin et al. 2019, Wu et al. 2016). CVD contributes to the progression of AD pathology by activating the cGAS‐STING pathway in the brain through vasogenic injury. Notably, in atherosclerosis, damaged endothelial cells and macrophages release DNA fragments and other damage‐associated molecular patterns (DAMPs), which activate the cGAS‐STING pathway and upregulate IRF3 expression (Rech and Rainer 2021, Liu et al. 2017). A similar activation of the cGAS‐STING‐TBK1‐IRF3 signaling axis occurs in cardiomyocytes during myocardial ischemia‐reperfusion injury (Zhai et al. 2024). Activated IRF3 drives the release of proinflammatory cytokines, including IFN‐β, TNF‐α, and IL‐6. These cytokines induce a proinflammatory phenotype in glial cells, exacerbating neuroinflammation and synaptic damage. Concurrently, IRF3 directly regulates the expression of AD‐associated genes such as Apoe and downstream targets like ZBP1, thereby promoting amyloid plaque‐related pathology (Joshi et al. 2024).

Diabetes, an endocrine and metabolic disorder characterized by chronic hyperglycemia, is among the fastest‐growing chronic diseases worldwide. The global number of adult diabetic patients is projected to reach 693 million by 2045 (Cole and Florez 2020). Epidemiological studies indicate that elderly diabetic patients are more susceptible to widespread vascular lesions and face a significantly elevated risk of developing AD compared to non‐diabetic individuals (Biessels et al. 2006, Ahtiluoto et al. 2010). Diabetes promotes AD pathology by activating the cGAS‐STING pathway via mitochondrial damage. Initially, diabetes‐associated metabolic stress induces mitochondrial dysfunction in the brain, resulting in the leakage of mtDNA into the cytoplasm. This cytoplasmic mtDNA acts as a DAMP, activating the cGAS‐STING innate immune pathway and driving a robust type I interferon response, notably IFN‐β production. This cytokine response activates microglia and promotes their polarization toward a pro‐inflammatory (M1) phenotype, which triggers chronic neuroinflammation. These events ultimately lead to synaptic protein loss and the development of tau pathology, thereby impairing cognitive function (Preeti et al. 2024). Separately, free fatty acid‐induced mitochondrial oxidative damage also causes mtDNA leakage and cGAS‐STING pathway activation. In peripheral systems, this activation promotes proinflammatory responses and cardiomyocyte pyroptosis through NLRP3 inflammasome‐dependent mechanisms (Yan et al. 2022). NLRP3 inflammasome‐mediated pyroptosis can indirectly exacerbate neuroinflammation and directly activate tau‐phosphorylating kinases such as GSK‐3β. This increases the formation of NFTs, leading to neuronal death and cognitive decline (Xia et al. 2022, Zhu et al. 2021) (Figure 2) (Table 1).

Neutrophil extracellular traps (NETs) are web‐like complexes released by stimulated neutrophils, primarily consisting of DNA, histones, and antimicrobial proteins derived from granules (Zuo et al. 2020). The migration of neutrophils into the central nervous system (CNS) is a central pathological hallmark of numerous neuroinflammatory disorders (Rossi et al. 2011). Importantly, such neutrophil infiltration is strongly implicated in the progression of AD and the accompanying cognitive decline (Zenaro et al. 2015). Clinical investigations have revealed that circulating NET markers are substantially higher in the plasma and serum of AD patients relative to cognitively normal older adults (Kretzschmar et al. 2021). In animal models of AD, reducing neutrophil levels—either by depletion or by blocking neutrophil migration through LFA‐1 inhibition—significantly improves memory performance and lessens neuropathology (Pietronigro et al. 2017). Together, these observations underscore the active role of neutrophils and NETs in mediating tissue injury in AD. The dsDNA within NETs can be detected by the cytosolic DNA sensor cGAS, resulting in STING activation and the triggering of downstream signaling. NETs have been demonstrated to stimulate the cGAS‐STING pathway, enhancing the secretion of type I IFNs and proinflammatory cytokines, which amplifies immune activation. In contrast, impairment of key enzymes such as deoxyribonuclease I (DNase I) or peptidylarginine deiminase 4 (PAD4), which are essential for NET generation and preservation, suppresses cGAS‐STING signaling (Wang et al. 2021). Additionally, studies in a mouse model of traumatic brain injury (TBI) revealed that NETs induce ER stress through STING activation, aggravating neuroinflammation and programmed neuronal death (Shi et al. 2023). In conclusion, NETs serve as a major endogenous source of ligands that trigger cGAS‐STING signaling under neuroinflammatory conditions.

mtDNA is a small, double‐stranded circular molecule enclosed within the double‐membrane system formed by the outer and inner mitochondrial membranes (OMM and IMM). The release of mtDNA into the cytosol has emerged as a central event in innate immune activation, triggering cGAS‐STING signal transduction (Kim et al. 2023). Under physiological conditions, mtDNA is strictly sequestered within the mitochondrial matrix by the OMM and IMM, preventing its recognition by cytosolic pattern recognition receptors (Li et al. 2024). However, upon cellular stress or mitochondrial dysfunction, mtDNA can translocate into the cytosol or extracellular space through mechanisms involving membrane damage or active transport. The mitochondrial protein transcription factor A (TFAM) facilitates cGAS dimerization by inducing DNA bending, thereby enhancing cGAS sensitivity to long DNA strands such as mtDNA. This TFAM‐mtDNA complex potently activates the cGAS‐STING pathway, inducing the production of proinflammatory cytokines and IFNs (Li et al. 2022, Lei et al. 2023). Mitochondrial dysfunction is a core driver of AD pathogenesis. For instance, the accumulation of APP‐CTF triggers mitochondrial structural alterations, overproduction of mitochondrial reactive oxygen species (mtROS), and impaired mitophagy in both AD mouse models and human brains (Vaillant‐Beuchot et al. 2021). During aging, exacerbated mtDNA oxidative damage acts synergistically with the secretion of SASP factors to promote inflammation (Li et al. 2024). Consequently, abnormal mtDNA dynamics during AD progression warrant particular attention. First, Hou et al. (Hou et al. 2021) observed a 3 to 6‐fold increase in cytosolic mtDNA enrichment in cellular models of AD compared to wild‐type controls, indicating a loss of mitochondrial membrane integrity and subsequent mtDNA leakage. Second, deficiency in aldehyde dehydrogenase 2 (ALDH2) exacerbates mtDNA damage and its cytosolic accumulation, leading to cGAS‐STING pathway activation and impaired mitophagy (Wang et al. 2020). Further mechanistic studies revealed that tau protein localizing to mitochondria can trigger mtDNA leakage and cGAS activation in microglia. This process inhibits the myocyte enhancer factor 2C (MEF2C)‐mediated transcriptional network in neurons, ultimately impairing cognitive function (Udeochu et al. 2023). Additionally, phospholipase D3 (PLD3) is a key lysosomal 5′–3′ exonuclease that primarily degrades mtDNA. In PLD3‐deficient cells, impaired lysosomal function leads to mtDNA leakage into the cytosol, activating the cGAS‐STING pathway and inducing the accumulation of APP‐CTF and cholesterol (Van Acker et al. 2023). In summary, the mtDNA‐cGAS‐STING signaling axis constitutes a crucial mechanism in the pathogenesis of AD.

The ER is the primary site for protein biosynthesis and processing, responsible for critical functions including the post‐translational modification, folding, and assembly of newly synthesized proteins (Nagar et al. 2023). The sustained accumulation of misfolded proteins induces ER stress, triggering the adaptive UPR to restore proteostasis by facilitating the degradation of aberrant proteins (Chen et al. 2023). In AD, the persistent accumulation of Aβ and p‐tau disrupts ER calcium homeostasis, impairs protein folding, and consequently induces profound ER stress. Studies have demonstrated that reducing pathological tau phosphorylation can alleviate ER stress‐mediated neurotoxicity (Park et al. 2009, Song et al. 2024). Furthermore, Aβ can increase intracellular calcium influx by modulating calcium channels, leading to disrupted cytoplasmic calcium homeostasis, ER stress, and subsequent memory impairment (Ghanbari‐Maman et al. 2019). Collectively, these findings establish ER stress as a critical contributor to AD pathogenesis.

A growing body of evidence indicates a robust interplay between ER stress and STING signaling activation. On one hand, ER stress can act as a potent inducer of STING signaling. For example, the ER stress inhibitor 4‐phenylbutyric acid (4‐PBA) has been shown to suppress STING‐IRF3 pathway activation (Li et al. 2023). Pharmacological studies indicate that UPR inducers can activate STING signaling and TBK1 phosphorylation, leading to IRF3 phosphorylation and its nuclear translocation (Liu et al. 2012). Furthermore, STING signaling activation is functionally linked to ER‐regulated Ca²⁺ homeostasis (Kwon et al. 2018). The stromal interaction molecule 1 (STIM1) binds to the calcium channel protein Orai1 to form a Ca²⁺ release‐activated Ca²⁺ (CRAC) channel, promoting extracellular Ca²⁺ influx that facilitates STING signaling activation (Srikanth et al. 2019). Conversely, activation of the cGAS‐STING signaling pathway can also induce ER stress. Studies confirm that cGAS‐STING activation exacerbates ER stress‐induced damage, while inhibition of cGAS with RU.521 alleviates ER stress (Huang et al. 2022). Moreover, cGAMP‐activated STING can interact with protein kinase R‐like ER kinase (PERK), a key UPR sensor that phosphorylates eukaryotic initiation factor 2α (eIF2α) to regulate ER homeostasis (Zhang et al. 2022, Wan et al. 2024). Critically, PERK activation promotes Aβ production and plaque deposition in AD models by upregulating the expression of activating transcription factor 4 (ATF4) and β‐site APP‐cleaving enzyme 1 (BACE1) (Hugon et al. 2017). Conversely, PERK knockout significantly ameliorates AD‐related pathology (Ma et al. 2013).

In mammalian cells, genomic DNA is confined within the nuclear compartment. When cells die, nuclear DNA can be released into the extracellular space. Under normal physiological conditions, this extracellular DNA is efficiently removed by serum deoxyribonuclease I (DNase I) or, after phagocytic uptake, is broken down within lysosomes via DNase II, thus avoiding unintended immune stimulation (Kawane et al. 2001). Strong evidence indicates that self‐DNA released during cell death serves as a major trigger for the cGAS‐STING signaling pathway. For instance, in mouse models of myocardial infarction, widespread death of heart muscle cells liberates large amounts of self‐DNA, which in turn stimulates cGAS‐STING signaling, prompting interferon production and enhancing inflammation (King et al. 2017). In AD, multiple pathological drivers—such as Aβ deposition (Naderi et al. 2023), hyperphosphorylated tau (Thal and Tomé 2022), glutamate excitotoxicity (Wang and Reddy 2017), and neuroinflammation (Choi et al. 2023)—cause extensive neuronal loss in susceptible regions, including the hippocampus and cerebral cortex. As a result, nuclear DNA from dying neurons activates the cGAS‐STING pathway. Notably, brains of Alzheimer's patients contain thirty times more cells with fragmented DNA than those of healthy individuals (Lassmann et al. 1995), and this accumulation of damaged DNA strongly potentiates cGAS‐STING signaling and downstream molecular effectors. In senescent cells, reduced DNase expression leads to nuclear DNA buildup, provoking aberrant cGAS activation that drives the SASP via IFN‐β induction (Takahashi et al. 2018). Similarly, loss of DNase II function results in undegraded DNA escaping from lysosomes, which activates STING‐dependent signaling (Ahn et al. 2012). Relevant to AD, microglial deficiency in DNase II enhances cGAS‐STING activation, accelerates Aβ and tau pathology, and worsens cognitive impairment in mouse models (Li et al. 2025). Moreover, cGAS can detect nuclear DNA that enters the cytosol due to retrotransposon activity. Long Interspersed Nuclear Element‐1 (LINE‐1), for example, reverse transcribes its RNA into cDNA that activates the cGAS‐STING pathway. Inhibiting LINE‐1 reverse transcription lowers cytoplasmic DNA levels and thereby dampens cGAS‐STING activity (Gamdzyk et al. 2020). Cells also employ intrinsic mechanisms to restrain cGAS activation by self‐DNA. One such mechanism involves barrier‐to‐autointegration factor 1 (BAF), which sequesters nuclear DNA upon loss of nuclear envelope integrity, preventing cGAS recognition (Guey et al. 2020).

Inhibition of cGAS activity represents a core therapeutic strategy for suppressing the cGAS‐STING pathway. Developed cGAS inhibitors primarily fall into two mechanistic categories: those that act as competitive antagonists by binding directly to the enzyme's active site, and those that function allosterically by disrupting the interaction between cGAS and dsDNA (Zhou et al. 2023).

Early studies identified that antimalarial drugs, including hydroxychloroquine and quinacrine, can bind to dsDNA and prevent its interaction with cGAS (An et al. 2017). I The inhibitory oligodeoxynucleotide A151 suppresses cGAS catalytic activity by binding to its dsDNA‐binding domain, as demonstrated in THP‐1 human monocytes (Steinhagen et al. 2018). High‐throughput screening has further identified RU.365 and its benzothiazole analog RU.332 as compounds with significant inhibitory efficacy against murine cGAS (Vincent et al. 2017). Additionally, the acetyl‐donor aspirin can reduce cGAS enzymatic activity by promoting its acetylation (Dai et al. 2019).

In AD models, cGAS inhibitors have demonstrated efficacy. For instance, Udeochu et al. (Udeochu et al. 2023) reported that the small‐molecule inhibitors TDI‐6570 (a murine cGAS inhibitor) and TDI‐8246 (a human cGAS inhibitor) phenocopied the beneficial effects of cGAS knockout in both in vitro cultures and PS19 tauopathy mice. RU.521, an inhibitor that directly targets the cGAS catalytic site, blocks cGAMP synthesis, thereby inhibiting STING activation and downstream inflammatory cascades and ultimately alleviating chronic inflammation (Wiser et al. 2020, Shao et al. 2023). Animal studies have confirmed that RU.521 treatment reduces IFN‐β1 expression in mouse macrophages and significantly ameliorates Aβ pathology in 5xFAD mice (Xie et al. 2023). (Table 2)

As a core regulatory factor of innate immunity, STING is a promising therapeutic target, and its inhibitors have demonstrated significant potential for treating inflammatory diseases. To date, two primary strategies have been employed to develop STING inhibitors. The first strategy involves designing molecules that occupy the cyclic dinucleotide (CDN) binding site, acting as competitive antagonists. The second strategy focuses on identifying compounds that bind to cysteine residues (Cys88 or Cys91) near the transmembrane domain. These inhibitors disrupt STING palmitoylation, thereby reducing the recruitment of downstream signaling factors and preventing STING activation (Decout et al. 2021).

A major focus of STING inhibitor research targets its palmitoylation. Many compounds covalently bind to Cys88 or Cys91 in the transmembrane region to disrupt this essential post‐translational modification. Additionally, inhibitors targeting other sites on STING have also been identified. For example, the cyclic peptide Astin C, isolated from the aster plant, binds specifically to the CDN‐binding pocket in the STING C‐terminal domain. This binding prevents IRF3 recruitment while maintaining the STING‐TBK1 interaction, thereby inhibiting downstream innate immune responses (Li et al. 2018). Similarly, nitrofuran derivatives (e.g., compounds 20–23) inhibit STING palmitoylation and potently reduce STING‐mediated IFN‐β reporter activity (Haag et al. 2018). Notably, the antiviral drug remdesivir has been shown to inhibit STING expression, regulating lipid metabolism and inflammation in hepatocytes and alleviating non‐alcoholic fatty liver disease (NAFLD) in high‐fat diet models (Li and Su 2020).

The irreversible covalent inhibitor C‐176, which binds to the Cys91 site, significantly reduces serum type IFN‐I and IL‐6 levels induced by STING agonists (Li et al. 2024). In microglial cells treated with Aβ_25‐35, C‐176—both as a monotherapy and in combination with the cGAS inhibitor RU.521—demonstrated potent anti‐neuroinflammatory effects (Haag et al. 2018, Wang et al. 2023). Mass spectrometry confirmed that H‐151 forms a covalent bond with Cys91, effectively inhibiting the type I interferon response, TBK1 phosphorylation, and human STING (hsSTING) palmitoylation. Both in vitro and in vivo studies have shown that H‐151 inhibits IFN‐β production and significantly suppresses the cytokine response to STING agonists (Haag et al. 2018). H‐151 treatment significantly ameliorates Aβ pathology in 5xFAD transgenic mice (Xie et al. 2023, Sharma et al. 2025). Furthermore, combination therapy with H‐151 and the soluble epoxide hydrolase inhibitor TPPU produces a synergistic anti‐AD effect, enhancing early‐stage Aβ phagocytosis and promoting late‐stage Aβ degradation (Fiala et al. 2025). Collectively, this research provides a strong rationale for the clinical translation of STING inhibitors. (Table 3)

As a non‐canonical IKK family kinase, TBK1 interacts with the C‐terminal tail (CTT) of oligomerized STING. This association facilitates the phosphorylation of both STING and the transcription factor IRF3, initiating the production of IFN‐I and related cytokines (Zhang et al. 2019). Targeting TBK1 has therefore become a significant therapeutic approach for suppressing downstream STING signaling. The selective TBK1 inhibitor GSK8612 potently blocks Toll‐like receptor 3 (TLR3)‐mediated IRF3 phosphorylation in Ramos cells and reduces IFN‐I release in human primary monocytes. In monocyte‐derived macrophages, GSK8612 markedly attenuates the secretion of IFN‐β induced by dsDNA and cGAMP (Zeng et al. 2022).

In addition to synthetic compounds, a range of natural compounds and nanomaterials have shown promise in alleviating AD pathology through modulation of the cGAS‐STING pathway. For instance, nicotinamide adenine dinucleotide (NAD⁺) suppresses neuroinflammatory responses and cellular senescence by inhibiting cGAS‐STING signaling (Hou et al. 2021). Tetrahydroxystilbene glucoside (TSG) mediates neuroprotection via regulation of the cGAS‐STING pathway and suppression of NLRP3 inflammasome activation (Gao et al. 2023). Several natural compounds demonstrate therapeutic potential for AD by targeting the cGAS‐STING pathway. For instance, silymarin confers neuroprotection in sporadic AD models by mitigating iron‐induced injury and attenuating downstream STING‐mediated neuroinflammation (Liu et al. 2023). Similarly, icariin has been shown to inhibit glial activation and neuroinflammation via the cGAS‐STING pathway, thereby ameliorating cognitive deficits (Lu et al. 2025). Additionally, Platycodin D alleviates behavioral deficits, reduces Aβ accumulation, and mitigates mitochondrial damage in an AD model by suppressing cGAS‐STING signaling (Saida et al. 2025). Zhang et al. (2025) developed a nanozyme based on a covalent organic framework (COF) responsive to copper ions, which selectively targets and inhibits cGAS‐STING signaling in the AD brain, presenting an innovative therapeutic approach. While no cGAS‐STING‐targeting drugs are currently approved for AD treatment, this growing body of research provides a crucial foundation for the development of novel disease‐modifying therapies.