The cGAS-STING axis begins with cyclic GMP-AMP synthase, or cGAS, an enzyme that binds double-stranded DNA without sequence specificity. When DNA appears in the cytosol, cGAS synthesizes a second messenger called 2’3′-cGAMP, which activates STING, an adaptor protein resident on the endoplasmic reticulum. STING then traffics through the ER-Golgi intermediate compartment and recruits TANK-binding kinase 1, which phosphorylates both STING and interferon regulatory factor 3. Phosphorylated IRF3 enters the nucleus and, together with NF-κB, drives transcription of type I interferons and proinflammatory cytokines. In the brain, this circuit translates cellular and metabolic stress into innate immune activation, and its consequences depend heavily on which cell is responding and how long the signal persists.

The review, authored by Cunlong Zhang and Qingqing Li of Pengzhou Hospital of Traditional Chinese Medicine, organizes preclinical findings along two independent dimensions: pathological context and site of action. This framework allows herbal medicines, natural compounds, and acupuncture-based interventions to be compared without assuming a uniform target or direction of modulation. The authors emphasize that mechanistic attribution is strongest when genetic or pharmacological perturbation links a pathway change to a functional outcome, and that direct binding has been demonstrated for only a small group of compounds. Most findings remain preclinical and model-dependent, but the patterns they reveal are consistent enough to support a context-sensitive framework for interpreting modulation direction.

In cerebral ischemia and reperfusion injury, the initiating insult is mitochondrial oxidative stress that drives mitochondrial DNA into the cytosol, activating microglial cGAS-STING. Tetramethylpyrazine preserves mitochondrial dynamics and integrity in permanent cerebral ischemia models, with improvements in mitochondrial DNA content and reduced STING activation and TBK1 phosphorylation. Icariside II binds triose phosphate isomerase 1 and restores mitochondrial cristae after middle cerebral artery occlusion, improving oxidative phosphorylation and suppressing microglial mtDNA-STING signaling. A mitochondria-targeted quercetin-Fe3+ nanozyme reaches the same upstream node by scavenging reactive oxygen species. Buyang Huanwu Decoction is associated with lower p53/cGAS/STING protein levels and improved cognition in chronic hypoperfusion, and the p53 activator Nutlin-3 counteracts its protective effects in oxygen-glucose-deprived BV2 cells, supporting an upstream p53-related mechanism.

Other interventions in stroke act at the sensor, adaptor, or downstream outputs. Molecular docking predicts an interaction between puerarin and cGAS. Geniposide is associated with lower cGAS-STING and inflammatory readouts after stroke. In chronic cerebral hypoperfusion, resveratrol reduces STING-pathway activation and IRF3 nuclear translocation. Studies of subarachnoid hemorrhage provide more direct pathway-perturbation results: a STING agonist reverses the antiferroptotic effect of ginsenoside Rd, and in the astragaloside IV study, STING knockout reduces early brain injury while the compound lowers pathway-related inflammation and improves neurological outcomes. Electroacupuncture at governor-vessel points reduces cortical cGAS/STING signaling and NLRP3 activation after ischemia-reperfusion, and other models connect cognitive improvement with ferritinophagy regulated through cGAS-STING or the FTO-YTHDF1 axis.

In Alzheimer’s disease models, amyloid-β and hyperphosphorylated tau compromise mitochondrial integrity and release mtDNA for microglial cGAS sensing. PQBP1 recognizes extracellular tau in microglia and recruits cGAS, activating STING-related inflammation. In 5 × FAD mice, cGAS deficiency reduces glial inflammation and amyloid-β burden, and in a separate model, STING inhibition reduces microglial synapse engulfment. Icariin reduces microglial cGAS-STING responses in an Aβ1-42 model, with the cGAS inhibitor RU.521 used to examine pathway involvement. Tetrahydroxystilbene glucoside lowers cGAS-STING activity and NLRP3 signaling with inhibitor experiments included. Bisdemethoxycurcumin reduces RIPK/MLKL-dependent necroptosis. Quercetin has been linked to the pathway through estrogen-receptor signaling. The Huanglian Jiedu Decoction study combined machine-learning-guided multiomics with pharmacological STING testing using H-151 and DMXAA. Silibinin restores mitochondrial homeostasis in LPS-stimulated microglia through a TREM2-related mechanism, and resveratrol blocks STING movement from the ER to the Golgi, reducing subsequent TBK1 activation and microglial senescence.

Parkinson’s disease studies use two preclinical strategies: suppression of STING-related inflammation and clearance of damaged DNA. Astragaloside IV reduces neuronal loss and motor deficits in rotenone and mutant α-synuclein models, with agonist and inhibitor experiments placing α-synuclein aggregation upstream of a γH2AX-STING-IFN-I sequence. Protocatechuic acid enhances autophagy in microglia, promotes clearance of damaged DNA, and reduces STING activation; microglia-specific Atg5 loss abolishes the benefit. Acupuncture reduces cGAS-STING signaling and improves cognition in 6-OHDA-treated rats, and cGAS overexpression reverses this improvement, supporting pathway involvement. In depression models, monotropein lowers cGAS and phosphorylated STING levels, Jiao-tai-wan attenuates STING-related signaling with microglial STING knockdown supporting pathway involvement, and Zhi-Zi-Hou-Po Decoction restores neuronal TFAM-dependent mitophagy to promote clearance of damaged mitochondria and cytosolic mtDNA, with TFAM knockdown abolishing the effect.

Viral encephalitis creates two therapeutic targets that change as infection progresses. In early herpes simplex virus type 1 infection, restoration of suppressed antiviral signaling improves host defense. Luteolin binds the substrate pocket of cGAS and promotes oligomerization, raising type I interferon production and reducing HSV-1 replication in cellular and murine encephalitis models. Rosmarinic acid acts on STING trafficking: stress-induced ALOX15 activity generates 4-HNE, which modifies STING and blocks its movement from the ER to the Golgi, and rosmarinic acid inhibits this sequence to restore antiviral signaling. When inflammatory injury dominates infection, STING inhibition becomes protective. In sepsis-associated encephalopathy, microglial STING activation drives pyroptosis, and gelsevirine lowers STING phosphorylation and improves survival and cognition, with its suppression of microglial pyroptosis weakened after STING knockout. Resveratrol disrupts the HSP90β-STING interaction and promotes STING degradation, reducing proinflammatory microglial responses in HSV-1 encephalitis.

Pain models reveal perhaps the most surprising pattern. In acute postoperative pain, electroacupuncture at ST36 and SP6 increases the pain threshold and reduces neuroinflammation while restoring STING-IFN-I signaling, and intrathecal C-176 weakens its analgesic and anti-inflammatory effects. In cancer-induced bone pain, electroacupuncture at L3-L5 Jiaji points increases spinal STING and IRF3 expression, with IFNAR and IFN-α/β measures rising while TNF-α and IL-1β fall during analgesia. STING immunoreactivity is predominantly colocalized with Iba1-positive microglia and, to a lesser extent, NeuN-positive neurons, and C-176 partly attenuates electroacupuncture-induced analgesia. The same study reports analgesia with cGAMP and with C-176 alone, indicating that analgesia and anti-inflammatory effects should be interpreted separately. In spinal cord injury, celastrol-loaded layered double hydroxide nanoparticles scavenge reactive oxygen species and improve microglial clearance of apoptotic debris, reducing cGAS-STING signaling, while dihydromyricetin inhibits microglial STING and restores autophagy, with STING overexpression reversing its effect on NLRP3-related pyroptosis.

The review identifies four levels at which traditional Chinese medicine-related interventions enter the pathway: upstream mitochondrial protection and reduction of cytosolic DNA burden, direct modulation of cGAS or STING, regulation of STING trafficking and degradation, and restriction of downstream inflammatory and cell-death output. Direct-action claims differ sharply by study design. Molecular docking places puerarin near cGAS, biochemical assays show luteolin binding cGAS, and surface plasmon resonance with mutational analysis establishes cryptotanshinone binding to STING. Resveratrol blocks ER-Golgi trafficking and can disrupt the HSP90β-STING interaction, promoting STING degradation. Artesunate restores lysosomal clearance. Rosmarinic acid acts in the opposite infection context, reducing 4-HNE modification and restoring STING trafficking. The authors stress that pathway change and direct target engagement are separate conclusions, and that protein expression or phosphorylation alone establishes pathway association, not direct action on cGAS or STING.

Clinical translation will depend on reproducible formulation composition, effective brain exposure, and safety. Polyphenol absorption and metabolism reduce systemic availability, while blood-brain barrier passage determines the concentration that reaches neural tissue. Delivery systems can improve distribution, but the carrier and its surface properties also change cellular uptake, and protein-corona formation adds an immune variable. Formula studies need chemical composition and in vivo exposure data before individual constituents can be assigned a mechanistic role. Human studies currently document disease-associated pathway changes, while intervention studies remain predominantly preclinical. The authors argue that a decisive experiment must connect intervention, pathway perturbation, and functional outcome in the same model, with cell-specific manipulation identifying the responding compartment and repeated time points showing when the direction changes. Until such validation chains are completed, the cGAS-STING axis offers a powerful organizing framework for comparing traditional Chinese medicine-related interventions across neuroinflammatory disorders, but the benefit-risk balance for each indication remains to be defined.

Subject of Research: A review of preclinical evidence on how traditional Chinese medicine-related interventions modulate the cGAS-STING axis in neuroinflammation across disease contexts and mechanisms of action.

Article Title: Traditional Chinese medicine-related interventions and the cGAS–STING axis in neuroinflammation: disease contexts and mechanisms of action

Article References: Zhang, C., & Li, Q. (2026). Traditional Chinese medicine-related interventions and the cGAS–STING axis in neuroinflammation: disease contexts and mechanisms of action. Frontiers in Neurology, 17, Article 1979957. https://doi.org/10.3389/fneur.2026.1979957

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1979957

Keywords: cGAS-STING, neuroinflammation, traditional Chinese medicine, electroacupuncture, microglia, mitochondrial DNA, type I interferon, stroke, Alzheimer's disease, Parkinson's disease, Traditional, Chinese