The pathway at the center of the work is the cGAS-STING axis, one of the innate immune system’s most fundamental alarm circuits. Cyclic GMP-AMP synthase, or cGAS, is a pattern recognition receptor that patrols the cytoplasm of cells for double-stranded DNA, whether it comes from an invading pathogen or from the cell’s own genome leaking out of damaged nuclei and mitochondria. When cGAS detects such DNA, it manufactures a second messenger called 2’3′-cGAMP, which binds to STING, an adaptor protein anchored in the endoplasmic reticulum. Activated STING autophosphorylates and recruits the kinase TBK1, which in turn phosphorylates the transcription factors IRF3 and NF-kB, driving the production of type I interferons and other inflammatory mediators. In the nervous system, this pathway has been implicated in brain injury, Alzheimer’s disease, Parkinson’s disease, and neuropathic pain, but its precise role in spinal cord injury had remained ambiguous.

To resolve that ambiguity, the researchers established a highly standardized mouse model of compressive spinal cord injury. Using calibrated forceps with a one-millimeter tip and a 0.55-millimeter spacer, they delivered a consistent five-second compression to the T10 spinal cord segment, producing near-complete paralysis of the hindlimbs. Behavioral scoring with the Basso Mouse Scale confirmed complete motor loss on the first day after injury, followed by gradual, spontaneous recovery that decelerated noticeably after day fourteen. Time-series transcriptomic sequencing of the lesion core at three and fourteen days after injury revealed thousands of differentially expressed genes, and gene set enrichment analysis pinpointed the cytosolic DNA sensing pathway, whose core component is cGAS-STING, as a central hub of the post-injury inflammatory response.

Crucially, the team then asked which cells were actually running this alarm system. Spatial transcriptomics and double immunofluorescence staining showed that after injury, cGAS and STING shifted from a mixed neuronal and phagocytic pattern to one dominated by myeloid cells. Staining for TMEM119, a microglia-specific marker, and CD14, a macrophage marker, revealed that the bulk of STING signal colocalized with resident microglia rather than with infiltrating macrophages. Western blotting confirmed that protein levels of cGAS and STING rose significantly after injury, accompanied by enhanced phosphorylation of TBK1, the pathway’s downstream workhorse. In other words, the injured spinal cord’s innate immune response was being orchestrated largely by its resident immune sentinels.

The first therapeutic attempt was deliberately conventional: continuous pharmacological modulation. The researchers administered STING agonists, including ADU-S100 and 2’3′-cGAMP, or antagonists, including C-176 and H-151, locally at the injury site immediately after injury and then daily by intraperitoneal injection for twenty-eight days. The compounds worked as intended, elevating or suppressing phosphorylated TBK1 as expected. Yet the behavioral results were disappointing. ADU-S100 produced only a transient improvement at three days, and by twenty-one days neither continuous activation nor continuous inhibition improved gait parameters relative to vehicle controls. Sustained manipulation of the pathway, it seemed, could not deliver lasting recovery.

That failure pointed toward a biphasic hypothesis. Inflammation after spinal cord injury is known to follow a two-phase arc: an early acute-to-subacute phase, roughly days zero to fourteen, during which inflammatory responses serve defensive and reparative functions such as clearing debris, and a later phase in which persistent inflammation becomes an obstacle to regeneration. Reanalyzing a publicly available transcriptomic dataset spanning fifteen minutes to forty-two days after injury, the team confirmed that day fourteen marks a genuine turning point, with pro-inflammatory cytokines peaking around days three and seven to fourteen before declining, and anti-inflammatory cytokines such as IL-4 and IL-10 showing a transient rise at day fourteen. Accounting for the twenty-four to forty-eight hour lag in drug bioavailability after intraperitoneal injection, they selected day eleven as the practical switching point.

The resulting sequential regimen was elegantly simple. Immediately after injury, mice received a local injection of either the agonist ADU-S100 or the inhibitor C-176, followed by daily systemic injections of the same compound through day ten. On day eleven, treatment crossed over: the agonist group switched to the inhibitor, and the inhibitor group switched to the agonist, continuing until day twenty-eight. Pharmacodynamic verification confirmed the switch worked, with phosphorylated TBK1 rising during the agonist phase and falling within three days of the crossover. The behavioral payoff was unambiguous. From day fourteen onward, the activation-then-inhibition group, ADU-S100 followed by C-176, showed significantly better hindlimb motor recovery than vehicle controls, while the reverse sequence, inhibition followed by activation, showed no benefit at all. Footprint analysis confirmed longer stride lengths in the winning group, and histology revealed more preserved Tuj1-positive axonal fibers and smaller glial scars.

Myelin protection told an equally compelling story. Luxol fast blue staining showed a smaller demyelinated area at the lesion center in the activation-then-inhibition group, and western blotting for myelin basic protein and a pathologically exposed myelin epitope confirmed that the reverse sequence actually worsened demyelination. Transmission electron microscopy of more than two hundred axons per group revealed that the sequential activation-then-inhibition regimen preserved the highest proportion of thickly myelinated axons and the lowest proportion of unmyelinated ones, with a g-ratio distribution approaching that of uninjured tissue. RNA sequencing added mechanistic texture: early STING activation was associated with upregulation of granzymes and chemokines such as CXCL10, consistent with enhanced immune clearance and recruitment of monocytes and macrophages, while late inhibition corresponded to elevated IL-10, resolution of inflammatory markers, and increased expression of axon growth genes including Tubb3, Map2, and doublecortin, a marker of nascent axonal growth confirmed by immunohistochemistry.

The team then asked whether microglia were indispensable to this effect. Using PLX5622, a brain-penetrant CSF1R inhibitor that depleted microglia by more than eighty-five percent, they showed that both early and late myeloid cell depletion completely abolished the functional benefits of the sequential strategy. The protective effect on neuronal apoptosis, demonstrated by caspase activation staining, also vanished upon depletion, and demyelination increased across all groups. In vitro experiments filled in the cellular logic: injured motor neurons released signals, likely including double-stranded DNA, that activated STING in cocultured microglia, driving them toward a phagocytic phenotype that engulfed fluorescent beads and tdTomato-labeled neuronal debris more avidly when STING was activated. Yet conditioned medium from STING-activated microglia inhibited neurite outgrowth in injured motor neurons, revealing the pathway’s double edge. In purified motor neuron cultures, STING activation directly promoted axon growth and regeneration after microfluidic axotomy, but in the inflamed in vivo environment, those intrinsic pro-growth effects are apparently overridden by microglia-derived inhibitory signals, which is why late inhibition proved essential.

The authors are candid about the study’s limitations. The pharmacological agents lack cell-type specificity, the depletion strategy targets multiple myeloid populations, and the transcriptomic analysis was performed on whole-lesion tissue, so some observed changes may reflect shifts in cellular composition rather than intrinsic reprogramming. The work was conducted exclusively in male mice, direct in vivo evidence of enhanced debris clearance remains indirect, and the feasibility, safety, and optimal switching window in humans are unknown. Still, the convergence of behavioral, histological, ultrastructural, and transcriptomic evidence supports a genuinely novel concept the authors call chrono-immunomodulation: interventions synchronized with the evolving pathology of central nervous system trauma rather than applied as static blocks. If the activation-then-inhibition logic can be translated, it may reshape not only spinal cord injury therapy but the broader approach to neuroinflammatory disease, where the difference between healing and harm may ultimately be measured in days.

Subject of Research: Sequential pharmacological modulation of the cGAS-STING innate immune pathway to promote repair after spinal cord injury

Article Title: Sequential modulation of the cGAS-STING pathway promotes spinal cord injury repair

Article References: Peng, Q., Luo, L., Yuan, M., Li, H., Chen, L., Wu, X., Han, M., Wen, S., Zhu, C., & Gu, Y. (2026). Sequential modulation of the cGAS-STING pathway promotes spinal cord injury repair. Journal of Advanced Research. https://doi.org/10.1016/j.jare.2026.10.023

Image Credits: AI Generated

DOI: 10.1016/j.jare.2026.10.023

Keywords: spinal cord injury, cGAS-STING pathway, neuroinflammation, microglia, innate immunity, STING agonist, STING inhibitor, myelination, axon regeneration, chrono-immunodulation, neurorepair, ADU-S100