Semaglutide targets the Isg15-FUNDC1 axis: suppressing IFN-β overactivation and attenuating blood-brain barrier injury in diabetic intracerebral hemorrhage

When a brain bleed (intracerebral hemorrhage, ICH) occurs in someone with diabetes, the damage to the brain's protective blood-brain barrier (BBB) is especially severe and few targeted treatments exist. This study found that a small protein called ISG15 is paradoxically reduced in the brain's blood vessel cells during peak injury, even though the broader immune alarm system is blaring loudly. Without ISG15, a key mitophagy receptor called FUNDC1 gets degraded by the cell's own protein-recycling machinery, which breaks down the process of clearing damaged mitochondria. Those damaged mitochondria then leak their DNA into the cell interior, triggering a runaway type-I interferon inflammatory response that tears apart the BBB. The diabetes drug semaglutide — a GLP-1 receptor agonist widely used for blood-sugar control and weight loss — was found to restore ISG15 levels, rescue FUNDC1 stability, restabilize mitophagy, and significantly reduce BBB leakage and neurological deficits in rats. Crucially, ISG15 protects FUNDC1 without using its classical 'ISGylation' tagging mechanism, pointing to a previously unrecognized non-canonical function. The findings open a plausible translational rationale for repurposing semaglutide in diabetic stroke patients, although the work is currently limited to male rodent and cell-line models.

• Single-cell RNA sequencing of rat ICH brain tissue revealed that brain microvascular endothelial cells (BMECs) show paradoxically reduced ISG15 expression at peak injury (day 3) despite robust activation of type-I interferon (IFN-I) signaling pathways, with ISG15 suppression further exacerbated under diabetic (high-glucose) conditions.
• ISG15 deficiency destabilized the mitophagy receptor FUNDC1 through enhanced proteasomal degradation (confirmed by cycloheximide chase and co-immunoprecipitation assays), impairing mitophagy, causing cytosolic mitochondrial DNA (mtDNA) leakage, and amplifying IFN-β-driven vascular inflammation — a mechanistic cascade independently of canonical ISGylation.
• Semaglutide administration in the DM-ICH rat model significantly upregulated ISG15 expression, restored FUNDC1 protein stability, reduced brain water content (cerebral edema marker), decreased Evans blue dye extravasation (BBB permeability marker), and improved neurological deficit scores, reversing the Isg15-FUNDC1-mitophagy pathological axis.

The study used a two-pronged experimental design: (1) an in vivo male rat DM-ICH model with semaglutide treatment, evaluated via brain water content, neurological scoring, Evans blue extravasation, and transmission electron microscopy of BBB ultrastructure; and (2) an in vitro model using the human cerebral microvascular endothelial cell line hCMEC/D3 exposed to high glucose plus hemin to simulate diabetic hemorrhagic conditions. Mechanistic dissection employed scRNA-seq on dynamic rat ICH tissue samples, adeno-associated virus (AAV) overexpression and siRNA knockdown of ISG15 and FUNDC1, co-immunoprecipitation to map protein–protein interactions, and cycloheximide chase assays to quantify proteasomal FUNDC1 degradation rates. The study does not specify exact animal group sizes or precise semaglutide dosing regimens in the available abstract.

The entire in vivo dataset derives from male rats only; the authors explicitly acknowledge that sex-specific effects on the ISG15-FUNDC1 axis remain uncharacterized, which is a critical gap given known sex differences in stroke biology and GLP-1 receptor signaling. The model relies on hemin exposure and collagenase-induced ICH in rodents, which may not fully recapitulate the heterogeneous hematoma environment in human diabetic ICH. All findings are preclinical, and neither pharmacokinetic data for semaglutide in brain tissue nor dose–response relationships are reported, making direct clinical translation premature. The next clarifying experiments would include (a) female-cohort replication, (b) a non-human primate or patient-derived endothelial cell model, and (c) structural mapping of how ISG15 physically shields FUNDC1 from ubiquitin-proteasome targeting without ISGylation.

This study extends [§136], which demonstrates that promoting mitophagy in neural tissue (peripheral nerve) via hirudin suppresses NLRP3 inflammasome activation, by showing an analogous mitophagy-rescue strategy (via ISG15-FUNDC1) that attenuates IFN-β-driven inflammation rather than NLRP3 — suggesting mitophagy failure is a convergent node for multiple inflammatory cascades across CNS-adjacent tissues. It parallels [§115], where dexmedetomidine attenuates organ-level barrier injury (lung in ALI) by modulating mitochondrial autophagy and innate immune pathways, reinforcing the broader theme that pharmacological mitophagy restoration can protect vascular barrier integrity in critical-injury contexts even when the organs and drugs differ. The mechanistic demonstration that mtDNA cytosolic leakage drives type-I interferon signaling also resonates with [§99], which identifies mitochondrial RNA release during aging as a driver of inflammation and senescence, collectively strengthening the corpus view that mitochondrial nucleic acid containment is a central regulator of sterile innate immune tone. The GLP-1 agonist angle adds a metabolic-disease dimension that is otherwise underrepresented in this mitochondrial-biology corpus, positioning the article as a bridge between metabolic pharmacology and mitochondrial quality-control research.

• Hirudin Promotes Peripheral Nerve Repair by Enhancing Mitophagy to Suppress NLRP3 Inflammasome Activation
  Both articles demonstrate that pharmacological restoration of FUNDC1-mediated mitophagy suppresses innate immune inflammasome/interferon activation in neural tissue injury, making direct mechanistic comparison of the upstream regulators (ISG15 vs. hirudin targets) highly informative.
• Dexmedetomidine attenuates LPS-induced acute lung injury via activation of mitochondrial autophagy
  Both studies use a mitochondrial autophagy axis to protect an organ-level vascular/epithelial barrier during acute sterile injury, but in different organs (BBB in brain vs. alveolar barrier in lung) and with different drug classes, allowing readers to assess how generalizable mitophagy-based barrier protection really is.
• Mitigating Mitochondrial RNA Release During Aging to Control Inflammation and Senescence
  This article on mitochondrial RNA release as a driver of age-related inflammation directly contextualizes the current study's finding that cytosolic mtDNA leakage activates IFN-β, offering a mechanistic parallel from a different biological trigger (aging vs. hemorrhagic-diabetic stress) and more mature evidence tier.
• Cardiomyocyte BMAL1 deficiency worsens obesity-related cardiomyopathy with heightened PINK1/Parkin-linked mitophagy and mitochondrial dysfunction
  Both studies examine mitophagy (PINK1/Parkin vs. FUNDC1 pathways) in the context of a metabolic disease (obesity-related cardiomyopathy vs. diabetic ICH), providing a useful cross-tissue, cross-receptor comparison of how metabolic dysfunction co-opts mitophagy receptors to amplify injury.
• Exercise hormone irisin alleviates rheumatoid arthritis by removing dysfunctional mitochondria
  Irisin's attenuation of rheumatoid arthritis via removal of dysfunctional mitochondria parallels semaglutide's restoration of mitophagy here, and comparing two circulating hormone/peptide-class molecules that converge on mitochondrial quality control may help delineate shared versus drug-specific downstream effectors.