When a blood vessel ruptures inside the brain, the resulting injury is serious under any circumstances. When that event happens in a person who also has diabetes, the damage tends to be substantially worse. Brain swelling is greater, the protective lining of blood vessels breaks down faster, and clinical outcomes are poorer. Researchers have long suspected that the combination of high blood sugar and active bleeding sets off an unusually destructive chain of events, but the precise molecular steps have been unclear.
A study published in the Journal of Neuroinflammation set out to map those steps, using single-cell gene sequencing in a rat model together with human brain endothelial cells grown in the laboratory. The team also tested whether a GLP-1 receptor agonist peptide could interrupt the damage. What they found was a previously undescribed protective axis inside the tiny cells that line brain blood vessels, and evidence that the peptide could restore it.
The setting: diabetes and brain hemorrhage
Intracerebral hemorrhage is a stroke subtype in which blood leaks directly into brain tissue. It carries a high risk of death and disability even in otherwise healthy people. Diabetes adds a second layer of harm. High blood sugar stiffens blood vessel walls, fuels inflammation, and appears to weaken the blood-brain barrier, the tightly sealed layer of endothelial cells that normally prevents harmful molecules from crossing from the bloodstream into brain tissue.
The researchers built a diabetic intracerebral hemorrhage model in male rats to study this overlap in a controlled setting. They also exposed a human cerebral microvascular endothelial cell line to high glucose and a compound called hemin, which mimics the toxic effects of blood pooling in brain tissue. Together, these two systems let the team study both the living animal and the individual cells most relevant to barrier function.
A surprising finding in the gene data
Using single-cell RNA sequencing, the team mapped which genes were active in brain microvascular endothelial cells at different time points after the hemorrhage. At the moment of peak injury, three days after the bleed, these cells showed a strong activation of type I interferon pathways. Type I interferons are signaling proteins the body normally uses to fight viruses, but here they appeared to be driving damaging inflammation rather than protecting tissue.
What made this finding unusual was a paradox in the data. Even though interferon pathways were highly active, a gene called Isg15, which normally acts as a brake on interferon signaling, was sharply reduced. Under diabetic conditions, that reduction was even more pronounced. The cells were producing a strong pro-inflammatory signal while simultaneously losing the molecule that would ordinarily keep that signal in check. The researchers identified this loss of Isg15 as a central feature of the injury pattern.
The Isg15-FUNDC1 axis explained
Isg15 is a small protein that can attach to other proteins and change how they behave, a process called ISGylation. The study uncovered a separate and previously undescribed function: Isg15 physically stabilizes a mitophagy receptor called FUNDC1.
Mitophagy is the process by which cells identify damaged mitochondria and dispose of them safely. Mitochondria are the energy-producing structures inside cells, and when they are damaged they can leak their own DNA into the surrounding cell interior. That leaked mitochondrial DNA acts as an alarm signal, triggering a strong type I interferon response.
The chain the researchers described works as follows. When Isg15 levels fall, FUNDC1 becomes unstable and is broken down by the cell's protein disposal system, the proteasome. Without enough functional FUNDC1, the cell cannot clear damaged mitochondria efficiently. Those damaged mitochondria leak DNA. The leaked DNA activates interferon-beta pathways. The resulting inflammation damages the tight junctions between endothelial cells, and the blood-brain barrier begins to fail. The study used a range of molecular tools, including gene silencing, viral gene delivery, protein interaction assays, and protein stability tests, to map and confirm each step of this cascade.
Where the GLP-1 peptide fits in
GLP-1 receptor agonists are a class of peptides that mimic glucagon-like peptide 1, a naturally occurring hormone involved in blood sugar regulation and a range of other physiological processes. The team administered a GLP-1 receptor agonist to the diabetic rat model and applied it to the cell culture system to see whether it could interrupt the injury cascade.
In both settings, the peptide raised Isg15 expression. That single upstream effect appeared to be enough to stabilize FUNDC1, restore mitophagy, reduce mitochondrial DNA leakage, and bring interferon-beta signaling back toward normal levels. Measurements of blood-brain barrier integrity, including a dye-leakage test and microscopic examination of vessel structure, showed substantially less disruption in treated animals. Neurological scoring also improved compared to untreated animals with the same injury.
The authors describe this as the first evidence that a GLP-1 receptor agonist acts on the Isg15-FUNDC1-mitophagy axis in this context. The finding is notable because it suggests the peptide may have a specific molecular target within brain endothelial cells, rather than exerting only a general anti-inflammatory or blood-sugar-lowering effect.
Study limitations and open questions
The researchers note several important constraints on their conclusions. All in vivo experiments used male rats only. The authors explicitly state that whether these findings apply to females is unknown and warrants further research. Hormonal and immune differences between sexes are well established, and a protective pathway that functions one way in males may behave differently in females.
The model also involves a specific and relatively severe experimental setup. How well the findings translate to the more varied circumstances of human diabetic intracerebral hemorrhage remains to be demonstrated. The study is mechanistic in nature, meaning it traces a biological pathway rather than evaluating a clinical treatment. No human trials are described in this abstract.
The research also used a single GLP-1 receptor agonist peptide. Whether other peptides in this class would produce the same Isg15-related effects, or whether the dose and timing used in rats would translate to useful parameters in humans, are questions the study does not answer.
What this research adds to the field
The study contributes a new mechanistic explanation for why diabetic intracerebral hemorrhage causes unusually severe blood-brain barrier damage. Prior research had established that interferon signaling played a role in vascular inflammation, but the specific involvement of Isg15, FUNDC1, and mitophagy failure in brain endothelial cells had not been reported in this context.
The finding that Isg15 stabilizes FUNDC1 independently of ISGylation is also described as novel. Most research on Isg15 has focused on its role as a tag that attaches to other proteins. Here, it appears to protect a critical receptor simply by being present, through a mechanism the authors linked to preventing proteasomal degradation rather than chemical modification.
For researchers working on neuroinflammation, metabolic disease, or peptide biology, this axis represents a potentially important target. Early data from this kind of preclinical work often takes years to develop into testable clinical approaches, but clearly mapping the pathway is an essential first step. The literature now has a more complete picture of how diabetic conditions transform a brain bleed into an especially destructive event at the cellular level.
