Epilepsy affects over 50 million people worldwide, making it one of the most common neurological disorders. Although medication helps many patients achieve seizure control, approximately one-third continue to experience seizures despite treatment. Seizures often arise when the brain’s excitation–inhibition (E/I) balance breaks down. In healthy conditions, specialized inhibitory neurons act as a natural braking system, releasing a neurotransmitter called gamma-aminobutyric acid (GABA) that helps prevent excessive electrical activity. When this inhibitory control is weakened, abnormal bursts of activity can spread through the brain and trigger seizures.

For years, researchers have viewed inhibitory neurons as an attractive target for gene therapy, which aims to treat diseases by introducing modified genetic material into cells. However, delivering therapeutic genes specifically to these neurons has proven difficult. The most widely used delivery vehicles for gene therapy are engineered viruses called adeno-associated vectors (AAVs). While certainly versatile, AAVs can only carry a limited amount of genetic cargo, roughly around 4.7 kilobases. Currently available genetic switches (or promoters, DNA sequences that control when and where genes are expressed) that specifically target inhibitory neurons are too large, taking more than half of that space. They are also too weak to drive a significant therapeutic effect, especially when delivered through the bloodstream.

Against this backdrop, a research team led by Professor Hirokazu Hirai, Director of the Viral Vector Core Center, Gunma University Initiative for Advanced Research, Japan, set out to find an effective solution. In their latest study, which was made available online in the journal Molecular Therapy on June 25, 2026, the researchers describe a compact genetic switch called compact mouse glutamic acid decarboxylase (cmGAD67) promoter that selectively targets inhibitory neurons while leaving enough room inside AAV vectors for therapeutic genes. The researchers used this novel strategy to develop and test a gene therapy-based approach designed to suppress seizures in multiple mouse models of epilepsy. The paper was actively co-authored by Assistant Professor Yuuki Fukai of the Viral Vector Core Center, Gunma University Initiative for Advanced Research, and Dr. Ayumu Konno of the Department of Neurophysiology & Neural Repair, Gunma University Graduate School of Medicine, Japan.

Sharing the novelty of their findings, Prof. Hirai explains, “Because cmGAD67 is highly compact, it also helps overcome one of the major technical limitations of AAV vectors, which is their restricted cargo capacity. Therefore, our work expands opportunities for therapeutic gene delivery and innovation in the gene therapy industry.”

The team discovered cmGAD67 by analyzing DNA sequences that naturally regulate the activity of the GAD67 gene, which is expressed in inhibitory neurons and plays a key role in GABA production. The developed promoter is only 410 base pairs long, making it substantially smaller than available alternatives. Despite its compact size, experiments in mice showed that it can drive strong and highly selective gene expression in inhibitory neurons in multiple brain regions. Notably, the promoter showed particularly strong expression in parvalbumin-positive inhibitory neurons, a subtype that plays a central role in controlling excessive brain activity.

The cmGAD67 promoter was used in an AAV vector to deliver the gene for GAD65, another enzyme involved in the production of GABA. The researchers reasoned that increasing GABA production in inhibitory neurons could strengthen the brain’s natural ability to suppress runaway electrical activity, thus preventing seizures. This cmGAD67-based construct, called AAV-GAD65, was tested in two complementary seizure models in mice.

In a ‘chemical kindling’ model, where repeated drug injections progressively sensitized the brain to seizures, systemic administration of AAV-GAD65 reduced abnormal electrical discharges, suppressed the slow-wave oscillations associated with overexcitability, restored brain network activity toward normal levels, increased GABA levels in the brain, and normalized anxiety-like behavior. In an even more severe version of the model, the treated mice exhibited markedly better survival rates. Meanwhile, in a focal seizure model, where a specific brain region is artificially driven to become hyperactive, local delivery of AAV-GAD65 to the affected brain region greatly reduced seizure severity, with three of the five treated mice showing no observable seizures.

Taken together, these findings suggest that enhancing inhibitory neuron function can restore the E/I balance across different severities of epilepsy. Notably, the researchers believe this technology could have broader applications beyond epilepsy. By enabling selective genetic modulation of inhibitory neurons, cmGAD67 may provide a versatile platform for developing gene therapies for neurological disorders associated with disrupted E/I balance.

Though further work will be needed to validate this approach in chronic epilepsy models and assess long-term safety, the present results are very encouraging. “My laboratory originally developed viral vector technologies to manipulate E/I balance in the brain and study its influence in brain development, learning, and memory. It is particularly rewarding to see a tool originally developed for basic neuroscience research evolve into a potential therapeutic strategy for patients with drug-resistant epilepsy,” concludes Prof. Hirai.


Reference
Fukai, Y., Konno, A., Hosoi N., Miyakawa K., Kaneko R., and Hirai, H. A compact GAD67 promoter enables inhibitory neuron-targeted AAV gene therapy for seizure suppression. Molecular Therapy (2026). DOI: https://doi.org/10.1016/j.ymthe.2026.06.007.