Back

A compact glutamic acid decarboxylase 67 promoter enables inhibitory neuron-targeted AAV gene therapy for treatment-resistant epilepsy

Fukai, Y.; Konno, A.; Hosoi, N.; Miyakawa, K.; Kaneko, R.; Hirai, H.

2025-12-10 neuroscience
10.64898/2025.12.07.692482 bioRxiv
Show abstract

Epilepsy often becomes treatment-resistant, partly due to impaired inhibitory neurotransmission and reduced {gamma}-aminobutyric acid (GABA) function. Enhancing inhibitory neuron activity via gene therapy may restore excitation-inhibition (E/I) balance. We developed a compact 410-bp glutamic acid decarboxylase 67 promoter (cmGAD67) that enables strong, selective transgene expression in inhibitory neurons while preserving AAV packaging capacity. When delivered systemically, AAV vectors carrying cmGAD67 preferentially targeted parvalbumin interneurons and supported effective circuit manipulation. To evaluate therapeutic potential, we expressed glutamic acid decarboxylase 65 (GAD65) under cmGAD67 (AAV-GAD65) in pentylenetetrazole (PTZ) epilepsy models. Systemic AAV-GAD65 suppressed abnormal delta oscillations, reduced seizure-like events, normalized anxiety-like behavior, and improved survival in a severe PTZ paradigm. Biochemical analyses confirmed increased cortical and hippocampal GABA levels, linking behavioral and electrophysiological improvements to enhanced inhibitory neurotransmitter synthesis. Prior clinical evidence indicates that AAV-GAD65 delivery to the subthalamic nucleus is safe and effective in Parkinsons disease. Building on this foundation, our findings establish the cmGAD67 promoter as a powerful platform for inhibitory neuron-targeted AAV gene therapy and highlight AAV-cmGAD67-GAD65 as a promising approach for treatment-resistant epilepsy and other disorders involving disrupted E/I balance.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.