Intracranial delivery of neuron-preferential lipid nanoparticles for gene-editing activity in mouse brain

Lipid nanoparticles (LNPs) are promising non-viral vectors for central nervous system (CNS) gene therapy, but effective delivery is limited by the blood-brain barrier and cell-type specificity. Here, we combined intracranial delivery with high-throughput, cluster-based screening and machine learning to identify LNP formulations optimized for functional neuronal gene-editing activity. We screened 720 LNP formulations for their ability to transfect neurons and deliver genetic payloads efficiently. Following this, cluster-mode intracranial screening enabled targeted delivery to specific brain regions, including the ventral posteromedial nucleus (VPM) and hippocampus, and efficiently narrowed hundreds of candidates to single optimized formulations. Optimized LNPs achieved approximately 20% gene editing, measured as functional activity, in targeted areas and exhibited neuron- and astrocyte-enriched transfection patterns. These results demonstrate that intracranial delivery, combined with machine-learning-guided optimization, can identify LNPs capable of precise cell- and region-preferential gene delivery, supporting the development of targeted therapies for neurological disorders.