Development of Hematopoietic Stem Cell-targeted LNPs through Lipid Composition Optimization

Hematopoietic stem cell (HSC)-targeted gene editing holds significant potential for treating hereditary hematopoietic disorders, yet the efficient and safe delivery of gene editing tools into HSCs remains a critical challenge. Lipid nanoparticles (LNPs) have emerged as a promising platform for nucleic acid delivery; however, achieving high transfection efficiency in HSCs remains challenging. In this study, we developed HSC-targeted LNPs by integrating Bayesian optimization with our functional amino lipids. The optimized LNPs exhibited markedly improved transfection efficiency while preserving cell viability, surpassing earlier formulations. Using these LNPs, we achieved ex vivo TP53 gene editing in cord blood (CB) CD34⁺ cells with up to 40% on-target editing efficiency. Additionally, one LNP demonstrated efficient RNA delivery into primary human monocytic leukemia cells. These results highlight the potential of machine learning-guided LNP design for advancing HSC-targeted therapies and underscore the promise of LNP-based gene editing platforms to treat hereditary and malignant hematopoietic disorders.