Engineering Lipid Nanoparticles through Integrated Compositional and Ligand Targeting Enhances β Cell-Directed RNA Delivery

Abstract

Beyond their deployment as COVID-19 vaccines, lipid nanoparticles (LNPs) have emerged as versatile vehicles for therapeutic nucleic acid delivery. However, achieving efficient and cell-targeted transfection in extrahepatic tissues, particularly pancreatic β cells, remains a major challenge. Here, we develop a dual-targeting LNP engineering strategy that integrates high-throughput compositional screening with surface conjugation of β cell-specific targeting ligands to enable selective gene delivery to pancreatic β cells. Compositional optimization identified LNP formulations that achieved over a 148-fold increase in β cell transfection efficiency in vitro and more than an 8-fold increase in pancreatic selectivity in vivo compared to the Moderna LNP formulation. Surface conjugation of the ZnT8-specific monoclonal antibody (mAb43), which recognizes the zinc transporter ZnT8 highly expressed on murine β cells, further increased pancreatic transgene expression by more than 2-fold and achieved over 70% β cell transfection in murine models. To improve translational potential, we conjugated a high-affinity camelid single-domain antibody (4hD29 nanobody) targeting dipeptidyl peptidase-6 (DPP6), a biomarker enriched on human β cells, to compositionally optimized LNPs to deliver human STAT2-siRNA. These dual-targeting LNPs reduced STAT2 expression in human β cells under IFN-α stimulation to below baseline levels observed in unstimulated controls and induced > 4-fold increase in PDL1 expression. Together, this integrated LNP design for β cell-directed gene delivery establishes a versatile platform for RNA therapeutics and gene-editing applications in a pro-inflammatory type 1 diabetes context.