UCSF

In the area of receptor-targeted lipid nanoparticles for drug delivery, efficiency has been mainly focused on cell-specificity, endocytosis, and subsequently effects on bioactivity such as cell growth inhibition. Aspects of targeted liposomal uptake and intracellular sorting are not well defined. This dissertation assessed a series of ligands as targeted functional groups against HER2 and EGFR for liposomal drug delivery. Receptor-mediated uptake, both mono-targeted and dual-targeted to multiple receptors of different ligand valence, and the intracellular sorting of lipid nanoparticles were investigated to improve the delivery of drugs to cancer cells.

Lipid nanoparticles were functionalized through a new sequential micelle transfer - conjugation method, while the micelle transfer method was extended to growth factors. Through a combination of both techniques, anti-HER2 and anti-EGFR dual-targeted immunoliposomes with different combinations of ligand valence were developed for comparative studies. With the array of lipid nanoparticles, the uptake and cytotoxicity of lipid nanoparticles in relationship to ligand valence, both mono-targeting and dual-targeting, were evaluated on a small panel of breast cancer cell lines that express HER2 and EGFR of varying levels. Comparable uptake ratios of ligand to expressed receptor and apparent cooperativity were observed. For cell lines that express both receptors, additive dose-uptake effects were also observed with dual-targeted immunoliposomes, which translated to marginal improvements in cell growth inhibition with doxorubicin delivery. Colocalization analysis revealed that ligand-conjugated lipid nanoparticles settle to endosomal compartments similar to their attached ligands. Pathway transregulation and pathway saturation were also observed to affect trafficking. In the end, liposomes routed to the recycling endosomes were never observed to traffic beyond the endosomes nor to be exocytose like recycled ligands.

Based on the experimental data, models were developed to help interpret and predict the binding and trafficking of lipid nanoparticles. The crosslink multivalent binding model of lipid nanoparticles to monovalent receptors was able to predict ligand valence for optimum binding, cell association concentrations, offer explanations to the antagonistic effects observed from high ligand valence, and predict the binding limitations of both ligand valence and ligand affinity. Hopefully, the models will serve as valuable tools for future optimizations in targeted liposomal drug delivery.