Novel Extracellular Vesicle Modalities for Delivery of Genome Editor Ribonucleoprotein Complexes

Abstract

Genome editing reagents have undergone rapid advancements in terms of specificity and the types of edits that can be executed, but the hurdle of delivery in vivo and ex vivo can still limit the efficiencies of gene editing therapeutic modalities. Current vehicles for delivering genome editing therapeutics include viral vectors (e.g., adeno-associated virus (AAV) and lentivirus) and DNA/RNA nanoparticles (e.g., lipid nanoparticles). Although these modalities can achieve robust expression of therapeutic cargo, they carry several inherent limitations, including constraints on the sizes of cargos, varying levels of immunogenicity, increased off-target effects for DNA/RNA delivery compared to protein delivery, risk of DNA integration, and persistence in non-dividing cells. To abrogate size constraints, minimize off-target effects, and eliminate prolonged expression, I sought to develop a robust method for delivering gene editors to human and other mammalian cells as ribonucleoprotein complexes (RNPs). However, RNP delivery poses unique challenges because CRISPR-based genome editors are comprised of proteins that interact with small guide RNAs, and this complex must remain intact and translocate to the nucleus of a recipient cell. Here, I describe the development and optimization of methods for packaging genome editor RNPs into pleckstrin homology ectosome-like particles (Phe-LPs). Phe-LPs are delivery particles produced in human cells by leveraging a novel method for non-covalently tethering CRISPR-Cas protein cargos complexed with guide RNAs to the inner plasma membrane using human pleckstrin homology domains that bind to specific phospholipids. I generated first-generation Phe-LPs by screening and optimizing its various components and show that these particles are at least as efficient as viral-like particles (VLPs), another cell-based particle that uses multiple viral components and that has been used in research settings to deliver gene editor RNPs. I also describe novel methods for scaled production, purification, and concentration of Phe-LPs. Finally, I describe initial work to develop second-generation Phe-LPs that do not use any viral-derived components and that therefore should have minimal to no immunogenicity. Phe-LP technology provides an important new modality with important advantages over existing methods for efficiently delivering CRISPR-based technologies to human cells and that should extend the research and therapeutic applications of gene editing.