Signal Transduction and Targeted Therapy | 2021
Membrane-destabilizing ionizable phospholipids: Novel components for organ-selective mRNA delivery and CRISPR–Cas gene editing
Abstract
In a new study published in Nature Materials, Liu et al. report a novel design of lipid nanoparticles (LNPs) in which multi-tailed ionizable phospholipids (iPhos) constitute the active component, and which facilitates endosomal escape and thus improves delivery of mRNA and/or single-guide (sg)RNA for in vivo gene editing. LNPs composed of the best-performing iPhos and different helper lipids—zwitterionic lipids, ionizable cationic lipids and permanently cationic lipids—achieved selective organ targeting (SORT) and organ-specific CRISPR-Cas9 gene editing in spleen, liver, and lungs of mice, respectively. A main challenge in DNA/RNA-based gene therapy is the delivery of nucleic acid molecules into target cells. To enter cells, these molecules need to be encapsulated into specialized vectors as the cell membrane is inherently not penetrable by naked DNA or RNA. Over the last decade, non-viral vectors have attracted increasing attention owing to their ability to deliver and/or codeliver different cargos for gene therapy (DNA, siRNA, mRNA, etc), as well as their ease of manufacturing, mild immunogenicity, relatively low toxicity, and compatibility with repeated dosing. Among the available non-viral vector variants, LNPs may be the most developed. They are typically composed of multiple entities, including (1) cationic lipids containing ionizable amines. The latter is positively charged at low pH during LNP manufacturing to facilitate encapsulation of negatively charged DNA/RNA. In contrast, they are relatively neutral at physiological pH to avoid the formation of large complexes (previously observed in LNPs using permanent cationic lipids), and protonated again in the endosome to induce endosomal escape. Moreover, LNPs comprise (2) zwitterionic phospholipids as helper or structure lipids that mimic lipids in cell membranes, (3) polyethylene glycol (PEG) lipid to provide a hydrating layer surrounding the nanoparticle, and (4) cholesterol to stabilize the nanoparticle (Fig. 1a). Despite recent progress in LNP design, the efficiency of RNA/DNA delivery by the current LNP generation remains rather low. LNPs enter cells by endocytosis and are then routed into the endosome. From there, only 1–4% of LNPs can escape and release DNA/RNA into the cytosol, largely explaining the low LNP efficiency. A proposed mechanism of endosomal escape of LNPs is that interaction between cationic lipids in the LNP with anionic lipids of the endosome membrane induces non-bilayer hexagonal HII phase formation, leading to disruption of the endosome membrane (Fig. 1b). Previous studies have mainly focused on optimizing ionizable cationic lipids, while zwitterionic phospholipids were merely regarded as helper lipids needed to form LNPs and remained unexplored. Intriguingly, though, zwitterionic phospholipids resemble the lipids forming the endosome membrane, implying their potential to fuse into the endosome membrane and thus trigger membrane disruption. Guided by this hypothesis and by their experience from cationic lipid optimization, Liu et al. rationally designed a series of multitailed, ionizable zwitterionic phospholipids called iPhos, which contain an ionizable amine, a phosphate group and three hydrophobic alkyl tails (Fig. 1c). Like ionizable cationic lipids, ionizable amines in iPhos mediate pH-dependent membrane disruption. At physiological pH, the amine group is not protonated and the iPhos with the negatively charged phosphate groups cannot fuse into the anionic biological membrane. In contrast, at the low pH in the endosome, the amine group is positively charged to form a zwitterionic head together with a phosphate group, which is able to interact with the endosome membrane and to form the HII phase (Fig. 1b, c). 1