Summary
A bioRxiv preprint used giant unilamellar vesicles to visualise how two ionisable-lipid nanoparticles interact with model endosomal membranes. The results suggest that apparent pKa alone does not predict the pH at which functional membrane engagement occurs.
A bioRxiv preprint has used a cell-free membrane model to visualise a key step in lipid-nanoparticle delivery: how nanoparticles disrupt the membrane of an endosome, the compartment that traps them after entering a cell.
The researchers compared nanoparticles built with two ionisable lipids, LP01 and BiP-20. Although both lipids had the same apparent pKa of 6.6 in the researchers’ measurements, the nanoparticles interacted most strongly with the model membrane at different acidity levels. LP01 nanoparticles showed maximal membrane engagement at pH 5.8, while BiP-20 nanoparticles required pH 5.1.
The finding indicates that a bulk pKa measurement alone does not identify the pH at which a nanoparticle will functionally engage with an endosomal membrane.
A model for watching endosomal escape
Lipid nanoparticles, or LNPs, are delivery particles that can carry mRNA into cells. After uptake, an LNP is enclosed in an endosome. For its cargo to reach the cell interior, the particle must interact with and cross the endosomal membrane—a process commonly called endosomal escape.
The study used giant unilamellar vesicles, or GUVs, designed to mimic an endosomal membrane. These are large, cell-free membrane compartments that can be observed with microscopy. Using confocal microscopy, the researchers followed the interaction between the model membrane and LNPs containing either LP01 or BiP-20.
Both lipids are described in the preprint as benchmark ionisable lipids for in-vivo gene-editing applications in the liver. The experiment focused on the physical steps of membrane interaction rather than on gene editing in cells or animals.
The observed sequence of membrane disruption
The researchers describe a series of stages. First, the LNPs were attracted to the model endosomal membrane through electrostatic interactions. The particles then formed clusters at the membrane surface.
Next, the membrane developed highly curved structures. These changes progressed to disruption and ultimately to rupture of the complete membrane surrounding the model endosome.
A central observation was that the mRNA cargo was released only when the model membrane had undergone complete lysis. In this context, lysis means that the membrane was fully broken apart. The result separates two events that are often considered together: destabilising the membrane and releasing the nanoparticle’s cargo.
That distinction matters because a nanoparticle can begin reshaping or weakening a membrane without immediately releasing its mRNA. A delivery system therefore needs to be studied across the individual stages of membrane engagement, deformation and rupture rather than being characterised only by a single overall escape measurement.
Why the pH result matters for nanoparticle design
Ionisable lipids are designed to change their charge depending on acidity. This behaviour helps LNPs remain suitable for formulation and circulation while allowing them to interact with the more acidic environment inside an endosome. The apparent pKa is commonly used to describe the pH at which the lipid’s charging behaviour changes.
In this experiment, the identical apparent pKa measured for LP01 and BiP-20 did not correspond to identical membrane-engagement behaviour. The difference between pH 5.8 and pH 5.1 shows that other properties of the nanoparticle–membrane system also influence when productive interaction begins. These may include the arrangement of lipids in the particle and the physical structure of the membrane interface, although the supplied preprint abstract does not identify the contribution of each factor.
The authors present the GUV system as a cell-free platform for separating the sub-steps of endosomal escape. Such a platform can make it possible to examine membrane interaction directly, before adding the biological complexity of whole cells or an animal model.
The study is a preprint, and its evidence comes from model membranes rather than a cellular, animal or clinical delivery study. Its immediate contribution is therefore a mechanistic framework for investigating how LNP composition and membrane properties shape cargo release, not a demonstrated improvement in gene-editing performance in living systems.