Toward understanding lipid reorganization in RNA lipid nanoparticles in acidic environments

Abstract

The use of lipid nanoparticles (LNPs) for therapeutic RNA delivery has gained significant interest, particularly highlighted by recent milestones such as the approval of Onpattro and two mRNA-based SARS-CoV-2 vaccines. However, despite substantial advancements in this field, our understanding of the structure and internal organization of RNA-LNPs -and their relationship to efficacy, both in vitro and in vivo- remains limited. In this study, we present a coarse-grained molecular dynamics (MD) approach that allows for the simulations of full-size LNPs. By analyzing MD-derived structural characteristics in conjunction with cellular experiments, we investigate the effect of critical parameters, such as pH and composition, on LNP structure and potency. Additionally, we examine the mobility and chemical environment within LNPs at a molecular level. Our findings highlight the significant impact that LNP composition and internal molecular mobility can have on key stages of LNP-based intracellular RNA delivery.

Keywords: MD simulations; RNA delivery; endosomal escape; lipid nanoparticles; mRNA therapeutics.

Fig. 1.

Coarse-grained molecular dynamics (MD) simulations: setup. In a direct coexistence (DC) simulation, a rectangular segment of the nanoparticle is set in coexistence with the external aqueous medium (top right). The rectangular segment length (Lz) is chosen to ensure that its Surface-Area-to-Volume ratio (SA/V =2/Lz) is equivalent to that of the LNP (SA/V =3/R), where R is the radius of the LNP and D its diameter.

Fig. 2.

Concentration and charge density profiles for LNP1 and LNP2. (A) Molar concentration profiles as a function of the distance to the LNP center for LNP1 (thick lines) and LNP2 (thin lines); (B) charge density profile as a function of the distance to the LNP center for both LNP1 and LNP2; (C and D) concentration profiles for LNP1 and LNP2, respectively (1 nm in a spherical LNP is equivalent to 0.33 nm in a DC simulation); (E) close-up snapshot of the core-shell-water interphases for the studied LNPs.

Fig. 3.

Deciphering the relationship between the (modeled) LNP structures and their function on different cells. (A) HepG2 and (B) HeLa cells constitutively expressing fluorescently labeled Gal9 were treated with 1 µg/mL LNP containing Cy5-labeled, GFP-encoding mRNA and followed by live-cell imaging (1 picture per hour) for 24 h; (C) mRNA entry and translation was measured by quantification of Cy5 or GFP response in the region of the cells, and endosomal escape was assessed by quantification of Gal9 spots per region of cells. All data were normalized to t = 0 of the negative control (PBS) in the respective readout. Error bars = SD. Exemplary data from two independent experiments with similar outcome. (Scale bar (parts A and B), ~40 µm.)

Fig. 4.

Characterization of the internal nanostructures and chemical environment around RNA and water within LNPs. (A) Visualization of the reverse micellar LNP interior from a direct coexistence simulation (one extra periodic neighbor is included, perpendicular to the interface). The render highlights exterior water, DSPC (cyan tails) to visually reference the LNP surface, RNA (green), and its bound CIL (purple), whereas other components have been rendered transparent to enhance clarity; (B) snapshots of reverse water micelles (left: DSPC-rich; right: CIL-rich) taken from a simulation of the interior of an LNP; (C) snapshot of a reverse micelle around RNA taken from a simulation of the interior of an LNP; (D) radial distribution function of water molecules to LNP-components; (E) radial distribution function of RNA molecules to LNP-components. In the snapshots, only molecules in contact—within one bead’s diameter—with the water cluster or RNA molecule, respectively, are shown. The surface of water molecules has been rendered transparent for visualization.

Fig. 5.

The role of DSPC for LNP stability and RNA encapsulation. Variations of pH ranging from fully protonated ionizable lipid (purple) to unprotonated (A: acidic pH; B: neutral pH) did not lead to the release of RNA into the aqueous phase in the presence of DSPC (yellow). Conversely, removal of the DSPC layer (C) in an unprotonated environment led to RNA (green) leaving the LNP.

Fig. 6.

Molecular mobility and structural flexibility inside LNPs. (A) Effect of RNA concentration on molecular mobility in LNPs by comparison of the mean squared displacement (MSD) for the molecules found inside LNPs. Weight ratios: 7% RNA (filled), and 0.7% RNA (striped); (B) LNP encapsulation drives compaction of RNA and expansion of CIL molecules. Comparison of the radius of gyration (Rg) distributions for RNA and CIL molecules in the interior of an LNP (continuous) and in a free aqueous environment (striped).