Characterizing lipid nanoparticle self-assembly and structure using coarse-grained simulations

Approach

Schrödinger scientists carried out coarse-grained (CG) simulations using the Schrödinger Materials Science Suite, Desmond molecular dynamics engine, and OPLS4 force field, following the steps below:

1. Automated parameterization of CG model for mRNA-containing LNP components using a 1.5 microsecond atomistic reference simulation employing the OPLS4 force field
2. Constructed a homogeneous system containing CG mRNA, lipids (with molar composition 45.6/9.4/43.4/1.6 mol % of ALC-0315/DSPC cholesterol/ALC-0159), and water using the new CG model
3. Relaxed the system and conducted Langevin dynamics (LD) simulations to investigate evolution of LNP morphologies
4. Conducted LD simulations with progressively higher levels of ionizable lipid (ALC-0315) neutralization starting from the final self-assembled LNP, in order to study the changes in internal morphology as the pH rises from the acidic LNP formation conditions to the pH characteristic of blood

Outlook

This work by Grzetic et al. demonstrates that self-assembly of LNPs for nucleic acid delivery can be simulated. The predictive power of the presented model opens the possibility of simulating larger systems with longer RNA sequences and enables investigation of the dependencies of LNP structure on lipid composition and other factors, rapidly providing information that can complement both experimental efforts as well as atomistic simulations.

Characterization of the internal structure of the LNP with 90% of the ALC-0315 protonated at low pH (panels a-c) and with ALC-0315 completely neutralized at neutral pH (panels d-f). Panels b and e show radial density profiles measured from the LNP center, and panels c and f show radial density profiles measured from RNA backbone sites.*

*Reprinted with permission from Mol. Pharmaceutics 2024, 21, 4747-4753. Copyright 2024 American Chemical Society.