Interaction of Choline-Based Ionic Liquids with Model Lipid Membranes: Force-Field Parametrization and Membrane Partitioning

In
recent years, ionic liquid (IL)-based formulations have gained
attention for their potential use in drug delivery and antibacterial
and antiseptic applications. Molecular dynamics simulations can provide
insights into complex interaction mechanisms, serving as valuable
tools to guide experimental efforts to design novel formulations.
However, to study interactions that involve micellar aggregates and
membrane partitioning dynamics, simulations should be able to access
time scales of several microseconds, requiring reliably parametrized
coarse-grained molecular models. In this study, we investigate the
interaction of choline–geranic acid (CAGE)-based ILs with a
model phospholipid membrane. In order to develop a coarse-grained
CAGE model, we carried out atomistic simulations with the GROMOS54a7,
CHARMM36m, OPLS, and OPLS-R force fields. The OPLS-R force field was
found to accurately predict experimental structural and dynamic properties
of CAGE molecules and was therefore used to parametrize the coarse-grained
models within the Martini 2 (M2) and Martini 3 (M3) frameworks. The
M3 model was in better agreement with both experimental observations
and atomistic simulations and captured the reported micellar phase
transition composition with increasing water content. In contrast,
the M2 model was found to overestimate the density, with a greater
tendency to form a lamellar phase. Using the newly parametrized M3
force field, free energy computations with a dipalmitoylphosphatidylcholine
(DPPC) lipid bilayer revealed a favorable free energy of partitioning
for geranic acid compared to geranate ions, while choline partitioning
was unfavorable. Geranate ion partitioning increased with a higher
concentration of geranic acid in the CAGE solution. Micellar aggregates
of geranic acid either released molecules in the extracellular space
for subsequent membrane uptake or underwent direct fusion with the
membrane. Both the area compressibility and order parameters decreased
with increasing geranic acid content, which also resulted in an increase
in the lipid area. The coarse-grained model developed in this study
allows us to study membrane partitioning, micellar breakup, and membrane
fusion events which occur on microsecond times scales. These models
can potentially be utilized to investigate the influence of CAGE-based
chemistry on membrane partitioning, thereby aiding in the development
of novel IL-based therapeutic formulations.