(642f) Computational Modeling of Liposome Formation in a Continuous Jet Flow Process Using CFD and MD

Purpose: Liposomes are spherical vesicles composed of a bilayer and/or a concentric series of multiple bilayers separated by aqueous compartments formed by amphipathic molecules such as phospholipids that are commonly used as carriers for drug delivery purposes. The current study focused on the continuous liposome processing system developed at UConn, where two liquid flows are mixed under a controlled manner through a coaxial jet in co-flow arrangement, of which the center flow consists of an ethanol solution of lipids and the exterior flow is an aqueous solution. It was suspected that the intermolecular forces among various components (e.g., ethanol-lipid-water) along with the unique fluid dynamics (e.g., jet flow) caused lipid aggregation and subsequent liposome formation. However, the underlying mechanism as well as the process dynamics are yet fully understood. Accordingly, we have carried out a multi-scale computational study of liposome formation in coaxial turbulent jet flow to probe the underlying mechanism and to quantitatively predict the formation of liposomes.

Methods: Both computational fluid dynamics (CFD), as a macro-scale simulation, and coarse-grained molecular dynamics (CG-MD), as a micro-scale investigation, have been conducted to not only reveal the detailed mechanism of liposome formation, but also implement multi-scale case studies for the process. CFD simulations were verified by comparison of flow pattern as well as formation temperature with experiments.

Results: The CG-MD simulations revealed that MARTINI force field (FF) could not capture a realistic behavior of lipids and cholesterol in ethanol solution due to lipids aggregation, which caused deviation from both the experimental results and all-atom MD simulations. Further optimization of MARTINI FF with reference to all-atom MD simulations was performed using a versatile object-oriented toolkit for coarse-graining applications (VOTCA). The multi-scale simulations were found to be most accurate when compared to the experimental data and trends. The CFD simulation results suggested that including heat of mixing in the energy equation would be desired to obtain a formation temperature comparable with the experiments.

Conclusions: Default MARTINI potential energy parameters require optimization with reference to an all-atom MD simulation for properly modeling lipids in ethanol solution. Discrete and continuum computational modeling of liposome formation were found to be two viable complementary approaches which cover micro-scale and macro-scale, respectively.

Acknowledgements: FDA Grant# 1U01FD005773-01.

Disclaimer: This article reflects the views of the authors and should not be construed to represent FDA’s views or policies.