Category: Manufacturing and Bioprocessing
Purpose: Liposomes are used as a carrier for drugs because of their important properties for both formulation stability and pharmaceutical application. The continuous manufacturing of liposomes avoids many issues associated with batch processing; therefore, we have focused on the continuous process in which two liquid flows are controllably mixed using a jet in a coaxial flow. The inner flow consists of a solution of lipids in ethanol and the exterior flow is an aqueous phase. The intermolecular interactions, in micro-scale, and fluid flow effects, in macro-scale, determine the properties of produced liposome. We have carried out multi-scale computational studies to understand the underlying mechanism as well as details of the liposome formation process.
Methods: We used both computational fluid dynamics (CFD) and molecular dynamics (MD) to model the process by continuum and molecular approaches. For the CFD work, both RANS and LES models were developed to model the turbulent jet in co-flow using COMSOL. For the MD models, we compared MARTINI, DRY-MARTINI, all-atom and force fields developed using MagiC (a multi-scale coarse-grained modelling package). All MD simulations were run using GROMACS.
Results: CFD simulation: A variety of RANS and LES models were used for modeling the jet flow and were compared with the corresponding experimental observations. LES models have shown comparable results with our experimental data.
MD simulation: Initially, we used MARTINI and DRY-MARTINI force fields to study lipid interactions by coarse-grained MD in both ethanol and aqueous solvent; however, these simulations showed deviation (e.g. lipid clustering in ethanol) when compared with the all-atom MD simulations. Consequently, MagiC was used for calculating coarse-grained force fields, and the simulation showed solubility of lipids in ethanol and pro-liposome formation, which indicates the successfully development of reliable force fields.
Conclusion: Discrete and continuum computational modeling streamline the understanding of the detailed mechanism of liposomal formation. The simulations are very helpful in analyzing parameters and effectors that are elusive from experimental means. CFD modeling can simulate temperature and concentration profiles (Fig. 1); whereas MD modeling reveals the molecular interactions via liposomal formation (Fig. 2). The formation temperature predicted by CFD calculation has less than 5% deviation from our experiment.
Antonio Costa– Assistant Research Professor, University of Connecticut, Storrs Mansfield, Connecticut
Xiaoming Xu– Senior Staff Fellow, U. S. Food and Drug Administration, Silver Spring, Maryland
Celia Cruz– Division Director, FDA/CDER/OPQ/OTR/DPQR, Silver Spring, Maryland
Su-Lin Lee– Science Staff, USFDA, Silver Spring, Maryland
Diane Burgess– Distinguished Professor of Pharmaceutics, University of Connecticut, Storrs Mansfield, Connecticut
Bodhi Chaudhuri– Associate Professor, University of Connecticut, Storrs, Connecticut