Shrink-wrap vesicles

Abstract

We describe a simple approach to the controlled removal of molecules from the membrane of large unilamellar vesicles made of fatty acids. Such vesicles shrink dramatically upon mixing with micelles composed of a mixture of fatty acid and a phospholipid (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC)), as fatty acid molecules leave the vesicle membrane and accumulate within the mixed micelles. Vesicle shrinkage was confirmed by dynamic light scattering, fluorescence recovery after photobleaching of labeled vesicles, and fluorescence resonance energy transfer between lipid dyes incorporated into the vesicle membrane. Most of the encapsulated impermeable solute is retained during shrinkage, becoming concentrated by a factor of at least 50-fold in the final small vesicles. This unprecedented combination of vesicle shrinkage with retention of contents allows for the preparation of small vesicles containing high solute concentrations, and may find applications in liposomal drug delivery.

Figure 1

Turbidity of MA vesicles after addition of MA/POPC (89:11) micelles (solid line) compared to addition of pure MA micelles (dotted line). The mixture contained 20 mM vesicles and 25 mM micelles. Micelles were added at t = 4 min.

Figure 2

Decrease in size of MA vesicles (3.2 mM, unless otherwise indicated) after mixing with MA/POPC micelles (89% MA, 11% POPC; 4 mM lipid, unless otherwise indicated). Solid lines show curve fits with parameters given in the text. (A) Diameter measured by dynamic light scattering. Empty circles: vesicles alone; filled circles: mixture. (B) Diameter relative to initial diameter, measured by FRET assay. Empty circles: vesicles alone; filled circles: mixture. (C) Diameter relative to initial diameter, measured by the calcein self-quenching assay. Data were compiled from seven separate shrinking experiments with initial [calcein] = 0.2, 2, or 20 mM. Vesicles alone showed <8% dye leakage and change in quenching efficiency after 22 hours. (D) Diameter of labeled MA/POPC micelles determined by FRAP. Empty circles: 40 mM micelles in buffer only; filled circles: mixture (20 mM micelles and 32 mM vesicles). Micelles in water have an apparent diameter of 8 nm by this assay. (E) Diameter of vesicles determined by FRAP (final diameter = 28 nm; k_FRAP = 0.004 min⁻¹). Empty circles: 32 mM vesicles alone; filled circles: mixture (20 mM micelles and 32 mM vesicles).

Figure 3

Shrinking of MA vesicles initially containing 0.5 mM HPTS. Filled circles: MA vesicles mixed with MA/POPC micelles; empty circles: vesicles alone. The average vesicle diameter decreased over time, as measured by DLS (A; solid line shows a single exponential decay curve fit, with k ∼ 0.0044 min⁻¹ and final diameter ∼23 nm). Encapsulated HPTS was released slowly (B), with a timescale much slower than the decrease of vesicle size.

Figure 4

Dependence of initial rate of vesicle shrinking on the concentration of reactants. (A) Van't Hoff plot indicating inhibition of the reaction rate by vesicles (slope ∼ −1). (B) Van't Hoff plot showing the effect of micelles on initial rate. The dotted lines are meant to guide the eye (slope ∼ 0.6 and 4), suggesting the presence of more than one reaction order. (C) Diagram of the transfer of MA (black) from vesicles encapsulating calcein (green) to micelles containing POPC (red). Transfer of MA to micelles might create a local membrane-like geometry, relieving strain in the micelles caused by the geometry of POPC.