Crowding-induced morphological changes in synthetic lipid vesicles determined using smFRET

Abstract

Lipid vesicles are valuable mesoscale molecular confinement vessels for studying membrane mechanics and lipid-protein interactions, and they have found utility among bio-inspired technologies, including drug delivery vehicles. While vesicle morphology can be modified by changing the lipid composition and introducing fusion or pore-forming proteins and detergents, the influence of extramembrane crowding on vesicle morphology has remained under-explored owing to a lack of experimental tools capable of capturing morphological changes on the nanoscale. Here, we use biocompatible polymers to simulate molecular crowding in vitro, and through combinations of FRET spectroscopy, lifetime analysis, dynamic light scattering, and single-vesicle imaging, we characterize how crowding regulates vesicle morphology. We show that both freely diffusing and surface-tethered vesicles fluorescently tagged with the DiI and DiD FRET pair undergo compaction in response to modest concentrations of sorbitol, polyethylene glycol, and Ficoll. A striking observation is that sorbitol results in irreversible compaction, whereas the influence of high molecular weight PEG-based crowders was found to be reversible. Regulation of molecular crowding allows for precise control of the vesicle architecture in vitro, with vast implications for drug delivery and vesicle trafficking systems. Furthermore, our observations of vesicle compaction may also serve to act as a mechanosensitive readout of extramembrane crowding.

Keywords: FRET; TIRF; lipid vesicle; membrane mechanics; molecular crowding; single-molecule.

FIGURE 1

Sorbitol induces conformational changes in freely diffusing vesicles. (A) Normalized fluorescence emission spectra of DiI–DiD LUVs in the absence and presence of sorbitol (λ_ex = 532 nm). Inset: schematic illustration of the vesicles, where R corresponds to the mean dye–dye separation distance. (B) Corresponding variations in E _FRET and R. Heuristic fits shown are quadratic (solid black lines), determined with Python 3 using NumPy’s polyfit routine. (C) Amplitude weighted average lifetime of DiI and (D) FliptR as a function of sorbitol. Insets correspond to the time-resolved fluorescence decays in the absence (blue) and presence of 3M sorbitol (red). Solid black lines represent biexponential fits to the raw data, and the solid gray lines represent the instrument response functions.

FIGURE 2

Sorbitol induces compaction and undulations in single-lipid vesicles. Quantitative comparison of diameter distributions of POPC vesicles in (A) the absence and (B) presence of 3 M sorbitol. Insets: representative SEM images of immobilized vesicles under the respective conditions. Scale bars = 1 μm and 100 nm in the larger and smaller insets, respectively.

FIGURE 3

Morphological characterization of lipid vesicles by Cryo-TEM. Representative examples of (A–D) unilamellar (white arrows), (E–F) bilamellar (black arrows), and (G–H) multilamellar (yellow arrows) vesicles in the absence of sorbitol. Also shown are examples of (I) encapsulated vesicles (green arrows), (J–K) bowling pin-shaped vesicles, and (L) large conglomerates. (M–O) Representative examples of vesicles in the presence of 0.5 M sorbitol and (P) example of vesicle displaying irregular bulging. (Q–R) Representative examples of vesicles in the presence of 1 M sorbitol with (S) examples of regions of membrane damage (purple arrows).

FIGURE 4

Sorbitol induces irreversible structural changes in single surface-tethered vesicles. (A) Representative wide-field TIRF image of surface tethered vesicles composed of DiI and DiD. Donor and acceptor emission channels are shown on the left- and right-hand side of the dashed line, respectively. Inset: surface immobilization scheme. Single vesicles containing biotinylated lipids are immobilized onto a BSA-Biotin coated glass coverslip via NeutrAvidin. (B) Fluorescence intensity population histograms of DiI (green) and DiD (magenta), obtained from surface-tethered vesicles in the absence of sorbitol. (C) Representative time traces of DiI (green) and DiD (magenta), obtained from single surface-tethered vesicles with 0 mM, 1 M, and 3 M sorbitol. (D) Representative variation in the peak probe separation distance, <d>, as a function of sorbitol concentration. Also shown are fitting errors associated with the application of single Gaussian distributions to histograms of the probe-separation distance. (E) Corresponding variations in the FRET efficiency histograms obtained for N > 2,000 vesicles. (F) FRET efficiency histograms obtained from N > 2,000 vesicles at 0 mM sorbitol, after a 3 M sorbitol rinse step (middle panel), and after vigorous washing of the sample with imaging buffer (lower panel). The dashed red lines in (E) and (F) correspond to the peak positions of the FRET efficiency histograms obtained in the absence of sorbitol. The solid black lines represent single Gaussian fits.

FIGURE 5

High molecular weight crowders induce reversible vesicle compaction. Representative variations in the FRET efficiency histograms obtained for N > 2,000 vesicles in the absence and presence of (A) PEG 200, (B) Ficol 400, (C) PEG 400, and (D) PEG 8000 at 5%, 10%, 15%, and 20% (w/w) in 50 mM Tris buffer (pH 8), respectively. The dashed lines correspond to the peak positions of the FRET efficiency histograms obtained in the absence of a crowder. The solid black lines represent single Gaussian fits. (E) Comparative bar plot summarizing the maximum variation in FRET efficiency observed under different crowding conditions. (F) Representative FRET efficiency histograms obtained from N > 2,000 vesicles in the presence of PEG400 (blue, top panel) and PEG8000 (blue, lower panel) and after vigorous washing with 50 mM Tris (pH 8) buffer (purple). (G) Injection of high molecular weight PEG crowders leads to reversible vesicle compaction.