FRAP to Characterize Molecular Diffusion and Interaction in Various Membrane Environments

Abstract

Fluorescence recovery after photobleaching (FRAP) is a standard method used to study the dynamics of lipids and proteins in artificial and cellular membrane systems. The advent of confocal microscopy two decades ago has made quantitative FRAP easily available to most laboratories. Usually, a single bleaching pattern/area is used and the corresponding recovery time is assumed to directly provide a diffusion coefficient, although this is only true in the case of unrestricted Brownian motion. Here, we propose some general guidelines to perform FRAP experiments under a confocal microscope with different bleaching patterns and area, allowing the experimentalist to establish whether the molecules undergo Brownian motion (free diffusion) or whether they have restricted or directed movements. Using in silico simulations of FRAP measurements, we further indicate the data acquisition criteria that have to be verified in order to obtain accurate values for the diffusion coefficient and to be able to distinguish between different diffusive species. Using this approach, we compare the behavior of lipids in three different membrane platforms (supported lipid bilayers, giant liposomes and sponge phases), and we demonstrate that FRAP measurements are consistent with results obtained using other techniques such as Fluorescence Correlation Spectroscopy (FCS) or Single Particle Tracking (SPT). Finally, we apply this method to show that the presence of the synaptic protein Munc18-1 inhibits the interaction between the synaptic vesicle SNARE protein, VAMP2, and its partner from the plasma membrane, Syn1A.

Fig 1. Various lipidic platforms used in this work to study the mobility and interaction of lipids and proteins within membranes.

Supported lipid bilayers are formed by the Langmuir-Blodgett deposition technique, which allows the formation of asymmetrical membranes (inner and outer monolayers can be of different lipid compositions). Giant liposomes are free standing membranes of 10–100 μm diameter. Sponge phases consist of interconnected bilayers forming aqueous channels whose diameter can be varied from 6 to 30 nm.

Fig 2

Fluorescence recovery of (A) DOPE-Rho or (B) DOPE-NBD lipids in the outer monolayer of a supported DOPC lipid bilayer following photobleaching of (A) a fluorescent disk of d = 20 μm diameter or (B) a pattern of fluorescent fringes separated by f = 20 μm. The recovery curve (result of 1 bleaching experiment on the same region in A and average of 3 bleaching experiments on the same region in B) can be fitted with (A) a Bessel function or (B) an Exponential function. (C) The contribution of the intrinsic photobleaching occurring during fluorescence reading (bleach) in the case of DOPE-NBD lipids can be removed in the fringe system by using the average of fluorescence gain in the dark fringes and fluorescence loss in the bright fringes. No intrinsic photobleaching was observed with DOPE-Rho lipids.

Fig 3

Diffusion coefficient (D) of DOPE-Rho lipids in the outer leaflet of a supported DOPC lipid bilayer deduced by varying the area of the bleached region (disks of diameters d = 2, 5, 10 or 20 μm in A; networks of fringes with an interfringe f = 5, 10 or 20 μm in B). The linear relationship between the bleaching area s² (d² or f²) and the recovery time (τ) proves that lipid diffusion is controlled by Brownian motion. The diffusion coefficient (D) is calculated from the slope of this straight line: D = d² / 16 τ in the case of the disk system, and D = f² / 4π² τ in the case of the fringe system. Each data point in A or B is the average of N = 3 independent bleaching experiments on different regions of the same bilayer, and the error bars correspond to the standard deviations from these averages. These experiments were reproduced with 4 independent (freshly prepared) bilayers using either the disk- or the fringe-shaped bleaching geometry, leading to D = (1.9 ± 0.3) μm²/s.

Fig 4

Experimental set-up used to perform FRAP experiments on (A) giant liposomes or (B) sponge phases. (A) To prevent liposome movement during FRAP measurements, giant liposomes are confined in a closed chamber of ~500 μm height (not shown here) or ideally (as shown here) held through micromanipulation. Fluorescence bleaching and recovery measurements are performed at the top (pole) of the giant liposome, which appears as a fluorescent disk in the confocal microscope. In this system, FRAP measurements are performed using the disk-shaped geometry. To ensure that the bleached spherical cap can be treated as a disk, the diameter of the bleaching disk should not exceed 25% of the giant liposome diameter. The confocal picture on the right shows the result of bleaching a 7 μm diameter fluorescent disk in the membrane of a DOPC:DOPE-Rho (99:1) giant liposome of 45 μm diameter. (B) Sponge phases (3 μL solution) are sandwiched between two glass surfaces (1.5 cm diameter) to form a liquid layer of ~20 μm height; this allows bleaching at all z-values within the sponge phase and thus measuring 2-dimensional fluorescence recovery. Here, bleaching is performed using the fringe-shaped geometry (interfringe f = 12 μm). The two systems were used to study the mobility of DOPE-Rho lipids. In both cases, DOPE-Rho lipids diffused faster than in supported lipid bilayers (D = 3.7 ± 0.5 μm²/s in the membrane of giant liposomes and D = 4.1 ± 0.4 μm²/s in sponge phases), which we attribute to the reduction of friction forces between the outer and the inner leaflets when the membrane is not linked to a solid substrate.

Fig 5. Interaction between SNARE proteins in a sponge phase.

The full length Syn1A protein and the cytoplasmic domain of FITC-labeled VAMP2 protein (cdVAMP2*) were reconstituted into two separate sponge phases (at the respective lipid-to-protein molar ratios of 20,000 and 80,000). The Syn1A sponge was pre-incubated for 1 hour at room temperature with or without Munc18-1 protein (1:1 molar ratio between Syn1A and Munc18-1). The Syn1A ± Munc18-1 sponge and the cdVAMP2* sponge were then mixed and allowed to interact for 1 hour at room temperature. In the absence of Munc18-1 (blue fitting curve), cdVAMP2* displays two diffusion coefficients: a slow diffusion coefficient (2.0 ± 0.1 μm²/s) corresponding to cdVAMP2* bound to Syn1A in the sponge membrane and a fast diffusion coefficient (9.7 ± 1.4 μm²/s) corresponding to free (unbound) cdVAMP2* in the sponge channels. In the presence of Munc18-1 (red fitting curve), cdVAMP2* displays a single, fast, diffusion coefficient (6.4 ± 1.1 μm²/s), as observed when it is added to a protein-free sponge phase, showing that Munc18-1 inhibits the interaction between Syn1A and cdVAMP2* (see also Table 3).