Fusion of single proteoliposomes with planar, cushioned bilayers in microfluidic flow cells

Abstract

Many biological processes rely on membrane fusion, and therefore assays to study its mechanisms are necessary. Here we report an assay with sensitivity to single-vesicle, and even to single-molecule events using fluorescently labeled vesicle-associated v-SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) liposomes and target-membrane-associated t-SNARE-reconstituted planar, supported bilayers (t-SBLs). Docking and fusion events can be detected using conventional far-field epifluorescence or total internal reflection fluorescence microscopy. In this assay, fusion is dependent on SNAP-25, one of the t-SNARE subunits that is required for fusion in vivo. The success of the assay is due to the use of: (i) bilayers covered with a thin layer of poly(ethylene glycol) (PEG) to control bilayer-bilayer and bilayer-substrate interactions, and (ii) microfluidic flow channels that present many advantages, such as the removal of nonspecifically bound liposomes by flow. The protocol takes 6-8 d to complete. Analysis can take up to 2 weeks.

Figure 1

The experimental setup. (A) Schematic of a v-SUV and a t-SBL. (B) Schematic drawing of the microfluidic flow system. (C) Photograph of the assembled microfluidic device on the microscope stage. Various components are labeled. (D) Close-up view of the assembly.

Figure 2

Verifying the fluidity of an SBL. Inset shows a sequence of images acquired before (1^st frame) and after bleaching a region delimited by the field diaphragm (closed to its minimum, ~34 μm across). The bleached region is automatically detected from the pre-bleach picture and its normalized fluorescence as a function of time after bleaching is plotted. Three other measurements (two from different channels on the same coverslip and one from a different coverslip) are also shown to illustrate the level of reproducibility. A representative fit to a function of the form f = a{exp(−2τ /t)·[I_o(2τ /t)+I₁(2τ /t)]}, where I_o and I₁ are modified Bessel functions and τ = w² /(4D) is the characteristic diffusion time over a circle of radius w⁵³, is shown. The best fit parameters are a=1.13 and τ = 30s (R²=0.99), implying a diffusion coefficient D = 2.4 μm²/s.

Figure 3

SUV-SBL fusion. (A) Far-field epifluorescence. Frames are 100 ms apart. (B) TIRFM. Frames are 17 ms apart. Note that individual LR-PE lipid labels become discernible as they diffuse sufficiently apart from one another. Their mobility can be quantified from single-molecule tracking, which yields a diffusivity of D ≈ 2 μm²/s.

Figure 4

Microfabrication of the SU-8 template for the PDMS flow cell. (A) Design of a 4-channel flow cell. On a 10cm wafer substrate (outline indicated by circle) 6 identical designs are placed. The outlines of 24mm × 60mm coverslips are shown as dashed lines. (B) Close-up view of one of the 4-channel designs. The distance between the inlet and outlet of a channel is 1cm and the width of each channel is 300μm. The channels are staggered and a reservoir is inserted to minimize the strain near the inlets and outlets. (C) A finished product. The template features are ~75μm high.

Figure 5

Making of the PDMS block. (A) The template from Figure 4 is placed in a petri dish and covered by a ~5mm thick PDMS that is cross-linked. A piece of PDMS is cut out (B) to be punched holes at the inlets and outlets (C). The hole has to be drilled all the way and the PDMS piece that enters the puncher's bore must be removed before retrieving the puncher (D). (E) Insertion of the tubing. The tubes must be inserted only to ~1/3 of the height of the PDMS to avoid excessive strain at the inlet and outlets (which leads to leakage).

Figure 6

Coverslip cleaning. (A) The coverslips are placed into a custom-made all-teflon holder (no glue used) which is placed inside a modified pyrex beaker with lid. (B, C) Detail of the teflon holder. Notice the wide notches and holes on the sides to avoid liquid stagnation spots. (C) Handling of the teflon holder using curved tweezers.

Figure 7

Analysis of data using PointPicker and MatLab. The data shown here were acquired using far-field epifluorescence with a time resolution of 100 ms. (A) Screenshot of PointPicker. Clicking the mouse on a vesicle marks it with a cross whose x, y and frame coordinates are saved into a text file. PointPicker files from many movies are read and analyzed in batch using the MatLab programs EK_SUVSBL_Fdotbatch.m and EK_SUVSBL_Ddotbatch.m which calculate the survival probability beyond a given delay after docking (B), the cumulative docking and fusion events as a function of time (C) and the averaged docking and fusion rates (D). See text and the supplementary material for more information.

Figure 8

Analysis of TIRFM data (17ms time resolution) using SpeckleTrackerJ and the MatLab program EK_Analyze_EKtraj_F_batch.m provided as supplementary material. (A) Screenshot of SpeckleTrackerJ. (B) The survival probability after docking, in an experiment where the lipid composition (in mole %) was PC/PS/PE/chol/LR-PE/mPEG2000PE=22.5/11.6/15.4/46/0.6/3.9 for the v-SUVs and PC/PS/PE/chol/PI(4,5)P2/NBD-PE/mPEG2000PE=18.9/11.6/15.4/46/3.9/0.4/3.9 for the t-SBL.