One SNARE complex is sufficient for membrane fusion

Abstract

In eukaryotes, most intracellular membrane fusion reactions are mediated by the interaction of SNARE proteins that are present in both fusing membranes. However, the minimal number of SNARE complexes needed for membrane fusion is not known. Here we show unambiguously that one SNARE complex is sufficient for membrane fusion. We performed controlled in vitro Förster resonance energy transfer (FRET) experiments and found that liposomes bearing only a single SNARE molecule are still capable of fusion with other liposomes or with purified synaptic vesicles. Furthermore, we demonstrated that multiple SNARE complexes do not act cooperatively, showing that synergy between several SNARE complexes is not needed for membrane fusion. Our findings shed new light on the mechanism of SNARE-mediated membrane fusion and call for a revision of current views of fusion events such as the fast release of neurotransmitters.

Figure 1

Characterization of the SNARE-containing liposomes. (a) Synaptobrevin 2 (1–96; black arrow), synaptobrevin 2 (green) and the synaptobrevin 2 (49–96) stabilized acceptor complex (blue) consisting of SNAP-25 (purple) and syntaxin 1A (183–288, 289C; red) were free of contaminants when analyzed by SDS-PAGE (Coomassie Blue staining). (b–c) The absorption spectrum of 18 µM synaptobrevin (b) and the acceptor complex (c; green curves), both labeled with Texas Red. The absorption of 18 µM unlabeled protein (black) and Texas-red (red) and the sum of both (pink) are also shown. The overlap between the pink and the green curves indicates stoichiometric labeling. (d) Negative staining electron microscopy (inset; scale bar 100 nm) showed that the proteins were reconstituted in monodisperse liposomes with an average radius of 17.9 ± 5.8 nm (s.d.). (e–f). All proteins were fully incorporated in the membrane as verified by density gradient flotation on a 30–80% nycodenz gradient. Partial proteolysis was used to determine the orientation of the proteins in the liposomes. In the absence of detergent, only correctly oriented SNAREs (i.e. with their cytoplasmic domains outside) are cleaved by trypsin, whereas all proteins are proteolysed in the presence of Triton X-100. About 50% of the synaptobrevin (e) and 80% of the acceptor complex (f) were correctly oriented in the liposomes. Electron microscopy, flotation gradients, and trypsin digestion were described previously.

Figure 2

SNARE distribution over the liposomes. (a) Two-photon excitation microscopy image of liposomes containing Texas-red labeled synaptobrevin. (b) Cross-section of the image (red line in a). The liposomes were selected using an offset (dotted line) as described. (c) Liposome recognition; each liposome was assigned a random colour. Scale bar, 5 µm. (d) Sequential photobleaching of the liposomes indicated in c (green). A Chung-Kennedy non-linear filtering technique corrected for fluorescence intermittency (black). These liposomes contained 4, 2, and 1 synaptobrevin molecules (red arrows). (e–f) Distribution of synaptobrevin (e) and syntaxin (f) for the p/l-ratios indicated in the figure (bars; left axis). Fitting the data with Poisson distributions (blue curves; right axis) indicated that the SNARE molecules were randomly distributed over the liposomes, with averages of 1.4, 0.8 and 0.4 synaptobrevin and 1.2, 0.7 and 0.35 syntaxin molecules per liposome, respectively.

Figure 3

Bulk lipid mixing as a function of SNARE density. (a) Scheme of the lipid mixing experiment where liposomes containing Texas-red-PE (red) and the acceptor SNARE complex (blue) fuse with liposomes containing Oregon-green-PE (green) and synaptobrevin (orange), resulting in quenching of the Oregon-green donor fluorophore. (b) Lipid mixing of 1:1,000 synaptobrevin liposomes with acceptor complex liposomes at the p/l-ratios indicated. The total lipid concentration in the cell was 50 µM, corresponding to approximately 4 nM liposomes as determined by fluorescence correlation spectroscopy. Recordings were made at 20.0°C. (c–d) Fusion of 1:16,000 synaptobrevin to (c) 1:1,000 and (d) 1:16,000 acceptor complex liposomes; in all cases lipid mixing was observed. (e) Lipid mixing experiment of 1:1,000 synaptobrevin to 1:1,000 syntaxin 1 (residues 183–288) liposomes. In the absence of SNAP-25, 10% fusion compared to the acceptor complex (panel b) was observed. Curves are normalized (left axis); real Oregon-green fluorescence is indicated as a fraction of Triton-X100 controls (right axis). For reference, the shape of the highest concentration curve is shown in all subsequent panels (red). Lipid mixing was SNARE specific and could be inhibited with 10 µM synaptobrevin 2 (residues 1–96; green curves) or with a combination of 10 µM SNAP-25 and the soluble SNARE domain of syntaxin 1 (residues 183–263; orange). Moreover, liposomes containing no acceptor SNARE complex or synaptobrevin did not show membrane lipid mixing (blue curves; b and c, respectively). Error bars indicate triplicates (s.d.).

Figure 4

Liposomes containing a single SNARE participate in membrane fusion. (a1) The fluorescence change after 25 min for the lipid mixing reactions of figure 3b and (a2) the number of acceptor complexes from figure 2f as a function of the protein-to-lipid ratio. Note that the figures are arranged counter-clockwise. (a3) The number of acceptor complexes relative to the fraction of liposomes containing less than 1 (blue), 2 (purple), or 3 (orange) SNAREs (data from Fig. 2f). The number of SNAREs for the 1:1,000 and 1:2,000 acceptor complex liposomes could not be determined with sequential photobleaching, because at these high protein-to-lipid ratios the error of the sequential photobleaching increases progressively. Therefore, these were estimated using Poisson distributions (a3, dotted lines) [correct? There are 3 dotted lines in a3: YES 3 dotted lines, one for each distribution (e.g. less than 1, 2, or 3)] with the averages linearly extrapolated from the 1:4,000 liposomes (a2, dotted line) [correct? There is 1 dotted line in a2: yes only 1 dotted line here]. Closed symbols indicate measured and open symbols indicate extrapolated data points. (a4) The fluorescence change as a function of the fraction of liposomes containing no (blue), 1 or less (purple), or 2 or less (orange) SNAREs. The data for multiple SNAREs (purple and orange) did not scale linearly (dotted lines; fits of measured data), whereas the data for empty liposomes (blue) scaled relatively well. Thus, the fluorescence change decreased linearly (dotted line) to the fraction of empty liposomes, indicating that all liposomes that contained one or more SNARE molecules participated in fusion. (b) Size distribution of a 1:1 ratio of 1:1,000 synaptobrevin and 1:16,000 acceptor complex liposomes determined with negative staining electron microscopy (n = 2,887). A slight increase in size was observed upon fusion (n = 1,774). The increase was only small, because the radius increases only 1.4-fold (√2) upon fusion and because only 30% of the acceptor complex liposomes contained SNARE proteins (Fig. 2f).

Figure 5

Fusion with purified synaptic vesicles, content mixing, and lipid mixing as a function of liposome concentration. (a–b) Lipid mixing experiment, where 1:1,000 and 1:16,000 acceptor SNARE complex liposomes (50 µM total lipid) containing both Texas-red-PE and Oregon-green-PE were fused to purified synaptic vesicles (8.5 µg total protein). Fusion results in lipid mixing and dequenching of the Oregon-green donor fluorophore. The experiment indicates that a single SNARE complex is sufficient to drive fusion of native biological membranes. (c) Scheme of the content mixing assay based on calcein (green) fluorescence dequenching upon membrane fusion. (d) At p/l-ratios of 1:1,000 and 1:16,000 of the acceptor SNARE complex and 1:1,000 of synaptobrevin, content mixing was observed. The experiment shows that a single SNARE complex is sufficient not only for lipid mixing but also for content mixing. (e–f) Lipid mixing experiment where 50 µM (total lipid) of Texas-red-PE synaptobrevin liposomes were fused with Oregon-green-PE acceptor SNARE complex liposomes at the lipid concentrations indicated in the figure. (g) Fusion of 50 µM Oregon-green-PE acceptor complex to 3.2 µM Texas-red-PE synaptobrevin liposomes. These experiments indicate that the rate of membrane fusion is only weakly dependent on the liposome concentration. For reference, the shape of the highest concentration curves is shown in all subsequent panels (red). Curves are normalized (left axis); real fluorescence is indicated as a fraction of Triton-X100 controls (right axis). Fusion was inhibited with 10 µM synaptobrevin 2 (residues 1–96; green curves). Error bars indicate triplicates (s.d.).

Figure 6

Formation of SNARE complexes monitored by C-terminal FRET. (a) Scheme of the FRET experiment to measure complex formation of Texas-red labeled syntaxin with Alexa-fluor 488 labeled synaptobrevin. (b) Complex formation of 1:1,000 synaptobrevin liposomes with the stabilized acceptor complex at the p/l-ratios indicated in the figure (50 µM total lipid). Data is presented as emission of the Texas-red acceptor fluorophore (sensitized emission), so initial fluorescence is proportional to the acceptor complex levels. (c) Reversing the label did not influence the curves. The experiment shows that the rate of complex formation is not dependent on the SNARE density and indicates that there is no cooperativity in SNARE complex formation. Curves are normalized (left axis); real fluorescent signals of the acceptor fluorophore are indicated (right axis). Fusion was inhibited with 10 µM 1–96 synaptobrevin (green curves). For reference, the shape of the highest concentration curves is shown in all subsequent panels (red). (d) The increase in the Texas-red emission after 25 min for the reactions of panel b as a function of the p/l-ratio. The linear correlation (dotted line) indicates that the total amount of core-complex formed was directly dependent on the concentration of SNAREs in the cell. Error bars indicate triplicates (s.d.).