Single-molecule studies of the neuronal SNARE fusion machinery

Abstract

SNAREs are essential components of the machinery for Ca(2+)-triggered fusion of synaptic vesicles with the plasma membrane, resulting in neurotransmitter release into the synaptic cleft. Although much is known about their biophysical and structural properties and their interactions with accessory proteins such as the Ca(2+) sensor synaptotagmin, their precise role in membrane fusion remains an enigma. Ensemble studies of liposomes with reconstituted SNAREs have demonstrated that SNAREs and accessory proteins can trigger lipid mixing/fusion, but the inability to study individual fusion events has precluded molecular insights into the fusion process. Thus, this field is ripe for studies with single-molecule methodology. In this review, we discuss applications of single-molecule approaches to observe reconstituted SNAREs, their complexes, associated proteins, and their effect on biological membranes. Some of the findings are provocative, such as the possibility of parallel and antiparallel SNARE complexes or of vesicle docking with only syntaxin and synaptobrevin, but have been confirmed by other experiments.

Figure 1. A typical smFRET experiment

A. Shown is the experimental setup for the smFRET experiments of the binary complex (29). Briefly, dual dye (donor/acceptor) labeled binary complex (syntaxin·SNAP-25) was reconstituted into a supported bilayer. Evanescent wave illumination was performed through total internal reflection. Laser light was chosen at two wavelengths to monitor donor and acceptor fluorescence. Synaptobrevin or other factors were injected and binding to binary complex monitored by a change in FRET from the dual labeled syntaxin·SNAP-25. A similar setup was used for docking and fusion experiments where synaptobrevin was reconstituted into liposomes that contained the soluble dye calcein that served as a content mixing indicator (30). B. Donor (green) and acceptor (red) dye labeling positions in the dual-labeled SNAP-25 molecule that forms the binary complex with syntaxin. Shown is a model of the three-helix bundle complex consisting of the SNARE core domains of syntaxin (orange), SNAP-25 SN1 (green), and SNAP-25 SN2 (red). C. Fields of view (50 × 100 μm) of donor (left) and acceptor (right) fluorescence arising from a dual-labeled SNAP-25 molecule in the syntaxin·SNAP-25 binary complex. D. Selected time trace of the donor and acceptor fluorescence arising from a co-localized spot (similar to the marked one in panel C). In this case, synaptobrevin is bound to the binary complex, so high FRET is observed until the acceptor photobleaches.

Figure 1. A typical smFRET experiment

A. Shown is the experimental setup for the smFRET experiments of the binary complex (29). Briefly, dual dye (donor/acceptor) labeled binary complex (syntaxin·SNAP-25) was reconstituted into a supported bilayer. Evanescent wave illumination was performed through total internal reflection. Laser light was chosen at two wavelengths to monitor donor and acceptor fluorescence. Synaptobrevin or other factors were injected and binding to binary complex monitored by a change in FRET from the dual labeled syntaxin·SNAP-25. A similar setup was used for docking and fusion experiments where synaptobrevin was reconstituted into liposomes that contained the soluble dye calcein that served as a content mixing indicator (30). B. Donor (green) and acceptor (red) dye labeling positions in the dual-labeled SNAP-25 molecule that forms the binary complex with syntaxin. Shown is a model of the three-helix bundle complex consisting of the SNARE core domains of syntaxin (orange), SNAP-25 SN1 (green), and SNAP-25 SN2 (red). C. Fields of view (50 × 100 μm) of donor (left) and acceptor (right) fluorescence arising from a dual-labeled SNAP-25 molecule in the syntaxin·SNAP-25 binary complex. D. Selected time trace of the donor and acceptor fluorescence arising from a co-localized spot (similar to the marked one in panel C). In this case, synaptobrevin is bound to the binary complex, so high FRET is observed until the acceptor photobleaches.

Figure 2. The neuronal SNARE complex

A. Primary structure diagram for syntaxin (red), SNAP-25 (green), and synaptobrevin (blue). The experiments referenced in this review refer to the following isoforms: syntaxin 1A, synoptobrevin II, and SNAP-25A which we simply refer to as syntaxin, synaptobrevin, and SNAP-25. TM: transmembrane domain. The SNARE core domains are defined through the 16 layers as found in the crystal structure of the neuronal SNARE complex (49). For SNAP-25, the palmitoylation sites are indicated by green lines. B. X-ray crystal structure of the core of the neuronal SNARE complex consisting of synaptobrevin (blue), SNAP-25 (green), and syntaxin (red) (PDB ID 1SFC) (49). This structure represents the fully folded post-fusion state of the complex, also referred to as the cis-state. The N- and C-terminal sides of the core complex are indicated. C. Model of the trans state of two SNARE complexes that dock a liposome to a supported bilayer in vitro. This model was obtained by modifying the membrane proximal end of the crystal structure of the neuronal SNARE complex in order to allow the transmembrane domains to enter into the juxtaposed membranes. The transmembrane domains were assumed to be helical (156). The connecting region between the transmembrane domains and the core complex are likely flexible. Two SNARE complexes are shown; the exact number is unknown, but 1–2 SNARE complexes suffice to dock liposomes (30).

Figure 3. Single molecule FRET (smFRET) studies of the binary complex

A. Selected time trace of the donor and acceptor fluorescence of the binary complex (syntaxin·SNAP-25, see Figure 1) arising from a co-localized spot (similar to the marked one in panel C) (29). Note, the switching between two different FRET states as indicated by the correlated changes in donor and acceptor fluorescence. B. FRET distributions of donor and acceptor dyes on the binary complex before (left panel) and after addition of synapbrevin (right panel). Note, that the intermediate FRET states have disappeared after addition of synaptobrevin (29). Below the FRET distributions, models of the binary complex conformations are shown. Before the addition of synaptobrevin, the binary complex exhibits three configurations: only the SN2 SNAP-25 domain bound to syntaxin (SX-SN2), only the SN1 domain of SNAP-25 bound to syntaxin (SX-SN1) or both SNAP-25 SNARE domains bound to the syntaxin SNARE domain (SX-SN1-SN2). Upon addition of synaptobrevin (right) or accessory proteins, these configurations collapse into the SX-SN1-SN2 configuration.

Figure 4. Alternative pathways for liposome fusion

Computer simulations of liposome-membrane fusion (88). Pathway 1 shows the canonical progression from an unfused starting state through a stalk intermediate and a hemifused diaphram intermediate to the fully fused state. Pathway 2 shows an alternative reaction pathway observed in ensemble molecular dynamics simulations (88): rapid fusion from the stalk intermediate to the fully fused state. All renderings are of snapshots from observed reaction trajectories; lipids are colored to distinguish the outer (red and green) and inner (gold and blue) lea ets of each vesicle. Explicit water was present in all simulations but not rendered.

Figure 5. Individual fusion event observed by single particle studies

Liposomes containing the content dye calcein were reconstituted with dye-labeled synaptobrevin molecules and then introduced above PC/PS bilayers with reconstituted syntaxin in complex with SNAP-25 (30). The images represent a single 11 μm by 11 μm patch of membrane with docked liposomes observed in two different spectral ranges to detect the content dye and synaptobrevin dye fluorescence. Two liposomes are docked to the bilayer in the field of view as indicated by the synaptobrevin dyes. A single fusion event occurs at 6 secs as indicated by the sudden appearance of a bright content dye signal. The increase of content dye fluorescence is due to dequenching. Fusion proceeds faster than the time resolution of the camera used in this experiment; in other words the fusion reaction is faster than 100 msec.