Rapid and efficient fusion of phospholipid vesicles by the alpha-helical core of a SNARE complex in the absence of an N-terminal regulatory domain

Abstract

A protease-resistant core domain of the neuronal SNARE complex consists of an alpha-helical bundle similar to the proposed fusogenic core of viral fusion proteins [Skehel, J. J. & Wiley, D. C. (1998) Cell 95, 871-874]. We find that the isolated core of a SNARE complex efficiently fuses artificial bilayers and does so faster than full length SNAREs. Unexpectedly, a dramatic increase in speed results from removal of the N-terminal domain of the t-SNARE syntaxin, which does not affect the rate of assembly of v-t SNARES. In the absence of this negative regulatory domain, the half-time for fusion of an entire population of lipid vesicles by isolated SNARE cores ( approximately 10 min) is compatible with the kinetics of fusion in many cell types.

Figure 1

Calibration of the fusion assay. (A) Schematic representation of SNARE-dependent liposome fusion reaction. Donor liposomes (v) contain an equimolar amount of NBD-PE (green beacons) and rhodamine-PE (red diamonds) and include the v-SNARE VAMP2. Each round of fusion decreases the concentration of fluorescent phospholipid in the bilayer, decreasing quenching of NBD (green beacons) and increasing the NBD fluorescence. (B) Calibration of percent of NBD fluorescence to rounds of fusion. Various lipid mixtures were prepared from the fluorescent donor lipids mix (used to prepare donor liposomes) and the nonfluorescent acceptor lipids mix in the following proportions (acceptor:donor): 0:1, 0.5:1, 1:1, 2:1, 4:1, and 8:1, mimicking the result (as shown in A) of 0, 0.5, 1, 2, 4, and 8 rounds of fusion of donor vesicles, respectively. These lipid mixtures were used to reconstitute VAMP2 into liposomes, and the NBD fluorescence in each set of re-isolated liposomes was expressed as a percent of the maximum fluorescence. The closed and open circles represent two independent experiments. (C) Normalized NBD fluorescence versus time of incubation. Shown is a standard fusion reaction (as described in A) measuring the increase in NBD fluorescence in which liposomes were either mixed and preincubated overnight a 4°C (closed circles), thereby allowing SNARE complexes to form before fusion at 37°C, or were not preincubated (open circles). (D) The same data as in C, now expressed as rounds of fusion of donor vesicles according to the calibration curve in B. (E) Consumption of predocked vesicles. The rounds of fusion of nonpredocked vesicles were subtracted from the rounds of fusion obtained from experiments in which predocked vesicles were used, and this difference was plotted versus time. To examine whether consumption of the predocked population behaved like a first order reaction, we represented the substrate of this reaction as 1 − [rounds of fusion (for the first round)] versus time (Inset).

Figure 2

Removal of the syntaxin H_ABC+hinge domain increases the rate of SNARE mediated liposome fusion. (A) Schematic representation of the membrane distal three-helical bundle formed by H_ABC domain (20), the hinge domain, the membrane proximal H₃ helix of syntaxin, and the helical and loop domains of SNAP-25. A thrombin site was engineered into syntaxin by replacing amino acids 176–180 (IFAS) with the thrombin recognition sequence (LVPR). (B) Coomassie blue-stained protein profile of tcSYN/SNAP-25 before and after thrombin treatment. After reconstitution, proteoliposomes containing thrombin-cleavable syntaxin/SNAP-25 were treated with either inactive thrombin (lanes 1 and 2) or active thrombin (lanes 3 and 4). This proteolysis resulted in the separation of the syntaxin H_ABC+hinge domain and the H₃ helix (lanes 3 and 4). An aliquot of tcSYN/SNAP-25 liposomes was treated with inactive thrombin (lane 2) and active thrombin (lane 4) in the presence of 0.2% TX-100 to access the lumenally oriented tcSYN/SNAP-25 complex. Thrombin was electrophoresed separately because it comigrates with syntaxin (lane 5). The asterisk (*) indicates a minor SNAP-25 cleavage product. AEBSF, 4-(2-aminoethyl)benzenesulfonyl fluoride. (C) Kinetic profile of membrane fusion of tcSYN/SNAP-25 and SYN H₃/SNAP-25 liposomes with v-liposomes. Donor liposomes containing VAMP were mixed with full length (closed circles) or SYN H₃/SNAP-25 (open circles) liposomes, and the increase in NBD fluorescence at 37°C was monitored for 2 hours. This increase in NBD fluorescence then was converted to rounds of fusion.

Figure 3

Excision of the SNAP-25 loop region has little effect on fusion. (A) Schematic representation of syntaxin, and the SNAP-25 H_A H_B helices and the loop domain. SNAP-25 amino acids 93–96 (NKLK) and 121–124 (VDER) were substituted with the thrombin recognition sequence LVPR. The arrow with asterisk indicates the position of the cryptic thrombin cleavage site in syntaxin (see below). (B) Coomassie blue-stained protein profile of SYN/SNAP-25 and SYN/tcSNAP-25 before and after thrombin treatment. SYN/SNAP-25 (lanes 1–4) and SYN/tcSNAP-25 (lanes 5–8) were purified and reconstituted into liposomes. These proteoliposomes were treated with either inactive thrombin (lanes 1, 2, 5, and 6) or active thrombin (lanes 3, 4, 7, and 8) for 4 hours at 37°C in the presence (lanes 2, 4, 6, and 8) or absence (lanes 1, 3, 5, and 7) of TX-100. Because syntaxin contains a cryptic thrombin cleavage site in its hinge domain, this proteolysis resulted in an unengineered cleavage of the H_ABC domain from the hinge H₃ domain in some syntaxin molecules. N-terminal sequencing of the SYN hinge+H₃ band revealed the sequence ¹⁵⁹TTTSEE, confirming that thrombin had severed the SYN H_ABC domain from the hinge+H3 domain. AEBSF, 4-2(aminoethyl)benzenesulfonyl fluoride. Shown are kinetic profiles of membrane fusion of SYN/SNAP-25 (C) or SYN/tcSNAP-25 (D) liposomes before and after thrombin treatment with v-liposomes. Donor vesicles containing VAMP were mixed with full length SYN/SNAP-25 or SYN/tcSNAP-25 (closed circles) or thrombin-treated SYN/SNAP-25 or SYN/tcSNAP-25 (open circles) liposomes, and the increase in NBD fluorescence at 37°C was monitored for 2 hours. This increase in NBD fluorescence then was converted to rounds of fusion, as described above. We repeatedly observed a slight lag in the initial rate of fusion when the loop region of SNAP-25 was excised.

Figure 4

The t-SNARE core domain is sufficient to efficiently drive membrane fusion. (A) Schematic representation of the engineered thrombin cleavage sites in both syntaxin and SNAP-25. The thrombin sites are at the same positions as described in Figs. 2 and 3. After thrombin proteolysis, the membrane proximal syntaxin H₃ and SNAP-25 H_A H_B helices were severed from all other t-SNARE domains. (B) Coomassie blue-stained protein profile of tcSYN/tcSNAP-25 before and after thrombin treatment. The tcSYN/tcSNAP-25 heterodimer was purified and reconstituted into liposomes and was treated with either inactive thrombin (lanes 1 and 2) or active thrombin (lanes 3 and 4) for 4 hours 37°C in the absence (lanes 1and 3) or presence (lanes 2 and 4) of 0.2% TX-100 to access the lumenally oriented tcSYN/tcSNAP-25 complex. AEBSF, 4-(2-aminoethyl)benzenesulfonyl fluoride. (C) Kinetic profile of membrane fusion of tcSYN/tcSNAP-25 and SYNH₃/SNAP-25 H_AH_B liposomes with v-liposomes. Donor liposomes containing VAMP were mixed with acceptor liposomes containing either full length (closed circles) or core t-SNAREs (open circles) and were allowed to fuse at 37°C without any preincubation. The increase in NBD fluorescence was monitored and converted to rounds of fusion. (D) Comparison of fusion efficiency of SYN H₃/SNAP-25, SYN/SNAP-25 H_AH_B, and SYN H₃/SNAP-25 H_AH_B liposomes with VAMP liposomes. The rounds of fusion after 2 hours at 37°C of the reaction for SYN H₃/SNAP-25 (n = 18, independent experiments), SYN/SNAP-25 H_AH_B (n = 5), and SYN H₃/SNAP-25 H_AH_B liposomes (n = 19) liposomes (filled histograms) were compared with their respective full length counterpart (open histograms) and were expressed as percent of full length protein signal. The SYN H₃/SNAP-25 and SYN H₃/SNAP-25 H_AH_B liposomes have a higher fusion efficiency than full length t-SNAREs (232 ± 32% and 216 ± 34%, respectively); SYN/SNAP-25 H_AH_B has a slightly higher fusion efficiency (112 ± 13%) compared with its full length counterpart.