Solution single-vesicle assay reveals PIP2-mediated sequential actions of synaptotagmin-1 on SNAREs

Abstract

Synaptotagmin-1 (Syt1) is a major Ca(2+) sensor for synchronous neurotransmitter release, which requires vesicle fusion mediated by SNAREs (soluble N-ethylmaleimide-sensitive factor attachment protein receptors). Syt1 utilizes its diverse interactions with target membrane (t-) SNARE, SNAREpin, and phospholipids, to regulate vesicle fusion. To dissect the functions of Syt1, we apply a single-molecule technique, alternating-laser excitation (ALEX), which is capable of sorting out subpopulations of fusion intermediates and measuring their kinetics in solution. The results show that Syt1 undergoes at least three distinct steps prior to lipid mixing. First, without Ca(2+), Syt1 mediates vesicle docking by directly binding to t-SNARE/phosphatidylinositol 4,5-biphosphate (PIP(2)) complex and increases the docking rate by 10(3) times. Second, synaptobrevin-2 binding to t-SNARE displaces Syt1 from SNAREpin. Third, with Ca(2+), Syt1 rebinds to SNAREpin, which again requires PIP(2). Thus without Ca(2+), Syt1 may bring vesicles to the plasma membrane in proximity via binding to t-SNARE/PIP(2) to help SNAREpin formation and then, upon Ca(2+) influx, it may rebind to SNAREpin, which may trigger synchronous fusion. The results show that ALEX is a powerful method to dissect multiple kinetic steps in the vesicle fusion pathway.

Figure 1

Sorting subpopulations of vesicle mixture by single vesicle assay, alternating-laser excitation (ALEX). (A) Schematic description of single-molecule alternating-laser excitation set-up (Supplementary Figure S1). Donor- and acceptor-excitation lasers are alternated faster (400 μs) than the transit time (∼5 ms) of a vesicle through the confocal excitation volume (∼1 femto-liter). Dilution to ∼100 pM vesicle concentration ensures that one vesicle passes through the excitation volume at a given time. The fluorescent emissions of donor and acceptor are detected separately, which results in fluorescence time traces. (B–E) Typical fluorescence time traces of vesicles. A random event of diffusing in and out of a vesicle generates a fluorescent burst (or spike) in time traces. Each time trace contains three different photon streams: I_D^D, fluorescent emissions of donor dyes (DiI) excited by donor-excitation laser (green line), I_D^A, fluorescent emissions of acceptor dyes (DiD) excited by donor-excitation laser, which are FRET signals (orange line), and I_A^A, fluorescent emissions of acceptor dyes excited by acceptor-excitation laser (red line). (B) Time trace of t-vesicle, reconstituted with t-SNARE and doped with DiI. (C) Time trace of v-vesicle, reconstituted with VAMP-2 and doped with DiD. (D) Time trace of docked-but-unfused vesicle. Both I_D^D and I_A^A present significant intensity. (E) Time trace of fused vesicle. Due to FRET, the intensity of I_D^D is reduced, while that of I_D^A is increased. (F) Two-dimensional E (FRET efficiency)-S (sorting number) graph. Three fluorescent intensities of a burst are used to calculate S and E (see ‘Definition and calculation of E and S of a single vesicle’). The unreacted, docked, and fused vesicles have characteristic values of S and E, and, as a result, occupy different areas in E-S graph: t-vesicle only (green box), v-vesicle only (red box), docked vesicle (purple box), and fused vesicle (orange oblique). (G) Comparison of bulk FRET assay with ALEX. For bulk FRET assay, T (reconstituted with t-SNARE) and V (reconstituted with VAMP-2) were mixed to be 50 μM lipid concentration at 35°C (black line). From single-exponential fitting, fusion rates were obtained; k_{fusion, bulk}=1.03±0.02 × 10⁻³ s⁻¹ and k_{fusion, ALEX}=1.19±0.08 × 10⁻³ s⁻¹, respectively. As a control, SNAP-25 was not added (grey line). For ALEX, the mixture was incubated at 35°C and diluted to be 3 μM lipid concentration at the selected time points and then measured for 5 min. The fractions of docked and fused vesicles (purple box and orange oblique, respectively) were obtained from three independent measurements. Dotted line denotes the level of random coincidence. (H–J) 2D E-S graphs obtained by ALEX. Each dot denotes a vesicle. (H) T-V mixture after 60 s incubation (blue arrow in G), and (I) after 2400 s incubation (red arrow in G). (J) No SNAP-25 control after 2400 s incubation (green arrow in G). Colour scale bar indicates the number of vesicles.

Figure 2

Syt1 is able to mediate vesicle docking by interacting with both t-SNARE and PIP₂. (A) Schematic description of vesicle docking measurement. S, v-vesicle without VAMP-2 but containing Syt1. T, t-vesicle with t-SNARE and 1% PIP₂. After mixing two vesicles, the mixture was analysed by ALEX. (B–E) 2D E-S graphs for mixture of t- and v-vesicles. (B) Mixture of S and T. Significant amounts of docked vesicles were observed (purple box). (C) Mixture of S_Y311N (v-vesicles incorporated with Syt1 Y311N mutant) and T. (D) Mixture of S and T_{no PIP2} (t-vesicle incorporated with t-SNARE but without PIP₂). (E) Mixture of S and T_syx-only (t-vesicle containing only syntaxin-1A but containing 1% PIP₂). (F) Positions of SNAP-25 mutations in SNARE complex. (G–I) Three SNAP-25 mutants of SNAP-25 D51K/E52K, D179K/D186K, and BoNT/A were used to reconstitute T (T_D51K/E52K, T_D179K/E186K, and T_BoNT/A, respectively), and each one of them was mixed with S. (G) T_D51K/E52K-S; (I) T_D179K/E186K-S; (H) T_BoNT/A-S. (J) Time-dependent T-S docking. For detailed method, see ‘ALEX measurements for the docking and fusion kinetics’ in Supplementary data. The error bars within 5 min were obtained from 10 independent measurements, while those after 5 min from three measurements. (K) Bar graph of the fraction of the docking subpopulations. Error bars (standard deviation) were obtained from more than three independent experiments, and >1000 vesicles were analysed in each measurement (^*P<0.005). (L) PIP₂ only control. We mixed protein-free liposome, containing 1% PIP₂, and S. (M) The t-SNARE/PIP₂ molar ratio determines docking efficiency of T_D51K/E52K-S. We kept 1% PIP₂, but varied the amounts of t-SNARE constructed with SNAP-25 D51K/E52K mutant, and then measured the docking efficiency of T_D51K/E52K-S. We used initial input molar ratio of t-SNARE/PIP₂ and the incorporation efficiency of t-SNARE into vesicle (60±11%; Supplementary Figure S5) to determine the t-SNARE/PIP₂ ratio. Error bars were obtained from more than three independent experiments.

Figure 3

Tethering by Syt1 enhances vesicle fusion. (A) 2D E-S graph of the mixture of T and V, (B) T and SV, (C) T_D51K/E52K-SV. T, t-vesicle reconstituted with t-SNARE and 1% PIP_2. T_D51K/E52K, the same as T except for using SNAP-25 D51K/E52K mutant. V, v-vesicle reconstituted with VAMP-2. SV, v-vesicle reconstituted with both VAMP-2 and Syt1. The data were obtained by incubating the mixture for 5 min at room temperature and measured for 5 min by ALEX. Purple and orange boxes denote docked and fused vesicles, respectively. (D) Bar graph for the fractions of the docked and fused subpopulations, analysed from . Error bars were obtained from at least three independent measurements (^*P<0.005). (E) Time-dependent measurement of docked and fused vesicles of T-SV. We used the same method for Figure 2J but using T-SV. The fractions of docked and fused vesicles for each time bin were obtained from 10 measurements. Purple, red, and black squares represent the fractions of docked, fused, and the sum of docked and fused vesicles, respectively. From double-exponential fitting, the average fusion rate was obtained, k_T-SV=3.7±0.9 × 10⁻³ s⁻¹. This value is similar to that of bulk FRET from (G); k_{T-SV, bulk}=2.9±0.1 × 10⁻³ s⁻¹. Expanded view is present in inset. (F) Time-dependent measurement of docked and fused vesicles of T-V. The same method for T-SV was used. Expanded view is present in inset. The dotted line indicates average fraction of random coincidence. (G and H) Bulk FRET measurements for T-SV and T-V, depending on vesicle concentration, respectively, at room temperature.

Figure 4

Soluble VAMP-2 reduces the docked subpopulation by Syt1. (A) Top, schematic description of soluble VAMP-2 (sVAMP-2) test. T and S were incubated together for 5 min at room temperature, and then 1 μM of sVAMP-2 was added to the mixture, followed by 15 min incubation. Bottom, 2D E-S graph of the mixture. (B) Top, T was pre-incubated with sVAMP-2 for 15 min, and then mixed with S for 5 min. Bottom, 2D E-S graph of the mixture. (C) Bar graph for the fraction of docked vesicles. T-S+sVAMP-2 and (T+sVAMP-2)-S from the results of Figure 4A and B; respectively. T-S+sVAMP-2₁₋₇₀ and T-S+sVAMP-2₁₋₅₉ denote the docking fraction after adding sVAMP-2₁₋₇₀ and sVAMP-2₁₋₅₉. As a control, we added 3 μM BSA instead of sVAMP-2 (T-S+BSA). Error bars were obtained from four independent measurements (^*P<0.005).

Figure 5

Ca²⁺ induces rebinding of Syt1 to ternary SNARE complex. (A) Schematic description of Syt1 rebinding experiments. To the mixture of Figure 4B, 10 μM Ca²⁺ was added at room temperature. (B) 2D E-S graph obtained from the mixture of Figure 5A. (C) 2D E-S graph obtained from the mixture of Figure 5A, but without PIP₂ on T. (D) 2D E-S graph obtained from the mixture of Figure 5A, but T was constructed with SNAP-25 D51K/E52K mutant. (E) Bar graph for the fraction of docked vesicles with and without Ca²⁺. We tested wild type and three mutants of SNAP-25, the case of no PIP₂ on T, and C-terminal truncations of sVAMP-2 (1–70 and 1–59). Error bars were obtained from four independent measurements (^*P<0.005).

Figure 6

Model for sequential actions of synaptotagmin-1 on SNAREs. In the absence of Ca²⁺, Syt1 interacts with t-SNARE·PIP₂, which brings vesicle to the plasma membrane in proximity. Because of this close geometry of two membranes, VAMP-2 can efficiently search and bind t-SNARE to form the SNARE complex. During this step, Syt1 loses its contact with SNAREs. When Ca²⁺ arrives, Syt1 rebinds to the SNARE complex together with PIP₂.