Morphofunctional changes at the active zone during synaptic vesicle exocytosis

Abstract

Synaptic vesicle (SV) fusion with the plasma membrane (PM) proceeds through intermediate steps that remain poorly resolved. The effect of persistent high or low exocytosis activity on intermediate steps remains unknown. Using spray-mixing plunge-freezing cryo-electron tomography we observe events following synaptic stimulation at nanometer resolution in near-native samples. Our data suggest that during the stage that immediately follows stimulation, termed early fusion, PM and SV membrane curvature changes to establish a point contact. The next stage-late fusion-shows fusion pore opening and SV collapse. During early fusion, proximal tethered SVs form additional tethers with the PM and increase the inter-SV connector number. In the late-fusion stage, PM-proximal SVs lose their interconnections, allowing them to move toward the PM. Two SNAP-25 mutations, one arresting and one disinhibiting spontaneous release, cause connector loss. The disinhibiting mutation causes loss of membrane-proximal multiple-tethered SVs. Overall, tether formation and connector dissolution are triggered by stimulation and respond to spontaneous fusion rate manipulation. These morphological observations likely correspond to SV transition from one functional pool to another.

Keywords: SNARE; cryo-electron tomography; synapse; synaptic vesicles.

Figure 1. Experimental models

1. Ai

   Glass atomizer used to disperse depolarizing solution on the EM grid milliseconds before the grid is vitrified.

2. Aii

   Spray droplets imaged with the GFP filter set. Scale bars, 20 μm.

3. Aiii

   Synaptosomes imaged with the DAPI filter set. Scale bar, 20 μm.

4. Aiv

   Overlay of spray droplets (green) and synaptosomes (red). Yellow circles show contact between droplets and synaptosomes in thin area, which is suitable for cryo‐ET. Scale bar, 20 μm.

5. Bi

   Schematic drawing of a six‐well Petri dish depicting astrocytes (pink) growing at the bottom of the Petri dish below EM grids (black, round grid overlaying the astrocytes), with neurons (blue) growing on top of the grids.

6. Bii

   Grid square overview with neurons growing over it. Scale bar, 5 μm.

7. Biii

   Medium‐magnification overview of neurons growing over R2/1 holes. Scale bar, 500 nm.

8. Biv

   One slice of a tomogram depicting a synapse. Scale bar, 100 nm.

Figure 2. Schematic representation of a spray‐mixing plunge‐freezing experiment

1. A–D

   In a single experiment, different synaptosomes get stimulated for between < 1 and 7 ms.** An EM grid is held by tweezers and is covered with synaptosomes in HBM solution. A magnified view of a grid square shows synaptosomes in blue and their synaptic state of three synaptosomes is represented on the rightmost part of each panel. Panel (A) represents the situation immediately after blotting off the solution excess before the grid is sprayed. The grid is accelerated towards the cryogen. Panel (B) shows a snapshot of the experiment when the grid crosses the spray, 7 ms before the freezing. Some fluorescently dyed droplets containing HBM with 52 mM KCl land on the grid and are depicted in green. At this time point, a synaptosome located at the impact point of a droplet is activated and is depicted in dark blue. Panel (C) shows a snapshot 5 ms later, that is, 2 ms before freezing. As KCl diffuses away from droplet impact points, another synaptosome gets activated because locally KCl concentration has reached a concentration to depolarize the synaptosome sufficiently so that voltage‐gated calcium channels open. Panel (D) shows a synaptosome at the time of immersion with ethane. 0.1 ms before freezing a third synaptosome got exposed to a high enough concentration of KCl and got stimulated.

Figure EV1. Representative slices through tomograms

1. A, B

   Tomographic slice without (A) and with (B) segmentation of synaptosome with late‐fusion events.

2. C, D

   Tomographic slice without (C) and with (D) segmentation of WT SNAP‐25 neurons.

Data information: segmentation colors: off‐white = cell outline; pink = active zone; blue = synaptic vesicles; green = mitochondria; yellow = connectors and red = tethers. Scale bar, 100 nm.

Figure EV2. 3‐D rendered segmented tomograms of neuron synapses

1. A–C

   (A) SNAP‐25 WT, (B) SNAP‐25‐4E, and (C) SNAP‐25‐4K. (left) Overview, (right) detail. Blue: synaptic vesicles; purple: active‐zone plasma membrane; green: endoplasmic reticulum‐like organelle; yellow: connectors; and red: tethers. Scale bars: 100 nm.

Figure 3. SV exocytosis morphology

1. A–H

   Tomographic slice of non‐stimulated (A) and stimulated rat synaptosomes (B–H). (A) Image of a 2.2‐nm‐thick tomographic slice showing a non‐stimulated with SVs at the AZ and a straight PM. (Bi) Membrane curvature event, 2.2‐nm‐thick tomographic slice. (Bii) Membrane curvature event, 6.5‐nm‐thick tomographic slice. (Biii) Membrane curvature event, 2.24 nm‐thick tomographic slice. Orange arrows showing membrane curvature event. (Ci, Cii) Lipid perturbations of PM and SV, 22‐nm‐thick tomographic slices. The space between SV and PM is denser than in the non‐stimulated synaptosomes (see pink arrow). (D–F) Vesicles with a pore opening that might be on the way to full‐collapse fusion, 33‐nm‐thick tomographic slice thickness: 22 nm (D), 30.8 (E), and 33 nm (F). (G) Wide pore opening, most likely on the way to full‐collapse fusion, 2.2 nm tomographic slice. (H) Remaining bump at the end of full‐collapse fusion, 11‐nm‐thick tomographic slice. Scale bar, 50 nm. Total number of observations of each type of exocytosis events: (B) 8; (C) 3; (D) 3; (E) 2; (F) 3; (G) 1; and (H) 11. Events of type (B–E) were classified as early, while events of type (F–H) were classified as late.

Figure 4. SV distribution

1. A, B

   Vesicle occupancy expressed as fraction of cytosol volume occupied by vesicles as a function of distance to AZ in (A) cultured neurons and (B) synaptosomes. Each bar is the average value of a distance group of all tomograms of the same treatment/genotype. Each dot represents the occupancy of a distance group in a single tomogram. Statistical test: multiple all against reference pairwise ANOVA comparisons with Benjamini–Hochberg correction. *P < 0.05 and **P < 0.01 after Benjamini–Hochberg correction. The reference distance group was the intermediate one. The reference experimental conditions were the WT genotype (A) and non‐sprayed synaptosomes (B), respectively. Comparisons between distance groups were performed only within the reference experimental conditions.

2. C, D

   Distance of proximal SVs from the AZ. Each dot represents the value of an individual SV. Horizontal line: mean; whiskers: 2×SEM interval. Statistical test: multiple all‐against‐reference pairwise ANOVA comparisons with Benjamini–Hochberg correction; *P < 0.05 after Benjamini–Hochberg correction.

Figure 5. Proximal SV tethering

1. A, B

   Fraction of proximal SVs that are triple tethered. Each bar shows the overall fraction of all proximal SVs from a given experimental condition. Each dot represents the value of an individual active zone. Statistical test: multiple all‐against‐control pairwise χ²‐test comparisons with Benjamini–Hochberg correction. **P < 0.01 after Benjamini–Hochberg correction. The reference was the WT genotype (A) or non‐sprayed synaptosomes (B).

2. C, D

   Number of tethers per proximal SV. Each dot represents an individual SV. The vertical line represents the mean value, and the horizontal whiskers correspond to the 95% confidence interval. Statistical test: multiple all‐against‐control pairwise ANOVA comparisons with Benjamini–Hochberg correction; *P < 0.05 and ***P < 0.001, after Benjamini–Hochberg correction.

Figure EV3. Additional SV tethering and connectivity data

1. A, B

   Histogram of the number of tethers per proximal SV. Statistical test: pairwise χ² test between control and each experimental condition in the zero‐tether group with Benjamini–Hochberg correction. *P < 0.05.

2. C, D

   Histogram of connected SV among tethered or non‐tethered proximal SVs.

3. E, F

   Histogram of connected SV among proximal non‐RRP or RRP SVs.

Data information: (A, C, E) Synapses in mouse cultured neurons. (B, D, F) Rat synaptosomes.

Figure 6. SV connectivity

1. A, B

   Number of connectors per SV as a function of their distance to the AZ PM for mouse neurons (A) and rat synaptosomes (B). Each bar represents the mean value of all SVs of the subgroup. The lines represent 2×SEM intervals. Statistical tests: multiple all‐against‐reference pairwise ANOVA comparisons with Benjamini–Hochberg correction. Within a single experimental condition, the reference was the proximal distance group; within a single distance group, the reference was the WT genotype (A) or no‐sprayed synaptosomes (B).

2. C, D

   Fraction of connected vesicles as a function of distance to the AZ PM for mouse neurons (C) and rat synaptosomes (D). Each bar shows the overall fraction of all SVs in a given distance group and a given experimental condition. Each dot represents the corresponding value of an individual active zone. Statistical test: multiple all‐against‐reference pairwise χ² test with Benjamini–Hochberg correction; references were defined as in (A) and (B).

3. E, F

   Number of connectors per proximal SV not belonging or belonging to the RRP for mouse neurons (E) and rat synaptosomes (F). Horizontal line: mean; whiskers: 2×SEM interval.

4. G, H

   Number of connectors per non‐tethered or tethered proximal SV for mouse neurons (G) and rat synaptosomes (H). Horizontal line: mean; whiskers: 2×SEM interval.

Data information: Statistical tests in (E–H): multiple all‐against‐control pairwise ANOVA comparisons with Benjamin––Hochberg correction. Control was WT genotype or non‐sprayed synaptosomes. In all statistical tests, *P < 0.05, **P < 0.01, and ***P < 0.001 after Benjamini–Hochberg correction.

Figure EV4. Tethered connected SVs

1. A, B

   Tomographic slices showing tethered connected vesicles. Blue arrows highlight the connectors. Scale bar, 50 nm.

Figure 7. Model depicting a synapse transitioning from resting state to early‐ and late‐fusion states

Tethering and connectivity changes upon synapse stimulation are depicted. Proximal non‐triple‐tethered vesicles (black proximal SVs) gain additional tethers, and some of them become triple tethered (yellow SVs) shortly after stimulation. Primed vesicles then fuse with the plasma membrane (late fusion) and leave an empty space in the AZ cytoplasm. The number of connectors (depicted in blue) per proximal SV decreases in late‐fusion tripled‐tethered vesicles. The red arrow shows a vesicle initially located in the intermediate region, which diffuses to the proximal region in the late‐fusion state. Tethers are shown in green.