Ultrastructural and functional fate of recycled vesicles in hippocampal synapses

Abstract

Efficient recycling of synaptic vesicles is thought to be critical for sustained information transfer at central terminals. However, the specific contribution that retrieved vesicles make to future transmission events remains unclear. Here we exploit fluorescence and time-stamped electron microscopy to track the functional and positional fate of vesicles endocytosed after readily releasable pool (RRP) stimulation in rat hippocampal synapses. We show that most vesicles are recovered near the active zone but subsequently take up random positions in the cluster, without preferential bias for future use. These vesicles non-selectively queue, advancing towards the release site with further stimulation in an actin-dependent manner. Nonetheless, the small subset of vesicles retrieved recently in the stimulus train persist nearer the active zone and exhibit more privileged use in the next RRP. Our findings reveal heterogeneity in vesicle fate based on nanoscale position and timing rules, providing new insights into the origins of future pool constitution.

Figure 1. Ultrastructural visualization of vesicles retrieved after RRP stimulation in acute hippocampal slice.

(a) Schematic illustrating experimental protocol for labelling recycled vesicles. Extracellular stimulation (40 APs 20 Hz) of Schaffer collaterals is combined with FM-dye application to CA1. (b) Representative image of fluorescence from dye-labelled synapses in CA1. Scale bar, 10 μm. (c) Schematic illustrating approach for photoconversion of labelled vesicles. Dye-loaded fixed target region is photo-illuminated using blue light focused through an objective lens in the presence of DAB (top), producing an electron-dense spot in the slice tissue (bottom). Orange square indicates trimmed target region used for ultrastructural analysis. Scale bar, 200 μm. (d) Low magnification electron micrograph showing presynaptic terminals (blue) and postsynaptic structures (brown). Scale bar, 500 nm. (e) Typical images showing terminals with PC+ vesicles (arrowheads). Scale bar, 100 nm. (f) (Top) High magnification images of PC+ vesicles with electron-dense lumen and PC− vesicles with clear lumen. Scale bar, 25 nm. (Bottom) Cross-sectional density profiles of PC+ and PC− vesicles (n=6, mean±s.e.m.) illustrating a quantitative approach that allows for the differentiation of vesicle classes. (g) 3D reconstruction showing PC+ vesicles (dark spheres) and PC− vesicles (transparent spheres). Active zone appears red. Scale bar, 100 nm. (h) Frequency histogram of PC+ pool sizes based on ultrastructural analysis of n=187 synapses from 11 slices from 10 animals, expressed as % of total pool. Red line shows gamma function fit. (i) Summary histogram of median±IQR PC+ pool sizes for synapses fixed at different times after loading (1 min: 3.1% (2.0–4.9), n=73 PC+ vesicles from 28 synapses (including n=8 full serial reconstructions) from 3 slices from 3 animals; 5 min: 3.6% (2.5–5.6), n=138 PC+ vesicles from 59 synapses (including n=4 full serial reconstructions) from 3 slices from 3 animals; 20 min: 3.3% (2.0–5.4), n=455 PC+ vesicles from 100 synapses (including n=10 full serial reconstructions) from 5 slices from 4 animals, not significant, Kruskal–Wallis one-way ANOVA, P=0.378).

Figure 2. Organization of retrieved vesicle pool at different post-stimulus time points.

Representative 3D serial reconstructions for synapses from 1, 5 and 20 min after the loading stimulus. PC+ vesicles are shown as dark spheres and PC− vesicles as transparent spheres. Active zone appears red. Grey cross hairs indicate boundaries of vesicle cluster from active zone centre. Scale bar, 100 nm.

Figure 3. Vesicles retrieved after RRP stimulation recycle to random positions in the terminal.

Normalized and smoothed spatial frequency density plots for the three time points (1, 5 and 20 min, n=73 PC+ vesicles from 28 synapses from 3 slices, n=138 PC+ vesicles from 59 synapses from 3 slices, n=455 PC+ vesicles from 100 synapses from 5 slices) for PC− (left panels), PC+ (middle panels) and merged (right panels) showing the positions occupied by vesicles with respect to the active zone and cluster boundaries. Active zone centres (AZ) are at the bottom middle in each plot. In merged plots, green corresponds to PC− and red to PC+ vesicles. See Supplementary Fig. 2 for approach used to generate plots.

Figure 4. Quantification of recycled vesicle pool spatial dynamics.

(a) Schematic of distance measure. Filled circles, PC+; empty circles, PC−. (b) Cumulative frequency distance plots (1 min, n=28; 5 min, n=59; 20 min, n=100). Lines and circles are mean values and shaded areas are s.e.m. Red lines/circles: PC+, blue lines/circles: PC−. Grey dashed lines: 1 min PC+ data set profile for comparison. Cumulative fraction is fraction of total data set represented by a given distance. (Inset) Median±IQR distance to active zone for PC+ (red) and PC− vesicles (blue). (c) Stacked bar charts showing %PC+ vesicles in edge versus core (left) and edge (excluding active zone) versus core (right) with summary schematics. (d) Summary of analysis of docked vesicle population. (Left) mean±s.e.m. of PC+ vesicles in pool of docked vesicles as % of total docked vesicle number (mean±s.e.m.: 14.4±4.2%, 9.1±2.1%, 4.5±1.0, n=28, n=59, n=100, ANOVA, P<0.01, Bonferroni's: 1 versus 5 min, NS; 5 versus 20 min, NS; 1 versus 20 min, P<0.05). (Right) mean±s.e.m. of docked PC+ vesicles as % of all PC+ vesicles (mean±s.e.m.: 31.5±7.9%, 15.5±3.9%, 11.4±2.7, n=28, n=59, n=100, ANOVA, P<0.01, Bonferroni's multiple comparisons: 1 versus 5 min, NS; 5 versus 20 min, P<0.05; 1 versus 20 min, P<0.05). (e) Sample micrographs of synapses. PC+ vesicles are marked by arrowheads. Examples of boundaries used to quantify edge and core are illustrated in left panel. Scale bar, 100 nm. (f) Schematic of cluster analysis approach based on measuring PC+ fractions in concentric bins around individual PC+ vesicles. (g) Circular frequency density plots. Colour look-up table represents fraction of PC+ vesicles. (h) Line plots show clustering ratio normalized to ratio for whole synapse. Error bars are omitted for clarity. Asterisks indicate radial distances colour coded for each time point where clustering significantly exceeds baseline values (based on one-sample t-tests against 1, see Methods section). NS, not significant.

Figure 5. Heterogeneity in functional use of recycled vesicles.

(a) (Top) Schematic illustrating experimental approach. The pool recycled from recruitment of the RRP (40 APs) is labelled and 20 min later subjected to two rounds of stimulation (destaining) in the absence of dye while imaging; 40 APs to release current RRP and 600 APs to release TRP. The ratio between dye loss after 40 APs stimulus versus total dye loss (after 600 APs) provides a measure of subsequent use of the labelled pool. Note: this is not an absolute measure; although 600 APs and 40 APs should together access >90% of the available pool, residual vesicles or those lost to a non-recycling pool could lead to this value being an overestimate. Accepting these possibilities, normalizing with respect to total release allows comparisons of relative use under different conditions. (Middle) Representative images of 40 APs loading and the two phases of activity-driven dye loss. Scale bar, 2 μm. (Bottom) Plot showing three sample destaining curves (grey lines) each fitted with two single exponentials (red lines). %fluorescence is total fluorescence lost during the two-phase destaining protocol, allowing us to make an estimate of the relative proportion of dye loss arising from the 40 APs versus the larger 600 APs destaining protocol. (b) As in (a) for 600 APs loading. Scale bar, 2 μm. (c) Mean±s.e.m. histogram summary (right) and cumulative frequency distribution plot (left) of % use for 40 APs stimulation after 40 APs loading (n=119 from 4 cultures, red) and 600 APs loading in culture (n=130 from 3 cultures, blue). Green bar (right) and line plot (left) shows a third protocol (tail load) in which only the tail end (<200 APs) of a 600 APs stimulus train had FM-dye present to allow labelling (n=131 from 4 cultures). Mean±s.e.m. use fraction was as follows: 40 APs: 37.2±0.9%, 600 APs: 27.1%±0.6%, tail load: 31.7%±0.9% (one-way ANOVA, P<0.001, Bonferroni's multiple comparison: 40 APs versus 600 APs, P<0.05, 40 APs versus tail load, P<0.05).

Figure 6. Preferential use of vesicles previously recruited by a limited loading protocol.

(a) (Top) Schematic illustrating experimental approach. Vesicles that were FM-dye labelled with 10 APs stimulation received two further stimulation steps after 20 min; 40 APs to release current RRP and 600 APs to release TRP. The ratio between dye loss after 40 APs stimulus versus total dye loss provides a measure of preferential use of the original loaded pool. (Bottom) Representative image of 10 APs loading and the two phases of activity-driven dye loss. Scale bar, 2 μm. (b) Plot showing sample destaining curves for three synapses each fitted with two single exponentials (red lines). Fraction of 10 APs-loaded vesicles used in subsequent stimulus was significantly higher than 40 APs loading (10 APs: 41.2%±1.6%, 40 APs: 37.2±0.9%, n=74 from 7 cultures, n=119, two-tailed unpaired t-test, P<0.05). (c) (Left) Example image of sypHy2x expression and schematic illustrating mechanism of action as fusion reporter. Image was collected at 40 AP peak response. White squares show typical small discrete synapses used for analysis. Scale bar, 10 μm. (d) Sample sypHy2x images showing fluorescence change at synapse in response to stimulation (1 AP, left; 10 AP, right). Scale bar, 2 μm. (e) Traces showing fluorescence intensity profiles for same single synapse from (d) to a range of stimuli. (f) Traces showing average fluorescence responses (3 synapses) with repeated 10 APs stimulations (interval 1 min) before (trials 1 and 2) and after (trials 3–10) the addition of 1 μM bafilomycin, a vATPase inhibitor that prevents vesicle re-acidification. Values are normalized relative to the first bafilomycin trial response amplitude. The reduction in response amplitude after bafilomycin treatment is consistent with the use of alkali-trapped vesicles. A larger maintained stimulus (1,200 APs) can recruit additional non-recycled vesicles, observed as a significant rise in fluorescence intensity. (g) Mean±s.e.m. responses (105 synapses from 4 cultures), normalized to first bafilomycin trial. Asterisks indicate significant outcomes (P<0.05) of two-tailed one-sample t-tests for each time point versus 1. Note 1,200 APs response amplitude is shown on different scale (right).

Figure 7. Segregated organization of preferentially used vesicles.

(a) Representative electron micrograph after 10 APs loading (PC+ vesicle, arrow). Scale bar, 100 nm. (b) Representative 3D reconstructions showing PC+ vesicles (dark spheres) and PC− vesicles (transparent spheres). Active zone appears red. Grey cross hairs indicate vesicle cluster boundaries. Scale bar, 100 nm. (c) Normalized, smoothed spatial frequency density plots for 10 APs condition (n=59) for PC− (left panel), PC+ (middle panel) and merged (right panel). Active zone centres (AZ) are bottom middle in each plot. In merged plots, green is PC−, red is PC+. (d) Cumulative frequency plots for vesicle-active zone distances (n=59 synapses (including n=11 full serial reconstructions) from 3 slices from 2 animals. Lines and circles are mean values and shaded areas are s.e.m. Red lines/circles: PC+, blue lines/circles: PC−. (Inset) Histogram shows median distance±IQR to active zone for PC+ vesicles (red bar) and PC− vesicles (blue bar) (PC+ versus PC−: 99 (50–164) nm, 142 (119–184) nm, n=59, Wilcoxon, P<0.05). (e) (Top) Circular frequency density plot summarizing cluster analysis. Colour look-up table represents fraction of PC+ vesicles. (Bottom) Line plot (orange) showing clustering ratio normalized to ratio for whole synapse. Grey line shows 40 APs 20 min data (from Fig. 4h). Asterisk indicates radial distance where clustering significantly exceeds baseline values (two-tailed one-sample t-test, P<0.05).

Figure 8. Activity advances recycled vesicles to the active zone.

(a) Schematic illustrating experimental approach. (b) Representative electron micrographs for 20 min+stimulation condition. Scale bar, 100 nm. (c) Example of 3D reconstruction for same condition (dark spheres: PC+, transparent spheres: PC−). Active zone appears red. Grey cross hairs indicate boundaries of vesicle cluster from active zone centre. Scale bar, 100 nm. (d) Normalized and smoothed spatial frequency density plots for 20 min+stimulation condition (n=81 synapses (including n=14 full serial reconstructions) from 4 slices from 3 animals) for PC− (left panel), PC+ (middle panel) and merged (right panel) showing the vesicle positions with respect to the active zone and cluster boundaries. Active zone centres (AZ) are bottom middle in each plot. In merged plots, green is PC−, red is PC+. (e) Cumulative frequency plots (n=81). Lines and circles are mean values and shaded areas are s.e.m. Red lines/circles: PC+, blue lines/circles: PC−. (Inset) Histogram shows median distance±IQR to active zone for PC+ vesicles (red bar) and PC− vesicles (blue bar) (PC+ versus PC−: 104 (46–195) nm, 154 (129–186) nm, n=81, Wilcoxon test, P<0.0001). (f) (Top) Circular frequency density plot showing vesicle clustering, with colour look-up table corresponding to fraction of PC+ vesicles. (Bottom) Line plot (red) showing clustering ratio with increasing distance from vesicle centre, normalized to ratio for whole synapse. Grey line shows 20 min data (from Fig. 4h) for comparison. Asterisk indicates radial distance where clustering significantly exceeds baseline values (two-tailed one-sample t-test, P<0.05).

Figure 9. Activity-driven progression of recycled vesicles is inhibited by actin stabilization.

(a) Schematic illustrating experimental approach. (b) Normalized and smoothed spatial frequency density plots for 20 min+stimulation condition in presence of 1 μM jasplakinolide (n=48 synapses (including n=5 full serial reconstructions) from 3 slices from 1 animal) for PC− (left panel) and PC+ (right panel) showing the vesicle positions with respect to the active zone and cluster boundaries. Active zone centres (AZ) are bottom middle in each plot. (c) Representative electron micrographs for same condition. Scale bar, 100 nm. (d) Example of 3D reconstruction for same condition (dark spheres: PC+, transparent spheres: PC−). Active zone appears red. Grey cross hairs indicate boundaries of vesicle cluster from active zone centre. Scale bar, 100 nm. (e) Cumulative frequency plots (n=48). Lines and circles are mean values and shaded areas are s.e.m. Red lines/circles: PC+, blue lines/circles: PC−. (Inset) Histogram shows median distance±IQR to active zone for PC+ vesicles (red bar) and PC− vesicles (blue bar) (PC+ versus PC−: 159 (108–223) nm versus 142 (115–168) nm, n=48, Wilcoxon test, P=0.090).