Mechanical actions of dendritic-spine enlargement on presynaptic exocytosis - Nature

a, Strong pushing of a presynaptic bouton labelled with iGluSnFR (green) by a glass pipette (magenta). White dashed-lines indicate the outline of the boutons. Orange lines represent the axial line of the axon. Pushing was applied over the axial line. Scale bar 0.5 µm. b, Fluorescence profiles along the yellow line (inset) before (black), during (magenta) and after (cyan) the strong pushing of the terminal shown in a. c, d, The Pr measurements before and after the strong pushing of a bouton (c) and their averaged traces (d). Orange bars (LED) indicate optogenetic stimulation (5 ms). e, The mean time courses of Pr (left) and the peak of the averaged traces (right). n = 12 boutons (8 slices, 8 rats). For the baseline fluctuations (Baseline), 3 consecutive values (5 min apart) from all boutons were obtained. For Pr, two-sided Kruskal–Wallis test (H = 22.2, **P = 1.85 × 10⁻⁴) followed by Dunn’s multiple comparison test (vs the baseline). 6.5 min, **P = 2.62 × 10⁻³; 11.5 min, **P = 4.2 × 10⁻⁴; 20 min, **P = 1.9 × 10⁻³. For the averaged trace, (H = 13.9, **P = 7.65 × 10⁻³) followed by post-hoc Dunn’s multiple comparison test (vs the baseline). 6.5 min, **P = 3.13 × 10⁻³; 11.5 min, *P = 0.015; 20 min, *P = 0.0194. f, Time courses of ΔFWHM caused by the fine pushing (black) and strong pushing (orange). For fine pushing, the trace is the replicate of Fig. 1g. For strong pushing, n =13 boutons (11 slices, 10 rats). Magenta bars in e and f indicate pushing. For strong pushing two-sided Kruskal–Wallis test (H = 28.6, **P = 6.22 × 10⁻⁷) followed by Dunn’s multiple comparison test (vs the baseline). Push (10 s), **P = 4.3×10⁻⁵; push (110 s), **P = 4.77 ×10⁻⁶. g, ΔFWHMs obtained from the same data as f. Two-tailed Mann−Whitney U test, U = 232, **P = 7.28 × 10⁻³. Error bars indicate s.e.m.

Source data

a, EPSCs recorded from the whole-cell clamped CA1 neurons. Optogenetic stimulations (5 ms, orange bars) were given to CA3 neurons, expressing FRET probes (Control) or FRET probes and the light chain of BoNTE (BoNTE). b, f, FRET intensities (A1%, see Methods) of the boutons that were finely pushed in the presence of BoNTE expression (b), or that were treated with K252a (2 µM, f). Scale bar indicates 0.5 µm. c, Time courses of ΔFWHM during fine pushing from HF side (black), LF side (open circle), in the absence of calcium (EGTA 100 µM, BAPTA-AM 30 µM, blue), in the presence of K252a (orange) or LatA (cyan) from the same data as Fig. 2e. The magenta bar indicates the duration of pushing. d, ΔFWHM of data in c during pushing. For the baseline fluctuations (Baseline), 2 consecutive FWHM values (20 s apart) from all boutons were plotted from the same data as Fig. 2c. Two-sided Kruskal–Wallis test (H = 127, **P = 6.41 × 10⁻¹⁵) followed by Dunn’s multiple comparison test (vs Baseline). HF side, n = 17, **P = 2.32 × 10⁻¹⁰; LF side, n = 10, **P = 2.3 × 10⁻⁸; BoNTE, n = 12, **P = 4.7 × 10⁻¹⁰; Ca^2+-free, n = 13, **P = 2.8 × 10⁻¹⁰; K252a, n = 8, **P = 7.07 × 10⁻⁷; LatA, n = 10, **P = 1.02 × 10⁻¹⁰. Kruskal–Wallis test (H = 0.492, P = 0.992) followed by Dunn’s multiple comparison test (vs HF side). LF side, P = 0.943; BoNTE, P = 0.758; Ca^2+-free, P = 0.819; K252a, P = 0.628; LatA, P = 0.563. e, FRET images of a pushed bouton and in a neighbouring bouton. Scale bar 1 µm. Error bars indicate s.e.m. g, h, Replots of the same data shown in e and f using the average lifetime, instead of A1%

Source data.

a, GCaMP7s fluorescence images of a bouton (green) before and during pushing with a glass pipette (magenta). Scale bar 0.5 µm. b, The mean time course of GCaMP7s or 6s intensities before and during pushing (magenta bar) among 9 boutons. c, Evoked calcium transients measured with GCaMP7s fluorescence from a bouton before, during and after pushing. Orange arrowheads and dashed lines indicate the timing of optogenetic stimulation with a duration of 5 ms. d, The peak values of evoked calcium responses (Exp#1–5) and their average (black). n = 5 boutons (5 slices, 3 rats). Two-sided Kruskal–Wallis test (H = 3.46, P = 0.177) followed by Dunn’s multiple comparison test (vs Before). Pushing, P = 0.752; after, P = 0.095. Error bars indicate s.e.m

Source data.

a, mEPSCs recorded from a whole-cell clamped CA1 neuron. By adding 20, 40, 60 or 150 mM sucrose to ACSF (320 mOsm), we made a series of sucrose solutions with the osmolarities of 340 mM, 360 mM, 380 mM and 470 mM. They were sequentially applied, each followed by 10 min wash periods with ACSF. b, The mean frequencies of mEPSC in the sucrose solutions normalized by the one in ACSF. n = 5 cells (5 slices, 3 rats). Two-sided Kruskal–Wallis test (H = 41.5, P = 2.11 × 10⁻⁸) followed by Dunn’s multiple comparison test (vs the baseline). 20 mM, P = 0.656; 40 mM, **P = 1.81 × 10⁻³; 60 mM, ** P = 7.15 × 10⁻⁴; 150 mM, **P = 3.50 × 10⁻⁸. c, Shrinkage of a bouton (left) by 20 mM sucrose quantified by the fluorescent profile (right). The profiles were aligned at the right side. n = 9 boutons (6 slices, 3 rats). Two-sided Kruskal–Wallis test (H = 17.3, **P = 1.76 × 10⁻⁴) followed by Dunn’s multiple comparison test (vs the baseline). 5 min, **P = 5.09×10⁻⁵. d, Evoked EPSCs recorded from CA1 neurons by optogenetic stimulation of the CA3 region (orange bar, 5 ms) in ACSF, isotonic sucrose (10 mM NaCl was replaced with 20 mM sucrose to yield an osmolarity the same as ACSF, 320 mOsm) and hypertonic sucrose (20 mM). We used the EPSCs from samples whose PPRs, with an interstimulus interval of 70 ms, were larger than 1.0 (see Methods). e, The mean time course for evoked EPSC amplitudes. Brown and magenta bars indicate stimulation with isotonic sucrose (n = 6) and hypertonic (n = 7) sucrose, respectively. f, Mean EPSC amplitudes during sucrose stimulations. For the baseline fluctuations (Baseline), 2 consecutive EPSC values (5 min apart) from all boutons were plotted. n = 6 cells (6 slices, 6 rats) for isotonic sucrose, n = 7 cells (7 slices, 7 rats) for hypertonic sucrose. Two-sided Kruskal–Wallis test (H = 16.1, **P = 3.28 × 10⁻⁴) followed by Dunn’s multiple comparison test (vs the baseline). Isotonic, P = 0.542, Hypertonic, **P = 6.5 × 10⁻⁵. Error bars indicate s.e.m

Source data.

a, b, iGluSnFR fluorescence intensity traces (a) in response to optogenetic (orange bars) paired-pulse stimulation and their averaged traces (b) before, during and after application of the ACSF solution containing 20 mM sucrose. The inter-stimulus interval was 70 ms. c, d, The mean time courses for PPR (a) and ΔPPR (d) for 5 boutons, in which PPR was calculated as the ratio of the peak values of the second and first pulses from the averaged traces (b). Magenta bars indicate sucrose application. Green horizontal bars at the x-axis indicate the time windows for the Pr analysis shown in a. n = 5 boutons (5 slices, 3 rats). Two-tailed Mann−Whitney U test, ACSF−sucrose (c), U = 6, *P = 4.78 × 10⁻²; ACSF−sucrose (d), U = 2, **P = 7.99 × 10⁻³. e, ΔPPR during ACSF and sucrose solution. For the baseline fluctuations (Baseline), 2 consecutive PPR values (5 min apart) from all boutons were plotted. n = 4 boutons. Two-tailed Mann−Whitney U test, U = 2, **P = 7.99 × 10⁻³. Error bars indicate s.e.m

Source data.

a, The design of the FRET probe for measuring the cis-SNARE assembly. Venus (Vn) and mTurquoise (mTq1) were fused with the C-terminals of the SNAREs, syntaxin-1A and VAMP2. b, mEPSCs induced by the 300 mM sucrose solution (620 mOsm, HSS) recorded from the whole-cell clamped CA1 neuron. c, FRET images of cis-SNARE probe during application of LSS (20 mM sucrose) and after application of HSS (300 mM sucrose). Imaging was not possible during the application of 300 mM sucrose solution, because it caused an increase in the refractive index. Scale bar 0.5 µm. d, The mean time course of ΔFRET, representing cis-SNARE formation (a). Magenta and purple bars indicate 20 mM and 300 mM sucrose solutions. b1–b4 indicate time points when the images in c were obtained. n = 11 boutons (4 slices, 3 rats) for both 20 mM sucrose and 300 mM sucrose. e, ΔFRET of the cis-SNARE probe by the 20 mM and 300 mM sucrose solutions from the same dataset analysed in d. For the baseline fluctuations (Baseline), 3 consecutive FRET values (5 min apart) from all boutons were plotted. Two-sided Kruskal–Wallis test (H = 48.27, **P = 8.27 × 10⁻¹⁰) followed by Dunn’s multiple comparison test (vs Baseline). 20 mM sucrose, P = 0.394; 300 mM sucrose, **P = 1.33 × 10⁻⁷. Error bars indicate s.e.m

Source data.

a, The values of FWHM of the bouton at 6 min after the onset of STDP from the data shown in Fig. 4e. n = 14 pushing spines (14 slices, 11 rats) and n = 9 non-pushing spines (9 slices, 8 rats). Two-sided Wilcoxon signed-rank test vs 0 µm, W = 4, **P = 2.58×10⁻³ and W = 41, *P = 3.3×10⁻²). b, Correlation between ΔFWHM of the bouton and ΔLength of the spine at 6 min after STDP induction from the data shown in a. Red and black circles indicate pushing and non-pushing spines, respectively. Spearman’s correlation with two tails, rho = 0.406, **P = 1.38 ×10⁻³. c, Correlation between ΔFRET and ΔFWHM of the boutons at 6 min after STDP induction from the data shown in a. Spearman’s correlation with two tails, rho = 0.557, **P = 6.8 ×10⁻⁵. d, Correlation between ΔFRET of the bouton and ΔV of the spine at 6 min after STDP induction from the data shown in a. Spearman’s correlation with two tails, rho = 0.002, P = 0.848. e, f, Absence of spine enlargement (e) and FRET changes (f) when uncaging was applied without postsynaptic spikes (green, n = 7 boutons, 7 slices, 5 rats). The control STDP traces are the replicates from the red traces in Fig. 4d, g. Two-sided Kruskal–Wallis test (H = 4.39, P = 0.356) for (e) and (H = 0.48, P = 0.975) for (f). Error bars indicate s.e.m

Source data.

a, d, Fluorescence images (single XY- and XZ-planes) of the bouton-spine pairs, and dendritic spines before and after stimulation with STDP protocol for a pushing (a) and a non-pushing (d) spine. Horizontal scale bars indicate 1 µm, vertical scale bars indicate 2 µm. In all bouton-spine pairs shown in this figure, the criteria (Fz > 0.8 and Fx > 0.6) were met (Methods). b–f, The Pr measurements (b, e) and the averaged iGluSnFR traces (c,f) before and after induction of STDP. Orange bars indicate optogenetic stimulation (5 ms). g–j, The time courses of spine enlargement (g), spine length (h), Pr (i) and the peak of the averaged iGluSnFR traces (j) in pushing (red) and non-pushing (black) spines for the same set of data shown in k–n. k–n, The increases in spine volume (ΔV, k), spine length (ΔL, l), Pr (ΔPr, m) and average release (ΔRelease, n) in the synapses with pushing and without pushing. n = 6 (6 slices, 6 rats) and 5 (5 slices, 5 rats) for pushing and non-pushing spines at 10 min after STDP induction. Two-sided Mann–Whitney U test (U = 11, P = 0.537 [k]). Two-sided Wilcoxson signed-rank test (W = 21, *P = 0.036 [l]; W = 21, *P = 0.036 [m]; W = 21, *P = 0.036 [n]) for pushing synapses and (W = 3, P =0.281 [l]; W = 1, P = 0.104 [m]; W = 1, P = 0.106 [n]) for non-pushing synapses. o, p, Correlation for ΔPr (o) and ΔRelease (p) vs ΔV from the same dataset as k–n. Spearman’s correlation, rho = 0.257, P = 0.116 (o) and rho = 0.132, P = 0.273 (p). q, r, Correlation for ΔPr (q) and ΔRelease (r) vs ΔL with (red) and without (black) spine elongation from the same dataset as k–n. Spearman’s correlation, rho = 0.469, *P = 0.025 (q) and rho = 0.746, **P = 1.25×10⁻³ (r). Error bars indicate s.e.m

Source data.