Article Figures & Data
Figures
- Figure 1.
mSCaTs measured by GCaMP6f imaging reflect NMDAR activation at individual synapses following spontaneous single vesicle release. A, Cultured hippocampal neuron infected with AAV-GCaMP6f. Left, GCaMP6f average (green); middle, ΔF calculated by subtracting GCaMP6f average from GCaMP6f maximum projection (magenta); right, merge of GCaMP6f average (green) and ΔF (magenta). Bottom, Zoom in of boxed spine from cell in top left. First panel is GCaMP6f average, the 2nd through 4th panels are individual frames showing a single mSCaT at 0, 400, and 800 ms, respectively. Red circles indicate ROIs for data traces shown in B. B, ΔF/F traces from spine and dendrite regions circled in red in A. C, Frequency histogram of mSCaT amplitude for individual synapses across 923 spines from 8 cells. D, Frequency histogram of mSCaT frequency for individual synapses across 923 spines from 8 cells. E, Representative GCaMP6f traces demonstrating that treatment with ryanodine, thapsigargin, DNQX, and nifedipine (blockers) did not alter mSCaT amplitude compared with vehicle treatment. F, Quantification of effect of blockers on mSCaT amplitude compared with vehicle treatment revealed that blockade of non-NMDAR sources of Ca2+ did not impact mSCaT amplitude. G, Representative GCaMP6f traces demonstrating that treatment with APV eliminated mSCaTs compared with vehicle treatment. H, Treatment with APV eliminates 94% of events. I, Raising extracellular Ca2+ causes increased mSCaT amplitude, whereas application of 30 μm, 100 μm, and 1 mm Mg2+ reduce mSCaT amplitude. For example traces, see Extended Data Figure 1-1. ns, not significant; ****p < 0.0001.
- Figure 2.
GluN2B-NMDARs mediate majority of response to spontaneous glutamate release. A, Average projection of GCaMP6f (green) at baseline and following treatment with vehicle or ifenprodil. Active synapses are shown in magenta. Following treatment with ifenprodil there is a clear reduction in the number of active spines compared with baseline. Scale bar, 10 μm. B, Example traces from spines treated with either vehicle (black) or ifenprodil (red). C, Ifenprodil treatment leads to a reduction in mSCaT amplitude at both 3 and 5 weeks. Outliers removed for data display. Solid line represents median and dashed lines indicate 1st and 3rd quartile. Second plot is cumulative probability for all spines. For immunoblotting data, see Extended Data Figure 2-1 and Extended Data Figure 2-2. D, Ifenprodil treatment leads to a reduction in mSCaT frequency at both 3 and 5 weeks. Outliers removed for data display. Solid line represents median and dashed lines indicate 1st and 3rd quartile. Second plot is cumulative probability for all spines. E, Post-treatment mSCaT amplitude versus baseline mSCaT amplitude has a slope of 0.6167 ± 0.033 for vehicle-treated cells and 0.327 ± 0.025 for ifenprodil-treated cells and these slopes are significantly different (p < 0.0001). Post-treatment mSCaT amplitude is correlated with baseline amplitude for both vehicle-treated synapses (R2=0.42, p < 0.0001) and ifenprodil-treated synapses (R2=0.22, p < 0.0001). F, Baseline mSCaT frequency versus post-treatment mSCaT frequency reveals that nearly all synapses show a reduction in event number with ifenprodil treatment (vehicle: slope= 0.7984 ± 0.02368; ifenprodil: slope = 0.108 ± 0.005452; p < 0.0001). Post-treatment mSCaT amplitude is correlated with baseline amplitude for both vehicle-treated synapses (R2=0.56, p < 0.0001) and ifenprodil-treated synapses (R2=0.31, p < 0.0001). G, Normalized amplitude post-ifenprodil treatment is negatively correlated with baseline amplitude. H, Normalized mSCaT frequency post-ifenprodil treatment is negatively correlated with baseline mSCaT frequency. I, Within spine CV decreases following ifenprodil treatment. J, Three micromolar ifenprodil (blue) treatment leads to a significant reduction in mSCaT frequency. Outliers removed for data display in violin plots. Second plot is cumulative probability for all spines. K, Three micromolar ifenprodil (blue) treatment leads to a significant reduction in mSCaT amplitude. Outliers removed for data display in violin plots. Second plot is cumulative probability for all spines. L, Ro 25-6981 (green) treatment leads to a significant reduction in mSCaT frequency. Outliers removed for data display in violin plots. Second plot is cumulative probability for all spines. M, Ro 25-6981 (green) treatment leads to a significant reduction in mSCaT amplitude. Outliers removed for data display in violin plots. Second plot is cumulative probability for all spines. ns, not significant; *p < 0.05; **p < 0.01; **** p < 0.0001.
- Figure 3.
NMDAR activation is independent of spine area. A, Example stretch of dendrite from post-Ca2+ imaging GCaMP6f z-stack. Scale bar, 10 μm. B, Zoom in of spines indicated by white arrowheads from A paired with their respective Ca2+ traces. Scale bar, 1 μm. C, mSCaT amplitude weakly, negatively correlates with spine area. First plot is all spines, second plot binned data. D, Effect of ifenprodil on amplitude (light red) and frequency (dark red) does not correlate with spine area. Left plot is all spines; right plot is binned data. E, GCaMP6f max projection acquired directly following Ca2+ imaging (scale bar, 20 μm) of dSTORM imaged neuron. White box indicates area where super-resolution imaging was performed. F, Zoom in on the region from E. Max projection of GCaMP6f stack acquired at time of STORM imaging (white). Scale bar, 5 μm. Super-resolved shank localizations are shown in red. Right, Zoom-in of the shank localizations from the spine indicated with the yellow arrowhead. G, Zoom-in of a mSCaT in the spine indicated by yellow arrowhead in F. H, Ca2+ trace from spine indicated by yellow arrowhead in F. I, Amplitude does not correlate with PSD area.
- Figure 4.
Mapping synapse position and mSCaT characteristics with Imaris. A, Ca2+ imaged spines along traced dendrite from Imaris. Scale bar, 30 μm. First panel shows distance from soma, second panel shows branch depth, and third panel shows mSCaT amplitude. Colors are warmer as spines that are farther from the cell body, have higher branch depth, or have larger mSCaT mean amplitudes respectively. For example of Imaris tracing see Extended Data Figure 4-1. B, Zoom-in of boxed areas from A. C, mSCaT amplitude did correlate with distance from soma for proximal spines. D, mSCaT amplitude did correlate with distance from soma when distal spines are included. E, Binned data from D demonstrating relationship between spine distance from the soma and mSCaT amplitude F, Magnitude of ifenprodil effect on mSCaT amplitude (light red) and frequency (dark red) does not correlate with distance from the soma. G, mSCaT amplitude did not correlate with the number of branch points away from the soma the synapse is (branch depth) for proximal spines. H, mSCaT amplitude does correlate with branch depth when distal spines are included. I, Mean mSCaT amplitude at each branch depth. J, Magnitude of ifenprodil effect on mSCaT amplitude (light red) and frequency (dark red) did not correlate with branch depth.
Extended Data
Figure 1-1
Example mSCaT traces from cells in Figure 1. Example mSCaT traces for data from Figure 1I. Download Figure 1-1, PDF file.
Figure 2-1
Chronic expression of GCaMP6f does not alter GluN2B-NMDAR contribution to mSCaTs or GluN2B developmental shift. A, Ifenprodil has the same effect on mSCaT amplitude for transiently transfected cells as infected cells (Normalized amplitude: Veh-infected: 0.98 ± 0.016, n = 1196/20; Ifen-Infected: 0.622 ± 0.017, n = 565/10; Veh-transfected: 1.07 ± 0.035, n = 103/4; Ifen-transfected: 0.633 ± 0.051, n = 81/4; p < 0.0001, one-way ANOVA). B, Ifenprodil has the same effect on mSCaT frequency for transiently transfected cells as infected cells (Normalized frequency: Veh-infected: 0.82 ± 0.022, n = 961/20; Ifen-Infected: 0.137 ± 0.006, n = 788/10; Veh-transfected: 0.855 ± 0.063, n = 110/4; Ifen-transfected: 0.20 ± 0.048, n = 142/4; p < 0.0001, one-way ANOVA). C, Hippocampal neurons were infected with AAV-GFP or AAV-GCaMP6f at 0 DIV and harvested at 12 or 19 DIV. Representative immunoblots for GluN1, GluN2A, GluN2B, and α-tubulin are shown. Molecular weight markers (kD) indicated for each immunoblot. Sample from adult rat hippocampus (HC) included as positive control. D, Quantitation shows decrease in GluN2B/GluN2A ratio from 12 to 19 DIV. Graph depicts normalized mean GluN2B/GluN2A ratios ± SEM (DIV factor: p < 0.0001; Infection factor: p = 0.5009; Interaction: p = 0.812; two-way ANOVA). Download Figure 2-1, TIF file.
Figure 2-2A–D
Representative full immunoblots from Figure 2-1. E, Full blot of REVERT protein stain. Download Figure 2-2, TIF file.
Figure 4-1
Imaris tracing examples. A, Max projection of GCaMP6f stack taken following GCaMP6f imaging. Zoom-in of white box is on the right. Scale bars, 30 μm. B, Semiautomatic spine detection on max projection of GCaMP6f. Blue spines are spines that do not have Ca2+ imaging data, whereas spines in red do. Zoom-in of white box is on the right. Scale bars, 30 μm. A, Semiautomatic dendrite and spine detection in Imaris detects dimensions of cell features based on fluorescence. Download Figure 4-1, TIF file.
In this issue
