Supplementary Figure 7: Monte Carlo simulations of SYP-1, vGlut1, Bassoon and PSD-95 synaptic nanomodules.

Analysis was conducted as in Fig. S6 except as described (a-b) Super-resolved maximum intensity projection images of PSD-95 (green), vGlut1 (red), SYP-1 (blue) nanomodules as in Fig. 2 and Bassoon nanomodules as in Fig. 3. Images were acquired with a Leica SP8 3X instrument using gated detectors for vGlut1 (AlexaFluor-594, 0.3 to 6 nsec), SYP-1 (Atto-647N, 0.2 to 6 nsec) and Bassoon (Atto-647N, 0.2 to 6 nsec) in combination with a pulsed 775 nm depletion beam for a maximum resolution of ~50 nm. Scale bar for a and b, 5 µm. Magenta circles indicate simulated spine sizes reflecting spine sizes from our data in Fig. 2 and Fig. 3 at 12.5, 25, 37.5, 50, 62.5, 75, 87.5 and 99 percentiles. The simulations were repeated using images from three independent transfection experiments from Fig. 2 and 3. (c-f) Quantification of the proportion of simulated spines of the indicated size that did not colocalize with (c) PSD-95 nanomodules (n = 30375 spines), (d) vGlut1 nanomodules (n = 26131 spines), (e) SYP-1 nanomodules (n = 27511 spines) and (f) Bassoon nanomodules (27091 spines). (g-j) Quantification of the proportion of (g) PSD-95, (h) vGlut1, (i) SYP-1 and (j) Bassoon nanomodules in the remaining spines that contain clusters. The majority of simulated spines that associate with puncta contain only a single nanomodule with significantly fewer multi-nanomodule containing spines. (g) PSD-95 simulations (single, 96.8%; two, 3.1%; three, 0.09%; four and five, 0%, p < 0.0001, F(4, 10) = 11750, one-way ANOVA, n = 1124 spines). (h) vGlut1 simulations (single, 78.1%; two, 17%; three, 3.6%; four, 1% and five, 0.1%, p < 0.0001, F(4, 10) = 381, one-way ANOVA, n = 5415 spines). (i) SYP-1 simulations (single, 90.6%; two, 8.3%; three, 0.8%; four, 0.05% and five, 0%, p < 0.0001, F(4, 10) = 852.2, one-way ANOVA, n = 3958 spines). (j) Bassoon simulations (single, 88.5%; two, 9.1%; three, 0.7%; four, 0.09% and five, 0%, p < 0.0001, F(4, 10) = 8162, one-way ANOVA, n = 4332 spines). Graphs in (c-j) represent mean +/− SEM. (k-m) Pearson’s correlations between simulated spine sizes and the number of (k) PSD-95 (Sim. 1, n = 629, Sim. 2, n = 336, Sim. 3, n = 159 PSD-95-containing spines), (l) vGlut1 (Sim. 1, n = 1291, Sim. 2, n = 2148, Sim. 3, n = 1976 vGlut1-containing spines), (m) SYP-1 (Sim. 1, n = 1962, Sim. 2, n = 759, Sim. 3, n = 1237 SYP-1-containing spines) and (n) Bassoon (Sim. 1, n = 1800, Sim. 2, n = 1317, Sim. 3, n = 1215 Bassoon-containing spines). Only spines that contained clusters were included in this analysis. Averages from three independent simulations for PSD-95, vGlut1 and SYP-1 are shown in Fig. 2f–h and for Bassoon in Fig. 3h.