Fig. 8: Simulations of vesicular release in synapses with different active zone geometries (model with endogenous buffers).

a Results of three representative simulations for active zone geometry cases ii, v and viii. Left, Simulated release. Grey circles depict randomly assigned locations of vesicular release sites. Blue and red dots correspond to synchronous and asynchronous release events respectively, which occurred during a simulated train of 51 action potentials. Middle, Blurred position. To mimic the experimental data, multi-vesicular release events were excluded and locations of the remaining events were blurred by adding a random value corresponding to the accuracy of SF-iGluSnFR event localisation (75 nm). Right, Sync/async areas. Comparison of convex hull areas circumventing simulated blurred locations of synchronous or asynchronous events. b–e Relationships between computed functional presynaptic parameters for nine modelled active zone geometries (averages of n = 250 simulations in each case): dependency of the asynchronous release fraction n_A/n_T on the overall release efficacy n_T (b) and dependencies of the probability of synchronous release p_s (c), of asynchronous release p_a (d) and of model-predicted asynchronous release fraction \({p}_{a}/({p}_{s}+{p}_{a})\) on active zone size, calculated for single RRP vesicles. Colour codes are shown on the right. f, g Comparison of the model’s output to the experimental results. Combined outputs from all nine considered geometry cases (n = 10 simulations per case) were processed using the same analysis routine as for the experimental data set. f Model-predicted relationship between asynchronous release fraction and the overall vesicular release efficacy n_T. Similar to analysis in Fig. 3a horizontal dashed lines depict the average fraction of asynchronous release across all simulated boutons, continuous line – moving average, black dots – average values of n_T and n_A/n_T in each tercile (mean ± SEM). g Results of reshuffling analysis (similar to Fig. 6d). Data points and box-and-whisker plots: centre, median; notch 95% confidence interval for median; box, 25th–75th percentiles; whiskers, the most extreme data points not considered outliers (within 1.5 times the interquartile range from the bottom or the top of the box). ***p < 0.001, two-tailed Mann–Whitney U test (for exact p values see SourceData.xlsx file).