Fig. 4: Synchronized sEPSCs are similar in amplitude and kinetics indicative of multi-directed transmission.

Cartoons of the tripartite complex (Fig. 1e) highlighting the differences between conventional “spillover” (a)^7,39,42,43 and multi-directed forms of ACh transmission (b). a ACh released from the starburst terminal rapidly activates low-affinity nAChRs mainly at null connections (n) giving rise to quantal EPSCs (red trace on right). A small fraction of ACh molecules also diffuse to preferred connections (p) where they mediate slower, weaker responses that generally fall below the detection limit of electrophysiological recordings (blue trace on the right). However, during more intense activity, ACh from multiple neighboring synapses can “spillover” and generate sizable responses at preferred sites, which could in theory explain the observed discrepancy between the anatomical and functional ACh connectivity^25,29,31. b The central proposal here is that ACh signaling occurs over a broad spatial scale giving rise to “multi-directed” transmission. Unlike spillover transmission, ACh release from a single vesicle is equally efficacious at preferred and null sites (right), likely mediated by receptors with high binding affinity (triangle). A variety of factors including the geometry of the synaptic cleft, the concentrations of ACh packaged in single vesicles and the precise biophysical properties of nAChRs (e.g., low vs. high binding affinity) could dictate whether spillover or multi-directed transmission occurs (Supplementary Fig. 5). c Synchronous cholinergic sEPSCs observed in DSGC pairs indicate that ACh release from a single vesicle activates two sites with comparable efficacy, supporting the multi-directed transmission model (b). d The peak amplitudes (left), decay constants (middle) and rise times (right) of the synchronous events plotted against each other (241 events from 6 DSGC pairs, across 5 retinas). The red line indicates the line of best fit, and r indicates the Pearson’s correlation coefficient which is significantly higher than a shuffled distribution (*p = 0.043, **p = 0.0055, ***p = 9.7 × 10⁻¹⁴; two-tailed z-test; see methods). The data were standardized to the mean and standard deviation (SD) to avoid artefacts that could be introduced by cell-to-cell variability in these parameters. Source data are provided as a Source Data file for Fig. 4c, d.