Back

Synaptic vesicle undocking induces low frequency depression

Silva, M.; Trigo, F.; Llano, I.; Marty, A.

2026-01-15 neuroscience
10.64898/2026.01.15.699619 bioRxiv
Show abstract

Synaptic depression is often interpreted as reflecting depletion of the readily releasable pool (RRP) following exocytosis. Such a mechanism predicts little or no depression at low stimulation frequency, as RRP replenishment should then offset the loss of vesicles by exocytosis. Nevertheless, in several types of mammalian central synapses, repetitive presynaptic stimulation at low frequency (< 5 Hz) elicits synaptic depression (low frequency depression, or LFD). In the present work we count the number of synaptic vesicles released at individual active zones to study the RRP and its replenishment during LFD. Contrary to depletion models of synaptic depression, we find that LFD does not depend on previous SV consumption. We find that LFD displays a long recovery time course (tens of seconds) when challenged by isolated stimulations but is immediately reversed by a high frequency train. We suggest that LFD results from undocking, a shift between two classes of synaptic vesicles organized sequentially inside the RRP (replacement vs. docked vesicles) in favor of the upstream (replacement) state. While undocking is apparent hundreds of milliseconds after a stimulation, calcium dependent docking takes only a couple of milliseconds, explaining the fast LFD recovery when stimulating at high frequency. Consistent with the undocking model, we find that double presynaptic stimulations alleviate LFD as they favor vesicular docking and RRP replenishment. Finally, we expand our model to explain how stimulation frequency shapes short-term synaptic depression, changing from depression at low frequency to a facilitation-depression sequence at medium or high frequency trains.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.