Information processing in the central nervous system depends, to a large extent, on the rate at which signals are transmitted at chemical synapses. For signals to be conveyed faithfully from presynaptic to postsynaptic neuron, every action potential arriving at the presynaptic terminal should elicit the release of neurotransmitter — to maintain synaptic transmission during periods of high-frequency firing (of up to several hundred action potentials per second), the presynaptic neuron should be capable of transmitter release at a similarly high rate. The maximum possible rate of release will depend, among other things, on the number of presynaptic release sites (active zones) and the rate at which synaptic vesicles can fuse with the cell membrane.

So, how fast can a conventional synapse release transmitter from vesicular stores? Previous estimates have been surprisingly low, at around 20 vesicles per second per release site; too slow, in fact, to follow the high-frequency firing observed at many central synapses. One possible explanation for this lies in the method used to determine the rate of transmitter release. By monitoring postsynaptic responses as a proxy for presynaptic exocytosis, a number of factors could have led to underestimates of the rate of release; for example, the saturation or desensitization of postsynaptic receptors. A more direct measure of the number of vesicles fusing with the presynaptic membrane can be made at synapses where the nerve terminal is large; in this case, individual fusion events can be detected as an increase in the presynaptic membrane capacitance. Would this approach lead to higher estimates of the speed of release at central synapses?

To address this question, Sun and Wu studied glutamate release at one of the largest synapses in the rat brainstem — the calyx of Held. By recording from the large presynaptic terminal at this synapse, which fires at high frequency in vivo, they were able to make presynaptic capacitance measurements, while simultaneously monitoring the excitatory postsynaptic currents (EPSCs) arising from glutamate release. As they report in Neuron, Sun and Wu found that EPSC amplitudes were saturated when capacitance measurements had only reached 35% of their maximum value, supporting the prediction that measuring presynaptic vesicular fusion does indeed lead to higher estimates of the amount of transmitter released — more than 300 vesicles per second per active zone in the case of this preparation.

These findings show that conventional central synapses can liberate transmitter at a much higher rate than was previously estimated; high enough to match the rate of incoming action potentials during periods of high-frequency stimulation. The ability to monitor synaptic transmission simultaneously from both the presynaptic terminal and the postsynaptic neuron at the calyx of Held provides a means of learning more about the features of transmitter release at a central site.