Probabilistic transformation of temporal information at individual synapses

The temporal structure of synaptic activation is critical to the induction of long-term synaptic plasticity. Experimentally, action potentials in presynaptic axons are reliably triggered by electric stimulation. The frequency of presynaptic input is the same as the stimulation frequency and is defined here as "input frequency". Upon arrival of presynaptic action potentials, transmitter release at individual synapses is probabilistic. We define the average frequency of transmitter release at individual synapses as "release frequency". Computer modeling has been widely used to simulate the input frequency-dependent elevation of calcium in dendritic spines, which is believed to trigger the induction of synaptic plasticity. In these models, input frequency has been used indiscriminately as release frequency to simulate synaptic events. Here, using both empirical and theoretical methods, we demonstrated that the release frequency is significantly smaller than the input frequency due to the probabilistic nature of transmitter release. We propose that this temporal transformation be taken into account in future synaptic plasticity models.