| Summary: | The function of a neural network is dependent on more than the static patterns of connections between neurons. For example, in response to time-varying stimuli in the environment, sensory neurons are activated in specific temporal patterns. Thus, to extract relevant information the neurons of a neural network must have the appropriate temporal properties. We have studied the dynamic properties of the synapses and of action potential generation in the cricket cercal system, a system that detects low-frequency air disturbances. First, we show that the electrotonic properties of the interneurons in the cricket cercal sensory system are preserved during postembyronic development, thereby maintaining their frequency filtering properties. Also, we found that the frequency of the spikes in the interneurons is linearly related to the level of injected current and that the interneurons show very little spike adaptation. Last, we combine a biophysical model of the spiking mechanism of an interneuron with the known temporal dynamics of the synaptic input in a computer simulation. We found that the rate of synaptic depression, a function of the fractional release rate of the sensory neuron synapses, is the most important factor in determining the temporal response of the interneurons to sound stimuli. In contrast, synaptic facilitation had a relatively weak effect on the temporal properties of the interneurons.
|
|---|
