FROM BIOPHYSICS TO MODELS OF NETWORK FUNCTION

Neurons and synapses display a rich range of time-dependent processes. Which of these are critical to understanding specific integrative fun ctions in the brain? Computational methods of various kinds are used t o understand how systems of neurons interact to produce behavior. Howe ver, these models often assume that neuronal dynamics and synaptic str engths are fixed. This review presents some recent models that illustr ate that short-term synaptic plasticity mechanisms such as facilitatio n and depression can have important implications for network function. Other features of synaptic transmission such as multi-component synap tic potentials, cotransmission, and neuromodulation with obvious poten tial computational implications are presented. These examples illustra te that synaptic strength and intrinsic properties in networks are con tinuously varying on numerous time scales as a function of the tempora l patterns of activity in the network. Thus, both firing frequency of the neurons in a circuit, and the modulatory environment determine the intrinsic and synaptic properties that produce behavior.