The role of the striatum in the control of movements and in the proces
sing of cortical information has received much attention in the recent
years. We set out a simple biophysical model for the medium-spiny neu
ron (msn), the most abundant cell in striatum. This neuron receives tw
o main kinds of inputs, namely, cortical excitatory inputs and dopamin
ergic inputs coming from the substantia nigra pars compacta. The msn a
xon impinges directly onto the globus pallidus and onto the substantia
nigra pars reticulata neurons and onto striatal neurons through recur
rent branches of the axon. The msn is characterized by spiny dendritic
trees with a high density of spines (1 to 4 spines/mum) and the proba
ble existence of dendritic spikes. The model predicts that the neuron
can integrate excitable inputs in a linear or a nonlinear mode. In the
nonlinear mode, the neuron allows the detection of simultaneous (or a
lmost simultaneous) synaptic inputs; it facilitates either a slowing d
own or a speeding up of the information transfer between the synaptic
input location and the soma and is sensitive to inhibition-excitation
pairing. Conversely, in the linear integrative mode, the somatic volta
ge is determined by a weighted summation of the synaptic inputs. Sever
al geometrical, electrical, or temporal factors can control the switch
between these behaviors: the density of excitable dendritic elements,
the dendritic radius, the resistance of the spine stem, the membrane
resistance, the time between excitations, and the distance between syn
aptic sites. Finally, the signification of this behavior is discussed
in connection with the putative role of dopamine and with the striatal
net organization. (C) 1994 Wiley-Liss, Inc.