Cell assembly dynamics of sparsely-connected inhibitory networks : a simple model for the collective activity of striatal projection neurons

The striatum is a brain structure involved in many functional roles, from action selection to learning and motor control. Some noticeable experiments demonstrated that the striatum encodes information received from sensory areas by the alternating activation of small assemblies of Medium Spiny Neurons firing synchronously. We propose a simple mathematical model of inhibitory neurons which is able to capture some of the most relevant features of the striatal dynamics. The simplicity of the model allows to identify the minimal ingredients required by a neural network to display the rich behavior observed in the striatum. The derivation of an easily tractable model can allow a deeper understanding of the fundamental mechanisms at the basis of the striatal dynamics and eventually to elucidate in a near future how neural diseases, like Parkinson and Huntigthon disease, alter the normal behaviour of Medium Spiny Neurons. Striatal projection neurons form a sparsely-connected inhibitory network, and this arrangement may be essential for the appropriate temporal organization of behavior. Here we show that a sparse inhibitory network of artificial Leaky-Integrate-and-Fire neurons can reproduce key features of striatal population activity, as observed in brain slices [Carrillo-Reid et al., J. Neurophysiology 99 (2008) 1435--1450]. In particular we develop a new metric to determine the conditions under which sparse inhibitory networks form anti-correlated cell assemblies with variable firing rates of individual cells. We find that in this parameter range, the network displays an input-specific sequence of cell assembly switching, and can optimally discriminate between similar inputs. Our results support the proposal [Ponzi and Wickens, PLoS Comp Biol 9 (2013) e1002954] that striatal network topology is set up to allow stimulus-selective, temporally-extended sequential activation of cell assemblies.