Despite a wealth of anatomical and electrophysiological data, network mechanisms underlying the role of striatum in cognitive and motor functions have remained obscure. To understand the functional properties of this network it is important to characterize how incoming feedforward excitatory inputs interact with the ongoing activity dynamics. Recent experimental data [1-3] and computational models [4] suggest that purely inhibitory recurrent connectivity of striatal neurons could support transient neuronal assemblies.

We found that the activity dynamics in networks with a Gamma-shaped connectivity profile were largely determined by the mean and variance of the external input: weak external input or high input variance induced unstructured asynchronous-irregular activity (AI), whereas stronger external inputs or low input variance induced stable ‘winner-takes-all’ (WTA) dynamics. In an input regime close to the noise threshold the activity organized into unstable, transient spatial bumps, referring to as a ‘transient activity’ state (TA), resembling the experimentally observed neuronal clusters [5]. Finally, we showed that a small asymmetry in the spatial structure of the recurrent connectivity or in the synaptic strengths was able to make the activity bump pattern moving in one direction [6]. Such moving patterns could form the basis of sequence generation in the striatum.

In summary, we conclude that a Gamma-shaped spatial connectivity provides the striatal network with a rich dynamical repertoire, enabling the bump activity to produce task related activity sequences.

Acknowledgements
This work was funded in parts by the NeuroSeeker Foundation, the EU Erasmus PhD program NeuroTime, and the Carl-Zeiss Foundatation. All simulations were carried out with NEST (http://www.nest-initiative.org).

References
[1] Adler et al. (2013) JNeurosci 33(11):4854-66 doi: 10.1523/jneurosci.4791-12.2013
[2] Carrillo-Reid et al. (2011) JNeurosci 31(42):14972-83 doi: 10.1523/jneurosci.3226-11.2011
[3] Lopez-Huerta et al. (2013) JNeurosci 33(11): 4964-75 doi: 10.1523/jneurosci.4721-12.2013
[4] Ponzi & Wickens (2010) JNeurosci 30(17):5894-911 doi: 10.1523/jneurosci.5540-09.2010
[5] Barbera et al. (2016) Neuron 92(1):203-213 doi: 10.1016/j.neuron.2016.08.037
[6] Koetter, R., & Wickens, J. (1995). JComp Neurosci, 2(3), 195-214. pmid: 8521287