Linking structure and function in striatum using algebraic topology, digital microcircuit reconstruction and simulations of the healthy and diseased network

The relationship between the structure and network dynamics within the striatum is currently not well understood. We have applied algebraic topology to investigate the local structural connectivity in the striatum, and then used simulations to predict how structure shapes network dynamics. We used a full-scale digital reconstruction of the mouse striatal microcircuitry: both healthy and at different stages of Parkinson’s Disease (PD). These stages are characterized by successively modified healthy morphologies of the striatal projection neurons (SPN), including changes in dendritic spine count. We compared the distribution of topological motifs, in the form of directed cliques, between these microcircuits. The distribution of directed cliques in the healthy striatal microcircuits showed that striatal interneurons, despite only accounting for 5%, are crucial for the construction of high dimensional directed cliques. In PD networks the presence of directed cliques drastically decreased with the disease progression. We then used simulations to investigate whether these changes in structural connectivity affect functional connectivity. Signal transfer, especially correlation transfer, in the corticostriatal system was affected. We also found that the resulting changes in intrastriatal inhibition altered the correlations between the striatal projection neurons. Directed cliques already provided insight on structural and functional properties of neocortical micrucircuitry. Here we applied this topological approach to investigate striatal networks and highligthed important differences with respect to neocortex. Combining theory with simulations using data-driven in silico reconstructions will allow us to form quantitative predictions on how structure and network dynamics relate in health and disease.