Modeling the Actin Organization in Dendritic Spines using a Minimal Stochastic Model

Synaptic plasticity is a complex process, involving the intricate coupling between biochemical and biophysical events. Actin remodeling plays an important role in this coupling of biochemical and biophysical events at the synapse. For example, changes in size and shape of dendritic spines, small membrane protrusions which form the postsynaptic component of most excitatory synapses, affect the geometry of the synapse and thereby neurotransmitter dynamics. While it is known that actin plays a key role in synaptic plasticity, the relationship between spine size and shape and actin organization remains poorly understood. In this work, we pose an inverse problem -- given a spine shape, what configurations of the actin network can be accommodated? To answer this question, we develop a minimal model of actin dynamics and simulate it using the agent-based modeling framework, Cytosim. Our simulations show that different spine geometries fundamentally accommodate different actin networks, characterized by the distribution of Arp2/3 branch points, numbers of barbed ends, network topology and other metrics. Furthermore, we find the synergy between Arp2/3 mediated branching and cofilin mediated severing for barbed end generation is dependent on the spine shape.