What is the information capacity of a synapse?

Although synapses have been already characterised at the functional level, knowledge of the operational mechanisms which lie beneath their function remain limited. Among the mechanisms which play a significant role in storage information is the movement of regulatory proteins to and from synapses, such as the calcium-dependent kinase (CaMKII), and this is translated into changes in synaptic strength connection. Here, a stochastic simulator of synaptic function at the molecular level is presented. The synaptic model which is used is based on published physiological and biochemical data and experimentation. The simulator treats presynaptic cells as a sequence of action potentials and thus, simulates the chemical reactions in the post-synaptic cell. The simulator is validated using a small-artificial-pool of five chemical species participating in one reversible and one enzymatic chemical reaction. Furthermore, the input transduction variability is calculated in terms of mutual information between the input and the first chemical species receiving input. Finally, the metabolic cost per action potential, i.e. the cost for operating this synapse, is measured.