Integrated Calendar

Since the discovery of place cells in the hippocampus, a variety of experimental observations have pointed to the complexity of hippocampal circuit dynamics and their importance in memory related tasks. During spatial navigation, place cell activity predicts the upcoming animal location within the short time scale of individual cycles of theta oscillations. Sudden changes of the spatial context are followed by a bistability between population coding of past and current context, paced by the theta rhythm. During immobility, brief sequences of place cell activation encode spatial trajectories, which have been linked to learning in spatial memory tasks and goal-directed navigation. Finally, when the animal is engaged in a delayed memory task, hippocampal cells fire at specific time intervals within the delay period and the activity of a population of cells is predictive of the behavioral outcome. I will present a unified attractor network model that accounts for this wide range of experimental observations. A critical component of the model is the use of realistic synapses that exhibit short-term plasticity driven by presynaptic activity. Complexity in the network dynamics emerges due to the effect of history dependent synaptic states on the network activity. Model predictions, possible extensions of the model and its relationship to dynamics observed in other cortical areas will be discussed.

Neuronal networks can generate complex patterns of activity that depend on membrane properties of individual neurons as well as on functional synapses. To decipher the impact of synaptic properties and connectivity on neuronal network behavior, we studied using a combination of electrophysiological recordings and the synaptic depression-facilitation model, the responses of neuronal ensembles from small (between 5-30 cells in a restricted sphere) and large (acute hippocampal slice) networks to single electrical stimulation.
Interestingly, in both cases, a single stimulus generated a synchronous long-lasting bursting activity. We characterized this activity in neuronal populations using electrophysiological recordings and we also extract the network time constant parameters using the mean-field model based on synaptic facilitation/depression. While the initial spikes triggered a reverberating network activity that lasted 2-5 seconds for small networks, it lasted only up to 300 milliseconds in slices, a phenomena that was also present in our simulations. We found here that the reverberation time has a bell shaped relation with the synaptic density. In addition, before reaching its maximum, this reverberation time increased sub-linearly with the network connectivity parameter.
We conclude that synaptic properties and the network connectivity shape the mean burst duration, which persists across various network scales. This synchronization is an inherent property of sufficiently connected neural networks based on synaptic depression and facilitation.