UCLA

Higher order functions of cognition rely on the cooperative computations of multiple brain regions. Understanding how these spatially and temporally distributed neural circuits work together remains an important question. Attempts have been made at a computational understanding of these brain circuits and their coordinated activity, but usually involve models requiring thousands of parameters. The explosion of dimensionality trades biophysical accuracy with predictive power and analytic understanding. In this dissertation, I outline a theory of cortical interaction that can explain a variety of phenomena within the in vivo brain using only two parameters, and I study the implications of this theory using the cortico-entorhinal-hippocampal circuit in vivo.

Briefly, a theory of coordinated interaction between generic cortical networks is developed using only excitation, inhibition, and the adaptation of excitation. By constraining this model onto a 2D parameter space by modulating only the recurrent excitation of the efferent network and the external drive coming in from the afferent network, the theory is able to reproduce over a dozen experimental observations during spontaneous activity in vivo. The theory reproduces previously observed phenomena, namely spontaneous persistent activity, which occurs when the efferent networks decouple from the afferent drive and remain active while the afferent network has shut down. The theory goes further to predict a novel phenomena, spontaneous persistent inactivity, when the efferent network remains inactive while the afferent excitation turns on. This has never before been observed experimentally or computationally. We test our theory using simultaneously recording local field potential within the neocortex and membrane potential measurements from individual neurons in the entorhinal cortex. We find that while MEC layer 3 neurons show persistent activity, both MEC and LEC neurons show persistent inactivity.

The advantage of our mean field theory is that its simplicity gives insights into the mechanistic principles behind these decoupling events. Both persistent activity and inactivity arise from the non-linear amplification of afferent network excitation.

Given the ubiquity of connections between the entorhinal cortex and the hippocampus, we further analyze hippocampal activity during these persistent activity and inactivity events. Our results indicate a new way of understand persistent activity, persistent inactivity, and the functional connectivity between large brain networks during spontaneous activity in the absence of sensory stimulus.

Finally, we show the presence of a unique, 3-Hz oscillation in cells of the lateral entorhinal cortex, dentate gyrus, and hippocampal inhibition. This is not present in neocortical and MEC neurons, suggesting that it is distinct from the theta and delta oscillations studied by others. The findings suggest a functional link between inhibition in the hippocampus and LEC circuits.