Synapse-to-network plasticity in the hippocampus

Resumen

The hippocampal formation has been known for decades for its central role in the formation of new episodic memories. Although synaptic plasticity is a well accepted and dominant mechanism underlying the process of learning, and current theories include a communication between hippocampus and cortical structures during such process, the truth is that not much is known regarding the gap between the synapse and the network; i.e. how local plasticity at the level of the neurons can affect a whole network nearby and at distant regions of the brain as well. It has been previously demonstrated that indeed this occurs with Long-Term Potentiation (LTP) of the perforant path, but the how still remains to be clarified. The aim of this work was to study in depth the specific phenomena that turn synaptic plasticity into a network emergent property, and which could be its role in learning and memory. We studied the hippocampus of rats in vivo using electrophysiological techniques and pharmacological manipulations, as well as functional magnetic resonance imaging (fMRI). We recorded the extracellular Local-Field Potentials (LFP), and since these signals are a summation of several different neuronal sources or ¿generators¿, we analyzed them using independent component analysis as a method of source separation; we then studied separately the spontaneous and the evoked activity of the different hippocampal LFP-generators during the induction of LTP in the perforant path. Synaptic potentiation produced a decrease of the feed-forward inhibition on granule cells in the dentate gyrus and an increase in their excitatory input, and thus a net change in the local excitation-inhibition balance. Moreover, de functional coupling between the local inhibitory and excitatory network decreased, and in turn there was an enhanced coordination of the entorhinal input with the activity at CA1 area. This evidences a rearrangement of the hippocampal network starting from the dentate gyrus, and therefore suggests that feed-forward inhibition in this area is acting as a gating mechanism for information propagation. Moreover, we could demonstrate that the reach of these functional changes is larger than the hippocampal network and it affects also distant regions of the brain; we observed with fMRI that a pharmacological blockade of GABAergic activity in the dentate gyrus provoked an increased activation of new cortical and subcortical areas. We reproduced the electrophysiological preparation to perform similar experiments in awake behaving rats, substituting LTP for a physiological behavioral task, in which animals were confronted with a novelty in a familiar environment. The results revealed equivalent changes to those found with anesthetized animals in the hippocampal network, further supporting our theory that this functional reorganization may indeed subserve learning and memory, regarded as a routing mechanism to preferentially process novel and relevant environmental information.