Learning alters theta-nested gamma oscillations in inferotemporal cortex

Keith Kendrick,Hanno Fischer,Xuejuan Zhang,Alister Nicol,Yang Zhan,Jianfeng Feng

doi:10.1038/npre.2009.3151.2

Keith Kendrick, Hanno Fischer + Show 4 more

Open Access

https://doi.org/10.1038/npre.2009.3151.2

Copy DOI

Abstract

AbstractHow coupled brain rhythms influence cortical information processing to support learning is unresolved. Local field potential and neuronal activity recordings from 64- electrode arrays in sheep inferotemporal cortex showed that visual discrimination learning increased the amplitude of theta oscillations during stimulus presentation. Coupling between theta and gamma oscillations, the theta/gamma ratio and the regularity of theta phase were also increased, but not neuronal firing rates. A neural network model with fast and slow inhibitory interneurons was developed which generated theta nested gamma. By increasing N-methyl-D-aspartate receptor sensitivity similar learning-evoked changes could be produced. The model revealed that altered theta nested gamma could potentiate downstream neuron responses by temporal desynchronization of excitatory neuron output independent of changes in overall firing frequency. This learning-associated desynchronization was also exhibited by inferotemporal cortex neurons. Changes in theta nested gamma may therefore facilitate learning-associated potentiation by temporal modulation of neuronal firing.

Highlights

The functions of both low and high frequency oscillations in the brain have been the subject of considerable speculation[1]
Since our data confirmed the presence of cross-frequency coupling between theta and gamma oscillations, similar to that reported in human EEG studies[7], we investigated whether the correlation between theta phase and gamma amplitude was a consequence of theta nested-gamma activity
Overall our results have demonstrated for the first time that theta nested gamma in the inferotemporal cortex (IT) is both influenced by learning and may serve an important function in the amplification and discriminability of inputs converging onto downstream neurons through a temporal desynchronization effect

Summary

Introduction

The functions of both low and high frequency oscillations in the brain have been the subject of considerable speculation[1]. Some human EEG recording studies have reported that theta phase rather than amplitude is correlated with cognitive processes, the so-called phase reset model[1,11], while others in the frontal and temporal lobes have placed more importance on the correlation between theta amplitude and gamma frequency[7]. The magnitude of both theta and gamma oscillations during encoding appears to predict the efficacy of subsequent recall[12] and the theta rhythm can both modulate gamma amplitude[13] and the firing of single neurons[2]. We developed a neural network model reproducing our electrophysiological findings to infer the functional consequences of observed learning associated changes

Methods

Results

Conclusion