Role of NMDA conductance in average firing rate shifts caused by external periodic forcing

TL;DR

This paper develops a theoretical framework to understand how NMDA synapses influence the interaction between neural oscillations and firing rates, revealing conditions under which oscillations can stably modulate neural activity.

Contribution

It introduces a mean-field model showing the critical role of NMDA synapses in enabling oscillations to control firing rates without destabilizing neural activity.

Findings

NMDA synapses are essential for stable oscillation-induced firing rate increases.

The effect depends on the balance of fast and slow synaptic weights.

Parameter regions for significant firing rate shifts are identified.

Abstract

A widely accepted view of computations in the brain relies on population coding, where the neural ensemble firing rate is modulated in a stable manner to transmit information and perform various cognitive tasks. At the same time, oscillatory neural activity is specifically modulated in frequency, coherence and power during cognitive performance. How the firing rate and oscillations interact remains a salient question. In this paper, we develop a theory for the interactions between oscillatory signals and the firing rate of neural populations based on activity of non-linear voltage-dependent NMDA synapses. Notably, we show under which conditions oscillatory inputs can control the mean firing rate without loss of stability. Using mathematical analysis and simulations of mean-field models, we demonstrate that presence of NMDA synapses on both the excitatory and the inhibitory neurons is critical for sinusoidal oscillations to significantly and stably increase the firing rate. We characterize the oscillation-induced mean firing rate shift as a function of the fast and slow synaptic weights and demonstrate the parameter region, in which the effect under investigation is mostly pronounced. Results of our work may help identify the properties of neural circuits that allow for constructive control of the firing rate codes by large-scale neural oscillations.