Irregular persistent activity induced by synaptic excitatory feedback.

Area of Science:

• Computational neuroscience
• Systems neuroscience
• Neural dynamics

Background:

• Neurophysiological studies in monkeys reveal irregular persistent activity in prefrontal neurons during oculomotor delayed-response tasks.
• During the delay period, the coefficient of variation (CV) of interspike intervals (ISI) exceeds 1, indicating higher variability than during fixation.

Purpose of the Study:

• To replicate the observed irregular persistent activity in a computational network model.
• To identify the necessary conditions within neural networks that generate this specific firing pattern.

Main Methods:

• Simulations of a homogeneous network of excitatory leaky integrate-and-fire neurons with excitatory feedback.
• Incorporation of a post-spike reset close to threshold and non-linear synaptic efficacies via short-term depression (STD).
• Validation in a more complex network featuring selective excitatory neurons and inhibition.

Main Results:

• The model successfully reproduces the high coefficient of variation (CV) of interspike intervals (ISI) characteristic of experimental data.
• Persistent activity with irregular firing statistics is achieved within a broad range of synaptic efficacy values.
• The findings hold true for both simplified and more complex network architectures.

Conclusions:

• Excitatory feedback, combined with specific neuronal reset properties and rate-dependent synaptic depression, can explain irregular persistent activity in prefrontal networks.
• The study provides a mechanistic explanation for observed neural variability during cognitive tasks.
• This computational approach offers insights into the neural basis of working memory and decision-making processes.