Sensory cortical dynamics

Abstract

Sensory cortical networks are commonly regarded as stable, changing only in the face of prolonged alteration of sensory input. There is increasing evidence, however, that the functional connectivity of cortical networks changes significantly, but reversibly, in response to conditions of sensory stimulation similar to those encountered in everyday life. In this review, we provide examples of sensory cortical dynamics at the single neuron and neural population levels. The dynamics detected at both levels of experimental observation suggest that a brief exposure (tens of milliseconds to tens of seconds) to sensory stimulation is accompanied by changes in the capacity of cortical networks to process and represent environmental stimuli. Candidate cellular mechanisms and the potential benefits of such stimulus-driven, rapid, and fully reversible sensory cortical dynamics are discussed.

Introduction

Although understanding of the neural basis of perception is far from complete, it is clear that the computations that underlie perception are carried out quickly, a necessity if the challenges posed by an ever-changing environment are to be met successfully. For instance, Thorpe et al. [47] demonstrated that humans can make sensory decisions—determining whether or not a complex scene contains a particular element—in less than 200 ms. Given the rapidity of this decision, it seems likely that cells involved in processing the information needed for this task rely primarily on rapid synaptic neurotransmission (e.g. AMPA and NMDA receptor mediated excitatory neurotransmission, and GABA_A receptor mediated inhibitory transmission). Traditionally, fast neural computation is viewed as occurring in essentially stable or ‘hard-wired’ networks; that is, networks whose processing characteristics are independent of their recent history of activation. On a much longer time scale (e.g. days or years), it has been established that cortical networks can undergo substantial changes in functional connectivity in response to prolonged alterations in the pattern of sensory drive (i.e. experience-based cortical plasticity, [5]). These long-term changes in functional connectivity are believed to involve changes in synaptic efficacy and/or axonal sprouting.

These two views of the information processing characteristics of primary sensory cortex—a fixed circuit capable of rapid processing, and a plastic network capable of long-term change—are conventionally regarded as distinct features of cortical networks. The two extremes, however, may not be as dichotomous as is commonly believed. Rather, we propose that they form opposing ends of a continuum of functional states that can be expressed by primary sensory cortex, each state having information processing capacities maximally consistent with the properties of the stimuli the network processed in a preceding temporal interval. Central to this alternative view is the idea that a stimulus applied either continuously or intermittently at a sufficient rate evokes an orderly temporal continuum of functional states in primary sensory cortex. We refer to the hypothesized temporal progression of functional network states associated with temporally extended sensory stimulation as ‘cortical dynamics’. The available experimental evidence leads us to believe that natural sensory stimulation evokes cortical dynamics which (1) occur over a time scale of tens of milliseconds to many minutes; (2) are associated with an orderly temporal progression of changes in the receptive field properties of primary sensory cortical neurons; and (3) involve a repertoire of linked cellular mechanisms, each requiring unique conditions for its activation, and each exhibiting a distinct time course of operation.

In this brief review, we present evidence obtained using both single neuron and neuronal population recording methods which supports our concept of sensory cortical dynamics. We also introduce a substantial literature describing cellular mechanisms with attributes compatible with short-term, reversible, use-dependent changes in sensory cortical networks. We conclude with a discussion of the potential functional benefits of cortical network dynamics.