Striatal function scrutinized through the PAN-TAN-FSI triumvirate

Abstract

Understanding the information encoded by distinct components of the neuronal circuitry in the striatum represents an avenue for elucidating the role of this subcortical region in adaptive behavior and its dysfunction in pathological conditions. In behaving animals, conventional single neuron recordings generally differentiated between three main electrophysiologically identified neuron subtypes in the striatum, referred to as phasically active neurons (PANs), tonically active neurons (TANs), and fast-spiking interneurons (FSIs), assumed to correspond to GABAergic spiny projection neurons, cholinergic interneurons, and parvalbumin-containing GABAergic interneurons, respectively. Considerable research has been devoted to exploring the behavior-related activities of neurons classified electrophysiologically into PANs, TANs, and FSIs in animals engaged in task performance, mostly monkeys. Although precise neuron identification remains a major challenge, such electrophysiological studies have provided insights into the functional properties of presumed distinct striatal neuronal populations. In this review, we will focus on current ideas about the functions subserved by these neuron subtypes, emphasizing their link to specific aspects of behaviors. We will also underline the issues that are yet to be resolved regarding the classification of striatal neurons into distinct subgroups which emphasize the importance of considering the potential overlap among electrophysiological characteristics and the molecular diversity of neuron types in the striatum.

Keywords: basal ganglia; behavior; monkey; neuron types; single-unit.

Figure 1

Examples of behavior-related changes in activity of the three main categories of neurons recorded in the monkey striatum. Neuronal activity is represented as raster plots (bottom) and perievent time histograms (top) during the performance of a visuomotor task in which the monkey reacted to a visual signal (red marker) by releasing a resting bar (green marker) and reaching a target (blue marker) to obtain a liquid reward. Each dot corresponds to the time of a neuronal impulse or spike and each line of dots to one trial. Raster plots are aligned on the onset of the stimulus serving as a trigger for arm-reaching movements and trials were ordered off-line according to the latency of movement. Vertical calibration is in spikes/s for all histograms. Baseline firing rates (i.e., activity during the period immediately preceding stimulus onset) vary among the three neuronal populations. PANs are usually silent or have a baseline firing rate < 1 spike/s, and display transient or sustained increases in discharge rate occurring in distinctive forms during different periods of the task, reflecting different processes that participate in movement generation. TANs fire more regularly than PANs, with firing frequencies ranging from 3 to 12 spikes/s, and exhibit quite homogeneous task-related changes in activity consisting mostly of a short lasting pause in firing in response to the visual signal. FSIs are characterized by a high firing variability with complex task-related modulations that combined increases and decreases in firing rate. Unlike TANs, FSI activity did not exhibit coordinated changes at specific moments within the context of the task used here.