The role of the lateral prefrontal cortex in inhibitory motor control

Abstract

Research on inhibitory motor control has implicated several prefrontal as well as subcortical and parietal regions in response inhibition. Whether prefrontal regions are critical for inhibition, attention or task-set representation is still under debate. We investigated the influence of the lateral prefrontal cortex (PFC) in a response inhibition task by using cognitive electrophysiology in prefrontal lesion patients. Patients and age- and education-matched controls performed in a visual Stop-signal task featuring lateralized stimuli, designed to challenge either the intact or lesioned hemisphere. Participants also underwent a purely behavioral Go/Nogo task, which included a manipulation of inhibition difficulty (blocks with 50 vs. 80% go-trials) and a Change-signal task that required switching to an alternative response. Patients and controls did not differ in their inhibitory speed (stop-signal and change-signal reaction time, SSRT and CSRT), but patients made more errors in the Go/Nogo task and showed more variable performance. The behavioral data stress the role of the PFC in maintaining inhibitory control but not in actual inhibition. These results support a dissociation between action cancellation and PFC-dependent action restraint. Laplacian transformed event-related potentials (ERPs) revealed reduced parietal activity in PFC patients in response to the stop-signals, and increased frontal activity over the intact hemisphere. This electrophysiological finding supports altered PFC-dependent visual processing of the stop-signal in parietal areas and compensatory activity in the intact frontal cortex. No group differences were found in the mu and beta decrease as measures of response preparation and inhibition at electrodes over sensorimotor cortex. Taken together, the data provide evidence for a central role of the lateral PFC in attentional control in the context of response inhibition.

Figure 1. Paradigms and patients

A Stop-Signal task featuring lateral stimuli as go- (white bar) and stop-signals (red bar). B Go/Nogo task with green square as go-signal and red square as nogo-signal. C Change-signal task with arrows indicating a right or left button press. D Overlay of lesion reconstructions of the PFC patients. The color coding indicates the number of patients with damaged tissue in that area. Note that right-sided lesions are flipped.

Figure 2. Behavioral results

A Results of the Stop-signal task showing an enhanced reaction time variability in the go task (intraindividual coefficient of variance, ICV; left side), but no significant differences between patients (PFC) and controls (CTR) in the stop-signal reaction time (SSRT; right side). B SSRT (left: contralesional stop-stimuli; right panel: ipsilesional stop-stimuli) of every patient relative to their lesion extent of the inferior frontal gyrus (BA 44/45; determined based on probabilistic Jülich brain atlas; (Eickhoff et al., 2005)), separately for right (triangles) and left (dots) lesion patients showing no correlation between right or left IFG damage and SSRT. Dashed lines depict the average controls’ SSRT and +/− one standard deviance. C Rate of commission errors in the Go/Nogo task in 50 and 80 % go blocks. Participants made more errors in the more demanding conditions and patients made overall more errors. D No group difference was found in the inhibitory speed (change-signal reaction time, CSRT) in the Change-signal task.

Figure 3. Event-related potentials in go-trials

A PFC patients had a reduced N1 to contralesional go-stimuli over the lesioned hemisphere (left side), but no significant differences were detected after ipsilesional go-stimuli (right side). The analyzed time-window between 160 – 190 ms is indicated with a grey bar. B shows the topographic map of the average amplitude between 160 – 190 ms after the contralesional (left side) and ipsilesional stimuli (right side) for controls (CTR) and patients (PFC). In all figures, the circle in the patients’ map indicates the lesion site.

Figure 4. Event-related potentials in stop-trials

A Depicted are the ERPs after the contralesional stop-stimuli in successfully inhibited trials. Patients (PFC) had a reduced negativity in parieto-occipital sites over the lesioned hemisphere and enhanced activity in a frontolateral cluster over the intact hemisphere. B Average amplitude in the frontolateral cluster over the intact hemisphere after contra- (CL) and ipsilesional stop-stimuli in patients and controls. Patients and controls differed significantly in ERPs to the contralesional stimuli. C Correlation between frontolateral activity (z-normalized) after contralesional stop-stimuli and the performance variability (intraindividual coefficient of variance, ICV) in PFC patients.

Figure 5. Fronto-parietal theta phase-coupling

A For controls, the percentage change of the phase-locking value (PLV) relative to the pre-stimulus baseline in go-trials is shown, separately for ipsi- (solid line) and contralateral (dashed) stimuli. Note that ipsi- and contralateral refers to relative to the response hand (corresponding to the left side here). On the left, intra-hemispheric phase-locking changes in the left hemisphere are plotted, on the right the PLV for the right hemisphere. B Depicted is the percentage PLV change for PFC patients, shown separately for ipsi- (solid) and contralateral (dashed) stimuli and for intra-hemispheric coupling ipsi- and contralateral to the response, i.e. the patients’ lesioned and intact hemisphere.

Figure 6. Power changes in mu and beta band in stop-trials

Percentage power change relative to the baseline before the go-stimulus (at time 0; average delay of stop-signal: around 350 ms) in patients (red) and controls (black) in inhibited trials (SI; solid line) and stop-errors (SE; dashed line). Shown is the time-course at C4 (contralateral to the response) for trials after ipsilesional stimuli. The grey bar indicates the analyzed time-window between 400 – 700 ms after the go-signal. A Results for the mu band (10 – 13 Hz) with a stronger decrease in errors compared to inhibited trials and contralateral to the response more than over ipsilateral sites. The maps on the right side show the average power change between 400 – 700 ms in error trials for controls and patients. B Results for the beta band (15 – 22 Hz) with a stronger decrease in errors compared to inhibitions and contralateral to the response more than ipsilateral (maps of error trials on the right for patients and controls).