Levodopa administration modulates striatal processing of punishment-associated items in healthy participants

Abstract

Rationale: Appetitive and aversive processes share a number of features such as their relevance for action and learning. On a neural level, reward and its predictors are associated with increased firing of dopaminergic neurons, whereas punishment processing has been linked to the serotonergic system and to decreases in dopamine transmission. Recent data indicate, however, that the dopaminergic system also responds to aversive stimuli and associated actions.

Objectives: In this pharmacological functional magnetic resonance imaging study, we investigated the contribution of the dopaminergic system to reward and punishment processing in humans.

Methods: Two groups of participants received either placebo or the dopamine precursor levodopa and were scanned during alternating reward and punishment anticipation blocks.

Results: Levodopa administration increased striatal activations for cues presented in punishment blocks. In an interaction with individual personality scores, levodopa also enhanced striatal activation for punishment-predictive compared with neutral cues in participants scoring higher on the novelty-seeking dimension.

Conclusions: These data support recent indications that dopamine contributes to punishment processing and suggest that the novelty-seeking trait is a measure of susceptibility to drug effects on motivation. These findings are also consistent with the possibility of an inverted U-shaped response function of dopamine in the striatum, suggesting an optimal level of dopamine release for motivational processing.

Fig. 1. Experimental design

Trial sequence for the study phase, shown exemplarily for a rewarded trial from a reward block. A cue picture was presented indicating whether participants could win money on that trial. Participants made a category decision on the picture, waited for the following number task, and then indicated quickly whether the number was higher or lower than five. In rewarded trials, they received win feedback (green upward arrow) after correct decisions made within a time limit and no-win feedback (yellow downward arrow) in incorrect trials. In neutral trials, they received uninformative feedback (black horizontal arrow). In punishment blocks, the cue category predicted punishment (red downward arrow) or neutral (yellow upward arrow) outcomes.

Fig. 2. Anticipation of reward and punishment across both groups

Higher activations (p < 0.05, FWE whole-brain corrected) to anticipation of rewards and punishments compared to neutral cues in (A) bilateral ventral striatum (MNI peak coordinates: left VS -12,10,-6; right VS 10,10,-4) and anterior cingulate (MNI peak coordinates 8,16,44), and (B) right SN/VTA (MNI peak coordinates 10,-14,-12). Colour bars indicate t values. To better localize SN/VTA activations, the two corresponding panels display an overlay onto an MT image (cf. methods section).

Fig. 3. Drug effects across all participants

Cue processing. In the levodopa group compared to placebo, A) both types of cues elicited higher striatal activation (p < 0.05, SVC) in punishment compared to reward blocks (MNI peak coordinates 22,10,-8). This effect is composed of B) higher striatal activation to anticipation of punishment compared to reward (MNI peak coordinates 26, 14, 0) and C) higher striatal activation to neutral cues in punishment compared to reward blocks (MNI peak coordinates 22, 10, -10). Clusters are shown at p<0.005, uncorrected, k>5 voxels. Colour bars indicate t values. Panels on the right of each activation map illustrate group differences in mean parameter estimates for peak voxels from the corresponding regions on the left.

Fig. 4. Drug interaction with personality

After levodopa administration compared to placebo, striatal activation (p < 0.05, SVC) correlated with the novelty-seeking score for punishment compared to neutral cues (MNI peak coordinates 14,14,2). Clusters are shown at p<0.005, uncorrected, k>5 voxels. Colour bar indicates t values. Panel on the right presents mean (+SEM) parameter estimates separately for participants with low (black) and high (grey) novelty-seeking scores grouped by median split.