Cortical dopamine reduces the impact of motivational biases governing automated behaviour

Abstract

Motivations shape our behaviour: the promise of reward invigorates, while in the face of punishment, we hold back. Abnormalities of motivational processing are implicated in clinical disorders characterised by excessive habits and loss of top-down control, notably substance and behavioural addictions. Striatal and frontal dopamine have been hypothesised to play complementary roles in the respective generation and control of these motivational biases. However, while dopaminergic interventions have indeed been found to modulate motivational biases, these previous pharmacological studies used regionally non-selective pharmacological agents. Here, we tested the hypothesis that frontal dopamine controls the balance between Pavlovian, bias-driven automated responding and instrumentally learned action values. Specifically, we examined whether selective enhancement of cortical dopamine either (i) enables adaptive suppression of Pavlovian control when biases are maladaptive; or (ii) non-specifically modulates the degree of bias-driven automated responding. Healthy individuals (n = 35) received the catechol-o-methyltransferase (COMT) inhibitor tolcapone in a randomised, double-blind, placebo-controlled cross-over design, and completed a motivational Go NoGo task known to elicit motivational biases. In support of hypothesis (ii), tolcapone globally decreased motivational bias. Specifically, tolcapone improved performance on trials where the bias was unhelpful, but impaired performance in bias-congruent conditions. These results indicate a non-selective role for cortical dopamine in the regulation of motivational processes underpinning top-down control over automated behaviour. The findings have direct relevance to understanding neurobiological mechanisms underpinning addiction and obsessive-compulsive disorders, as well as highlighting a potential trans-diagnostic novel mechanism to address such symptoms.

Fig. 1. Motivational Go-NoGo task design and overview of main task effects.

A Go NoGo task trial sequence for each of the four cue categories: Go-to-Win, Go-to-Avoid, NoGo-to-Win, and NoGo-to-Avoid. Go-to-Win and NoGo-to-Avoid are bias congruent cue categories, as their action requirement is in line with the stimulus-response coupling strengthened by the motivational bias. Go-to-Avoid and NoGo-to-Win are bias-incongruent response-stimulus couplings, which are usually harder to execute for participants. On each trial, a cue was presented for 1300 milliseconds (ms) and subjects could decide to make a Go response by pressing a button or choosing a NoGo response by withholding a response. After this, subjects were presented with the outcome (reward, neutral, punishment) for 750 ms, the valence of which was determined by the cue category and the probabilistic feedback schedule. The inter-trial-interval (ITI) was 2200–3400 ms, in steps of 200 ms. B The feedback contingencies for this task version were 80%: 20%. C Trial-by-trial behaviour. Depiction of the probability of making a Go response, P(Go), (± SEM) and plotted with a sliding window of 5 trials for Go cues (solid lines) and NoGo cues (dashed lines) across trials per cue category, here collapsed across both treatments (tolcapone and placebo). Choice biases are evident from the first trial onwards, as the green lines characterising P(Go) for Win cues are always above the red lines depicting the probability of making a go response for cues requiring a NoGo response as optimal action choice. D Probability of making a Go response for each cue condition, grouped by required action. Learning is evident from the increased proportion of ‘Go’ responses to Go cues. Motivational/ Pavlovian biases is evident from the reduced probability of Go responses to Avoid cues. E Probability of making a correct response (i.e. 1-pGo for NoGo cues), reorganised so that now bias-congruent and bias-incongruent cues are grouped together. Note that this means that the data plotted here are the same as in panel D for Go cues, and the inverse for NoGo cues. This more clearly illustrates the reduced accuracy on bias-incongruent cues, regardless of action requirement. Cue categories abbreviated as follows: G2W = Go to Win, G2A = Go to Avoid Punishment, N2W = NoGo to Win, N2A = NoGo to Avoid Punishment.

Fig. 2. Hypothesised and measured effect of tolcapone administration.

A Illustration of task design to capture motivational biases- through coupling of the orthogonalized axes of motivational valence (Reward, Punishment) and action (motor activation | Go) or (motor inhibition | NoGo). Yellow: valence-action bias-congruent responses is required; White: bias-incongruent responses is required. Predicted change in choice accuracy following tolcapone administration relative to placebo, for each of the 4 conditions. The right 2 panels represent the hypothesised effects of tolcapone. Hypothesis 1: Tolcapone enhances adaptive control, i.e. suppresses Pavlovian bias on incongruent trials, thereby increases the proportion of correct responses (accuracy) on incongruent trials. Speculatively, performance on congruent trials may improve also. Hypothesis 2: Tolcapone promotes a general shift away from automated responding, reducing bias overall. This would lead to improved choice accuracy on incongruent trials (as for Hypothesis 1), but crucially, to reduced choice accuracy for congruent trials (highlighted in yellow). B Data: Mean (±SED) accuracy, i.e. proportion of correct responses, under tolcapone relative to placebo, shown across all trials, for the first half of the trials (block 1) only, and for the 2^nd half of the trials (block 2) only. In line with hypothesis 2, performance on congruent trials is reduced, while performance on incongruent trials is reduced. This is particularly evident for block 1. C Mean probability of making a correct response under placebo versus tolcapone administration. The tolcapone-induced reduction in bias leads to reduced performance for both action-valence congruent cues, but increased performance for incongruent Avoid cues. D Mean probability of making a biased response (Go for Win cues, Nogo for Avoid cues), as a function of block and drug administration. For block 1, tolcapone clearly induces a global reduction in bias, while for block 2, tolcapone appears to have opposite effects depending on cue valence. G2W = Go to Win, G2A = Go to Avoid Punishment, N2W = NoGo to Win, N2A = NoGo to Avoid Punishment.

Fig. 3. Tolcapone induced changes in model parameter estimates.

A Full model comparison showing Model M4 including four parameters, namely feedback sensitivity, a learning rate, a Go bias and a motivational bias parameter to outperform the other three base models. As a small inset, the base model comparison is shown. Here, model M3 outperformed the simpler models M1 and M2. Model frequency and protected exceedance probability were employed as model fit indices. B The π parameter capturing effects of motivational biases was significantly reduced under tolcapone administration. The remaining parameters feedback sensitivity, learning rate, and Go bias were not significantly affected by tolcapone, indicated by p values > 0.1 for all main effects of condition or interaction terms.