Proefschrift_Holstein

Chapter 2

no effect in the group of subjects with already high dopamine levels. Furthermore, a number of studies have shown that the effects of a dopamine D2 receptor agonist can be explained by natural variation in the gene coding for the dopamine D2 receptor (Kirsch et al., 2006; Cohen et al., 2007). Subjects carrying the Taq 1A1 variant of the allele (A1+ subjects) have ~30% fewer dopamine D2 receptors in the striatum and exhibit impairments in reward processing, compared with those not carrying this allele (A1- subjects). Cohen and colleagues (2007) assessed whether this genetic predisposition could explain individual responses to a dopamine receptor agonist during reward processing. To this end they used a reversal learning paradigm, which requires flexible adaptation of behaviour when a previously rewarded stimulus is no longer rewarded and a newrule needs tobe learned. In the placebo condition, the lowdopamine receptor group (A1+ subjects) performed worse than the A1- group. However, administration of the dopamine D2 receptor agonist cabergoline improved rule-learning performance in the subset of subjects with genetically determined low dopamine receptor density (A1+), but the dopamine D2 receptor agonist impaired performance in those already performing well under placebo (the A1- group). This effect was accompanied by opposite effects in reward- related neural responses in regions of the reward network (the medial orbitofrontal cortex and striatum): Administration of the D2 receptor agonist increased activity in these regions in subjects with low reward-related activity under placebo (A1+), while it had the opposite effect in those with already high baseline reward-related activity (A1-). As was discussed in chapter 1 , in addition to predicting individual differences indrug response, natural (genetic) variation between individuals can also explain individual differences in task performance (e.g. (Frank et al., 2007; Dreher et al., 2009; Aarts et al., 2010; Colzato et al., 2010a; Stelzel et al., 2010) ( box 2.2c ). In chapters 3 and 4 , I exploited individual variability in the gene coding for the dopamine transporter (DAT) to account for inter-individual variability in task performance, neural signalling, and drug response. Task-related differences in performance or neural signalling as a function of variation in this genotype can be taken to suggest that dopamine is involved in the studied process. Evidence for a role for dopamine in the integration of reward motivation and cognitive control has been provided by a number of studies, and much of this evidence is reviewed in chapter 1 of this thesis. Previous work that also employed the rewarded task-switching paradigm presented in this thesis ( box 2.3 ) has shown that reward can modulate flexible control in the context of task switching (Aarts et al., 2010). The anticipation of a reward (i.e. high vs. low reward cue) increased neural responses in the ventral striatum, while the integration between reward and task switching was associated with increased signalling in the caudate nucleus. Interestingly, these signals correlated, suggesting that communication between the ventral and dorsal striatum may mediate the information transfer from reward regions to cognitive control regions ( figures 1.2c, d and 2.1, box 2.1 ). Dopamine-dependent effects in this latter study were revealed by showing that inter-individual differences in signalling in the caudate nucleus depended crucially on individual differences in dopamine signalling, measured by exploiting differences in the DAT1 genotype. Although these results provide

30