Proefschrift_Holstein

Chapter 8

to the loss of dopamine cells in the midbrain, which project to the striatum. Combined, the DAT1 -dependent effects in chapters 3 and 4 , the increased BOLD response in the striatum in chapter 4 and the age-related deficits in motivated cognitive control ( chapter 5 ) suggest a role for striatal dopamine in mediating motivated cognitive control. However, future work should further substantiate this claim (see future research ). Based on previous work, we anticipated that the ventral part of the striatum would be crucial for motivated cognitive control ( chapter 1 and 2 )(Mogenson et al., 1980; Floresco, 2015). This idea is also in line with previous work, suggesting that the depletion of dopamine neurons in healthy aging is more severe in the midbrain dopamine neurons that project to the ventral part of the striatum compared with those projecting to more dorsal parts of the striatum (Fearnley and Lees, 1991). We tested this hypothesis in chapter 6 by applying excitotoxic lesions to the rodent ventral striatum, i.e. the nucleus accumbens core. To achieve this, we developed a rewarded task-switching paradigm in rodents and we showed that rodents can increase their flexible cognitive control when a situation is associated with a high reward, compared with a low reward situation, like we previously observed in humans. Importantly, only animals with an intact striatum showed this reward-related improvement in cognitive control: After lesions of the ventral striatum animals no longer showed improved flexible behaviour in the high reward condition, suggesting that the ventral striatum is a crucial player when one needs to facilitate behaviour leading to a desired outcome (a high reward), while inhibiting irrelevant behaviour (yielding merely a low reward) (Floresco, 2015). Combined, the observed deficits in aging ( chapter 5 ), ADHD ( chapter 4 ) and rodents with lesions of the striatum ( chapter 6 ) suggest a role for the striatum in motivated cognitive control. One of the hypotheses formulated in chapter 1 suggested that input from the prefrontal cortex can modulate processing in the striatum during motivated cognitive control ( chapter 1 and 2 ). We set out to test the hypothesis that manipulation of the prefrontal cortex can alter processing in the striatum during the integration of reward, cognitive control and subsequent action selection ( figure 2.1 and chapter 1 ). To test this hypothesis (in chapter 7 ) we combined fMRI with offline brain stimulation (transcranial magnetic stimulation; TMS) to target the cortical regions involved in reward processing, task switching (cognition) and response switching (action), thereby using knowledge from previous work showing that stimulation of cortical regions can change processing in the regions of the striatum it is connected to (Strafella et al., 2001; Ko et al., 2008). In line with the existence of functionally specific corticostriatal circuits, we showed that stimulation of the cortical region involved in reward processing (the anterior prefrontal cortex) altered neural signalling in the anterior portion of the caudate nucleus, which has been implicated in reward processing in previous work (Cromwell and Schultz, 2003). Crucially however, we also showed that stimulating the same region in the anterior prefrontal cortex modulated signalling in the motor part of the striatum as a function of the integration between reward motivation, cognitive control and action control. Combined, these results show that the prefrontal cortex and the striatum interact during the integration of the functions necessary to execute the task (i.e. when

166