Executive control is a vital cognitive function that facilitates the focussing and shifting of attention, planning and working towards a goal, ignoring distractions, and flexibly responding to novel situations. Disruptions to executive control are seen in many psychiatric and neurodegenerative disorders, as well as healthy ageing, which can be profoundly detrimental. Despite having many effective and well-validated methodologies for detecting and quantifying these deficits, there are very few treatments — pharmacological or otherwise — for ameliorating executive dysfunction. This lack of progress can partly be blamed on difficulties associated with identifying drugs that enhance cognition in preclinical research. The work in this thesis aimed to expand our understanding of executive dysfunction — as well as the tasks that measure it — in rats. In results presented in chapter three, middle-aged rats demonstrated impaired reversal learning on the standard attentional set-shifting task, but this was treatable with a novel drug targeting the N-methyl-D-aspartate receptor. The age impairments seen in this experiment were similar to those previously found in young rats with orbital prefrontal cortex (OFC) lesions. The results of chapter four expanded on this similarity to show that, along with reversal deficits, young OFC-lesioned rats are impaired at forming attentional sets when tested on a modified task. In chapter five, another modified set-shifting task revealed that middle-aged rats also suffer from impaired set-formation, but their reversal learning impairments only manifest before attentional set has been formed — not after. Finally, in chapter six, the putative cognitive enhancer modafinil was found to exacerbate middle-aged rats' reversal learning deficit, but it also enhanced their subsequent ability to form attentional set. These experiments reveal that modifying the rat attentional set-shifting task can sometimes make it a more effective tool for testing cognitive enhancers in preclinical settings.

Precisely timed behavior and accurate time perception plays a critical role in our everyday lives, as our wellbeing and even survival can depend on well-timed decisions. Although the temporal structure of the world around us is essential for human decision making, we know surprisingly little about how representation of temporal structure of our everyday environment impacts decision making. How does the representation of temporal structure affect our ability to generate well-timed decisions? Here we address this question by using a well-established dynamic probabilistic learning task. Using computational modeling, we found that human subjects' beliefs about temporal structure are reflected in their choices to either exploit their current knowledge or to explore novel options. The model-based analysis illustrates a large within-group and within-subject heterogeneity. To explain these results, we propose a normative model for how temporal structure is used in decision making, based on the semi-Markov formalism in the active inference framework. We discuss potential key applications of the presented approach to the fields of cognitive phenotyping and computational psychiatry.

Les individus diffèrent les uns des autres dans leur manière générale de se comporter au fil du temps et entre les contextes, ainsi que dans leur habileté à ajuster leurs comportements lorsqu’un changement environnemental survient. Encore aujourd’hui, les causes proximales et ultimes de ces différences ne sont pas bien comprises. Certains défendent que différents compromis de risque liant les types comportementaux rapide/proactif-lent/réactif aux styles cognitifs rapide/inflexible-exact/flexible pourraient en partie expliquer le maintien de ces différences. Toutefois, il semble y avoir des écarts dans la littérature quant à l’existence et la nature d’un tel compromis vitesse-exactitude. Une explication serait que ce genre de relations entre la personnalité et la cognition peuvent être modérées par différents facteurs. Ici, nous explorons la possibilité que l’impulsivité d’action, l’impulsivité de choix et l’incertitude attendue interagissent sur les performances d’apprentissage et de renversement de l’apprentissage chez l’humain. Pour évaluer leur impulsivité, les participants devaient remplir le questionnaire d’impulsivité de Barratt et effectuer une tâche de signal de stop ainsi qu’une tâche de dévaluation temporelle expérientielle. Puis, leurs performances d’apprentissage et de renversement de l’apprentissage ont été mesurées avec une nouvelle tâche de renversement de l’apprentissage, sous différents niveaux d’incertitude attendue. Les résultats démontrent que les patrons de performances d’apprentissage et de renversement de l’apprentissage associés à l’impulsivité d’action dépendent de l’incertitude attendue et de l’impulsivité de choix. Ils appuient également l’idée que l’impulsivité ne serait pas inadaptée per se, ce qui pourrait en partie expliquer le maintien des différences interindividuelles d’impulsivité et de flexibilité. / Individuals differ in how they each generally behave across time and contexts, as well as in their ability to acquire new information, and flexibly adjust their behavioral responses when a change in contingencies occurs. Still today, the proximal and ultimate causes of these differences are not well understood. In recent years, some advocate that divergent risk-reward trade-offs linking fast/proactive-slow/reactive behavioral types to fast/inflexible-accurate/flexible cognitive styles could partly explain the maintenance of these differences. However, it seems that there is a discrepancy in the literature about the existence and nature of such a speed-accuracy trade-off. One explanation could be that the link between personality and cognition is moderated by different factors. Thus, we propose here an exploratory study on how impulsive actions, impulsive choices, and expected uncertainty may interact altogether on learning and reversal learning performances in humans. To assess their impulsivity, participants had to fill out the Barratt impulsiveness scale questionnaire and to complete both a stop signal task and an experiential discounting task. Then, their learning and reversal learning performances were measured in a new reversal learning task, under different levels of reinforcer uncertainty. Results show that learning and reversal learning performances patterns linked to action impulsivity depend on expected uncertainty and choice impulsivity. In addition, they also support the idea that these dimensions of impulsivity are not maladaptive per se, which may provide another line of explanation for the maintenance of variation in impulsivities and flexibility.

This thesis extends the understanding of the neural and neurochemical contributions to two forms of behavioural adaptation, reversal learning and contingency degradation, in which stimulus/action-reward contingencies are altered. The results are interpreted within the psychological framework of the compulsivity construct, and their implications for the pathological behaviour of obsessive-compulsive-disorder (OCD) are considered. The orbitofrontal cortex (OFC) and striatum are key brain structures involved in reversal learning, as are the neurotransmitters serotonin (5-hydroxytryptamine, 5-HT) and dopamine (DA) within those respective regions. However, there has been little empirical evidence of how these two structures and neurochemical systems interact, especially in the functional context of reversal learning. In Chapter Three, the impact of experimentally-induced reductions of 5-HT in the anterior OFC on monoamine levels in subcortical structures such as the striatum and amygdala was determined, DA being found to be significantly up-regulated in the amygdala. Functionally, 5-HT depletion of the OFC has previously been shown to induce deficits in reversal learning. To determine the possible causal significance of amygdala dopamine up-regulation for said reversal learning deficit, the effects of blocking the upregulation with the infusion of intra amygdala DA receptor antagonists following bilateral OFC 5-HT depletion were investigated in a reversal learning paradigm. In Chapter Four, the differential roles of regions of striatum were examined in visual reversal learning. Two recent investigations in non-human primates highlighted the role of the striatum in reversal learning,but pinpointed the critical region to be either the ventromedial caudate or the putamen. Marmosets were trained on a serial reversal task that allowed multiple acute neural manipulations, and the ventromedial caudate and putamen were then reversibly inactivated using the GABAA agonist muscimol. Results indicated dose-related impairments specifically in reversal learning within the putamen, with sparing of discrimination retention. By contrast, similar reversible inactivation of the caudate nucleus produced marked deficits in visual discrimination performance (retention). In Chapter Five, the neural basis of action-outcome contingency knowledge was investigated by inactivating distinct regions of the PFC, the perigenual ACC (pgACC; area 32) and the anterior OFC, and determining response sensitivity to the degradation of action-outcome contingencies. In previous work, excitotoxic lesions of either the pgACC or the OFC had been found to induce insensitivity to contingency degradation in marmosets. However, the design of that experiment did not allow specification of whether stimulus- or action-outcome associations were disrupted, and a precise neural locus could not be determined for the behavioural effects as the OFC lesions included parts of the lateral and medial OFC. I therefore developed a novel contingency degradation paradigm that distinguished between stimulus- and action-outcome associations to enable the study of acute pharmacological manipulations in both brain regions. The pgACC and OFC were reversibly inactivated using GABAA-GABAB agonists (muscimol-baclofen). Whereas the pgACC inactivation produced selective deficits in sensitivity to action-outcome contingency degradation, OFC inactivation reduced the suppressive effect of noncontingent reward on responding more generally but left intact sensitivity to degradation of the contingencies. These results are discussed in terms of different theories of the functions of the pgACC and OFC. In the final discussion the findings on the neural substrates of reversal learning and contingency degradation are drawn together in terms of their significance for theories of PFC involvement in cognitive control, and for the understanding of OCD and other neuropsychiatric disorders.

The purpose of this thesis was to develop methodology by which treatments for the cognitive impairments in Huntington's disease (HD) could be tested. As such, the thesis focused mainly on evaluating rats with quinolinic acid (QA) lesions of the striatum, as this manipulation mimics some aspects of the neural damage in Huntington's disease, to try to identify cognitive deficits of HD resulting from cell loss in the striatum. In the first part (Chapters 3-5), the role of the striatum in implicit memory was investigated. Chapter 3 compared the performance of rats and humans on a reaction time task that evaluated implicit memory by presenting visual stimuli with differing probabilities which change over time. Although rats made higher percentage of incorrect responses and late errors, both groups showed a similar pattern of reaction times. Chapter 4 investigated whether implicit memory (the computation of probabilities to predict the location of a stimulus) was affected by selective blockade of dopaminergic transmission at the D1 or D2 receptors by SCH-23390 and raclopride, respectively. Reaction times were slower with SCH-23390 and raclopride, but only SCH-23390 reduced errors to the least probable target location. Chapter 5 used the same task to evaluate implicit memory in rats with QA lesions of the dorsomedial striatum (DMS). Implicit memory was not affected by lesions of the DMS, which suggested that once a task that requires implicit memory has been learned, the DMS was not involved in sustaining the performance of the task. The second part of this thesis (Chapter 6), explored the contribution of the DMS in habit formation. DMS lesioned rats did not show habitual responding, and were not impaired in learning a new goal-directed behaviour. The third part (Chapters 7 and 8), investigated the role of the dorsal striatum in reversal learning, attentional set-formation, and set-shifting. Dorsal striatum lesioned rats were not impaired in reversal learning, but had a diminished shift-cost, which suggested that dorsal striatum lesions disrupted the formation of attentional sets. These results showed that although QA lesions of the dorsal striatum mimic some aspects of the neural damage in HD, they did not result in the same cognitive deficits observed in patients with HD, at least using the tasks presented in this thesis. However, other animal models of HD could be evaluated using the different tasks presented in this thesis to continue the search of a reliable animal model of HD in which treatments for the disease could be evaluated.