Manipulations of the context-response relationship reduce the expression of response habits.

Habitual instrumental behaviour is believed to rely on stimulus-response (S-R) associations. However, the method most commonly used to identify habitual behaviour, outcome devaluation, provides only indirect evidence of S-R control. Therefore, it is important to have a better understanding of the S-R association believed to underlie habitual responding. Under free-operant conditions, the context itself likely serves as at least part of the relevant stimuli in the association, and so modifications to the predictive power of the context should alter the expression of habits. The following experiments investigated how changes to the relationship between the training context and performance of the response, either by changing the context during testing or by exposing animals to the context alone, without the response lever present, impacted behavioural control during a devaluation test. We found evidence that the training context is important for the expression of habits; testing animals in a different context than where they were trained resulted in increased goal-directed control (Experiment 1). Furthermore, context alone exposure also increased goal-directed control with animals that received context alone exposure showing stronger devaluation effects, whether the context alone exposure happened on the last day of training (Experiment 2) or throughout training (Experiment 3). These findings are consistent with prior reports that the training context is important for the expression of habits and extends these findings by using sensory-specific satiety as a means for devaluation and by using context alone exposure to alter behavioural control.

Ghrelin receptor antagonism and satiety attenuate Pavlovian-instrumental transfer.

Animals rely on learned cues to guide their behaviour for rewards such as food. The Pavlovian-instrumental transfer (PIT) task can be used to investigate the influence of Pavlovian stimuli on instrumental responding. Ghrelin, an orexigenic peptide, and its receptor, growth hormone secretagogue receptor 1A (GHS-R1A), has received growing interest for its role in reward-motivated learning and behaviours. A significant population of GHS-R1A have been identified within the ventral tegmental area (VTA), a critical node in the mesolimbic reward circuit that is necessary for the expression of PIT. As ghrelin has been found to increase dopaminergic activity in the VTA, we predicted that GHS-R1A antagonism with JMV-2959 would attenuate PIT. Further, given the relationship between hunger levels and changes in ghrelin signalling, we sought to compare the effects GHS-R1A antagonism with those of satiety, hypothesizing parallel effects, with each attenuating PIT. Rats received daily sessions of Pavlovian and then instrumental training over 3 weeks. Across three experiments, we examined the effects of a shift to satiety, or treatment with the GHS-R1A antagonist JMV-2959, either peripherally or directly into the VTA. We found that presentations of a stimulus paired with food reward enhanced responding for food across all conditions, thus demonstrating the expected PIT effect. Further, GHS-R1A antagonism, both peripherally and within the VTA, as well as satiety significantly reduced the magnitude of the PIT effect compared to control conditions. These results clarify our understanding of ghrelin signalling in PIT and begin to elucidate the role of feeding-related peptides in the modulation of reward-related responding.