The amygdala and flavour preference conditioning: Crossed lesions and inactivation Dominic M. Dwyer ⁎, Mihaela D. Iordanova School of Psychology, Cardiff University, Tower Building, Park Place, Cardiff, CF10 3AT, UK abstract article info Article history: Received 27 January 2010 Received in revised form 24 June 2010 Accepted 14 July 2010 Keywords: Taste Learning Nucleus accumbens Motivation GABA Current studies examined whether temporary inactivation of the amygdala influenced the learning and/or expression of conditioned flavour preferences and whether interactions between the amygdala and the nucleus accumbens contribute to this learning. Experiments 1A and 1B examined temporary inactivation of the amygdala in rats, by the administration of muscimol through chronically implanted cannulae, given during acquisition and/or expression of flavour preferences based on a sucrose reinforcer. Despite differences in the number of training trials and control procedures, in both of Experiments 1A and 1B inactivation during training attenuated, but did not totally prevent, the acquisition of a preference for the CS+ (conditioned stimulus) flavour over the CS-. Inactivation during testing had no effect on the preference for the CS+. In Experiment 2A rats were given access to a CS+ flavour paired with fructose and a CS- flavour without fructose prior to testing the preference for the CS+ over the CS- in the absence of the reinforcer. In Experiment 2B the same rats were tested for their preference with another set of CS+ and CS- flavours and maltodextrin as the reinforcing solution. Contralateral unilateral lesions of the amygdala and nucleus accumbens attenuated, but did not totally prevent, flavour preference learning based on either fructose or maltodextrin compared to either ipsilateral or sham lesioned animals. These results suggest that the amygdala plays a role in the learning, but not expression, of flavour preferences and that this role is partially dependent on interactions with the nucleus accumbens. © 2010 Elsevier Inc. All rights reserved. 1. Introduction Omnivorous animals demonstrate unlearnt positive reactions to very few tastes and so learning plays a major role in food selection and preferences. Animals learn to prefer conditioned stimulus (CS) flavours that are associated with positive consequences (uncondi- tioned stimuli: US). Flavour preferences can be acquired when the US is a nutrient such as starch [e.g. 1] or a particularly palatable taste (i.e. another flavour) such as saccharin [e.g. 2] or fructose [e.g. 3]. Recent investigations have implicated the amygdala and the nucleus accumbens in both flavour–nutrient and flavour–flavour learning: Following lesion studies [4–6], more recent studies demonstrate that antagonising dopaminergic transmission in the amygdala or nucleus accumbens interfered with the acquisition but not expression of flavour-nutrient learning [7,8]. In the case of flavour–flavour learning disrupting dopaminergic transmission in either the amygdala or accumbens affected expression to a greater degree than acquisition [9, but see discussion, 10]. Recent theoretical accounts of the role of the amygdala in general appetitive learning have suggested that it is critical for the formation of associations between CSs and both the specific sensory properties and the general affective valence of USs [e.g. 11,12]. Moreover, lesions of the basolateral amygdala following Pavlovian conditioning disrupt the effects of US devaluation prior to test [13] suggesting that the amygdala is critically involved in the expression of CS–US associations as well as their formation. This result appears to contrast with demonstrations that antagonising dopaminergic processes in the amygdala has no effect on the expression of flavour–nutrient preferences and leaves flavour–flavour preferences at least partially intact [7,10]. This apparent dissociation may reflect the contribution of non-dopaminergic processes in the amygdala to the expression of learnt associations or it might reflect differential contributions of the amygdala to flavour preference learning and appetitive Pavlovian conditioning. In addition, these previous studies have examined the impact of manipulating amygdala function on acquisition and expression in separate studies, which leaves the possibility of an interaction between the processes underpinning acquisition and expression unexplored. Therefore, our first aim was to address the relationship between amygdala involvement in flavour preference conditioning and its role in appetitive conditioning more generally by using muscimol to functionally inactivate the amygdala. In order to make a factorial investigation of the effects of muscimol inactivation on the learning and expression of acquired flavour preferences we needed to make the total length of the protocol as short as possible. Thus orally consumed sucrose was used as the reinforcer as pilot studies had demonstrated that strong preferences could be acquired with brief training. The designs of Experiments 1A and 1B are shown in Table 1. Physiology & Behavior 101 (2010) 403–412 ⁎ Corresponding author. Tel.: + 29 20876285; fax: + 29 20874858. E-mail address: DwyerDM@cardiff.ac.uk (D.M. Dwyer). 0031-9384/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2010.07.004 Contents lists available at ScienceDirect Physiology & Behavior journal homepage: www.elsevier.com/locate/phb