Brain Research Bulletin 68 (2006) 233–248 Learning-dependent activation of Fra-1: Involvement of ventral hippocampus and SNc/VTA complex in learning and habit formation Alexis Faure a,∗ , Franc ¸oise Conde b , Fabrice Cheruel a , Nicole el Massioui a a Laboratoire de Neurobiologie de l’Apprentissage, de la M´ emoire et de la Communication, CNRS UMR 8620, Bat. 446, Universit´ e Paris Sud, 91405 Orsay, France b Unit´ e de Recherche Associ´ ee 2210, Commissariat ` a l’Energie Atomique, Centre National de la Recherche Scientifique, Service Hospitalier Fr´ ed´ eric Joliot, 91401 Orsay, France Received 29 November 2004; received in revised form 20 July 2005; accepted 16 August 2005 Available online 19 September 2005 Abstract Although the effect of overtraining on learning processes in rats has long been studied, only few studies have specifically assessed the differential involvement of brain areas in habit formation. We used the analysis of expression of the immediate early gene Fra-1 as a tool to differentiate the areas involved in training and overtraining. Behavioural experiments showed that instrumental performance (signalled and non-signalled instrumental tasks), but not pavlovian conditioned responses, were no longer under the control of the incentive value of the reward after overtraining. The number of Fra-1 expressing neurons was increased in SNc/VTA and ventral hippocampus after training in all groups independently of behavioural performance. After overtraining, the number of learning-induced Fra-1 immunoreactive neurons remained increased in the SNc/VTA. However, in CA1, it significantly decreased in the signalled instrumental group, whereas it further increased in the pavlovian group, with no modulation in non-signalled instrumental animals. The increase in the number of Fra-1 neurons observed after training in SNc/VTA and ventral hippocampus suggests that a general underlying incentive process regulates Fra-1. Moreover, the sustained increased expression of Fra-1 in the SNc/VTA after instrumental overtraining could reflect a possible role of dopaminergic neurons in habit formation. © 2005 Elsevier Inc. All rights reserved. Keywords: Fra-1 immediate-early gene; Ventral hippocampus; Dopaminergic neurons; Habit learning; Devaluation of the reinforcer; Rat 1. Introduction Two main forms of memory are usually distinguished by the associative representations upon which they rely: a declarative form which is flexibly accessible and results from the association between stimuli, reinforcer and behavioural actions and a proce- dural form based on automatic and inflexible stimulus–response (S–R) habits [21,76,82]. The transition from declarative to pro- cedural representations is thought to be mediated by extended training [1,16]. For example, it is well established that overtrain- ing rats to press a lever for a reward renders the performance autonomous of the goal of the action (lever-press) demonstrated by its insensitivity to post-training changes in the value of the reward [1,2]. ∗ Corresponding author. Tel.: +33 1 69 15 75 86; fax: +33 1 69 15 77 26. E-mail address: alexis.faure@ibaic.u-psud.fr (A. Faure). In the framework of learning theories, the limbic system is known to play a key role in processing the incentive value of conditioned stimuli and in controlling goal-directed behaviour. More specifically, the basolateral nucleus of amygdala, the pre- limbic cortex and the dorsal hippocampus seem to be partic- ularly involved in acquisition of action–outcome associations [5,6,11,17,18,20,22,61]. Similarly, the nucleus accumbens and particularly its core part, seems to be an interface between moti- vation and action, allowing the incentive value of reinforcement and stimuli to control instrumental performance [19]. The striatum and more broadly, the basal ganglia are anatom- ically and functionally linked to the limbic system. It has been recently proposed that specific parts of these systems could be selectively involved in training or overtraining [55,63]. More- over, the development of an S–R habit, assessed by a post- overtraining outcome devaluation procedure, has been shown to be dependent upon the integrity of the lateral part of dorsal stria- tum [86], but also upon the integrity of the lateral nigro-striatal 0361-9230/$ – see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.brainresbull.2005.08.017