Hippocampus, Temporal Context and Taste Memories

Abstract

Human research points to the relevance of the hippocampus for episodic memory, i.e. the ability to remember unique events that took place at a particular time and place (Tulving, 2002). The study of the hippocampal involvement in animal models of episodic memory faces several pitfalls (Eichenbaum and Fortin, 2003). As an alternative animal model, episodic-like taste memories may be studied using taste recognition memory tasks (Bermudez-Rattoni, 2004). As it is required for an animal model of episodic memory, taste memories are bound to the context in which learning occurred. This has been shown both for safe and aversive taste memories. On one hand, the contextual dependency of latent inhibition (LI), i.e. the fact that previous taste exposure without consequences interferes with later learning (Lubow, 1989), indicates that the safe taste memory includes contextual cues. A context change between preexposure and conditioning disrupts LI (Hall and Channell, 1986). On the other hand, retrieval of an aversive taste memory may be context-dependent (Bonardi et al., 1990; Loy et al., 1993), mainly when the taste was previously exposed without consequences (Puente et al., 1988; Boakes et al., 1997). This conditional discrimination task shows that context cues may facilitate retrieval either of the safe or the aversive taste memory trace. There is evidence using other learning tasks that supports a hippocampal involvement both in the contextual dependency of LI and in conditional learning (Honey and Good, 1993; Holland and Bouton, 1999; Maren and Holt, 2000). In these studies context frequently refers to a particular place. A good model of episodic memory, however, should include not only the external but also the temporal context (Bouton, 1993). We reported that the time of day may act as a context which modulates either LI (Manrique et al., 2004) or conditioned taste aversions retrieval (Moron et al., 2002) in Wistar rats. Both phenomena have been tested using the same behavioral procedure that included two non-reinforced saline preexposures before conditioning and testing (Table 1). Animals were assigned to one of two conditions, animals in the ‘Same’ condition were pre-exposed, aversively conditioned to saline ( i.p. injections of lithium chloride after saline intake) and tested (one-bottle test) in their home cages during the evening drinking session. Animals in the ‘Different’ condition were aversively conditioned to saline during the morning drinking session while preexposure and testing took place in the evening session. Subtle modifications of the procedure (Table 1) seem to drastically modify the role of the temporal context. By increasing the duration of the previous habituation training to drink water twice a day and also allowing the animals to drink freely during the conditioning session, the ability of a change of the time of day for disrupting LI becomes evident (Manrique et al., 2004). Thus, ‘Different’ groups exhibit stronger aversions than ‘Same’ groups because the change of time of day interferes with LI. However, if the procedure includes only two days of previous habituation and restricted drinking during the conditioning session, ‘Same’ groups show stronger aversions than ‘Different’ groups. This may be due to the fact that the same time-of-day of conditioning facilitates retrieval of the aversive taste memory (Moron et al., 2002). Thus, the results are opposite. Using these behavioral procedures, the effect of NMDA induced lesions of the dorsal hippocampus in both roles of the temporal context was explored. In a series of two experiments using only preexposed groups (see Table 1) the performance of male Wistar rats with hippocampal (experiment 1, n = 18; experiment 2, n = 20) and amygdala (experiment 1, n = 18; experiment 2, n = 19) lesions was compared. Amygdala lesioned rats were used as control groups. Although the amygdala has been related with taste recognition memory (Bermudez-Rattoni, 2004), there are conflicting results concerning the effect of amygdala lesions on taste aversion learning (Bermudez-Rattoni and Yamamoto, 1998). Damage to this area did not disrupt taste aversion learning using the lesion and behavioral procedure applied in these experiments. The results indicated that the excitotoxic damage of the amygdala did not impair any of the temporal context effects, showing similar patterns to those reported in intact rats. As can be seen in Figure 1, a long habituation training to the temporal contexts (experiment 1, upper panel) led to significantly higher aversion rates in the ‘Different’ group than in the ‘Same’ group. A short habituation training to the drinking temporal context (experiment 2, lower panel) led to significant differences in the opposite direction, i.e. higher aversion rates in the ‘Same’ group than in the ‘Different’ group. Excitotoxic lesions of the dorsal hippocampus, however, selectively disrupted the effect of a temporal context change in the long-