Dissociation between memory reactivation and its behavioral expression: scopolamine interferes with memory expression without disrupting long-term storage.

The reconsolidation hypothesis has challenged the traditional view of fixed memories after consolidation. Reconsolidation studies have disclosed that the mechanisms mediating memory retrieval and the mechanisms that underlie the behavioral expression of memory can be dissociated, offering a new prospect for understanding the nature of experimental amnesia. The muscarinic antagonist scopolamine has been used for decades to induce experimental amnesias The goal of the present study is to determine whether the amnesic effects of scopolamine are due to storage (or retrieval) deficits or, alternatively, to a decrease in the long-term memory expression of a consolidated long-term memory. In the crab Chasmagnathus memory model, we found that scopolamine-induced amnesia can be reverted by facilitation after reminder presentation. This recovery of memory expression was reconsolidation specific since a reminder that does not triggers reconsolidation process did not allow the recovery. A higher dose (5 µg/g) of scopolamine induced an amnesic effect that could not be reverted through reconsolidation, and thus it can be explained as an interference with memory storage and/or retrieval mechanisms. These results, showing that an effective amnesic dose of scopolamine (100 ng/g) negatively modulates long-term memory expression but not memory storage in the crab Chasmagnathus, are consistent with the concept that dissociable processes underlie the mechanisms mediating memory reactivation and the behavioral expression of memory.

Enhancement of long-term memory expression by a single trial during consolidation.

Before the memory trace is stored long term, it must undergo a phase of consolidation during which it remains susceptible to modifications. It has previously been proposed that during consolidation, memories are kept from being stored long term, and can therefore be modified with additional information resulting from ongoing behavior. The Chasmagnathus associative memory model is used here to test whether it is possible during consolidation to modify the long-term expression of a memory generated by a weak training procedure. In this memory model, long-term memory expression is achieved after strong training protocols, a 15-spaced trial procedure. After a weak training protocol (WTP, six spaced trials), crabs do not show memory retention when tested in the long term. Nevertheless, the WTP builds a long-term memory that it is indeed consolidated, but remains unexpressed. Here we show that memory can be modified by experience during this short period after learning: memory expression can be enhanced by a Single Trial Session, on the condition that this session takes place contingent upon the consolidation period. We also found that during this time, the memory built by the WTP is behaviorally expressed, in contrast with what occurs at long term. Our results support the idea that during consolidation memories can be evaluated in the background of concurrent experiences. In particular, we propose that during the consolidation period it is possible for crabs to assess which experiences, among those stored long term, will be expressed long term.

Angiotensin modulates long-term memory expression but not long-term memory storage in the crab Chasmagnathus.

Memory reconsolidation is a dynamic process in which a previously consolidated memory becomes labile following reactivation by a reminder. In a previous study in the crab Chasmagnathus memory model, we showed that a water-shortage episode, via angiotensin modulation during reconsolidation, could reveal a memory that otherwise remains unexpressed: weakly trained animals cannot reveal long-term memory (LTM) except when an episode of noticeable ethological meaning, water deprivation, is contingent upon reconsolidation. However, these results are at variance with two of our previous interpretations: weak training protocols do not build LTM and angiotensin II modulates the strength of the information storing process. A parsimonious hypothesis is that in Chasmagnathus angiotensins regulate LTM expression, but not LTM storage. Here, we tested three predictions of this hypothesis. First, the well-known retrograde amnesic effect of the angiotensin II antagonist saralasin is not due to interference on memory storage, but to modulation of memory expression. Second, the recovery of the LTM memory expression of the apparently amnesic retrograde effect produced by saralasin, through the water-shortage episode contingent upon reconsolidation, must be reconsolidation specific. Consequently, summation-like effects and retrieval deficits cannot explain these results because of the parametric conditions of reconsolidation. Third, weak training protocols build an unexpressed LTM that requires mRNA transcription and translation, a diagnostic characteristic of LTM. Results show that angiotensin modulates LTM expression but not LTM memory storage in the crab Chasmagnathus. The results lead us to suggest that, in Chasmagnathus, LTM expression - the process of gaining appreciable control over behavior of the reactivated trace in the retrieval session - may be considered a distinct attribute of its long-term storage. This strategy, a positive modulation during reconsolidation, is proposed to distinguish between memories that can be reactivated, labilized and are not expressed, and memories that are not stored long term, obliterated or altered in other retrieval mechanisms.

Updating contextual information during consolidation as result of a new memory trace.

Reconsolidation studies have led to the hypothesis that memory, when labile, would be modified in order to incorporate new information. This view has reinstated original propositions suggesting that short-term memory provides the organism with an opportunity to evaluate and rearrange information before storing it, since it is concurrent with the labile state of consolidation. The Chasmagnathus associative memory model is used here to test whether during consolidation it is possible to change some attribute of recently acquired memories. In addition, it is tested whether these changes in behavioral memory features can be explained as modifications on the consolidating memory trace or as a consequence of a new memory trace. We show that short-term memory is, unlike long-term memory, not context specific. During this short period after learning, behavioral memory can be updated in order to incorporate new contextual information. We found that, during this period, the cycloheximide retrograde amnesic effect can be reverted by a single trial in a new context. Finally, by means of memory sensitivity to cycloheximide during consolidation and reconsolidation, we show that the learning of a new context (CS) during this short-term memory period builds up a new memory trace that sustains the behavioral memory update.