One-trial conditioned taste aversion in Lymnaea: good and poor performers in long-term memory acquisition.

In the majority of studies designed to elucidate the causal mechanisms of memory formation, certain members of the experimental cohort, even though subjected to exactly the same conditioning procedures, remember significantly better than others, whereas others show little or no long-term memory (LTM) formation. To begin to address the question of why this phenomenon occurs and thereby help clarify the causal mechanism of LTM formation, we used a conditioned taste aversion (CTA) procedure on individuals of the pond snail Lymnaea stagnalis and analyzed their subsequent behavior. Using sucrose as an appetitive stimulus and KCl as an aversive stimulus, we obtained a constant ratio of ;poor' to ;good' performers for CTA-LTM. We found that approximately 40% of trained snails possessed LTM following a one-trial conditioning procedure. When we examined the time-window necessary for the memory consolidation, we found that if we cooled snails to 4 degrees C for 30 min within 10 min after the one-trial conditioning, LTM was blocked. However, with delayed cooling (i.e. longer than 10 min), LTM was present. We could further interfere with LTM formation by inducing inhibitory learning (i.e. backward conditioning) after the one-trial conditioning. Finally, we examined whether we could motivate snails to acquire LTM by depriving them of food for 5 days before the one-trial conditioning. Food-deprived snails, however, failed to exhibit LTM following the one-trial conditioning. These results will help us begin to clarify why some individuals are better at learning and forming memory for specific tasks at the neuronal level.

De Novo synthesis of CREB in a presynaptic neuron is required for synaptic enhancement involved in memory consolidation.

Interaction between the activator type of cyclic AMP response element binding protein (CREB1) and the repressor type (CREB2) results in determining the emergence of long-lasting synaptic enhancement involved in memory consolidation. However, we still do not know whether the constitutively expressed forms of CREB are enough or the newly synthesized forms are required for the synaptic enhancement. In addition, if the newly synthesized forms are needed, we must determine the time for translation of CREB from its mRNA. We applied the methods of RNA interference and real-time polymerase chain reaction (PCR) to CREB in the cerebral giant cells of Lymnaea. The cerebral giant cells play an important role in associative learning and employ a CREB cascade for the synaptic enhancement to neurons such as the B1 motoneurons. We injected the small interfering RNA (siRNA) of CREB1 or CREB2 into the cerebral giant cells and examined the changes in amplitude of excitatory postsynaptic potential (EPSP) recorded in the B1 motoneurons. The changes in the amounts of CREB1 and CREB2 mRNAs were also examined in the cerebral giant cells. The EPSP amplitude was suppressed 15 min after injection of CREB1 siRNA, whereas that was augmented 60 min after injection of CREB2 siRNA. In the latter case, the decrease in the amount of CREB2 mRNA was confirmed by real-time PCR. Our results showed that the de novo synthesized forms of CREB are required within tens of minutes for the synaptic enhancement in memory consolidation.

Altered gene activity correlated with long-term memory formation of conditioned taste aversion in Lymnaea.

The pond snail Lymnaea stagnalis is capable of learning conditioned taste aversion (CTA) and then consolidating that learning into long-term memory (LTM) that persists for at least 1 month. LTM requires de novo protein synthesis and altered gene activity. Changes in gene activity in Lymnaea that are correlated with, much less causative, memory formation have not yet been identified. As a first step toward rectifying this situation, we constructed a cDNA microarray with mRNAs extracted from the central nervous system (CNS) of Lymnaea. We then, using this microarray assay, identified genes whose activity either increased or decreased following CTA memory consolidation. We also identified genes whose expression levels were altered after inhibition of the cyclic AMP response element-binding protein (CREB) that is hypothesized to be a key transcription factor for CTA memory. We found that the molluscan insulin-related peptide II (MIP II) was up-regulated during CTA-LTM, whereas the gene encoding pedal peptide preprohormone (Pep) was down-regulated by CREB2 RNA interference. We next examined mRNAs of MIP II and Pep using real-time RT-PCR with SYBR Green. The MIP II mRNA level in the CNS of snails exhibiting "good" memory for CTA was confirmed to be significantly higher than that from the CNS of snails exhibiting "poor" memory. In contrast, there was no significant difference in expression levels of the Pep mRNA between "good" and "poor" performers. These data suggest that in Lymnaea MIP II may play a role in the consolidation process that forms LTM following CTA training.

Taste discrimination in conditioned taste aversion of the pond snail Lymnaea stagnalis.

Conditioned taste aversion (CTA) in the pond snail Lymnaea stagnalis has been widely used as a model for gaining an understanding of the molecular and behavioral mechanisms underlying learning and memory. At the behavioral level, however, it is still unclear how taste discrimination and CTA interact. We thus examined how CTA to one taste affected the feeding response induced by another appetitive food stimulus. We first demonstrated that snails have the capacity to recognize sucrose and carrot juice as distinct appetitive stimuli. We then found that snails can become conditioned (i.e. CTA) to avoid one of the stimuli and not the other. These results show that snails can distinguish between appetitive stimuli during CTA, suggesting that taste discrimination is processed upstream of the site where memory consolidation in the snail brain occurs. Moreover, we examined second-order conditioning with two appetitive stimuli and one aversive stimulus. Snails acquired second-order conditioning and were still able to distinguish between the different stimuli. Finally, we repeatedly presented the conditional stimulus alone to the conditioned snails, but this procedure did not extinguish the long-term memory of CTA in the snails. Taken together, our data suggest that CTA causes specific, irreversible and rigid changes from appetitive stimuli to aversive ones in the conditioning procedure.

The early snail acquires the learning. Comparison of scores for conditioned taste aversion between morning and afternoon.

The pond snail Lymnaea stagnalis acquires conditioned taste aversion (CTA) and maintains its memory for more than a month. Snails in our laboratory were cultured at 20 degrees C on a 12:12 light-dark cycle (light from 7 am to 7 pm). To examine the hours during which snails acquire CTA effectively, we trained some snails in the morning and others in the afternoon, and then compared their scores. CTA developed in both cases, but scores were significantly better in the morning than in the afternoon. To elucidate the cause of this difference in scores, we observed the voluntary activity of snails and found the circadian rhythm reflected in the snails' free-movement distances; distances at the circadian time 0-12 (daytime) were significantly longer than those at the circadian time 12-24 (nighttime). This rhythm was kept up for at least 3 days, even in constant darkness. In conclusion, L. stagnalis should be trained in the morning to acquire associative learning, possibly because of its greater propensity to roam about at that time as opposed to the afternoon.

The expression pattern of CREB genes in the central nervous system of the pond snail Lymnaea stagnalis.

To analyze the expression pattern of genes of cAMP responsive element binding protein (CREB), we performed in situ hybridization for the whole central nervous system (CNS) of the pond snail Lymnaea stagnalis. The CREB1 (activator) and CREB2 (repressor) homologues have already been cloned in L. stagnalis, and they are referred to as LymCREB1 and LymCREB2. Using the frozen sections and the whole mount preparations of the CNS, we mapped the distribution of LymCREB1 and LymCREB2 mRNA containing neurons. The present findings showed that the LymCREB1 mRNA containing neurons are a relatively few, whereas LymCREB2 mRNA is contained ubiquitously in the whole CNS of L. stagnalis.