Activation of MAPK is necessary for long-term memory consolidation following food-reward conditioning

Abstract

Although an important role for the mitogen-activated protein kinase (MAPK) has been established for memory consolidation in a variety of learning paradigms, it is not known if this pathway is also involved in appetitive classical conditioning. We address this question by using a single-trial food-reward conditioning paradigm in the freshwater snail Lymnaea stagnalis. This learning paradigm induces protein synthesis-dependent long-term memory formation. Inhibition of MAPK phosphorylation blocked long-term memory consolidation without affecting the sensory and motor abilities of the snails. Thirty minutes after conditioning, levels of MAPK phosphorylation were increased in extracts from the buccal and cerebral ganglia. These ganglia are involved in the generation, modulation, and plasticity of the feeding behavior. We also detected an increase in levels of MAPK phosphorylation in the peripheral tissue around the mouth of the snails where chemoreceptors are located. Although an increase in MAPK phosphorylation was shown to be essential for food-reward conditioning, it was also detected in snails that were exposed to the conditioned stimulus (CS) or the unconditioned stimulus (US) alone, suggesting that phosphorylation of MAPK is necessary but not sufficient for learning to occur.

Figure 1.

MAPK-like immunoreactivity in Lymnaea central nervous system. (A) Western blots showing immunoreactivity of anti-MAPK, anti-pMAPK, and anti-Aplysia MAPK antibodies. (B,C,D) Immunohistochemistry showing anti-pMAPK immunoreactivity in neuropile and neuronal nuclei of buccal (B) and cerebral (C,D) ganglia. (B) Section through a buccal ganglion showing neurons with nuclear staining (filled arrow) as well as some staining in the neuropile (open arrow). (C) Section through a cerebral ganglion showing staining in the nuclei of a few neurons and in a lip nerve (open arrow). (D) Section through a cerebral ganglion showing a cerebral giant cell (arrow) with nuclear staining. Bars, 50 μm.

Figure 2.

Single-trial reward conditioning paradigm induces associative memory formation in Lymnaea.(A) Schematic diagram of the behavioral procedures used during food-reward classical conditioning. Two groups of snails were presented with either a single pairing of amyl acetate and sucrose (CS+US group; n = 24) or with the two stimuli presented in the same order but separated by a 1-h interval (unpaired group; n = 22). On the following day, both groups were presented again with amyl acetate. The feeding responses to amyl acetate of both groups were measured on the day of conditioning before sucrose presentation and again 1 d after. (B) A graph showing the feeding responses to amyl acetate of CS+US and unpaired group before and after conditioning. During conditioning, the responses of both groups to amyl acetate were low and not significantly different from each other (P = 0.2). In contrast, 1 d after conditioning (test), the CS+US group responded to amyl acetate with a significantly higher number of feeding movements than did the unpaired group (P < 0.001). Within-group comparisons showed that while the behavior of the CS+US group after conditioning was significantly different from its behavior before conditioning (^***P < 0.001), the responses of the unpaired group on both days were not significantly different (P = 0.5).

Figure 3.

Inhibition of MAPK phosphorylation blocks memory formation. Schematic diagram of experimental procedure (top) and feeding responses of two groups of snails injected with either vehicle (n = 20) or U0126, a MAPK kinase inhibitor (n = 21) (bottom). The snails were injected 30 min before conditioning, and their feeding behavior was monitored during conditioning and 1 d after conditioning during amyl acetate presentation. The groups were conditioned with a single pairing of amyl acetate and sucrose. During conditioning, the responses of both groups to water, amyl acetate, and amyl acetate mixed with sucrose, were not significantly different from each other (water, P = 0.2; amyl acetate, P = 0.3; amyl acetate + sucrose, P = 0.2). One day after conditioning, the group injected with U0126 responded to amyl acetate with a significantly lower number of feeding movements than did the vehicle-injected group. ^**P < 0.001.

Figure 4.

Inhibition of MAPK phosphorylation does not impair feeding related sensory and motor abilities of the snails. (A) Schematic diagram of the experimental procedure (top) and a graph showing the feeding responses to amyl acetate during conditioning and 1 d after conditioning (bottom). Snails from two groups were injected with either vehicle (n = 18) or U0126 (n = 17), 30 min before test. No significant difference was observed when the responses to amyl acetate of both groups were compared (before conditioning, P = 0.6; after conditioning, P = 0.2). However, the feeding responses to amyl acetate of both the vehicle-injected and the U0126-injected groups were significantly higher after conditioning compared with the respective responses during conditioning (vehicle, ^*P < 0.05; U0126, ^**P < 0.01), suggesting that learning occurred. (B) Schematic diagram of the experimental procedure (top) and a graph showing the feeding response to amyl acetate during conditioning and 2 d after conditioning of two groups injected 1 d after conditioning and tested 1 d after injection with either vehicle (n = 20) or U0126 (n = 19) (bottom). No significant difference was observed when the responses to amyl acetate of the two groups were compared (before conditioning, P = 0.5; after conditioning, P = 0.8). However, the feeding responses to amyl acetate of both the vehicle-injected and the U0126-injected groups were significantly higher after conditioning compared with the responses before conditioning (vehicle, ^**P < 0.0001; U0126, ^*P < 0.05), suggesting that learning occurred.

Figure 5.

Inhibition of MAPK phosphorylation during the period of memory consolidation blocks long-term memory formation. Schematic diagram of the experimental procedure (top) and a graph showing the feeding responses to amyl acetate during conditioning and at test, 1 d after conditioning (bottom). Snails from two groups were injected immediately after conditioning with either vehicle (n = 31) or U0126 (n = 21). Before conditioning, the responses of both groups were not significantly different from each other (P = 0.2). However, 1 d after conditioning, the response of the U0126-injected group was significantly lower than was the response of the vehicle-injected group (^*P < 0.05). Within-groups comparisons showed that the feeding response of the vehicle-injected group after conditioning was significantly greater than before conditioning (P < 0.01), while the responses of the U0126-injected group before and after conditioning were not significantly different from each other (P = 0.2).

Figure 6.

MAPK phosphorylation levels are elevated 30 minutes after conditioning in both brain and lip tissue. (A) Western blots of cerebral-buccal extracts from naive and conditioned (CS+US) snails frozen in liquid nitrogen 30 min after conditioning. The immunoblots were probed with an anti-pMAPK antibody, stripped, and then reprobed with an anti-MAPK antibody. The corresponding densitometric analysis is shown below (n = 5). MAPK phosphorylation, but not the amount of total MAPK, is significantly increased 30 min after conditioning. ^*P < 0.05. (B) Western blots of lip tissue extracts from naive and conditioned (CS+US) snails frozen 30 min after conditioning. As above, the immunoblots were probed with an anti-pMAPK antibody, stripped, and then reprobed with an anti-MAPK antibody. The corresponding densitometric analysis is also shown (n = 4). Levels of MAPK phosphorylation are significantly higher 30 min after conditioning compared with naive levels. No significant difference was observed between the levels of total MAPK of conditioned and naive groups. ^**P < 0.01.

Figure 7.

Injection of U0126 blocks conditioning-induced MAPK phosphorylation in brain and lip tissues. Two groups of snails were injected with either vehicle or U0126 30 min before conditioning and frozen 30 min after conditioning. At the same time, a group of naive snails was also frozen. Levels of phosphorylated MAPK and total MAPK were monitored by Western blots. (A,B) Example of immunoblots of protein extracts from the combined cerebral and buccal ganglia (A) and lip tissue (B). The respective densitometric analysis is shown below each immunoblot example (cerebral-buccal, n = 4; lip, n = 3). In both tissues, protein extracts from the conditioned U0126-injected group contained significantly lower levels of pMAPK than did the conditioned vehicle-injected group, but levels of pMAPK of conditioned U0126-injected animals were not significantly different from naive levels. As expected, the conditioned vehicle-injected group showed significantly higher levels of pMAPK than did the naive group. No significant difference was observed in the levels of total MAPK between the three groups in both types of tissues that were analyzed (P > 0.05; ^*P < 0.05).

Figure 8.

Presentation of amyl acetate (the CS) alone or sucrose (the US) alone is sufficient to induce MAPK phosphorylation in the cerebral and buccal ganglia and lip tissue, although neither stimulus induces a change in the feeding response to amyl acetate. (A) Schematic diagram of the experimental procedure. The snails were divided into four groups. The naive group did not receive any stimulation. The CS+US group was presented with a single pairing of amyl acetate and sucrose. The CS alone group was presented with amyl acetate paired with water, and the US alone group was presented with water paired with sucrose. The groups were then divided into two subgroups each, one subgroup that was frozen 30 min after stimulation and another that was kept alive for behavioral testing the day after stimulation. (B) A graph showing the feeding responses to amyl acetate of the naive group (n = 21), the CS+US group (n = 21), the CS alone (n = 20), and the US alone (n = 21). ^*P < 0.05 relative to naive, CS alone, and US alone. (C,D) Example of immunoblot and densitometric analysis of pMAPK Western blot of cerebral-buccal extracts (C) and lip extracts (D). The levels of pMAPK of the groups CS+US (n = 10), CS alone (n = 10), and US alone (n = 10) were significantly higher than were levels of the naive group (n = 8 for cerebral-buccal extracts and n = 9 for lip extracts). There was no significant difference in pMAPK levels between any of the three stimulated groups. ^*P < 0.05 relative to naive.