Activation of hippocampal nuclear factor-kappa B by retrieval is required for memory reconsolidation

Abstract

Initially, memory is labile and requires consolidation to become stable. However, several studies support that consolidated memories can undergo a new period of lability after retrieval. The mechanistic differences of this process, termed reconsolidation, with the consolidation process are under debate, including the participation of hippocampus. Up to this point, few reports describe molecular changes and, in particular, transcription factor (TF) involvement in memory restabilization. Increasing evidence supports the participation of the TF nuclear factor-kappaB (NF-kappaB) in memory consolidation. Here, we demonstrate that the inhibition of NF-kappaB after memory reactivation impairs retention of a hippocampal-dependent inhibitory avoidance task in mice. We used two independent disruptive strategies to reach this conclusion. First, we administered intracerebroventricular or intrahippocampal sulfasalazine, an inhibitor of IKK (IkappaB kinase), the kinase that activates NF-kappaB. Second, we infused intracerebroventricular or intrahippocampal kappaB decoy, a direct inhibitor of NF-kappaB consisting of a double-stranded DNA oligonucleotide that contains the kappaB consensus sequence. When injected immediately after memory retrieval, sulfasalazine or kappaB decoy (Decoy) impaired long-term retention. In contrast, a one base mutated kappaB decoy (mDecoy) had no effect. Furthermore, we also found NF-kappaB activation in the hippocampus, with a peak 15 min after memory retrieval. This activation was earlier than that found during consolidation. Together, these results indicate that NF-kappaB is an important transcriptional regulator in memory consolidation and reconsolidation in hippocampus, although the temporal kinetics of activation differs between the two processes.

Figure 1.

Effect of intracerebroventricular administration of NF-κB inhibitors after reexposure to the training context. A, Effect of the IKK inhibitor sulfasalazine. Top diagram, Design of the experiment. TR, Training with a footshock; T1, reactivation session; T2–T5, successive testing sessions with 24 h intervals; the injections were performed immediately after T1 (memory reactivation). The graph represents the latencies to step-through expressed as medians and interquartile ranges of animals injected with vehicle (N = 10), 3.5 mm (N = 10), or 7 mm sulfasalazine (Sulfa) (N = 12). **p < 0.01; *p < 0.05 when compared with its respective test session of the vehicle-treated group (Mann–Whitney U test, two-tailed). B, Effect of indomethacin. Top diagram, Design of the experiment. The injections were performed immediately after T1 (memory reactivation). The graph is as in A. N = 10 for each group. C, Effect of Decoy. Top diagram, Design of the experiment. TR, Training with a footshock; T1, reactivation session; T2–T6, successive testing sessions with 24 h intervals; the injections were performed immediately after reactivation. Each bar represents the medians and interquartile ranges. Mice were injected either with vehicle (N = 7), mDecoy (N = 8), or Decoy (N = 12). **p < 0.01 compared with its respective test session of the vehicle and mDecoy-treated groups (Mann–Whitney U test, two-tailed).

Figure 2.

Effect of the IKK inhibitor sulfasalazine when injected intrahippocampally after reexposure to the training context. A, Top diagram, Design of the experiment. TR, Training with a footshock; T1, reactivation session; T2 testing sessions after 24 h interval; the injections were performed immediately after T1 (memory reactivation). The graph represents the latencies to step-through expressed as medians and interquartile ranges of animals injected with vehicle (N = 12), 0.2 μg (N = 8), 0.5 μg (N = 12), or 1 μg (N = 11) of sulfasalazine (Sulfa). **p < 0.01, when compared with its respective test session of the vehicle-treated group (Mann–Whitney U test, two-tailed). B, Effect of indomethacin injection after reexposure. Top diagram, Design of the experiment. The injections were performed immediately after T1 (memory reactivation). The graph is as in A. Two doses of indomethacin, 13 or 65 ng per hippocampus, were administrated. N = 10 for each group.

Figure 3.

Effect of Decoy in hippocampus after reexposure to the training context. A, Decoy administered immediately after reexposure. Top diagram, Design of the experiment. TR, Training with a footshock; T1, reactivation session; T2 and T3, successive testing sessions with 24 h intervals; the injections were performed immediately after reactivation. Graph, Each bar represents the medians and interquartile ranges. Mice were injected either with vehicle (N = 7), mDecoy (N = 7), or Decoy (N = 7). **p < 0.01 compared with its respective test session of the vehicle and mDecoy-treated groups (Mann–Whitney U test, two-tailed). B, Effect of Decoy administered 3 h after reexposure. Top diagram, Design of the experiment. The graph is as in A. N = 10 for each group.

Figure 4.

A, Effect of Decoy in hippocampus without reexposure to the training context. Top diagram, Design of the experiment. Graph, Each bar represents the medians and interquartile ranges. Mice were injected either with vehicle (N = 7), mDecoy (N = 7), or Decoy (N = 7). B, Effect of Decoy administered after reexposure on short-term reactivated memory. Top diagram, Design of the experiment. Animals were tested (T2) 3 h after injection and retested 24 h after injection. The graph is as in A. N = 8 for each group. **p < 0.01 compared with its respective test session of the vehicle and mDecoy-treated groups (Mann–Whitney U test, two-tailed).

Figure 5.

Temporal course of NF-κB activity after memory reactivation. A, Design of the experiment for shocked (S) and unshocked (U) groups. Animals were killed 5, 15, or 45 min after T1 or 48 h after training (S non-reexposed group). B, NF-κB activity relative to the naive (N) mean value estimated by densitometric analysis (ROD) of the p65/p50 EMSA band, obtained with hippocampal nuclear extracts from animals of the different groups. General ANOVA, F_(8,87) = 8.94, p < 0.01. Duncan's test, N versus S or U, p < 0.01. C, Representative EMSAs of the three groups (N, U, and S) for the different time points analyzed. N = 10 for each group.

Figure 6.

NF-κB activity after training when animals were preexposed to the context. A, Design of the experiment. Animals were killed 45 min after training. S group, Mice reexposed to the context for 6 d and shocked in the context at day 7. U group, Mice reexposed to the context for 7 d. N, Untreated naive mice. B, Representative EMSAs of the three groups. The arrowheads indicate specific bands. The black arrow indicates the complex measured. C, NF-κB activity relative to the mean value of N group for the three groups, 45 min after the last trial. General ANOVA, p < 0.01; Duncan's test, S versus N, p < 0.05; S versus U, p < 0.01; U versus N, p > 0.05. **p < 0.01 and *p < 0.05. N of each group is indicated in the respective bar.

Figure 7.

CS–US association activates NF-κB in hippocampus. A, Design of the experiment. Animals were killed 45 min after the last training or exposure. B, Densitometric analysis of the p65/p50 retarded band. C, Representative retarded bands obtained in gel shift for each group; the black arrowhead indicates the quantified band. General ANOVA, F = 3.15, p < 0.05. Duncan's test, CG versus ICG, SCG, CUG, and N, *p < 0.05. N of each group is indicated in the respective bar.

Figure 8.

Presence of Decoy in hippocampus inhibits NF-κB activity. A, Confocal micrographs show localization of the fluoresceinated Decoy (green) 15 min after injection. Costained with propidium iodide (red). The arrowheads show the deepest position of the needle. B, Coronal brain image is adapted from the atlas of Franklin and Paxinos (2001), indicating with dashed lines and black squares the position of the injection in the hippocampus. The last coronal section (Nissl stained) shows the trace of the needle. C, Top diagram, Design of the experiment for shocked groups receiving either mDecoy or Decoy intrahippocampal injection immediately after reexposure. TR, Training with a footshock; T1, reactivation session. Animals were killed 15 min after T1. The graph represents NF-κB activity relative to the Decoy group mean value 15 min after injection of mDecoy or Decoy. ANOVA, F_(1,10) = 11.36, p < 0.01. Duncan's test, mDecoy versus Decoy, **p < 0.01; N = 5 and 7, respectively. Inset, Representative EMSAs of the two groups. D, Administration of mDecoy or Decoy in the forelimb primary somatosensory cortex after reexposure, shows no effect on retention. The top diagram is as in C. Graph, Each bar represents the medians and interquartile ranges. Mice were injected either with mDecoy (N = 9) or Decoy (N = 9).