ERK-associated changes of AP-1 proteins during fear extinction

Abstract

Extensive research has unraveled the molecular basis of learning processes underlying contextual fear conditioning, but the mechanisms of fear extinction remain less known. Contextual fear extinction occurs when an aversive stimulus that initially caused fear is no longer present and depends on the activation of the extracellular signal-regulated kinase (ERK), among other molecules. Here we investigated how ERK signaling triggered by extinction affects its downstream targets belonging to the activator protein-1 (AP-1) transcription factor family. We found that extinction, when compared to conditioning of fear, markedly enhanced the interactions of active, phospho-ERK (pERK ) with c-Jun causing alterations of its phosphorylation state. The AP-1 binding of c-Jun was decreased whereas AP-1 binding of JunD, Jun dimerization protein 2 (JDP2) and ERK were significantly enhanced. The increased AP-1 binding of the inhibitory JunD and JDP2 transcription factors was paralleled by decreased levels of the AP-1 regulated proteins c-Fos and GluR2. These changes were specific for extinction and were MEK-dependent. Overall, fear extinction involves ERK/Jun interactions and a decrease of a subset of AP-1-regulated proteins that are typically required for fear conditioning. Facilitating the formation of inhibitory AP-1 complexes may thus facilitate the reduction of fear.

Fig. 1

Interaction of pERK and c-Jun during fear extinction. (A) The proposed pathway linking ERK and AP-1 proteins. (B) Extinction of conditioned fear, as revealed by a significant reduction of freezing behavior (F_1,4 = 7.23, p < 0.01 vs test 1), after repeated exposure to a context without shock. (C) Co-localization of ERK and c-Jun within hippocampal CA1 pyramidal neurons 1 hr after the 4^th extinction (E) test. i) Confocal image (overlay) of pERK (red) and c-Jun (green) immunofluorescence in CA1. ii-vii) Orthogonal projections along the pERK-positive cell marked with crosshair. ii-iv) Orthogonal projection (YZ plane) of pERK (red), c-Jun (green), and pERK + c-Jun (overlay). v-vii) Orthogonal projection (XZ plane) of pERK (red), c-Jun (green), and pERK + c-Jun (overlay). (D) pERK2 interacts with c-Jun but not JunB or JunD in nuclear extracts of the dorsal hippocampus (left panel). This interaction is significantly stronger in the E (middle and right panels) when compared to the other groups (F_2,15 = 11.47, p < 0.001 vs T and N). Statistically significant differences: *p < 0.01 vs test 1.

Fig. 2

Phosphorylation of c-Jun during extinction. (A) The levels of p-c-Jun Ser63/73 were undetectable in the N and T groups and significantly increased in the E group (F_2,15 = 9.65, p < 0.001). p-c-Jun Ser243 was decreased in the T group and returned to control, levels in the E group (F_2,15 = 4.12, p < 0.01). (B) Quantification of the immunoblot data. Statistically significant differences: *p < 0.01 vs N, #p < 0.01 vs E, ##p < 0.001 vs E.

Fig. 3

AP-1 binding of c-Jun during extinction. (A) The AP-1 binding activity of c-Jun was triggered in the T but not E group (F_2,15 = 11.23, p < 0.001 T vs E and N). (B) After inhibition of the MEK/ERK pathway, extinction enhances the AP-1 binding of c-Jun, as it is observed after training. Statistically significant differences: *p < 0.001 vs N and E/vehicle.

Fig. 4

AP-1 binding of JunD and JDP2 during extinction. (A) Contrary to c-Jun, the AP-1 binding of JunD (F_2,15 = 9.7, p < 0.001) and JDP2 (F_2,15 = 8.91, p < 0.001) was increased during extinction. (B) This enhancement was completely reversed by intrahippocampal inhibition of the MEK/ERK pathway. Statistically significant differences: *p < 0.001 vs N and E/vehicle.

Fig. 5

Biding of ERK to AP-1 and the c-Fos promoter. (A) ERK exhibits enhanced binding to the AP- consensus sequence after extinction. (B) This effect is specific because only addition of the consensus but not mutated AP-1 sequence displaces ERK from the immobilized AP-1 oligonucleotide. (C) Schematic representation of the c-Fos promoter. (D) ChIP assays showing amplification of the c-Fos promoter after the nested PCR (PCR2) from genomic DNA, input DNA and DNA co-immunoprecipitated by ERK but not control IgG (right half of the upper panel). PCR1 alone did not yield significant amplification. ERK loaded on the c-Fos promoter only in the E but not N and T groups (lower panel, representative of 6 independent replicates). (E) ERK did not show strong interaction with the c-Fos repressors mSin3A or HDAC1. A slight increase of ERK/HDAC1 (upper panel) was seen both in the T and E groups (lower panel). Statistically significant differences: *p < 0.001 vs all other groups.

Fig. 6

MEK-dependent decrease of c-Fos and GluR2 levels after fear extinction. (A) The activity of c-Fos in dorsohippocampal nuclear extracts was significantly lower in the E when compared to the T group (F_2,15 = 11.19, p < 0.001). The AP-1 binding and c-Fos levels were increased after inhibition of the MEK/ERK pathway. (B) The level of GluR2 increased in the T group (F_2,15 = 6.67, p < 0.01) but this increase was reversed in the E group. After inhibition of the hippocampal MEK/ERK pathway by U0126, the level of GluR2 remained elevated (F_2,15 = 15.67, p < 0.001. Statistically significant differences: *p < 0.001 all other groups.

Fig. 7

ERK-dependent signalling during conditioning and extinction of fear. Conditioning involves the ERK/MSK/RSK/c-Fos pathway. Extinction, on the other hand, induces nuclear pERK/c-Jun interactions and increases the inhibitory AP-1 complexes JunD and JDP2. The later changes may be involved, at least in part, in the down-regulation of c-Fos and GluR2.