The role of calsenilin/DREAM/KChIP3 in contextual fear conditioning

Abstract

Potassium channel interacting proteins (KChIPs) are members of a family of calcium binding proteins that interact with Kv4 potassium (K(+)) channel primary subunits and also act as transcription factors. The Kv4 subunit is a primary K(+) channel pore-forming subunit, which contributes to the somatic and dendritic A-type currents throughout the nervous system. These A-type currents play a key role in the regulation of neuronal excitability and dendritic processing of incoming synaptic information. KChIP3 is also known as calsenilin and as the transcription factor, downstream regulatory element antagonist modulator (DREAM), which regulates a number of genes including prodynorphin. KChIP3 and Kv4 primary channel subunits are highly expressed in hippocampus, an area of the brain important for learning and memory. Through its various functions, KChIP3 may play a role in the regulation of synaptic plasticity and learning and memory. We evaluated the role of KChIP3 in a hippocampus-dependent memory task, contextual fear conditioning. Male KChIP3 knockout (KO) mice showed significantly enhanced memory 24 hours after training as measured by percent freezing. In addition, we found that membrane association and interaction with Kv4.2 of KChIP3 protein was significantly decreased and nuclear KChIP3 expression was increased six hours after the fear conditioning training paradigm with no significant change in KChIP3 mRNA. In addition, prodynorphin mRNA expression was significantly decreased six hours after fear conditioning training in wild-type (WT) but not in KO animals. These data suggest a role for regulation of gene expression by KChIP3/DREAM/calsenilin in consolidation of contextual fear conditioning memories.

Figure 1.

Analysis of immunoreactivity for expression levels of relevant proteins in KChIP3 KO mice. (A) Western blotting was used to confirm genotyping in KChIP3 KO and WT littermate mice. The presence of a KChIP3 band indicated WT while the absence of a band indicated KChIP3 KO. (B) Representative Western blots for Kv4.2, Kv4.3, KChIP2, and β tubulin, which was used as a loading control. (C) Bar graph summary of immunoreactivity of protein expression levels in KO relative to WT of KChIP1 (KO n = 4; WT n = 3), KChIP2 (KO n = 6, WT n = 5), Kv4.2 (KO n = 10, WT n = 8), and Kv4.3 (KO n = 5, WT n = 5). There were no significant differences between WT and KO expression levels of any of the proteins assayed (P > 0.05, t-test).

Figure 2.

KChIP3 KO mice exhibit normal locomotor activity and shock sensitivity. (A) The accelerating rotarod task was used to assess motor learning and coordination in KChIP3 KO and WT littermate mice. Mice were trained four trials a day for 2 d for a total of eight trials, and time spent on the rod per trial was averaged across genotype and plotted per trial. No significant effect of genotype over trials was observed (KO n = 15; WT n = 8; Repeated Measures ANOVA; P = 0.28). (B) Response to foot shock in WT and KO mice. KChIP3 KO (n = 10) and WT littermate mice (n = 10) were tested for foot shock responses (see Materials and Methods) at increasing intensities (0.1–0.7 mA in 0.1 increments). There were no significant differences in the responses of the KChIP3 KO mice compared to WT littermates (P > 0.05, Repeated Measures ANOVA).

Figure 3.

Open field and elevated plus maze (EPM) performance in KChIP3 KO mice. KChIP3 KO and WT littermate mice were tested in the open field assay (OFA) to determine baseline locomotor activity and anxiety-like behavior. (A) Summary bar graph of exploratory activity of KChIP3 KO vs. WT mice in the OFA. Total exploratory activity and exploration of the center in the OFA exhibited by the KChIP3 KO mice (n = 11) was decreased but not significantly different compared to WT mice (n = 16; P > 0.05, t-test). (B) Bar graph representing the center/total distance ratio decreased in KChIP3 KO mice as compared to WT but not significantly different (P > 0.05; t-test). (C) KChIP3 KO and WT littermate mice were also tested in the EPM. KChIP3 KO mice (n = 6) spent approximately the same amount of time in the open arms as did the WT littermates (n = 10). (D) Crossings into the open arms by KO mice were decreased compared to the WT animals, but the differences were not significant (P > 0.05; t-test).

Figure 4.

Male KChIP3 KO mice exhibit enhanced memory of contextual fear conditioning at 24 h and 1 mo. The fear memory of KChIP3 KO and WT littermate mice was tested using cued and contextual fear conditioning 24 h after training. (A) Mice were trained on day 1 using a single-shock protocol. There were no significant differences between KChIP3 KO and WT littermates in freezing on the training day during the exploration period before the tone, during the tone (black bar), or after the foot shock (arrow; 0.5 mA, 2 sec). (B) KChIP3 KO mice (n = 7) exhibited similar freezing compared to WT littermates in the cued test (n = 6; P > 0.05, t-test). (C) KChIP3 KO mice (n = 10) showed significantly increased freezing behavior to context compared to WT littermates (n = 11) when tested at 24 h after training (t-test; ** P = 0.002). This effect was long-lasting as KChIP3 KO mice (n = 5) re-exposed to the box one month after training also showed enhanced freezing compared to WT littermates (n = 7, t-test; ** P = 0.009).

Figure 5.

Male KChIP3 KO mice exhibited no differences in cued or contextual fear conditioning when tested 1 h after training. (A) A separate group of mice were tested for freezing in the context and to the cue 1 h after training. KChIP3 KO (n = 4), and WT littermates (n = 8) showed no differences in freezing levels in the context (t-test, P > 0.05) or (B) the cued tests (t-test, P > 0.05) only 1 h after training.

Figure 6.

Changes in the immunoreactivity of hippocampus membrane expression levels of KChIP3, Kv4.2, and KChIP3 mRNA following fear conditioning in WT mice. Western blotting and real-time PCR were used to study expression of KChIP3 and Kv4.2 protein immunoreactivity as well as KChIP3 mRNA in the hippocampus at different time points (1, 6, and 24 h) following fear-conditioning training. Three groups of WT mice were trained as described in Materials and Methods: a context-trained group (CXT), a fear-conditioned group (FC), and a latent inhibition group (LI). (A) Representative Western blots of membrane KChIP3 and Kv4.2 protein levels at 1- and 6-h time points in all three groups. (B) Bar graph summarizing immunoreactivity for KChIP3 membrane association at different time points after FC and LI training. KChIP3 membrane association in FC1 (n = 5) hippocampus was not significantly different from CXT1 (n = 5) at 1 h after training, but was significantly higher than LI1 (** P < 0.01, n = 3, which was significantly different from context, * P < 0.05; One-Way ANOVA with post-hoc Tukey's test). KChIP3 membrane association was significantly decreased (P < 0.05) 6 h after FC training (n = 8) relative to CXT6 (n = 8). No significant differences between CXT24 (n = 3) and FC24 (n = 3) or LI24 (n = 3) were found at the 24-h time point. (C) Membrane Kv4.2 expression in the hippocampus at different time points after FC and LI training. There was no significant effect on immunoreactivity for Kv4.2 membrane expression at the 1-h time point (P > 0.05; FC1, n = 3, LI1, n = 3, CXT1, n = 3), 6-h time point (P > 0.05; FC6, n = 5, LI6, n = 5, CXT6, n = 5), or 24-h time point (P > 0.05; FC24, n = 4, LI24, n = 2, CXT24, n = 4). (D) There was no difference in KChIP3 mRNA expression at any of the time points; 1-h time point (P > 0.05; FC1, n = 3, LI1, n = 3, CXT1, n = 3), 6-h time point (P > 0.05; FC6, n = 5, LI6, n = 5), and 24 h time point (P > 0.05; FC24, n = 3, LI24, n = 3).

Figure 7.

KChIP3 and Kv4.2 interaction is decreased 6 h after FC training relative to CXT. (A) Co-immunoprecipitation with KChIP3 antibody and probed with the Kv4.2 antibody shows that the Kv4.2 and KChIP3 interaction is diminished in the FC-trained animals relative to the CXT trained animals (n = 3). The exposure time for the Kv4.2 input was shorter than the exposure time for IP. (B) Summary bar graph showing that there was no difference in immunoreactivity of Kv4.2 expression measured in the input.

Figure 8.

Levels of KChIP3 immunoreactivity are increased in the nucleus 6 h after FC training. Nuclear fractions were assayed for the presence of KChIP3 in the nucleus. (A) Representative Western blot of nuclear fraction with KChIP3 antibody from CXT6, LI6, and FC6 tissue. The KChIP3 antibody recognized a band in the nuclear fraction that was slightly smaller than the membrane KChIP3 band (∼25 kDa). (B) Summary bar graph showing percent change in immunoreactivity for this band in the nucleus of FC6 and LI6 relative to CXT6. This band increased significantly in the nucleus at the 6-h time point in the FC-trained animals (One-Way ANOVA, **, P < 0.01; n = 3) relative to CXT (n = 3), but not in the LI-trained (P > 0.05, n = 3).

Figure 9.

Prodynorphin mRNA levels in hippocampus following fear conditioning. Real-time PCR was used to determine if prodynorphin mRNA expression was regulated after FC training and LI training in WT mice. Prodynorphin mRNA was significantly decreased at the 6-h time point in both FC6 and LI6 hippocampus relative to CXT6 (One-Way ANOVA with post-hoc Tukey's test; FC6, n = 6, P < 0.05, LI6, n = 6, P < 0.01). Prodynorphin mRNA was also significantly increased at the 1-h time point in both FC1 and LI1 relative to CXT1 (FC1, n = 3, 142 ± 15%, P < 0.05; LI1, n = 3, P < 0.05), and prodynorphin mRNA was decreased in both FC and LI at the 24-h time point, although not significantly (FC24, n = 3, LI24, n = 3; P > 0.05). Prodynorphin mRNA expression was also tested in KChIP3 KO mice 6 h after FC and CXT. Prodynorphin mRNA was not significantly modulated in the KChIP3 KO mice 6 h after FC training relative to CXT (150 ± 28%, n = 6, P > 0.05).

Figure 10.

Schematic diagram of a proposed model for the role of KChIP3 in contextual fear conditioning. We propose that during consolidation (between 1 and 6 h after training) KChIP3 translocates from the membrane to the nucleus and regulates gene expression. We show that prodynorphin mRNA is regulated after training and KChIP3 plays a role in this regulation.