Contextual fear response is modulated by M-type K+ channels and is associated with subtle structural changes of the axon initial segment in hippocampal GABAergic neurons

Abstract

Background: In the fear memory network, the hippocampus modulates contextual aspects of fear learning while mutual connections between the amygdala and the medial prefrontal cortex are widely involved in fear extinction. G-protein-coupled receptors (GPCRs) are involved in the regulation of fear and anxiety, so the regulation of GPCRs in fear signaling pathways can modulate the mechanisms of fear memory acquisition, consolidation and extinction. Various studies suggested a role of M-type K+ channels in modulating fear expression and extinction, although conflicting data prevented drawing of clear conclusions. In the present work, we examined the impact of M-type K+ channel blockade or activation on contextual fear acquisition and extinction. In addition, regarding the pivotal role of the hippocampus in contextual fear conditioning (CFC) and the involvement of the axon initial segment (AIS) in neuronal plasticity, we investigated whether structural alterations of the AIS in hippocampal neurons occurred during contextual fear memory acquisition and short-time extinction in mice in a behaviorally relevant context.

Results: When a single systemic injection of the M-channel blocker XE991 (2 mg/kg, IP) was carried out 15 minutes before the foot shock session, fear expression was significantly reduced. Expression of c-Fos was increased following CFC, mostly in GABAergic neurons at day 1 and day 2 post-fear training in CA1 and dentate gyrus hippocampal regions. A significantly longer AIS segment was observed in GABAergic neurons of the CA1 hippocampal region at day 2.

Conclusions: Our results underscore the role of M-type K + channels in CFC and the importance of hippocampal GABAergic neurons in fear expression.

Keywords: GABAergic neurons; fear conditioning; hippocampus; neuronal plasticity; potassium channels.

Figure 1.. Contextual Fear conditioning. (a) Protocol of Contextual Fear conditioning. Day 0 is referred to the training day: after a habituation period of 2 min, animals received 3 foot shocks, each lasting 2 sec at one min interval. (b) Behavioral profile showing fear expression at day 1 and subsequent re-exposure to the context at days 2–4. No shocked animals are referred to control. Data are expressed as mean ± SEM. The total number of mice was 40, where each day a number of animals were sacrificed for IHC processing. Animals were randomly divided in control group and 4 experimental groups for sample collection. Control (n = 8), day 1 (n = 32), day 2 (n = 24), day 3 (n = 16) and day 4 (n = 8). ANOVA, F(4,82) = 13.2; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 with Dunnett's post-hoc test.

Figure 2.. Regulation of the contextual fear response by pharmacological modulation of M-type K+ channels. (a and c) Protocol of contextual fear conditioning upon single intraperitoneal (IP) drug injections of XE991, retigabine (RTG) and vehicle (sterile water). All animals underwent an adaptation training for injection handling, where animals received single IP injections of vehicle during 3 consecutive days, before the training session. In (a), animals received IP drug injection immediately after the training session. In (c), animals received IP drug injection 15 min before the training session. (b) Behavioral profile of the CFC when IP drug injection was carried out immediately after the foot shocks. Single IP injections were as follows: XE991 (1 mg/Kg), retigabine (4 mg/Kg) and vehicle in a total volume of 150 µL. No shock (n = 12 for each day from day 0 to day 4), for XE991, retigabine and vehicle groups (for each group n = 40 at day 0 and day 1, n = 30 at day 2, n = 20 at day 3 and n = 10 at day 4). Data are expressed as mean ± SEM. No statistical differences were found for any treatment in the shocked groups vs. vehicle (p > 0.05). Two-way ANOVA by treatments F(3, 589)= 40.16, p < 0.0001, and by days F(5, 589) = 35.63, p < 0.0001. Multiple comparisons by Fisher LSD's post-hoc test (vs. vehicle group): *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (d) Behavioral profile of the CFC when IP drug injection was carried out immediately before the training session. Single IP injections were as follows: XE991 (2 mg/Kg), retigabine (10 mg/Kg) and vehicle in a total volume of 150 µL. No shock (n = 12 for each day from day 0 to day 4), for XE991, retigabine and vehicle groups (each group n = 30 for each day from day 0 to day 4). Data are expressed as mean ± SEM. There is a statistical difference between the XE991 and the vehicle groups. Two-way ANOVA by treatments F(3, 554) = 26.3, p < 0.0001, and by days F(5, 554) = 41.01, p < 0.0001. Multiple comparisons by Fisher LSD's post-hoc test (vs. vehicle group): *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 3.. c-Fos expression in hippocampal neurons following contextual Fear conditioning. (a) c-Fos activation patterns during contextual fear conditioning in the CA1 and DG regions of the hippocampus. In the total CA1 neuronal population (blue dots), a significant increase of 27.6% in c-Fos expression was observed in day 1 (t-test t(12)=2.793, p= 0.0163), then 34.5% in day 2 (t-test t(11) = 2.686, p = 0.0212), which was followed by a return near control values during days 3 and 4 (n = 7–8; multiple t-test, p > 0.05). In the other hand, in the total DG neuronal population (pink dots), a significant increase in c-Fos expression of 38.8 % was observed in day 1 (t-test t(13) = 3.842, p = 0.0020), which then returned to control levels in days 3 and 4 (n = 7–8; multiple t-test, p > 0.05). *p < 0.05, **p<0.01, ***p < 0.001. (b) A proportion in the expression of c-Fos in vGAT and vGLUT cells in CA1 and DG region of the hippocampus is showed here. In CA1 region vGAT positive neurons, a significant increase of 29.05 % in c-Fos expression was observed at day 1, then 35.2 % at day 2, which was followed by a return to baseline levels at days 3 and 4 during contextual re-exposure. Meanwhile, c-Fos active VGLUT neurons do not change in proportion during any day in CA1 (n = 7–8; two-ways ANOVA F(1, 64) = 516.5, p < 0.0001). Furthermore, c-Fos activation patterns in the DG region of the hippocampus in vGAT positive neurons showed an expression significantly increased by 39.9 % at day 1, then by 24.8 % at day 2 and resumed back near control values during contextual re-exposure at days 3 and 4. Additionally, c-Fos active VGLUT neurons do not change in proportion any day in DG (n = 7–8; two-way ANOVA F (1, 64) = 590.3, p < 0.0001). Multiple comparisons by Fisher LSD's post-hoc test (vs. control group): *p < 0,05, **p < 0.01, ***p < 0.001. (c) Representative examples of c-Fos immunoreactive neurons of the DG and CA1 regions showing increased expression at day 1 versus control (no-shocked animals) and a subsequent decrease at day 4. (d) Representative examples of vGAT and c-Fos immunoreactive DG neurons showing increased expression at day 1 versus control (no-shocked animals). (e) Representative examples of vGAT and c-Fos immunoreactive CA1 neurons showing increased expression at day 1 versus control (no-shocked animals). The stars show c-Fos immunoreactive cells labeled with vGAT and the arrows show c-Fos immunoreactive cells not labeled with vGAT.

Figure 4.. Impact of the fear response on the axon initial segment of hippocampal neurons. (a) Distance from the soma of the AIS in CA1 and DG hippocampal GABAergic neurons immunolabeled with AnkG and vGAT antibodies during contextual fear conditioning. A significant increase in the distance Soma-AIS in CA1 was found in Day 1(n = 7–8; CA1: ANOVA F(4, 31) = 7.713, p = 0.0002, Dunnett's post-hoc test *p < 0.05, **p < 0.01; DG: ANOVA F(4, 32) = 2.873, p = 0.0386). (b) AIS length of CA1 and DG hippocampal GABAergic neurons immunolabeled with AnkG and vGAT antibodies during contextual fear conditioning. A significant longer AIS segment was observed in vGAT positive GABAergic neurons of the CA1 region at day 1 and day 2 (n = 7–8, ANOVA, F(4, 34) = 2.847, p = 0.0388, Dunnett's post-hoc test *p < 0.05, **p < 0.01). (c) Distance from the soma of the AIS in CA1 and DG hippocampal glutamatergic neurons immunolabeled with AnkG and vGluT1 antibodies during contextual fear conditioning (CA1: n = 7, ANOVA F(4, 29) = 0.3647, p=0.8317; DG: n = 7, ANOVA F(4, 28) = 1.639, p = 0.1923). (d) AIS length of CA1 and DG hippocampal glutamatergic neurons immunolabeled with AnkG and vGluT1 antibodies during contextual fear conditioning (CA1: n = 7–8, ANOVA F(4, 32) = 0.2256, p=0.9221; DG: n = 7–8, ANOVA F(4, 29) = 0.6468, p = 0.6336). (e) Representative examples of AIS in CA1 GABAergic neurons labeled with AnkG (blue color) and MAP2 (pink) antibodies in c-Fos positive cells. The arrows show the start and the endpoint of each AIS.