Nck1 activity in lateral amygdala regulates long-term fear memory formation

Abstract

Fear conditioning leads to long-term fear memory formation and is a model for studying fear-related psychopathological conditions such as phobias and post-traumatic stress disorder. Long-term fear memory formation is believed to involve alterations of synaptic efficacy mediated by changes in synaptic transmission and morphology in lateral amygdala (LA). Nck1 is a key neuronal adaptor protein involved in the regulation of the actin cytoskeleton and the neuronal processes believed to be involved in memory formation. However, the role of Nck1 in memory formation is not known. Here we explored the role of Nck1 in fear memory formation in lateral amygdala (LA). Reduction of Nck1 in excitatory neurons in LA enhanced long-term, but not short-term, auditory fear conditioning memory. Activation of Nck1, by using a photoactivatable Nck1 (PA-Nck1), during auditory fear conditioning in excitatory neurons in LA impaired long-term, but not short-term, fear memory. Activation of Nck1 immediately or a day after fear conditioning did not affect fear memory. The hippocampal-mediated contextual fear memory was not affected by the reduction or activation of Nck1 in LA. We show that Nck1 is localized to the presynapses in LA. Nck1 activation in LA excitatory neurons decreased the frequency of AMPA receptors-mediated miniature excitatory synaptic currents (mEPSCs). Nck1 activation did not affect GABA receptor-mediated inhibitory synaptic currents (mIPSCs). These results show that Nck1 activity in excitatory neurons in LA regulates glutamate release and sets the threshold for fear memory formation. Moreover, our research shows that Nck1 may serve as a target for pharmacological treatment of fear and anxiety disorders.

Fig. 1. Reduction of Nck1 in excitatory neurons in lateral amygdala impairs long-term memory formation.

A 21-mer short/small hairpin shRNA sequences that are selected specifically against the mouse Nck1 and are embedded in a human microRNA-30 (miR-E) were used in the study. B Nck1 shRNAs or control non-targeting (NT) shRNAs are expressed in the basolateral amygdala (BLA) as shown by the EGFP that is co-expressed with the shRNAs. C Nck1 shRNAs (n = 3) significantly reduces the level of Nck1 in the lateral amygdala when compared to NT shRNAs control (n = 3) (p = 0.048). A representative blot is shown. D F-actin level is reduced in Nck1 shRNA expressing neurons when compared to F-actin level in neurons expressing NT shRNA (t₍₁₀₎ = 2.517, p = 0.031). E Description of the long-term behavioral protocol. F Freezing during fear conditioning training is not different between the Nck1 shRNA and NT shRNAs groups (F_(1,19) = 0.556, p = 0.456). G Freezing during the long-term contextual fear memory test is not different between the Nck1 shRNAs and NT shRNAs groups (t_(0.523) = 0.19, p = 0.607). H Reduced Nck1 in LA (n = 10) enhanced long-term auditory fear memory formation in LA when compared to the control group (n = 11) (F_(1,19) = 4.492, p = 0.047). I Description of the short-term behavioral protocol. J Freezing throughout the tones presentations during training of animals used in the STM experiment that expressed Nck1 shRNAs or control NT shRNAs was analyzed. There is no significant difference between the groups (F_(1,12) = 0.576, p = 0.462). K Contextual fear conditioning tested 1 h after fear conditioning is not different between the Nck1 shRNAs (n = 7) and control NT shRNAs (n = 7) groups (t₍₁₂₎ = 1.456, p = 0.171). L Auditory fear conditioning tested 2 h after fear conditioning is not different between animals injected with shRNAs against Nck1 compared to NT shRNAs injected mice (F_(1,12) = 0.077, p = 0.786). Values are mean ± SEM. C, D Independent samples t-test; F–H and J–L Repeated measures ANOVA.

Fig. 2. Blue light stimulation of HEK293 cells and lateral amygdala neurons leads to the formation of PA-Nck1 and F-actin clusters.

A PA-Nck1 contains the 3 SH3 domains of Nck1 conjugated to Cry2olig and mCherry. Blue light stimulation leads to PA-Nck1 clustering. Studies in culture cells have shown that such clustering leads to actin polymerization. B HEK293 cells were transfected with PA-Nck1 under the control of the CMV promoter. Stimulation of the cells with light leads to a significant increase in PA-Nck1 clusters (t₍₂₂₎ = 3.367, p = 0.003) (normalized to fluorescence intensity) when compared with cells that were kept in a dark condition. Representative cells that were transfected with PA-Nck1 and activated by light or remained in the dark are shown. C PA-Nck1 is co-localized with F-actin in multiple spots (see arrows) in light stimulation HEK293 cells. F-actin clusters are increased (normalized to fluorescence intensity) significantly after light stimulation (n = 7) compared to cells stored in dark (n = 6) (t₍₁₁₎ = 2.367, p = 0.036). D Mice were microinjected with AAV expressing the PA-Nck1 under the control of the short CaMKII promoter into the lateral amygdala (LA) and stimulated with blue light through optic fibers. PA-Nck1 construct and light stimulating protocols are shown. Representative expression of PA-Nck1 in LA as observed by mCherry reporter protein expression is shown. E Stimulation of LA neurons with light leads to a significant increase (t₍₂₈₎ = 2.744, p = 0.01) in PA-Nck1 clusters (normalized to fluorescence intensity) in LA compare with neurons that were not stimulated with light. Representative LA neurons that express PA-Nck1 and are activated by light or remain in the dark are shown (single neurons for dark and light conditions are shown). F PA-Nck1 is co-localized with F-actin in multiple spots (see arrows) in the light-activated neurons in LA. Application of light to LA leads to a significant increase in PA-Nck1 and F-actin co-localization (t₍₁₈₎ = 2.158, p = 0.045). Values are mean ± SEM. B, C, E, F Independent samples t-test.

Fig. 3. Activation of Nck1 in LA during fear conditioning impaired auditory long-term, but not short-term, fear memory.

A Description of the behavioral protocol. Mice injected with AAV containing PA-Nck1 into LA were subjected to 3 tone-shock pairings. During each pairing, the animals received blue light (473 nm) illumination in LA to activate the PA-Nck1. Controls were animals injected with AAV, containing the PA-Nck1, into LA, implanted with optic fibers and subjected to fear conditioning but without light. Fear memory was tested 24 h and 48 h after training for contextual and auditory long-term fear conditioning memory (LTM), respectively. B Freezing during training in response to the tone CS is not different between light (n = 12) and no light (n = 13) animals (F_(1,21) = 0.138, p = 0.714). C Contextual fear conditioning tested 24 h after fear conditioning is not different between the light and no light groups (t₍₂₃₎ = 1.938, p = 0.065). D Auditory fear conditioning tested 48 h after fear conditioning is impaired in animals subjected to light activation of PA-Nck1 compared to no light mice (F_(1,21) = 10.369, p < 0.005). E Animals were subjected to light only (no PA-Nck1 expressed). F Freezing throughout the tones presentations during training of animals that were not injected with the virus but subjected to light (light only) or no light control (no light only) that were used in the LTM experiments was analyzed. There is no significant difference between light (n = 7) and no light (n = 7) groups (F_(1,12) = 0.109, p = 0.747). G Contextual fear conditioning tested 24 h after fear conditioning is not different between the light and no light groups (t₍₁₂₎ = 0.992, p = 0.341). H There is no significant difference between the animals that were subjected to light and not light in the long-term auditory fear conditioning memory test (F_(1,12) = 0.383, p = 0.548). I Schematic description of the protocol of behavior and light stimulation for the immediate post-training activation of PA-Nck1 experiment. J Freezing during training in response to the tone CS is not different between light (n = 6) and no light (n = 6) animals (F_(1,10) = 1.899, p = 0.198). K Contextual fear conditioning tested 24 h after fear conditioning is not different between the light and no light groups (t₍₁₀₎ = 0.601, p = 0.561). L Auditory fear conditioning tested 48 h after fear conditioning is not different between animals subjected to light activation of PA-Nck1 compared to no light mice (F_(1,10) = 0.026, p = 0.875). M Schematic description of the protocol of behavior and light stimulation a day after training. N Freezing throughout the tones presentations during training of animals used in the experiment. There is no significant difference between light and no light groups (F_(1,13) = 0.042, p = 0.841). O Contextual fear conditioning tested 1 day after light stimulation is not different between the light and no light groups (t₍₁₃₎ = 1.458, p = 0.169). P Auditory fear conditioning tested a day after light stimulation is not different between animals subjected to light activation of PA-Nck1 compared to no light mice (F_(1,13) = 0.097, p = 0.76). Q Schematic description of the protocol of behavior and light stimulation for the STM experiment. R Freezing throughout the tones presentations during training of animals used in the STM experiment subjected to light or no light control was analyzed. There is no significant difference between light and no light groups (F_(1,11) = 3.372, p = 0.093). S Contextual fear conditioning tested 1 h after fear conditioning is not different between the light and no light groups (t₍₁₁₎ = 1.631, p = 0.131). T Auditory fear conditioning tested 2 h after fear conditioning is not different between animals subjected to light activation of PA-Nck1 compared to no light mice (F_(1,11) = 0.673, p = 0.43). Values are mean ± SEM. C, G, K, O, S Independent samples t-test. B, D, F, H, J, L, N, P, R, T Repeated measures ANOVA.

Fig. 4. Nck1 is localized to presynapses in LA and its activation decreases the frequency of AMPAR-mediated mEPSCs.

A Lateral amygdala neurons were co-labeled with antibodies that bind to Nck1 and Bassoon (a presynaptic protein). Nck1 is colocalized with bassoon in presynapse. B Super-resolution microscopy reveals that Nck1 localization is almost perfectly aligned with Bassoon in presynapse. C PA-Nck1 is co-localized with Bassoon. Arrows show PA-Nck1, Bassoon and the colocalized PA-Nck1 and Bassoon. D Schematic description of the voltage patch-clamp preparation and mode of light stimulation. E AMPAR-mediated mEPSCs amplitude and frequency before lights stimulation (Baseline; black) and during the 3 light stimuli (blue) are shown. No significant differences in amplitude between the baseline and light stimuli were detected (F_(3,9) = 1.145, p = 0.382). Significant differences were detected in frequencies of mEPSCs between the baseline and light stimuli (F_(3,9) = 6.642, p = 0.012; n = 4). F Shining light alone on the lateral amygdala that does not express PA-Nck1 does not affect AMPA-mediated mEPSCs (n = 3) amplitude (F_{(1.181,2.362)} = 2.351, p = 0.254) and frequency (F_{(1.089,2.178)} = 0.804, p = 0.47). G GABAR-mediated mIPSCs amplitude and frequency before lights stimulation (Baseline; black) and during the 3 light stimuli (blue). No significant differences were detected (n = 4) (amplitude- F_(3,9) = 0.705, p = 0.573; frequency- F_(3,9) = 0.923, p = 0.468). Representative event traces of AMPAR-mediated mEPSCs and GABAR-mediated mIPSCs from neurons recorded in the LA are shown. H Lateral amygdala neurons were co-labeled with antibodies that bind to Nck1 and PSD-95. Nck1 is colocalized with PSD-95. Arrows show Nck1, PSD-95 and the colocalized Nck1 and PSD-95. Values are mean ± SEM. E–G Repeated measure ANOVA with LSD post hoc analysis.