Skip to main content
NCBI home page
Learning & Memory logoLink to Learning & Memory
. 2023 Oct;30(10):250–259. doi: 10.1101/lm.053833.123

Presence of a remote fear memory engram in the central amygdala

Robert J Hammack 1,2, Victoria E Fischer 1,3, Mary Ann Andrade 1, Glenn M Toney 1,2,✉
PMCID: PMC10561632  PMID: 37802546

Abstract

Fear memory formation and recall are highly regulated processes, with the central amygdala (CeA) contributing to fear memory-related behaviors. We recently reported that a remote fear memory engram is resident in the anterior basolateral amygdala (aBLA). However, the extent to which downstream neurons in the CeA participate in this engram is unknown. We tested the hypothesis that CeA neurons activated during fear memory formation are reactivated during remote memory retrieval such that a CeA engram participates in remote fear memory recall and its associated behavior. Using contextual fear conditioning in TRAP2;Ai14 mice, we identified, by persistent Cre-dependent tdTomato expression (i.e., “TRAPing”), CeA neurons that were c-fos-activated during memory formation. Twenty-one days later, we quantified neurons activated during remote memory recall using Fos immunohistochemistry. Dual labeling was used to identify the subpopulation of CeA neurons that was both activated during memory formation and reactivated during recall. Compared with their context-conditioned (no shock) controls, fear-conditioned (electric shock) mice (n = 5/group) exhibited more robust fear memory-related behavior (freezing) as well as larger populations of activated (tdTomato+) and reactivated (dual-labeled) CeA neurons. Most neurons in both groups were mainly located in the capsular CeA subdivision (CeAC). Notably, however, only the size of the TRAPed population distributed throughout the CeA was significantly correlated with time spent freezing during remote fear memory recall. Our findings indicate that fear memory formation robustly activates CeA neurons and that a subset located mainly in the CeAC may contribute to both remote fear memory storage/retrieval and the resulting fear-like behavior.

Creating associations is a fundamental brain function (Yuan et al. 2011). Formed associations pair contextual environmental cues with salient threats or rewards so that subsequent encounters with similar cues evoke the memory of the threat or reward and elicit appropriate behavioral responses (Kim and Jung 2006). Fear-related contextual memories are among the most widely studied forms of associative memory in rodents, as re-exposure to contextual cues produces predictable, gradable, and reproducible behavioral outputs, such as postural immobility (i.e., freezing) (Kim and Jung 2006). While classical studies used neuronal excitatory and inhibitory chemical stimuli in conjunction with brain ablations to identify brain regions participating in fear memory-related behaviors, these approaches were inadequate to precisely identify neuronal populations encoding fear memory and mediating attendant behavioral responses to memory recall (Burgos-Robles et al. 2009; Do-Monte et al. 2015; Kim and Cho 2017).

During the last decade, studies have advanced the concept of neuronal engrams as substrates of memory storage (Josselyn and Tonegawa 2020). Fear memory engram theory holds that perceived environmental threats activate ensembles of interconnected neurons whose activity induces persistent potentiating plasticity among synapses that encode memory storage. Heightened synaptic efficacy among fear memory engram neurons results in robust circuit reactivation upon re-exposure to contextual cues present during fear memory formation (Josselyn and Tonegawa 2020; Roy et al. 2022). More recent studies also suggest that, with time, neuronal ensembles participating in a memory engram reorganize, introducing the possibility that reactivation of the same neurons comprising the engram that formed during fear memory acquisition may not, in fact, result in fear memory recall, especially when recall occurs at more remote time points (Do-Monte et al. 2015; Grewe et al. 2017; DeNardo et al. 2019).

Previously, we reported that the anterior portion of the basolateral amygdala (aBLA)—an area strongly implicated in both recent and remote fear memory recall (Gale et al. 2004; Kitamura et al. 2017; Liu et al. 2022)—contains an ensemble of neurons reactivated at a remote time point, consistent with the aBLA participating in long-term fear memory storage (Hammack et al. 2023). Undetermined in that study were structures downstream from aBLA that couple memory formation with remote recall. Prior studies identified the central amygdala (CeA) as a major target of BLA output—one that contains anatomical subdivisions that mediate distinct aspects of fear learning (Kim et al. 2017). Specifically, the capsular division of the CeA (CeAC) has been implicated in negative valence fear-related behaviors, similar to the aBLA (Kim et al. 2016). In contrast, the medial and lateral CeA divisions (CeAML) contain functionally heterogeneous neuron populations associated with both positive and negative valence behaviors (Ciocchi et al. 2010; Li et al. 2013; Kim et al. 2017). We hypothesized that neuronal ensembles within CeA subdivisions are activated at the time of fear memory acquisition and reactivated during remote memory recall such that a CeA-inclusive fear memory engram is essential for remote fear memory recall.

We used adult offspring of second-generation targeted recombination in activated populations (TRAP2; Fos2A-iCreERT2) transgenic mice expressing a Cre-recombinase–estrogen receptor complex (iCreERT2) tied to induction of the immediate early gene c-fos (DeNardo et al. 2019) crossed with Cre-dependent tdTomato reporter (Ai14) mice (Madisen et al. 2010). Immediately following contextual fear conditioning or context-only conditioning, TRAP2:Ai14 mice were treated with short-acting 4-hydroxytamoxifen (4-OHT) to enable later identification of those CeA neurons persistently expressing tdTomato as a result of conditioning-induced c-fos activation; that is, the “TRAPed” population. Three weeks later, mice were re-exposed to the same context cues as during conditioning to test for remote fear memory recall. CeA neurons that underwent c-fos activation during recall testing were identified by Fos immunoreactivity (ir). Those neurons that were both activated by conditioning and reactivated during remote recall (tdTomato+ and Fos-ir; i.e., dual-labeled) were quantified as the CeA remote fear memory engram population. Like the aBLA (Hammack et al. 2023), our findings support the presence of a remote fear memory engram resident mainly in the CeAC subdivision. Unlike the aBLA, we found that freezing behavior during remote memory recall was positively correlated with the size of the fear memory acquisition ensemble (tdTomato+) but not that of Fos-ir recall ensemble. Collectively, our findings implicate the CeA and its capsular subdivision in fear-related memory processing but not memory storage per se. Our findings further raise the possibility that CeA neurons couple upstream fear memory neurons to circuitry that grades behavioral manifestations of fear in proportion to the strength of fear memory.

Results

Fear memory formation activates a CeA Fos neuronal ensemble

To investigate CeA neuronal ensemble activation during fear memory formation as well as activation and reactivation during remote memory recall, mice underwent fear (n = 5) or context (n = 5) conditioning followed by injection of 4-OHT (Fig. 1A,B). During memory acquisition, learning was observed in the fear but not the context group. Two-way ANOVA revealed a significant interaction between treatment (shock vs. no shock) and time bin (30 sec after each shock) on total time spent freezing (two-way ANOVA, F(5,40) = 6.733, P = 0.0001) (Fig. 1C). Fear-conditioned mice showed increased freezing time relative to baseline (4 sec ± 2 sec) after shock 3 (10 sec ± 3 sec, P = 0.0203), shock 4 (18 sec ± 3 sec, P < 0.0001), and shock 5 (15 sec ± 1 sec, P < 0.0001) and relative to context-conditioned mice after shock 4 (7 sec ± 1 sec, P = 0.0056) and shock 5 (4 sec ± 1 sec, P = 0.0044). Twenty-one days later, mice were re-exposed to the same contextual conditioning chamber for remote memory recall. Previously fear-conditioned mice spent significantly greater time freezing than context-conditioned mice (context: 75 sec ± 10 sec; fear: 261 sec ± 14 sec; unpaired t-test, t(8) = 11.16, P < 0.0001) (Fig. 1D). Collectively, our contextual fear conditioning protocol resulted in robust fear memory acquisition and remote fear memory recall.

Figure 1.

Figure 1.

Open in a new tab

Contextual fear conditioning promotes robust remote fear memory. (A) Mouse breeding strategy. (B) Experimental protocol: Immediately following conditioning, all mice received 100 mg/kg 4-OHT i.p. to “TRAP” the CeA neuronal Fos ensemble (tdTomato+) activated during conditioning and 21 d later were perfused 90 min after the remote memory recall session to capture the Fos neuronal ensemble activated during recall using Fos immunostaining. (C) Fear conditioning (shocks [blue circles]; n = 5) resulted in fear acquisition relative to baseline and relative to context-conditioned mice (no shocks [open squares]; n = 5). Values represent total time spent freezing during a 30-sec time bin immediately after each shock or dummy shock. (*) P = 0.0203, (****) P < 0.0001 versus shock group baseline (BL); (##) P = 0.0056, (#) P = 0.0044 between groups. Two-way ANOVA with Sidak's posttest. (D) Remote memory recall at day 21 after conditioning. (####) P < 0.0001. Unpaired t-test. Data are mean ± SEM. Biorender.com was used to generate mouse images.

Counts of c-fos-activated CeA neurons expressing tdTomato (TRAPed population) was significantly greater (t(8) = 8.224, P < 0.00001, unpaired t-test) in fear-conditioned (200 ± 13) than in context-conditioned (77 ± 7) mice (n = 5/group) (Table 1). TRAPed counts in both groups were dominantly located in the middle and caudal CeA subregions, with counts, especially in the fear group, being most numerous in the CeAC subdivision relative to the CeAML (Fig. 2A–C). Similar to our recent report (Hammack et al. 2023), we compared TRAPed count density across rostro–caudal subregions within groups as well as within each subregion between fear-conditioned and context-conditioned groups (Fig. 2D). Analysis yielded the main effects of treatment (F(1,24) = 80.76, P < 0.0001) and subregion (F(2,24) = 12.92, P = 0.0002), with count density being greater in fear-conditioned mice than in context-conditioned mice in all three rostro–caudal subregions (rostral: 177 ± 21 vs. 59 ± 8, P = 0.0027; middle: 283 ± 30 vs. 139 ± 20, P = 0.0003; caudal: 106 ± 16 vs. 330 ± 23, P < 0.0001). Among fear-conditioned mice, TRAPed count density was greater in the middle (P = 0.0070) and caudal (P = 0.0002) subregions compared with the rostral subregion, while context-conditioned mice similarly demonstrated greater TRAPed density in the middle than in the rostral subregion (P = 0.0484).

Table 1.

Quantification of CeA Fos ensembles

graphic file with name LM053833HAMTB1.jpg

Open in a new tab

Figure 2.

Figure 2.

Open in a new tab

Contextual fear conditioning activates a larger CeA fos ensemble than context conditioning. (A–C) Stereotaxic plate drawings of CeA rostro–caudal subregions (left; gray), along with representative examples of TRAPed neurons and corresponding summary distribution plots from context-TRAPed (middle) and fear-TRAPed (right) mice (n = 5/group) in the rostral (A), middle (B), and caudal (C) CeA subregions. Scale bars (context images: middle, left), 200 µm. (D) Average density of tdTomato+ counts in each CeA rostro–caudal subregion. (##) P = 0.0027, (###) P = 0.0003, (####) P < 0.0001 in context-TRAPed (squares) versus fear-TRAPed (circles) mice; (*) P = 0.0484 context group between subregions; (**) P = 0.0070, (***) P = 0.0002 fear group between subregions. Two-way ANOVA with Sidak's posttest. Data are mean ± SEM. (E) Average tdTomato+ count density in the rostral, middle, and caudal subregions of the CeAC and CeAML in context-TRAPed (pink) and fear-TRAPed (red) mice. (#) P = 0.0137, (####) P < 0.0001 between groups; (****) P < 0.0001 fear group between subdivision; (ϕϕϕϕ) P < 0.0001 fear group between rostro–caudal subregions. Three-way ANOVA with Tukey's posttest. Data are mean ± SEM. (F) Correlation between the amount of time spent freezing during memory recall and tdTomato+ counts in context (squares) and fear (circles) groups in the entire CeA (left), CeAC (middle), and CeAML (right). Least squares regression lines are shown along with corresponding Pearson's coefficients (r). Fisher's z-transformation was applied to compute r′ values and thereby compare the strength of correlations between groups for the entire CeA and its CeAC and CeAML subdivisions (see Supplemental Table S2).

We then evaluated count densities specifically within the CeAC and CeAML subdivisions in rostral, middle, and caudal subregions (Fig. 2E). Three-way ANOVA revealed significant effects of treatment (F(1,48) = 111.7, P < 0.0001), CeA subdivision (F(1,48) = 31.47, P < 0.0001), and rostro–caudal subregion (F(2,48) = 18.66, P < 0.0001). Between-group comparisons revealed within the CeAC that TRAPed count density was greater in fear-conditioned mice than in context-conditioned mice in both the middle (fear: 385 ± 49; context: 180 ± 30; P < 0.0001) and caudal (fear: 429 ± 36; context: 128 ± 20; P < 0.0001) subregions. In the CeAML, count density in the caudal subregion was likewise greater in fear-conditioned mice than in context-conditioned mice (fear: 230 ± 27; context: 85 ± 14; P = 0.0137). In the fear group, subdivision comparisons revealed greater TRAPed count density in the CeAC than in the CeAML in the middle (CeAC: 385 ± 49; CeAML: 180 ± 21; P < 0.0001) and caudal (CeAC: 429 ± 36; CeAML: 230 ± 27; P < 0.0001) subregions. In the CeAC of the fear-conditioned group, the comparison revealed that TRAPed count densities were greater in the middle (P < 0.0001) and caudal (P < 0.0001) subregions than in the rostral subregion (164 ± 20) (Fig. 2E; Supplemental Table S1). We also assessed whether TRAPed counts in the entire CeA and specifically within the CeAC and CeAML subdivisions as well as within their rostral–caudal subregions correlated with behavior during memory recall (Fig. 2F; Supplemental Table S2; Supplemental Fig. S1A). In the fear-conditioned group, freezing behavior during remote memory recall was significantly correlated with TRAPed counts in the entire CeA (Pearson's r = 0.8877, P = 0.0444) and in the caudal subregion of the CeAC (Pearson's r = 0.9255, P = 0.0241).

Collectively, we observed a larger TRAPed ensemble in the entire CeA of fear-conditioned mice than in mice exposed to context alone. The size of the CeA ensemble reliably predicted freezing behavior during remote memory recall. Furthermore, the TRAPed ensemble in fear-conditioned mice was largely localized within the CeAC, with counts in the caudal CeAC reliably predicting freezing behavior during remote fear memory recall.

Remote fear memory recall induces a neuronal ensemble in the CeA

To investigate neuronal ensemble activation during remote memory recall, we first compared Fos-ir counts between fear and context groups within the entire CeA and found that fear memory recall recruited a larger Fos-ir ensemble than contextual memory recall (fear: 95 ± 13; context: 60 ± 7; t(8) = 2.313, P = 0.0494) (Table 1). We next evaluated the rostral, middle, and caudal CeA subregions (Fig. 3A–D). A treatment effect on count density was detected (F(1,24) = 6.845, P = 0.0151, two-way ANOVA) but with no significant pairwise comparisons (Tukey post hoc test). We subsequently evaluated CeAC and CeAML rostro–caudal subdivisions (Fig. 3E) by three-way ANOVA, which similarly revealed a significant treatment effect on count density (F(1,48) = 8.402, P = 0.0056) but, again, with no significant pairwise comparisons (Fig. 3E; Supplemental Table S1). We next evaluated the correlation between freezing behavior during memory recall and Fos-ir counts in the entire CeA and separately in the CeAC and CeAML subdivisions (Fig. 3F; Supplemental Table S2). No between-group or within-group correlations were found.

Figure 3.

Figure 3.

Open in a new tab

Remote fear memory recall activates a Fos ensemble uniformly distributed across CeA subdivisions. (A–C) Stereotaxic plate drawings of CeA rostro–caudal subregions (gray; left) along with representative examples of Fos-ir and corresponding summary distribution plots from context-TRAPed (middle) and fear-TRAPed (right) mice (n = 5/group) in the rostral (A), middle (B), and caudal (C) CeA subregions. Scale bars (context images: middle, left), 200 µm. (D) Average Fos-ir count density in each subregion. Two-way ANOVA with Sidak's posttest revealed no significant pairwise comparisons. Data are mean ± SEM. (E) Average Fos-ir count density in the rostral, middle, and caudal subregions of the CeAC and CeAML in context-TRAPed (light green) and fear-TRAPed (green) mice. Three-way ANOVA with Tukey's posttest revealed no significant pairwise comparisons. Data are mean ± SEM. (F) Correlation between the amount of time spent freezing during memory recall and Fos-ir counts in context (squares) and fear (circles) groups in the entire CeA (left), CeAC (middle), and CeAML (right). Least squares regression lines are shown along with corresponding Pearson's coefficients (r). Fisher's z-transformation was applied to compute r′ values comparing the strength of correlations between groups for the entire CeA and its CeAC and CeAML subdivisions (see Supplemental Table S2).

Collectively, the results indicate that the counts of the CeA remote memory recall Fos-ir ensemble differs between the fear-conditioned and context-conditioned groups, but this ensemble does not preferentially localize to a specific rostro–caudal CeA subregion or to the capsular or medio–lateral CeA subdivision. Notably, remote memory recall ensemble sizes were not correlated with recall-related freezing behavior.

Remote fear memory recall induces a CeA activation–reactivation engram

Analysis revealed that counts of reactivated (dual-labeled) neurons was greater in the remote fear memory recall (17 ± 4) than in the remote context memory recall (3 ± 1; t(8) = 3.768, P = 0.0055) group (Table 1). Whereas the percentage of reactivated neurons between groups was not different (fear: 9% ± 2%; context: 4% ± 1%; t(8) = 2.105, P = 0.0684), reactivated counts comprised a larger percentage of the recall-activated (Fos-ir) population in the fear (18% ± 2%) than in the context (5% ± 1%; t(8) = 4.855, P = 0.0013) group. Evaluation of CeA rostro–caudal subregions revealed a treatment effect on count density (F(1,24) = 25.34, P < 0.0001) (Fig. 4A–D). Specifically, the reactivated ensemble count density was greater during memory recall in fear-conditioned than in context-conditioned mice in the middle subregion only (33 ± 9 vs. 6 ± 3; P = 0.0007). Further analysis of reactivated CeA ensemble count density by treatment, subdivision, and rostro–caudal subregion by three-way ANOVA detected significant effects only for fear versus context treatment (F(1,48) = 24.73, P < 0.0001) and CeA subdivision (F(1,48) = 9.071, P = 0.0041) (Fig. 4E). Specifically, the density of reactivated neurons in the middle subregion of the CeAC in the fear group (52 ± 17) was significantly greater than in the context group (6 ± 3; P = 0.0030). Similarly, within-fear-group analysis showed greater reactivated ensemble density in the middle subregion of the CeAC than in the CeAML subdivision (CeAC: 52 ± 17; CeAML: 13 ± 3; P = 0.0003). Moreover, the remote memory engram was larger in the middle than in the rostral subregion of the CeAC of fear-conditioned mice (middle: 52 ± 17; rostral: 18 ± 7; P = 0.0200) (Fig. 4E; Supplemental Table S1). We subsequently investigated whether reactivated ensemble counts or percentage of neurons reactivated in the entire CeA as well as in the CeAC and CeAML subdivisions correlated with freezing behavior during remote memory recall (Fig. 4F,G; Supplemental Table S2; Supplemental Fig. S1B). As with memory recall ensembles (Fos-ir neurons), no between-group or within-group correlations were detected.

Figure 4.

Figure 4.

Open in a new tab

Remote fear memory recall reactivates a larger CeA Fos engram than contextual memory recall. (A–C) Stereotaxic plate drawings of CeA rostro–caudal subregions (gray; left), along with representative examples of reactivated neurons (tdTomato+ and Fos-ir; i.e., dual-labeled) and corresponding summary distribution plots from context-TRAPed (middle) and fear-TRAPed (right) mice (n = 5 group) in the rostral (A), middle (B), and caudal (C) CeA subregions. Scale bars (context images: middle, left), 200 µm. (D) Average density of reactivated counts in each CeA subregion in context-TRAPed (squares) and fear-TRAPed (circles) mice. (###) P = 0.0007 context versus fear group. Two-way ANOVA with Sidak's posttest. (E) Average density of reactivation counts in rostral, middle, and caudal subregions of the CeAC and CeAML in context-TRAPed (white) and fear-TRAPed (yellow) mice. (**) P = 0.0030 fear group, CeAC versus CeAML; (###) P = 0.0003 CeAC, context versus fear group; (ϕ) P = 0.0200 fear group between rostro–caudal subregions. Three-way ANOVA with Tukey's posttest. Data are mean ± SEM. (F) Correlation between the amount of time spent freezing during memory recall and dual-labeled counts in context (squares) and fear (circles) groups in the entire CeA (left), CeAC (middle), and CeAML (right). (G) Correlation between the amount of time spent freezing during memory recall and the percent of tdTomato+ or Fos+ neurons that are dual-labeled in context (squares) and fear (circles) groups in the CeA. Least squares regression lines are shown along with corresponding Pearson's coefficients (r). Fisher's z-transformation was applied to compute r′ values comparing the strength of correlations between groups for the entire CeA and its CeAC and CeAML subdivisions (see Supplemental Table S2).

Collectively, we found greater neuronal reactivation in the CeAC subdivision of CeA during remote fear memory recall than during context memory recall and that the fear memory-reactivated CeAC population comprises a greater fraction of the entire recall-activated population. However, freezing behavior did not correlate with either the CeAC counts or the proportion of CeAC neurons reactivated during remote memory recall. Our findings suggest that remote memory-related fear behavior does not depend on the size of the CeA activation–reactivation engram or on the percentage of the activated ensembles (during either memory formation or recall) that is reactivated during recall. Hence, the CeA engram population appears to play a role in remote fear processing that is largely unrelated to long-term memory storage/recall.

Discussion

Using TRAP2:Ai14 mice, we investigated the participation of CeA neurons in a remote fear memory engram. Compared with context-conditioned mice, fear-conditioned mice showed greater CeA Fos activation during memory acquisition (tdTomato+ neurons) and, to a lesser extent, during remote memory recall (Fos-ir neurons). The fear memory acquisition ensemble was most densely localized in the middle and caudal portions of the CeAC subdivision. Notably, a relatively small subpopulation of mostly CeAC neurons that underwent Fos activation during fear memory acquisition was reactivated during remote fear memory recall (i.e., dual-labeled), possibly indicating their participation in a remote fear memory engram. Freezing behavior during remote memory recall correlated with the size of the fear memory formation (fear-TRAPed and tdTomato+) ensemble distributed across the entire CeA and that localized specifically within the caudal CeAC. Our findings collectively support a role for CeA Fos ensembles in fear memory formation and/or consolidation and suggest that a small population of CeA neurons, especially within the CeAC, might also participate in remote fear memory storage and recall.

Fear memory acquisition induces a Fos ensemble in the CeA

As a major output region of the amygdala, the CeA plays important roles in fear memory (Hitchcock and Davis 1991; Campeau and Davis 1995; Ciocchi et al. 2010). Investigations of the CeA, which is comprised of at least three functionally distinct subdivisions (medial, lateral, and capsular), have revealed some of its functional roles in fear-related behaviors (Ciocchi et al. 2010; Li et al. 2013), including induction of synaptic plasticity among output projections that contribute to multiple fear learning-associated processes ranging from memory acquisition (Li et al. 2013; Penzo et al. 2014) to memory expression (Ciocchi et al. 2010; Kim et al. 2016) and extinction (Whittle et al. 2021). The CeAM contains the majority of CeA output neurons, which are mostly GABAergic and exhibit significant neurochemical diversity (Kim et al. 2017; Wang et al. 2023). The CeAL is similar in that neurons are mostly GABAergic and exhibit overlapping neurochemical diversity with the CeAM (Kim et al. 2017; Wang et al. 2023). However, the CeAL is comprised largely of interneurons that inhibit CeAM output (Ciocchi et al. 2010; Keifer et al. 2015). Notable targets of CeAM efferents include the lateral and paraventricular hypothalamic nuclei (Prewitt and Herman 1998), nucleus tractus solitarius (Hopkins and Holstege 1978), rostral ventrolateral medulla (Hopkins and Holstege 1978), and periaqueductal gray (Johansen et al. 2010; Kim et al. 2013), which contribute to fear memory-related behavioral and autonomic responses (Ciocchi et al. 2010; Johansen et al. 2010; Kim et al. 2013, 2017; Li et al. 2013).

While the CeAM and CeAL are well studied, comparatively little is known about the involvement of CeAC neurons in fear memory processing. This likely reflects limited information on their neurochemical complexity and connectivity with known fear memory-encoding neurons (Kim et al. 2017; Wang et al. 2023). Here, neurons in the middle and caudal subregions of the CeAC were strongly activated during contextual fear conditioning but far less so during fear memory recall (Fig. 2). This is consistent with literature evidence that CeAC is strongly activated by inputs responsive to aversive sensory stimuli and projects heavily to the substantia innominata and other limbic regions linked to the generation of aversive behaviors (Cassell et al. 1999; Bourgeais et al. 2001; Cui et al. 2017). Although the notably greater CeAC activation by fear conditioning than by remote fear memory recall points strongly toward CeAC involvement in the processing of sensory inputs activated by unconditioned stimuli (i.e., footshocks), we cannot rule out a population of CeAC neurons participating in the remote fear memory engram that mediates freezing behavior during memory recall.

To our knowledge, this study is the first to identify a subpopulation of CeAC neurons possibly contributing to a remote fear memory engram. Although this finding requires functional confirmation, our use of TRAP2 mice and testing for remote, not recent, fear memory recall was likely pivotal in revealing this potential aspect of their function. Noteworthy is that the rostro–caudal distribution of fear conditioning-TRAPed CeAC neurons was strikingly similar to that of their counterparts in the adjacent aBLA, as we previously reported (Hammack et al. 2023). This raises the possibility that an aBLA-to-CeAC microcircuit previously implicated in negative valence signaling (Kim et al. 2016) might function in remote fear memory processing.

Advantages and limitations of the TRAP2 system

Here, our use of TRAP2 mice may have uncovered populational recruitment in the CeA that could not have been discovered using traditional Fos immunohistochemistry (IHC). In contrast to Fos IHC (Chowdhury and Caroni 2018), TRAP2 allows for quantification of Fos-activated neurons over a longer period after an inducing stimulus (6 h vs. ∼90 min) (DeNardo et al. 2019). As a result, neurons TRAPed by fear conditioning likely represent not only those acutely activated during memory acquisition, similar to Fos IHC, but also those activated later during the critical period of memory consolidation (Chowdhury and Caroni 2018; DeNardo et al. 2019). Indeed, fear conditioning-induced c-fos induction in the CeA is consistent with reports of CeA participation in memory consolidation associated with fear-induced synaptic plasticity (Li et al. 2013; Penzo et al. 2014). Fos-driven AP1 promoter activation is closely associated with signaling systems and protein turnover essential for synaptic plasticity (Yap and Greenberg 2018), and blocking protein synthesis specifically in the CeA during memory consolidation impairs recall of recently formed fear memory as indexed by decreased freezing behavior (Wilensky et al. 2006). Moreover, select neuronal populations within the CeA undergo fear-induced plasticity (Samson et al. 2005; Li et al. 2013; Penzo et al. 2014). Combined, these findings suggest that Fos expression may be a key factor inducing the synaptic plasticity essential for memory consolidation immediately following fear memory acquisition. In this regard, greater TRAPed counts in the CeA could reflect a heterogeneous population consisting of neurons involved in negative valence freezing behavior, memory consolidation processing, and possibly other fear memory-associated outcomes, such as autonomic activation (Hopkins and Holstege 1978; Prewitt and Herman 1998).

A remote fear memory engram in the CeA

Activation

Similar to the fear memory acquisition tdTomato+ Fos ensemble, the CeA of fear-conditioned mice did have a larger Fos-ir ensemble at the time of remote memory recall compared with context-conditioned mice; however, ensemble activation was not localized to a specific CeA subdivision or rostro–caudal subregion. This discrepancy in ensemble location might indicate functionally distinct roles for CeA neurons in fear memory acquisition/consolidation versus remote memory recall. Importantly, lack of recall ensemble localization to the CeAC suggests that putative negative valence-encoding aBLA–CeAC circuitry does not mediate memory-driven freezing behavior at a remote time point. This interpretation does not necessarily conflict with evidence for CeA participation in fear expression during recent memory recall but instead could suggest that inputs to CeA from brain regions other than the BLA, possibly from the paraventricular nucleus of the thalamus (Wilensky et al. 2006; Zimmerman et al. 2007; Do Monte et al. 2016), participate in remote fear memory. Therefore, while the CeA Fos ensemble during fear learning (tdTomato+) may contribute broadly to synaptic plasticity underlying fear memory acquisition/consolidation, this neuronal population likely plays a distinctly different role during remote memory recall. Concerning the latter, the present findings cannot rule out the possibility that CeA Fos ensemble neurons identified during remote memory recall participate in fear memory extinction learning rather than remote memory recall (Whittle et al. 2021).

Reactivation

Reactivation of the CeA during fear recall is not well established. Whereas at least some degree of CeAC reactivation during recent fear memory recall has been reported (Roy et al. 2022), to our knowledge reactivation stability over time has not been investigated. In our study, a larger reactivated (dual-labeled) ensemble was observed during remote fear memory recall than during context recall. Like the memory acquisition ensemble, the reactivation engram population was located dominantly within the CeAC subdivision. The relatively small size of the reactivation ensemble compared with either the acquisition or recall ensemble suggests that remote fear memory-related freezing behavior may rely largely on neurons outside the CeA, but reactivation of a small kernel of CeAC neurons within an extended negative valence circuit, possibly arising from the aBLA (Kim et al. 2017), could possibly contribute. Notable, however, is our finding that the percentage of reactivated CeA neurons that were previously activated during conditioning was not different between fear-conditioned and context-conditioned groups. Importantly, our reactivated population comprised a greater percentage of the recall-activated population in the fear group than in the context group. This suggests that although a small remote fear memory engram may reside within the CeA, and specifically within the CeAC, reactivation during remote recall testing could be unrelated to fear memory storage per se (Yap and Greenberg 2018). Functional studies are needed to determine the contribution that this reactivated population makes to fear memory-related behavior at a remote time point.

Behavioral correlates

Of particular interest is our finding that time spent freezing during memory recall was correlated with the size of the fear memory acquisition ensemble (tdTomato+ neurons) quantified across all analyzed CeA subregions and the caudal subregion of the CeAC. This suggests that spatial recruitment of the CeA during fear learning as well as subregional recruitment predict fear memory-related behavior at a remote time point. Moreover, to the extent that more widely distributed CeA neurons are functionally important for fear memory, more robust spatial recruitment of the CeA during fear memory formation per se could participate in “grading” perceived threat salience/imminence. If so, it will be important to establish the mechanisms of synaptic plasticity among activated neurons and the extent to which the types of plasticity occurring among these neurons predict their inclusion in the engram (reactivated) responsible for/participating in remote memory recall. It may be that increased recruitment of CeA during fear learning/memory acquisition encodes more enduring memory via enhanced signal propagation to downstream circuits that results in greater freezing behavior during recall testing. This may be especially important for recall at a remote time point. Notably, we previously reported, using a similar experimental approach, that Fos ensemble sizes in the aBLA (whether activated during memory acquisition, remote recall, or both) did not correlate with remote fear memory-related freezing behavior (Hammack et al. 2023). This suggests that the initial activation of the aBLA, in contrast to the CeA, encodes fear memory more through the induced state of electrophysiological plasticity than through the size of the recruited neuron population. In this regard, aBLA neurons upstream of the CeA may represent the computational engine of fear memory that determines the extent of CeA recruitment and thereby grades the intensity of behavioral manifestations of fear in proportion to the strength of remotely recalled fear memories. Furthermore, ensemble recruitment size within the CeA may be a common feature of intensity grading of negative valence encoding in general. While our fear conditioning paradigm results in a robust passive fear response (i.e., freezing), other test paradigms that favor an active (avoidance) over a passive (freezing) coping strategy (Salah et al. 2021) might also recruit a CeA ensemble in proportion to threat salience/imminence as a means of proportionally grading the behavioral response.

Conclusions

Using the TRAP2 (fos-iCre-ERT2) mouse model, we identified a relatively small fear memory engram population in the CeA that resides mostly in the CeAC subdivision. Together with our earlier report of an even larger engram population upstream in the aBLA (Kim et al. 2017), the present findings raise the possibility that a serially connected circuit extends from the aBLA to the CeA that encodes remote fear memory. The present findings make clear that the aBLA and CeA do not respond to fear conditioning with the same temporal patterns of engram activation. Whereas we previously reported that the strength of remote fear memory does not correlate with the size of aBLA fos ensembles, here we found that the size of the CeA fear conditioning ensemble specifically does correlate with the strength of remote fear memory. To the extent that ensemble neurons in the aBLA synaptically drive those downstream in the CeA, a possibility that requires further study, our findings suggest that serial communication between the aBLA and CeA couples the strength of remote fear memory with the intensity of behavior response. It is worth cautioning that functionally relevant communication between fear memory-related CeA neurons and their antecedents lying synaptically upstream in the aBLA (or elsewhere) may occur only among a relatively small fraction of fear memory encoding neurons. To the extent that this is indeed the case, the resolution of the TRAP2/Fos-ir approach might be insufficient to fully detect CeA engrams and their recruitment during remote memory recall such that engram neurons may be generally underestimated when only c-fos activation is used to index their population sizes.

Materials and Methods

Ethical approval

The Institutional Animal Care and Use Committee of the University of Texas Health San Antonio approved all experimental procedures, which conformed to the National Research Council Guide for the Care and Use of Laboratory Animals.

Animals

We crossed homozygous Fos2A-iCreERT2 knock-in mice (TRAP2; Jackson Laboratory 030323) with homozygous R26Ai14/ mice+ (Ai14; Jackson Laboratory 007914;) to produce male hemizygous TRAP2;Ai14 offspring. At P21, mice were group-housed in a 14:10-h light:dark cycle (lights on at 6:00 a.m.) in temperature-controlled (24°C) vivarium plastic cages (29 × 18 × 13 cm) containing rodent bedding (Sani-chips, Harlan Teklad). Food and water were accessible ad libitum. Experiments were conducted once mice turned 3 mo old.

Behavioral habituation

As previously described (Hammack et al. 2023), mice were habituated for five consecutive days immediately prior to conditioning. Briefly, each habituation session consisted of a 3-min exposure to the conditioning chamber containing the following contextual cues: patterned background and visible light (visual cue), 70% ethanol (olfactory cue), and metal grid floor (haptic cue). All mice were handled for 1 min prior to each session and “scruffed” for 10 sec following each session to mimic the handling that occurred during intraperitoneal (i.p.) injections.

Contextual fear conditioning and memory recall

Habituated mice were randomly assigned to undergo either fear conditioning (fear group) or context conditioning (context group). On the day of conditioning (day 0), fear group mice were allowed to explore the chamber for 2 min before receiving a series of five footshocks. Each shock was at an intensity of 0.75 mA and duration of 1 sec. Shocks were delivered at unpredictable intervals, but intervals were consistent across mice. Context-conditioned mice underwent the same protocol without delivery of footshocks. Immediately (<1 min) after completion of their conditioning protocol, all mice were given a systemic injection of 100 mg/kg 4-OHT i.p., prepared as previously described (DeNardo et al. 2019). On day 21 after conditioning, mice were re-exposed to the same conditioning chamber for remote memory recall. Postural freezing behavior was quantified as mice explored the chamber for a total of 5 min. Thereafter, mice were returned to their home cages.

Brain fixation and histology

Ninety minutes after the remote memory recall session, mice were deeply anesthetized (5% isoflurane in O2) and underwent transcardiac perfusion with 30 mL of 100 U/mL heparin in 0.01 M phosphate-buffered saline (PBS) followed by 100 mL of 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer. Brains were removed, postfixed for 6 h in 4% PFA at room temperature (∼22°C), and then placed in 30% sucrose–PBS for at least 2 d at 4°C. Next, 30-µm-thick coronal sections were cut on a freezing microtome (Leica Microsystems) and stored in polyvinylpyrrolidone (PVP) cryoprotectant at −20°C.

Immunohistochemistry

Immunostaining was performed as previously described (Mitchell et al. 2018; Maruyama et al. 2019; Hammack et al. 2023). Briefly, brain sections containing the CeA were washed in PBS for 20 min and incubated for 30 min in PBS containing sodium borohydride (0.5%) to remove any autofluorescent aldehydes generated during fixation. Following additional PBS washes, slices were incubated in blocking solution (3% goat serum, 0.05% Triton X-100 in PBS) for 2 h at ∼22°C followed by 72-h incubation at 4°C with c-Fos polyclonal antibody (1:1500; Synaptic Systems 226 003). After additional PBS washes, sections were incubated with a biotinylated goat antirabbit IgG secondary antibody (1:250; EMD Millipore AP132B) for 2 h at ∼22°C followed by washes with 0.05 M tris-buffered saline (TBS) and 0.1 M sodium acetate. Last, sections were incubated with streptavidin-Alexa fluor 488 (1:250; Invitrogen S11223) for 1 h in TBS-based blocking solution (3% goat serum, 0.05% Triton X-100 in TBS) and mounted on slides with Fluoromount-G (Invitrogen 00-4958-02).

Imaging and analysis

Images of CeA were captured as previously described (Hammack et al. 2023). Briefly, a Ziess LSM710 laser scanning confocal microscope with a 20× objective (NA 0.8) and a 47.5-μm pinhole size was used. Separate pixel maps of tdTomato (TRAPed; red in the figures) and Alexa-488 (Fos; green in the figures) were generated with ImageJ software (National Institutes of Health), and background fluorescence was subtracted by comparing experimental sections processed without Fos primary antibody (Mitchell et al. 2018; Maruyama et al. 2019). Counts were analyzed using the ImageJ plug-in EzColocalization (Stauffer et al. 2018).

Next, an outline of the CeA was produced at rostral–caudal subregions coextant with the aBLA and stored in ImageJ using the region of interest (ROI) manager. Generated ROIs were applied separately to images of the same rostral–caudal subregion from animals in the fear and context groups. Distribution scatter plots of CeA neurons singly labeled with tdTomato+ (conditioning-activated; i.e., TRAPed) or Fos immunoreactivity (recall-activated) were generated. Counts were obtained from counting numbers assigned by software to each identified CeA neuron. The numbers of images used to generate cumulative scatter plots for each treatment group were as follows: eight contexts and five fears for the rostral subregion, eight contexts and seven fears for the middle subregion, and nine contexts and seven fears for the caudal subregion. To correct for differences in the number of subregion images across groups, a correction factor of 0.62 (five out of eight), 0.88 (seven out of eight), and 0.77 (seven out of nine) was applied to the context group cumulative distribution plots for the rostral, middle, and caudal subregions, respectively. To avoid bias when excluding counts in scatter plots of the context group, Microsoft Excel was used to generate random numbers in the range of software-generated cell counts. Complete analysis included dividing the CeA into two subdivisions: the capsular subdivision (CeAC), defined as ≤170 µm medial to the external capsule, and the medial/lateral subdivision (CeAML). Counts between the rostro–caudal subregions and medio–lateral subdivisions were reported in surface density (counts per square millimeter).

Statistics

Statistical testing was performed with Prism 9.5.1 software (GraphPad). All data sets underwent testing for normality using either a Shapiro–Wilk or Kolmogorov–Smirnov test. Two-group comparisons for continuous data were made with a two-tailed unpaired Student's t-test. Multiple-group comparisons were then made with either a two-way or three-way ANOVA followed by a Sidak's and a Tukey's post hoc test, respectively, for pairwise comparisons that followed significant ANOVA interactions. Significant within-group correlations between counts and behavioral parameters were determined using Pearson's r coefficient and the resultant P-value. A Fisher's z-transformation was used to compare across-group correlations (Supplemental Table S2). Group data are expressed as mean ± SEM. Statistical significance was set at P < 0.05.

Competing interest statement

The authors declare no competing interests.

Supplementary Material

Supplement 1
Supplement 2
Supplemental_Table_2.docx (22.1KB, docx)
Supplement 3
Supplemental_Figure_1.docx (581.3KB, docx)

Acknowledgments

Images were generated using the University of Texas Health Science Center at San Antonio Optical Imaging Core Facility supported by National Institutes of Health/National Cancer Institute grant P30 CA54174. This work was supported by National Institutes of Health grants R01MH093320 and R01NS115072.

Author contributions: R.J.H. and G.M.T. designed the experiments. R.J.H. and M.A.A. performed the research. R.J.H., V.E.F., and G.M.T. analyzed the data. R.J.H., V.E.F., and G.M.T. wrote the manuscript.

Footnotes

[Supplemental material is available for this article.]

References

  1. Bourgeais L, Gauriau C, Bernard JF. 2001. Projections from the nociceptive area of the central nucleus of the amygdala to the forebrain: a PHA-L study in the rat. Eur J Neurosci 14: 229–255. 10.1046/j.0953-816x.2001.01640.x [DOI] [PubMed] [Google Scholar]
  2. Burgos-Robles A, Vidal-Gonzalez I, Quirk GJ. 2009. Sustained conditioned responses in prelimbic prefrontal neurons are correlated with fear expression and extinction failure. J Neurosci 29: 8474–8482. 10.1523/JNEUROSCI.0378-09.2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Campeau S, Davis M. 1995. Involvement of the central nucleus and basolateral complex of the amygdala in fear conditioning measured with fear-potentiated startle in rats trained concurrently with auditory and visual conditioned stimuli. J Neurosci 15: 2301–2311. 10.1523/JNEUROSCI.15-03-02301.1995 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cassell MD, Freedman LJ, Shi C. 1999. The intrinsic organization of the central extended amygdala. Ann N Y Acad Sci 877: 217–241. 10.1111/j.1749-6632.1999.tb09270.x [DOI] [PubMed] [Google Scholar]
  5. Chowdhury A, Caroni P. 2018. Time units for learning involving maintenance of system-wide cFos expression in neuronal assemblies. Nat Commun 9: 4122. 10.1038/s41467-018-06516-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Ciocchi S, Herry C, Grenier F, Wolff SB, Letzkus JJ, Vlachos I, Ehrlich I, Sprengel R, Deisseroth K, Stadler MB, et al. 2010. Encoding of conditioned fear in central amygdala inhibitory circuits. Nature 468: 277–282. 10.1038/nature09559 [DOI] [PubMed] [Google Scholar]
  7. Cui Y, Lv G, Jin S, Peng J, Yuan J, He X, Gong H, Xu F, Xu T, Li H. 2017. A central amygdala–substantia innominata neural circuitry encodes aversive reinforcement signals. Cell Rep 21: 1770–1782. 10.1016/j.celrep.2017.10.062 [DOI] [PubMed] [Google Scholar]
  8. DeNardo LA, Liu CD, Allen WE, Adams EL, Friedmann D, Fu L, Guenthner CJ, Tessier-Lavigne M, Luo L. 2019. Temporal evolution of cortical ensembles promoting remote memory retrieval. Nat Neurosci 22: 460–469. 10.1038/s41593-018-0318-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Do-Monte FH, Quinones-Laracuente K, Quirk GJ. 2015. A temporal shift in the circuits mediating retrieval of fear memory. Nature 519: 460–463. 10.1038/nature14030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Do Monte FH, Quirk GJ, Li B, Penzo MA. 2016. Retrieving fear memories, as time goes by. Mol Psychiatry 21: 1027–1036. 10.1038/mp.2016.78 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gale GD, Anagnostaras SG, Godsil BP, Mitchell S, Nozawa T, Sage JR, Wiltgen B, Fanselow MS. 2004. Role of the basolateral amygdala in the storage of fear memories across the adult lifetime of rats. J Neurosci 24: 3810–3815. 10.1523/JNEUROSCI.4100-03.2004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Grewe BF, Grundemann J, Kitch LJ, Lecoq JA, Parker JG, Marshall JD, Larkin MC, Jercog PE, Grenier F, Li JZ, et al. 2017. Neural ensemble dynamics underlying a long-term associative memory. Nature 543: 670–675. 10.1038/nature21682 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hammack RJ, Fischer VE, Andrade MA, Toney GM. 2023. Anterior basolateral amygdala neurons comprise a remote fear memory engram. Front Neural Circuits 17: 1167825. 10.3389/fncir.2023.1167825 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hitchcock JM, Davis M. 1991. Efferent pathway of the amygdala involved in conditioned fear as measured with the fear-potentiated startle paradigm. Behav Neurosci 105: 826–842. 10.1037/0735-7044.105.6.826 [DOI] [PubMed] [Google Scholar]
  15. Hopkins DA, Holstege G. 1978. Amygdaloid projections to the mesencephalon, pons and medulla oblongata in the cat. Exp Brain Res 32: 529–547. 10.1007/BF00239551 [DOI] [PubMed] [Google Scholar]
  16. Johansen JP, Tarpley JW, LeDoux JE, Blair HT. 2010. Neural substrates for expectation-modulated fear learning in the amygdala and periaqueductal gray. Nat Neurosci 13: 979–986. 10.1038/nn.2594 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Josselyn SA, Tonegawa S. 2020. Memory engrams: recalling the past and imagining the future. Science 367: eaaw4325. 10.1126/science.aaw4325 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Keifer OPJ, Hurt RC, Ressler KJ, Marvar PJ. 2015. The physiology of fear: reconceptualizing the role of the central amygdala in fear learning. Physiology 30: 389–401. 10.1152/physiol.00058.2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kim WB, Cho JH. 2017. Synaptic targeting of double-projecting ventral CA1 hippocampal neurons to the medial prefrontal cortex and basal amygdala. J Neurosci 37: 4868–4882. 10.1523/JNEUROSCI.3579-16.2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kim JJ, Jung MW. 2006. Neural circuits and mechanisms involved in Pavlovian fear conditioning: a critical review. Neurosci Biobehav Rev 30: 188–202. 10.1016/j.neubiorev.2005.06.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kim EJ, Horovitz O, Pellman BA, Tan LM, Li Q, Richter-Levin G, Kim JJ. 2013. Dorsal periaqueductal gray-amygdala pathway conveys both innate and learned fear responses in rats. Proc Natl Acad Sci 110: 14795–14800. 10.1073/pnas.1310845110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kim J, Pignatelli M, Xu S, Itohara S, Tonegawa S. 2016. Antagonistic negative and positive neurons of the basolateral amygdala. Nat Neurosci 19: 1636–1646. 10.1038/nn.4414 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kim J, Zhang X, Muralidhar S, LeBlanc SA, Tonegawa S. 2017. Basolateral to central amygdala neural circuits for appetitive behaviors. Neuron 93: 1464–1479.e5. 10.1016/j.neuron.2017.02.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kitamura T, Ogawa SK, Roy DS, Okuyama T, Morrissey MD, Smith LM, Redondo RL, Tonegawa S. 2017. Engrams and circuits crucial for systems consolidation of a memory. Science 356: 73–78. 10.1126/science.aam6808 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Li H, Penzo MA, Taniguchi H, Kopec CD, Huang ZJ, Li B. 2013. Experience-dependent modification of a central amygdala fear circuit. Nat Neurosci 16: 332–339. 10.1038/nn.3322 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Liu J, Totty MS, Melissari L, Bayer H, Maren S. 2022. Convergent coding of recent and remote fear memory in the basolateral amygdala. Biol Psychiatry 91: 832–840. 10.1016/j.biopsych.2021.12.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Madisen L, Zwingman TA, Sunkin SM, Oh SW, Zariwala HA, Gu H, Ng LL, Palmiter RD, Hawrylycz MJ, Jones AR, et al. 2010. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat Neurosci 13: 133–140. 10.1038/nn.2467 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Maruyama NO, Mitchell NC, Truong TT, Toney GM. 2019. Activation of the hypothalamic paraventricular nucleus by acute intermittent hypoxia: implications for sympathetic long-term facilitation neuroplasticity. Exp Neurol 314: 1–8. 10.1016/j.expneurol.2018.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mitchell NC, Gilman TL, Daws LC, Toney GM. 2018. High salt intake enhances swim stress-induced PVN vasopressin cell activation and active stress coping. Psychoneuroendocrinology 93: 29–38. 10.1016/j.psyneuen.2018.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Penzo MA, Robert V, Li B. 2014. Fear conditioning potentiates synaptic transmission onto long-range projection neurons in the lateral subdivision of central amygdala. J Neurosci 34: 2432–2437. 10.1523/JNEUROSCI.4166-13.2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Prewitt CM, Herman JP. 1998. Anatomical interactions between the central amygdaloid nucleus and the hypothalamic paraventricular nucleus of the rat: a dual tract-tracing analysis. J Chem Neuroanat 15: 173–185. 10.1016/S0891-0618(98)00045-3 [DOI] [PubMed] [Google Scholar]
  32. Roy DS, Park YG, Kim ME, Zhang Y, Ogawa SK, DiNapoli N, Gu X, Cho JH, Choi H, Kamentsky L, et al. 2022. Brain-wide mapping reveals that engrams for a single memory are distributed across multiple brain regions. Nat Commun 13: 1799. 10.1038/s41467-022-29384-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Salah H, Abdel Rassoul R, Medlej Y, Asdikian R, Hajjar H, Dagher S, Darwich M, Fakih C, Obeid M. 2021. A modified two-way active avoidance test for combined contextual and auditory instrumental conditioning. Front Behav Neurosci 15: 682927. 10.3389/fnbeh.2021.682927 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Samson RD, Duvarci S, Pare D. 2005. Synaptic plasticity in the central nucleus of the amygdala. Rev Neurosci 16: 287–302. 10.1515/REVNEURO.2005.16.4.287 [DOI] [PubMed] [Google Scholar]
  35. Stauffer W, Sheng H, Lim HN. 2018. EzColocalization: an ImageJ plugin for visualizing and measuring colocalization in cells and organisms. Sci Rep 8: 15764. 10.1038/s41598-018-33592-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wang Y, Krabbe S, Eddison M, Henry FE, Fleishman G, Lemire AL, Wang L, Korff W, Tillberg PW, Lüthi A, et al. 2023. Multimodal mapping of cell types and projections in the central nucleus of the amygdala. Elife 12: e84262. 10.7554/eLife.84262 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Whittle N, Fadok J, MacPherson KP, Nguyen R, Botta P, Wolff SBE, Muller C, Herry C, Tovote P, Holmes A, et al. 2021. Central amygdala micro-circuits mediate fear extinction. Nat Commun 12: 4156. 10.1038/s41467-021-24068-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wilensky AE, Schafe GE, Kristensen MP, LeDoux JE. 2006. Rethinking the fear circuit: the central nucleus of the amygdala is required for the acquisition, consolidation, and expression of Pavlovian fear conditioning. J Neurosci 26: 12387–12396. 10.1523/JNEUROSCI.4316-06.2006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Yap EL, Greenberg ME. 2018. Activity-regulated transcription: bridging the gap between neural activity and behavior. Neuron 100: 330–348. 10.1016/j.neuron.2018.10.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yuan Q, Isaacson JS, Scanziani M. 2011. Linking neuronal ensembles by associative synaptic plasticity. PLoS One 6: e20486. 10.1371/journal.pone.0020486 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zimmerman JM, Rabinak CA, McLachlan IG, Maren S. 2007. The central nucleus of the amygdala is essential for acquiring and expressing conditional fear after overtraining. Learn Mem 14: 634–644. 10.1101/lm.607207 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1
Supplemental_Table_1.docx (15KB, docx)
Supplement 2
Supplemental_Table_2.docx (22.1KB, docx)
Supplement 3
Supplemental_Figure_1.docx (581.3KB, docx)

Articles from Learning & Memory are provided here courtesy of Cold Spring Harbor Laboratory Press

RESOURCES