Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Aug 20.
Published in final edited form as: Neuron. 2014 Aug 7;83(4):906–918. doi: 10.1016/j.neuron.2014.07.019

Amygdala-dependent fear memory consolidation via miR-34a and Notch signaling

Brian George Dias 1,2, Jared Vega Goodman 1,2, Ranbir Ahluwalia 1,2, Audrey Elizabeth Easton 1,2, Raül Andero 1,2, Kerry James Ressler 1,2,3
PMCID: PMC4172484  NIHMSID: NIHMS616147  PMID: 25123309

Summary

Using an array-based approach after auditory fear-conditioning, and microRNA (miRNA) sponge-mediated inhibition, we identified a role for miR-34a within the basolateral amygdala (BLA) in fear memory consolidation. Luciferase assays and bioinformatics suggested the Notch pathway as a target of miR-34a. mRNA and protein levels of Notch receptors and ligands are down-regulated in a time- and learning-specific manner after fear conditioning in the amygdala. Systemic and stereotaxic manipulations of the Notch pathway indicated that Notch signaling in the BLA suppresses fear memory consolidation. Impairment of fear memory consolidation after inhibition of miR-34a within the BLA is rescued by inhibiting Notch signaling. Together, these data suggest that within the BLA, a transient decrease in Notch signaling, via miR-34a regulation, is important for the consolidation of fear memory. This work expands the idea that developmental molecules have roles in adult behavior, and that existing interventions targeting them hold promise for treating neuropsychiatric disorders.

Introduction

Forming associations about events and then consolidating memories of those associations is an important strategy for navigating our environments. However, in traumatic situations, these associations sometimes become overly consolidated and then, potentially, are resistant to extinction over time, resulting in Post-Traumatic Stress Disorder (PTSD) and other fear-related disorders in some individuals (Parsons and Ressler, 2013; Steckler and Risbrough, 2012). Future therapeutic intervention strategies for PTSD might therefore target this enhanced consolidation and/or lack of extinction to salient cues in the environment (Dias et al., 2013; Fitzgerald et al., 2013; Vanelzakker et al., 2013). To accomplish such goals, we need knowledge of the genetic and molecular mechanisms occurring within specific brain regions that underlie specific memory consolidation processes.

Auditory fear conditioning presents a framework within which to study the consolidation of cued fear memory (Johansen et al., 2011). In this classical conditioning paradigm, animals are first trained to associate auditory cues with mild footshocks. By subsequently presenting animals with only the auditory cues, the consolidation of associative learning can be assessed. Within the brain, the amygdala has received considerable attention for its role in consolidating these learned associations (Maren, 2003). As a result, the molecular and genetic landscape within the amygdala during the period of consolidation after conditioning undoubtedly holds clues to how memories come to be consolidated.

Regulation of both gene expression and protein synthesis are crucial to the consolidation of memory (Dudai, 2004; Maguschak and Ressler, 2012; Maren et al., 2003; Silva and Giese, 1994). In recent times, the epigenetic regulation of gene expression has received considerable attention within the realm of neuropsychiatric disorders (Gräff and Mansuy, 2008; Zovkic and Sweatt, 2013). Among these epigenetic mechanisms, microRNA (miRNA) mediated regulation of gene expression has been implicated in spatial memory, trace conditioning, extinction memory, and fear memory consolidation (Gao et al., 2010; Griggs et al., 2013; Konopka et al., 2010; Spadaro and Bredy, 2012; Wang et al., 2012; Zovkic et al., 2013). Of note, some of these studies uncovered miRNA that target specific genes whose functions are related to synaptic and structural plasticity (Griggs et al., 2013; Lin et al., 2011). Another molecular perspective suggests that signaling molecules canonically viewed as essential to development have functions in the adult that include synaptic plasticity (Ables et al., 2011; Pierfelice et al., 2011). In keeping with this perspective, we have previously uncovered a role for developmental signaling pathways (namely, Wnt signaling) within the amygdala in adult learning and memory (Maguschak and Ressler, 2011, 2012).

In this paper, we examined if activity of miR-34a, a target identified from an unbiased miRNA screening study, is required for cued fear memory consolidation. When we examined the targets of this miRNA, we found them to be components of the Notch signaling pathway normally involved in development, unearthing a role for a developmental molecular cascade in the consolidation of cued fear memory in the adult.

To our knowledge, this is the first study that links a miRNA (miR-34a) and developmental signaling cascade (Notch signaling) to a role in memory consolidation. In addition, this work adds to the growing body of literature that classically considered developmental molecules function outside of a developmental context.

Results

miR-34a is upregulated in the amygdala 30 mins after auditory fear conditioning

A microarray was performed on miRNA extracted from amygdala of adult mice, 30 minutes after fear conditioning. Results indicated a dynamic regulation of several non-coding RNA species during the period of fear memory consolidation (Fig. 1A). Two miRNA (miR-34a and miR-187) met the a priori criteria (p < 0.01 and dLMR = 0.5) as being significantly upregulated in the amygdala of adult mice, 30 minutes after fear conditioning (Fig. 1B). To independently verify whether these miRNA were indeed regulated in a learning-dependent manner, we replicated the aforementioned behavioral protocol with a completely independent set of animals, and used qPCR to quantify miRNA levels. We found that only miR-34a was significantly upregulated in the amygdala in a learning-dependent manner 30 minutes after fear conditioning (Fig. 1C) {ANOVA: F(2,19) = 4.42, p = 0.026, Home vs CS+US Paired: p < 0.05}. miR-187 levels in the amygdala were not significantly regulated 30 minutes after fear conditioning (Fig. 1D).

Figure 1. miRNA expression profiling and qPCR confirmation shows upregulation of miR-34a in the amygdala 30 minutes after fear conditioning.

Figure 1

Open in a new tab

(A) miRNA levels in the amygdala of fear conditioned adult male mice compared to Home cage controls were profiled by Exiqon using an array-based approach (Green-Black-Red: Low to High mRNA levels). (B) Two miRNA (miR-34a and miR-187) met the fold change threshold (p < 0.01 and dLMR = 0.5) (n = 4/group). qPCR analysis on a completely independent set of samples (n = 8/group) indicated that miR-34a (C) but not miR-187 (D) was significantly up-regulated in the amygdala 30 minutes after fear conditioning in a learning-dependent manner. Data are shown as Mean±SEM. * p < 0.05.

Inhibiting miR-34a function suppresses fear memory consolidation

We reasoned that an increase in miR-34a in the amygdala 30 minutes after fear conditioning would be permissive to fear memory consolidation. To test this idea, we infused viruses that expressed sponge sequences to inhibit miR-34a activity or scrambled controls into the BLA (Supplementary Fig. 1). The sponge construct contained seed sequences to which the miR-34a would bind and thereby not be able to exert its effect. Using traditional cell counting methods, we estimate that on average our infusions infected 200,000 cells/mm3 of tissue. Two weeks after infusion of the lentivirus into the BLA (Fig. 2A), animals were tested in an Elevated Plus Maze (EPM) to assay for anxiety-like behavior. We found no differences between groups in the EPM implying no effect of blocking miR-34a action on anxiety-like metrics (Supplementary Fig. 2). Both groups were then subjected to auditory fear conditioning and acquired fear similarly on the day of training (Fig. 2B). 24 hours later, when tested with 5 CS presentations, miR-34a sponge infused animals froze significantly less than miR-34a sponge scrambled control infused animals, suggesting that blocking miR-34a from exerting its effect suppressed fear memory consolidation {t=2.27, df=13, p = 0.040}. Freezing levels to individual CS presentations (CS 1, CS2, etc) are shown in Supplementary Fig. 7A. It is important to note that this suppression of fear memory consolidation was replicated in an independent experiment, where we found it to be specific to the consolidation of long term memory (LTM: tested 24 hours after training) (Fig. 6), with no differences detected within the timeframe of short term memory (STM: tested one hour after training) (Supplementary Fig. 2).

Figure 2. Blocking miR-34a action in the Basolateral amygdala (BLA) using a sponge-based strategy impairs the consolidation of cued-fear memory.

Figure 2

Open in a new tab

Lentiviral particles that expressed either a miR-34a sponge or a scrambled control under the control of a Ubc promoter were injected into the BLA. Two weeks later, animals were fear conditioned using 5 Tone-Shock pairings (CS−US pairings) and 24 hours later tested for the consolidation of cued-fear memory. (A) Based on mouse stereotaxic atlas by Paxinos and Franklin, representative expression of mCherry (also driven by the Ubc promoter) in the BLA indicates correct targeting of the injection (filled circles) versus incorrect targeting (solid triangles). Shown are only representative hits and misses. For final analysis, all LV-miR-34a sponge and LV-miR-34a sponge scrambled control hits targeting the BLA were included. Scale Bar: 1mm (B) No differences between groups were observed in the acquisition of fear responses as a function of CS presentation on training day. However, blocking miR-34a action using the sponge impaired the consolidation of the cued-fear memory as observed by a diminished freezing response during the 5 CS presentations in the fear expression test compared to the scrambled control animals. Data are shown as Mean±SEM. * p < 0.05.

Figure 6. miR-34a decreases Notch signaling and facilitates the consolidation of fear memory.

Figure 6

Open in a new tab

(A and B) Western Blotting using an antibody against Notch1 revealed a decrease in Notch1 protein level in the amygdala two hours after auditory fear conditioning (Home cage vs miR-34a sponge scrambled control) suggestive of a decrease in Notch signaling that mirrors the decrease in mRNA levels (Fig. 4). Infusion of the miR-34a sponge prevents this decrease emphasizing the in vivo functionality of the sponge construct used. (C) The consolidation of fear memory that is impaired when animals are tested 24 hours after conditioning (Fig. 2 and this figure) (miR-34a-Scrambled+Veh vs miR-34a Sponge+Veh) is no longer impaired when Notch signaling is decreased via administration of DAPT, a γ-secretase inhibitor. Data are shown as Mean±SEM. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Components of the Notch pathway are targets of miR-34a

Published work in the field of cancer biology has suggested that Notch1 is a bona fide target of miR-34a (Comery, 2005; Kashat et al., 2012). We wanted to extend our search for targets of miR-34a and turned to several algorithms (eg. miRDB, TargetScan, MicroCosm). While extremely useful, these algorithms generated many possible predicted targets. We decided instead to focus on a list of targets for miR-34a using a more directed approach that utilized the filter functions of Ingenuity Pathway Analysis (https://analysis.ingenuity.com/pa). Given that our previous work has implicated the developmental Wnt molecules in the formation of fear memory, and our desire to maximize the probability that miR-34a actually targets some of the genes in our bioinformatics query, we filtered results in IPA using the “developmental molecules” and “observed targets” filter functions. It is our understanding that with this approach the Ingenuity Pathway software collates from a variety of algorithms literature that has demonstrated genes to be experimentally validated targets of specific miRNA. Thus the gene list generated is not only the result of Ingenuity’s own algorithm (“Ingenuity Expert Findings”) but instead includes other commonly mined algorithms like TargetScan, TarBase and miRecords. This approach gave us a list of 13 genes. On further examination, we noticed that a number of both ligands and receptors involved in Notch signaling predominated this list and so chose to focus on the Notch pathway moving forward (Fig. 3A).

Figure 3. Notch1 is a target of miR-34a, and components of the Notch signaling pathway are also targets of miR-34a.

Figure 3

Open in a new tab

(A) Ingenuity Pathway Analysis filtered for Developmental Molecules and Experimentally Observed revealed several components of the Notch pathway to be targets of miR-34a. (B) A luciferase-based assay was used to demonstrate that Notch1 is a target of miR-34a. Co-transfection of HEK293T cells with a plasmid that contains the 3′UTR of Notch1 cloned downstream of the Firefly-Luciferase gene (FF-Luc-Notch1) and one that expresses miR-34a, decreased the luminescence relative to the FF-Luc-Notch1 control. Blocking miR-34a action using a sponge-based strategy (miR-34a sponge) prevents this decrease in luminescence. Data are shown as Mean±SEM. ** p < 0.01.

We then used cell culture and a luciferase-based system to ask whether Notch1 is indeed a molecular target of miR-34a (Supplementary Fig. 3). To begin, the 3′UTR of Notch1 was cloned downstream of luciferase and luciferase-mediated luminescence was normalized to 1. When HEK293T cells were co-transfected with a construct encoding miR-34a and the construct containing the 3′UTR of Notch1, luciferase-mediated luminescence was decreased. This suggests that miR-34a targeted the 3′UTR of Notch1 to suppress luciferase activity (Fig. 3B). This reduction was ameliorated when miR-34a activity was inhibited using the same sponge construct used for the in vivo experiment in Fig. 2. {ANOVA: F(3,11) = 5.414, p = 0.015. Post-hoc tests: FF-Luc-Notch1/Ren-Luc vs. FF-Luc-Notch1/Ren-Luc/miR-34a: p = 0.006, FF-Luc-Notch1/Ren-Luc vs. FF-Luc-Notch1/Ren-Luc/miR-34a/mirR-34a sponge: p = 0.09, FF-Luc-Notch1/Ren-Luc vs. FF-Luc-Notch1/Ren-Luc/mirR-34a sponge: p = 0.09}.

To ascertain that the miR-34a-sponge used was specific to miR-34a action and did not inhibit closely related miRNA activity, we turned to recent literature that comprehensively showed CRFR1 (Corticotrophin Releasing Factor Receptor1) to be a target of both miR-34c and miR-34a. As was shown by these authors, luciferase-mediated luminescence was decreased in our experiment when the 3′UTR of CRFR1 was targeted by miR-34c and miR-34a. While miR-34a-sponge co-transfected with the miR-34c construct did not rescue this decrease, this decrease was rescued when the miR-34a activity was inhibited by miR-34a-sponge (Supplementary Fig. 4)

mRNA levels of components of the Notch pathway are regulated in the amygdala after auditory fear conditioning

With miR-34a being regulated in the amygdala 30 minutes after auditory fear conditioning, and bioinformatics and luciferase analyses indicating that Notch pathway components are targets of miR-34a, we sought to examine whether Notch pathway mRNA levels are themselves regulated in the amygdala after fear conditioning. Included in this experiment were CS+US paired and CS−US unpaired groups, to control for unconditioned stress responses separate from associative memory formation. We then performed qPCR on mRNA isolated from the amygdala of mice that had been sacrificed 2, 6 and 12 hours after fear conditioning. More specifically, we chose to measure the mRNA levels of ligands (Jag1 and Dll1) and receptors (Notch1 and Notch2).

Analysis of the amygdala mRNA (Fig. 4) demonstrated a significant reduction in Notch pathway mRNA levels at primarily 2 and 6 hours after fear conditioning. By the 12-hour time point, all groups showed similar mRNA levels of the components queried.

Figure 4. mRNA levels of components of the Notch pathway in the amygdala are altered 2 and 6 hours after fear conditioning.

Figure 4

Open in a new tab

After subjecting animals to Paired and Unpaired conditioning paradigms, amygdala mRNA levels of Notch pathway components were measured 2, 6 and 12 hours after fear conditioning relative to Home cage controls (n = 8/group). mRNA levels of the ligands, Jag1 (A), Dll1 (B) and receptor, Notch1 (C) were significantly down-regulated 2 hours after fear conditioning in both the Paired and Unpaired groups, suggestive of a cued- and contextual-learning related regulation. In addition, Jag1 and Notch1 mRNA levels remained significantly down-regulated in the Paired group at 6 hours after fear conditioning compared to the Unpaired condition. Notch2 (D) mRNA levels remained unchanged throughout. All mRNA levels returned to baseline levels 12 hours after the behavioral paradigms. Data are shown as Mean±SEM. * p < 0.05, ** p < 0.01, *** p < 0.001.

Ligands

Jag1 mRNA levels (Fig. 4A) were significantly reduced at the 2 hour time point in both the paired and unpaired groups as compared to home cage controls {F(2,19) = 42.06, p<0.0001, Home vs Paired: p < 0.001, Home vs Unpaired: p < 0.001}. At 6 hours, Jag1 levels in the paired group were significantly down-regulated as compared to those in the Home cage and unpaired group {F(2,20) = 4.569, p = 0.023, Paired vs Unpaired: p < 0.05}. Dll1 mRNA levels (Fig. 4B) showed similar regulation at the 2 hour mark, as both the paired and unpaired groups were reduced as compared to home cage {F(2,19) = 6.13, p = 0.008, Home vs Paired: p < 0.05, Home vs Unpaired: p < 0.01}. However, there was no significant difference in mRNA levels among the groups at 6 hours.

Receptors

Notch1 mRNA levels (Fig. 4C) were also decreased in the paired and unpaired groups at the 2-hour time point as compared to home cage {F(2,19) = 17.56, p < 0.0001, Home vs Paired: p < 0.001, Home vs Unpaired: p < 0.001}. At 6 hours, the paired group displayed lower mRNA levels as compared to the unpaired group {F(2,19) = 3.67, p = 0.045, Paired vs Unpaired: p < 0.05}. No regulation was noted in Notch2 mRNA levels (Fig. 4D) between any of the groups. Together, these data suggest that the Jag1 ligand and Notch1 Receptor mRNAs are dynamically regulated during the consolidation period following fear conditioning in an associative learning-specific manner.

Systemic γ-secretase inhibition to inhibit Notch signaling facilitates fear memory consolidation

Our data so far suggest that mRNA levels of the Notch signaling components in the amygdala change in a time-dependent manner after fear conditioning, downstream of miR34a regulation. Next we chose to directly manipulate Notch signaling to test how consolidation of fear memory may be affected. Upon ligand binding to the Notch receptors, γ-secretase cleaves the Notch receptor so that the Notch intracellular domain can then translocate into the nucleus and exert its function (Sail and Hadden, 2012). DAPT, a γ-secretase inhibitor that crosses the blood-brain barrier, was administered i.p. to animals 15 minutes after the last conditioning trial to block γ-secretase during consolidation. When tested 24 hours later for expression of consolidated fear memory, DAPT-injected animals froze more than vehicle (DMSO) injected controls (Fig. 5A) {t=2.41, df=19, p = 0.026}. These data suggest that transiently inhibiting Notch signaling in adults creates a molecular environment that is permissive to the consolidation of fear memory. Freezing levels to individual CS presentations (CS 1, CS2, etc) are shown in Supplementary Fig. 7B.

Figure 5. Manipulating Notch activity affects the consolidation of cued-fear memory.

Figure 5

Open in a new tab

(A and B) Inhibiting Notch activity facilitates the consolidation of cued-fear memory. (A) The γ-secretase inhibitor (DAPT) or Vehicle (100% DMSO) was injected i.p. 15 minutes after the last CS+US pairing. Systemic inhibition of Notch signaling enhanced the consolidation of cued-fear memory when tested 24 hours later as seen by an increase in freezing behavior compared to the Vehicle group. (B) Anti-Jag1 Fc at a concentration known to inhibit Notch signaling was injected into the BLA 15 minutes after the last CS+US pairing. Such local inhibition of Notch signaling also facilitated the consolidation of cued-fear memory compared to the control infusion of anti-Human Fc when freezing responses to the tone were measured 24 hours later. (C and D) Increasing Notch signaling in the Basolateral amygdala (BLA) by overexpressing a downstream effector of Notch signaling (Hes1) impairs the consolidation of cued-fear memory. Lentiviral particles that expressed Hes1 or control GFP under the control of a EF1a promoter were injected into the BLA. Two weeks later, animals were fear conditioned using 5 Tone-Shock pairings (CS−US pairings) and 24 hours later tested for the consolidation of cued-fear memory. (C) Representative expression of GFP (driven by the CMV promoter in the dual promoter Hes1 plasmid) in the BLA indicates correct targeting of the injection (filled circles) versus incorrect targeting (solid triangles). Scale Bar: 1mm (D) Both groups acquired similar fear responses as a function of CS presentation on training day. However, increasing Notch signaling via Hes1 overexpression impaired the consolidation of the cued-fear memory as observed by a diminished freezing response compared to the GFP-control infused animals. Data are shown as Mean±SEM. * p < 0.05.

Inhibiting Notch signaling in the BLA facilitates fear memory consolidation

While the DAPT experiment demonstrates that γ-secretase inhibition and, by proxy inhibition of Notch signaling, facilitates fear memory formation, γ-secretase activity is also important for the processing of other peptides, including amyloid precursor protein (APP) that has been implicated in Alzheimer’s disease (Comery, 2005). To more specifically determine what effect inhibition of Notch signaling within the BLA has on fear memory consolidation, we utilized a Notch-inhibiting antibody with stereotaxic intracranial manipulation. Cannulae were implanted into the BLA at least 1 week prior to fear conditioning. On the day of fear acquisition, 30 minutes after the last training trial, animals were injected with an antibody (Jag1-Fc) at a concentration that has been shown to inhibit Notch signaling (Conboy, 2003). When tested 24 hours later for expression of consolidated fear memory, animals that received the Jag1-Fc antibody infused into the BLA froze more than Control antibody-injected controls (Fig. 5B) {t=2.36, df=12, p = 0.036}. These data further corroborate the idea that inhibition of Notch signaling within the BLA facilitates the consolidation of fear memory. While we cannot definitively rule out the possibility that the antibody-based Notch inhibition extended beyond the BLA, infusion of these antibodies into the BLA followed by sacrifice an hour later and immunohistochemistry to detect the antibody infusion indicated localization to the BLA (Supplementary Fig. 5). Freezing levels to individual CS presentations (CS 1, CS2, etc) are shown in Supplementary Fig. 7C.

Gain-of-function of Notch signaling in the BLA suppresses fear memory consolidation

Having demonstrated that inhibiting Notch signaling is permissive to the consolidation of fear memory, we asked whether overexpressing Hes1 in the BLA, and thereby increasing Notch signaling, might inhibit fear memory consolidation. While having functions outside of a Notch signaling context, Hes1 is one of the main effectors of Notch signaling and hence is used in our experiment to activate this pathway. Our desire to manipulate Notch signaling that encompasses a large family of receptors and ligands and prevent any compensatory effects from other Notch pathway components, led us to utilize a broader approach via the targeting of Hes1 instead of merely overexpressing Notch1.

For this purpose, lentiviral particles that contained either LV-EF1a-mHes1-CMV-GFP (Hes1) or LV-GFP (GFP) were infused into the BLA, followed by a two week incubation period prior to behavior (Fig. 5C). Animals were then tested in EPM to assay for anxiety-like behavior, and we again found no differences between groups (Supplementary Fig. 6). Both groups were then subjected to a 5 CS+US paired auditory fear conditioning paradigm and acquired fear similarly on the day of training (Fig. 5D). However, 24 hours later, when tested with 5 CS presentations, Hes1-infused animals froze significantly less than GFP-infused control animals (t=2.16 df=15, p = 0.043, Fig. 5D). Freezing levels to individual CS presentations (CS 1, CS2, etc) are shown in Supplementary Fig. 7D. These data suggest that increasing Notch downstream signaling suppresses fear memory consolidation.

miR-34a in the BLA decreases Notch signaling to facilitate fear memory consolidation

Our data thus far have separately implicated BLA miR-34a and Notch function in fear memory consolidation. We chose to ask if miR-34a and Notch interact to influence this consolidation (Fig. 6). In addition to the decrease in Notch1 mRNA levels in the amygdala, we also detected a decrease in Notch1 protein levels in the amygdala 2 hours after auditory fear conditioning (Fig. 6A and 6B) {ANOVA: F(2,10) = 42.50, p < 0.0001}. Evidence of this down-regulation after fear conditioning comes from comparing Notch1 protein levels in the amygdala of Home cage control animals to those in which miR-34a-sponge scrambled control was infused into the amygdala (Home vs miR-34a-sponge scrambled control: p < 0.01). By infusing miR-34a-sponge into the amygdala, we could also verify the efficacy of sponge activity in vivo. We find that miR-34a-sponge activity within the amygdala prevents any decrease in Notch1 levels after fear conditioning (Home vs miR-34a-sponge: p < 0.01, miR-34a-sponge scrambled control vs miR-34a-sponge: p < 0.0001).

For behavior, we first infused either miR-34a-sponge or miR-34a-sponge scrambled control into the BLA of animals. This allowed us to replicate our initial finding that inhibiting miR-34a activity in the BLA suppressed fear memory consolidation. The suppression of fear memory consolidation in miR-34a-sponge animals was rescued by administration of DAPT (Fig, 6C). This suggests that it was the increase in Notch signaling (based on the western blot data: Fig 6A and 6B) in the miR-34a-sponge animals that was responsible for the suppression, which is now overridden by administration of DAPT, a γ-secretase inhibitor, and consequently inhibitor of Notch activity {ANOVA: F(2,12) = 4.886, p = 0.02. Post-hoc tests: miR-34a-Scrambled control+Veh vs miR-34a-Sponge+Veh: p < 0.05}.

Discussion

Recently, microRNA belonging to the non-coding RNA family have begun to be investigated for their role in neuropsychiatric disorders. For example, single nucleotide polymorphisms (SNPs) in several human miRNAs were found to be associated with panic disorder in humans, and a specific miRNA was found to be associated with dementia (Muiños-Gimeno et al., 2011; Zovoilis et al., 2011).

Animal models are being increasingly utilized to address how specific miRNA may be related to such disorders and the mechanisms underlying the same. With specific brain regions known to be involved in anxiety that accompanies fear expression, as well as mechanisms underlying the consolidation and extinction of fear memories (Davis et al., 2010; Fanselow and Poulos, 2005), these studies have asked whether miRNA in these brain regions are involved in the behavioral endpoints. Of note are studies that have implicated miR-34c in the central amygdala in stress-induced anxiety (Haramati et al., 2011), miR-128b in prefrontal cortex in the formation of fear extinction memory (Lin et al., 2011), and miR-182 in the lateral amygdala in the consolidation of fear memory in rats (Griggs et al., 2013). By blocking miR-34a function in the mouse BLA using a sponge-based strategy, we demonstrated that this miRNA species is necessary for the consolidation of cued fear memory. Our luciferase assays indicated that the miR-34a sponge that we used was specific to miR-34a and did not inhibit the action of the closely related miR-34c. It must be noted that expression of the mCherry and sponge/sponge scrambled sequences are driven by the strong non-specific Ubc promoter and as such we cannot make any claims about the cellular identity of the infected cells. Our future studies would benefit from manipulating miR34a activity in specific cellular populations to enable circuit level analysis of the role of amygdalar miR-34a in fear memory consolidation. Our array and sponge-based functional studies and previously cited data taken together indicate that amygdala microRNA regulation and action accompanies the consolidation of memory across species, and present a framework to investigate how regulation of gene expression mediates this important behavioral process.

It is important to note that we observe an increase in miR-34a levels in the amygdala shortly (30 minutes) after auditory fear conditioning, and in a learning-specific manner. Our reasoning to choose this time point was based on the fact that consolidation typically occurs in the few hours following conditioning, and levels of regulatory miRNA should consequently be altered quickly so as to enable their action on target mRNA within the temporal window of consolidation. We do not observe an increase in miR-34a in a stressful, but non-associative context (CS+US unpaired), in contrast to the published report of a stress-induced increase in miR-34a (Haramati et al., 2011). Two parameters might explain this apparent discrepancy: the nature of the stressor being acute restraint stress in the prior study contrasted by multiple mild-footshocks in ours, and the arrays being performed 90 minutes after the acute restraint stress versus 30 minutes after in our paradigm.

While it is relatively straightforward to measure changes in the miRNA profile in specific brain regions and correlate them to a behavior of choice, understanding how a specific miRNA exerts its action is a complicated endeavor. Their short nature coupled with the fact that a very small seed region in the target mRNA is required for a miRNA to exert its effect mean that a single miRNA can target hundreds or thousands of genes (Lim et al., 2005). Using our previously described focused approach, it was reassuring and important that Wnt1 and CREB were on our list of miR-34a targets, with both having documented roles in the consolidation of fear memory (Cowansage et al., 2013; Frankland et al., 2004; Josselyn et al., 2001; Maguschak and Ressler, 2011). To our unexpected surprise, ligands and receptors of the Notch pathway dominated the list. It is worth noting that our report of miR-34a mediated regulation of Notch1, Jag1 and Dll1 has been previously suggested by cancer researchers (de Antonellis et al., 2011; Pang et al., 2010). As such, this presented us with a compelling rationale to focus on not just one gene in a developmental signaling cascade, but rather on multiple members of a pathway. Per the accepted role of miRNAs decreasing target mRNA levels, we hypothesized that an increase in miR-34a would result in a decrease in mRNA of Notch pathway components. More importantly, we predicted that a decrease in Notch signaling would be permissive to fear memory consolidation.

We did indeed observe decreases in both ligand and receptor mRNA levels of Notch pathway components in the amygdala that coincided with the timeframe of fear memory consolidation. At both the 2- and 6-hour time points, this decrease was observed in the paired versus control group indicating learning-dependent gene regulation. It should be noted that 6 hours after fear conditioning, we do not observe statistically significant differences in Jag1 and Notch1 mRNA levels of Home Cage and Paired groups, but instead detect significant differences between the Paired and Unpaired groups. While we appreciate that the Home cage group is the most important control in this case, we have found in our hands that Unpaired groups show negligible cue-specific learning, and as such we interpret our differences between Paired and Unpaired groups to be the result of learning-specific effects on mRNA levels. We also observed a decrease in mRNA levels of some components in our unpaired group at 2 hours after the unpaired protocol, and postulate that this could be related to consolidation of contextual fear memory that may occur in this unpaired condition. We know the amygdala to be involved in the consolidation of both cued- (our main focus) and contextual fear memory (Davis, 1992; Sauerhöfer et al., 2012). Future studies would do well to investigate the role of Notch signaling in this dichotomy. However, we cannot rule out the possibility that foot-shocks (US) or tones (CS) alone might be responsible for these effects, and in future studies, we will incorporate shock controls in which un-signaled foot-shocks are administered immediately upon entering the training arena followed by CS presentations; a manipulation resulting in no context conditioning (Landeira-Fernandez et al., 2006). In support of our results (albeit in a different memory task), Notch mRNA levels have also been shown to be down-regulated in the rat hippocampus during the window of memory consolidation in the hippocampus-dependent passive avoidance task (Conboy et al., 2007). Using western blotting, we were also able to show that Notch1 protein levels are decreased in the amygdala, two hours after fear conditioning, with this decrease prevented when miR-34a-sponge was infused into the BLA. Therefore mRNA and protein levels of components of the Notch pathway are regulated in the period that consolidation of fear memory occurs.

From a learning perspective, the role of Notch signaling in memory formation appears to be complicated. Consistent with our data, in the passive avoidance paradigm cited above, a transient attenuation of Notch signaling facilitates memory consolidation in rats, and forced activation of the Notch pathway impairs consolidation (Conboy et al., 2007). In apparent contrast to this finding is the observation that, mutant Notch+/− mice appear to have cognitive impairment, evidenced by poor performance on the hippocampal-dependent water maze task (Costa et al., 2003), with cued fear memory being unaffected. We propose that this difference between the Notch+/− mice and our data might be a consequence of the constitutive deletion of Notch1 vs a more acute manipulation of Notch signaling. It also appears that the involvement of Notch in learning and memory may be task-dependent. A most recent study using Jag+/− mice demonstrated that an impairment in Notch signaling affected spatial memory formation but not other forms of memory (Sargin et al., 2013). Using conditional inhibition of Notch signaling by deleting a downstream effector (RPBj) in a CamKII-Cre dependent manner, another recent study found no evidence for any impairment in learning and memory (Sato et al., 2012).

Using loss-of-function studies we have shown that decreasing cleavage of the Notch receptor by inhibiting γ-secretase facilitates cued fear memory consolidation. This is in keeping with the reports of such inhibition of γ-secretase facilitating the consolidation of contextual and spatial memory, which are both hippocampus dependent (Comery, 2005; Dash et al., 2005). However, an important caveat of our experiment is that γ-secretase is not only involved in Notch signaling but also plays a role in the processing of amyloid precursor protein and consequently has a role in Alzheimer’s disease. Despite this lack of specificity, we are confident that decreasing Notch signaling facilitates cued memory formation because intra-amygdala inhibition of Notch signaling using a previously published antibody-based strategy (Conboy, 2003) also facilitates memory consolidation in our design. Complementing this experiment is our finding that over-expressing Hes1, a key effector of Notch signaling, in the BLA suppressed the consolidation of cued fear memory. Utilizing an experimental design that is focused on cued-fear conditioning, and conditional approaches of manipulating Notch function at acute timescales, our data make a case for Notch signaling as being suppressive of the consolidation of cued-fear memory. We suspect that this focused and conditional manipulation may explain some of the differences between our conclusions and the studies that suggest Notch signaling to be permissive of memory consolidation.

The impairment of fear memory consolidation that results from inhibiting miR-34a action in the BLA could be rescued by inhibiting γ-secretase activity and consequently Notch activity. Our data over bioinformatics, expression, biochemical, and behavioral levels of analyses viewed together, lead us to postulate that Notch signaling at baseline conditions maintains a steady state of synaptic stability, repressing synaptic plasticity. A fear conditioning-mediated, transient increase in miR-34a in the amygdala then decreases Notch signaling, thereby creating an environment that is transiently permissive of plasticity and consequently consolidation of the cued-fear memory (Fig. 7). While we contribute to the body of work that suggests that Notch signaling impairs the consolidation of cued fear memory, further investigation of the role of Notch signaling in learning and memory will need to be nuanced to determine its task-, manipulation- (constitutive vs conditional), and region-specific roles.

Figure 7. Working model of miR-34a mediated regulation of Notch signaling as being permissive to fear memory consolidation.

Figure 7

Open in a new tab

(A) Baseline levels of Notch signaling in the Basolateral Amygdala (BLA) as a consequence of ligand (Jag1 and Dll1) and receptor (Notch 1 and Notch 2) interaction maintains status quo and the baseline state. (B) After training, miR-34a up-regulation in the amygdala results in down-regulation of components of the Notch pathway and a decrease in Notch signaling thereby creating a molecular and cellular environment that is permissive to the consolidation of cued-fear memory. Green: Inhibiting miR-34a action in the BLA by expressing miR-34a sponges via lentiviruses impairs fear memory consolidation, as does overexpressing Hes1 in the BLA and consequently increasing Notch signaling. Yellow: In contrast, inhibiting Notch signaling by suppressing the activity of γ-secretase using systemic DAPT administration, or using Anti-Jag1-Fc antibody in the BLA enhances fear memory consolidation.

In keeping with the idea that gene expression and protein synthesis are involved in fear memory consolidation, our study makes two important contributions to our understanding of how this occurs and a candidate gene pathway that is regulated during consolidation. First, using an array-based approach followed by loss-of-function in vivo manipulation, we show that miR-34a function in the amygdala allows for memory of a fearful stimulus to be consolidated. Second, using a targeted bioinformatics approach, complemented with loss- and gain-of-function studies, we establish a link for miRNA-mediated regulation of the Notch pathway as being involved in this consolidation. With Notch pathway function being traditionally viewed through the lens of stem cell differentiation during development (Carlson and Conboy, 2007), this second contribution indicates that signaling pathways normatively used in development are also parsimoniously utilized for adult behavior. Finally, Notch signaling is now being targeted in the treatment of various cancers (Rizzo et al., 2008; Sail and Hadden, 2012), with numerous new therapeutic tools available. We would do well to co-opt relevant therapeutic agents that regulate Notch signaling and use them in the treatment or prevention of disorders that have a substantial memory component including those with declarative memory loss, such as dementia, or those with substantial over-consolidation of emotional memory such as PTSD, OCD, and substance abuse disorders.

Experimental Procedures

Animals

All experiments were conducted with 2-month old C57Bl/6J male mice purchased from Jackson Laboratory (Bar Harbor, ME). Animals were housed on a 12 h light/dark cycle in standard group cages (≤ 5/cage) with ad libitum access to food and water. All experiments were conducted during the light half of the cycle. All procedures were approved by Emory University’s IACUC, and followed guidelines set by NIH.

Behavior

All behavior was performed in a double-blind manner and data were acquired using automated computer software programs.

Elevated Plus Maze

Mice were placed in the center of the plus maze arms with the maze elevated about 2 feet from the floor. Animals were recorded while exploring the plus maze for 5 minutes in a dimly lit room. The amount of time spent in the closed versus open arms is viewed as a measure of anxiety-like behavior (Gafford et al., 2012). The recording and analysis were both carried out using Limelight Software (Actimetrics).

Auditory Fear Conditioning

Auditory fear conditioning was carried out as previously described (Dias and Ressler, 2014). Briefly, mice were pre-exposed to sound attenuated conditioning chambers (San Diego Instruments) (grid floors, room light on, cleaned with Quatricide: Context A) for 3 consecutive days before training. On the day of auditory fear conditioning in Context A, mice received five CS−US pairings (CS: 30 second, 6 kHz, 75 db tone) (US: 500 ms, 0.6mA foot-shock) wherein the tone co-terminated with the mild foot-shock with a 5 minute inter-trial interval (ITI). Where an unpaired condition was used, the same CS and US parameters were used with no co-termination and presented in a random sequence. The percentage of time spent freezing during fear acquisition was measured by SR-LAB software (San Diego Instruments). The consolidation of fear memory was tested 24 hours after fear conditioning in a novel context (modular test chambers; Med Associates Inc.) (plexiglass floor, room light off, red chamber lights on, cleaned with EtOH: Context B) when mice were exposed to 5 CS tones with a 2 minute ITI. Freezing during the tone presentations was measured with FreezeView software (Coulbourn Instruments). All statistical analyses were conducted using a repeated-measure ANOVA design with Bonferroni correction.

miRNA microarray and data analysis

Animals were either subjected to the auditory fear conditioning protocol mentioned above or left undisturbed in their home cages. Thirty minutes after conditioning, both groups of animals were decapitated, brains rapidly frozen on dry ice, and then stored at −80°C. Fresh frozen brains were mounted on the Microm HM450 freezing microtome with Tissue-Tek O.C.T. compound with the tissue kept at −23.0°C. Using a 1.0 mm biopsy tool, bilateral punches of the brain were made to collect amygdala tissue (Bregma −0.94 mm to −2.3 mm)(Paxinos and Franklin, 2004). The punches were stored at −80°C until miRNA extraction. miRNA were extracted from this tissue using the miRNeasy kit (Qiagen). The miRNA array (miRCURY LNA microRNA array, 6th gen) was conducted by Exiqon using previously published techniques (www.exiqon.com for more information). To examine miRNA that were upregulated after fear conditioning, we set a priori thresholds (p < 0.01 and dLMR = 0.5) for follow-up studies. We then proceeded to verify whether specific miRNA that met those criteria were indeed regulated after fear conditioning using a completely different cohort of animals.

Independent verification of miRNA microarray hits using qRT-PCR

After miRNA were extracted from amygdala tissue punches obtained from animals 30 minutes after auditory fear conditioning (Home, Paired and Unpaired groups), the Exiqon cDNA synthesis kit was used to reverse transcribe the miRNA to cDNA. Quantitative real-time PCR (qRT-PCR) was then performed using the above cDNA as template in a SYBRGreen MasterMix+primer mixture. Plates were run in the Applied Biosystems 7500 Fast Real-Time PCR System under the Standard 7500 run mode (1 cycle 95°C, 10 mins; 40 cycles 95.0°C, 10 secs and 60°C, 1 min with fluorescence measured during 60°C step; 1 cycle 95.0°C, 15 secs, 60°C, 1 min, 95.0°C, 15 secs, 60°C, 15 secs). Data were then analyzed using the 2−ΔΔCT method (Livak and Schmittgen, 2001). All collected data were normalized to the Home Cage group, and statistical analysis involved ANOVA on the fold-change values with Bonferroni post-hoc correction.

Molecular cloning and virus production

The plasmids, ED156, ED350, ED245, and LIB01, were generously gifted by Dr. Dieter Edbauer (Munich, Germany). Primers used for all sub-cloning experiments are listed in Supplementary Figure 1. Schematic of constructs used for experiments are shown in Supplementary Figure 1.

Notch1 3′UTR downstream of Firefly-luciferase for luciferase assay

The 3′UTR of Notch1 and CRFR1 was sub-cloned into ED156 downstream of the firefly luciferase gene (FF-luc) with the FF-luc being driven by the Synapsin promoter (See Supplementary Figure 1 for Notch 1 primer sequences, CRF-R1 primers were same as used in Haramati et al., 2011). Another construct (ED350) containing Renilla luciferase (R-luc) downstream of the Synapsin promoter was used as an internal control.

miR-34a

The miR-34a sequence was sub-cloned into ED245 downstream of the Synapsin promoter. In addition, miR-34c was also sub-cloned into ED245 with primers used by Haramati et al., 2011.

miR-34a sponge and scrambled constructs

Seed sequences (4 concatenated seeds) to which miR-34a can bind, thus inhibiting its activity, were sub-cloned into LIB01 downstream of the Synapsin promoter. As a control, these sequences were scrambled (miR-34a-sponge scrambled control).

Hes1 overexpression

We obtained Plasmid # 17625 from Addgene, packaged it to produce lentivirus to overexpress GFP (driven by Ubc promoter), and Hes1 (driven by Ef-1a promoter) in the mouse brain.

Lentiviral production

The miR-34a-sponge, miR-34a-sponge scrambled control, and Hes1 overexpression constructs were packaged as lentiviruses by the Emory Viral Vector Core using previously published protocols (Lois et al., 2002).

Luciferase activity to validate Notch1 as a target of miR-34a

We used methodology referenced in Edbauer et al. 2010 (Edbauer et al., 2010). Briefly, luciferase assays (Supplementary Figure 3) were used to validate Notch1 as a target of miR-34a. To validate Notch1 as an actual target of miR-34a, HEK293T cells were co-transfected with (a) the FireFly-luc-Notch1 UTR and Renilla-luc constructs, (b) the FF-luc-Notch1 UTR, R-luc, and miR-34a constructs, (c) the FF-luc-Notch1 UTR, R-luc, miR-34a, and miR-34a sponge constructs, and (d) the FF-luc-Notch1 UTR, R-luc, and miR-34a sponge constructs. To determine the specificity of the miR-34a sponge, we also performed a luciferase assay wherein the 3′UTR of CRFR1 was sub-cloned downstream of FireFly luciferase. Seeing that CRFR1 is a target of miR-34c and miR-34a (Haramati et al., 2011), HEK293T cells were co-transfected with (a) the FireFly-luc-CRFR1 UTR and Renilla-luc constructs, (b) the FF-luc-CRFR1 UTR, R-luc, and miR-34c constructs, (c) the FF-luc-CRFR1 UTR, R-luc, miR-34c, and miR-34a sponge constructs, (d) the FF-luc-CRFR1 UTR, R-luc, and miR-34a sponge constructs, and (e) the FF-luc-CRFR1 UTR, R-luc, miR-34a, and miR-34a sponge constructs. HEK293T cells were harvested 2 days after the transfection and luminescence as a function of both FF-luc and R-luc was measured using the Promega Stop-n-Glo system.

mRNA quantification in the amygdala after fear conditioning using qRT-PCR

Male mice were subjected to auditory fear conditioning (Paired and Unpaired groups). Brains from these animals and Home Cage controls were collected 2, 6 and 12 hours after fear conditioning. After micro-punching the amygdala as outlined above, mRNA were extracted from the tissue punches using the RNeasy Kit (Qiagen). The SABiosciences RT2 First Strand Kit was used to reverse transcribe the mRNA to cDNA. Quantitative real-time PCR (qRT-PCR) was then performed using the above cDNA as template in a TaqMan® Universal PCR Master Mix mixture. The primers included Mouse Gapdh (GAPDH) as Endogenous Control, Mouse Notch1 (Mm00435249_m1), Mouse Notch2 (Mm00803077_m1), Mouse Dll1 (Mm01279269_m1), Mouse Jag1 (Mm00496902_m1). The plate was run in the Applied Biosystems 7500 Fast Real-Time PCR System under the Standard 7500 run mode (1 cycle 50.0°C, 2 mins; 1 cycle 95.0°C, 10 mins; 40 cycles 95.0°C, 15 secs and 60°C, 1 min with fluorescence measured during 60°C step). Data were analyzed as noted previously.

Western blot analysis of Notch1 protein levels in BLA of miR-34a sponge-infused animals

Notch1 protein levels in the amygdala were measured 2 hours after fear conditioning in Home cage controls vs LV-miR-34a-sponge, and LV-miR-34a-spongeSCR animals using Western Blotting. Briefly, 30μg of protein was run in a 4–20% MiniProtean TGX gel (BioRad) and then blotted on to nitrocellulose membrane. The membrane was then blocked with 1X TBST+5% milk and incubated overnight with 1:1000 of Rbt anti-Notch1 (D1E11) (Cell Signaling Technology) in blocking solution. After washing, the membrane was then incubated with 1:2000 Peroxidase anti-Rbt (Vector Laboratories), washed again and bands detected using WestPico Chemiluminescent Substrate (Pierce Biotechnology) and a BioRad ChemiDoc system. Two bands were detected for Notch 1 with the higher band known to be non-specific. The Notch 1 band at 110kDa was quantified using ImageJ software and normalized to the GAPDH bands detected using Mouse anti-GAPDH (Fitzgerald Laboratories).

Manipulation of miR-34a and memory consolidation

Using stereotaxic surgery, 1ul of LV-miR-34a-sponge or LV-mmu-miR-34a-spongeSCR was infused into the basolateral amygdala (BLA) over a period of 8 minutes using a Hamilton syringe pre-coated with 10% BSA (From Bregma: −1.4 A/P, −5.0 D/V, ± 3.4 M/L). The needle was left in place for 10 minutes after infusion and gradually withdrawn over 2 minutes. Viral titer was at least 2×109 iu/ml. Two weeks later, mice were habituated and fear conditioned in Context A as noted above. Consolidation of cued-fear memory was then tested 24 hours later in Context B as previously described. Following behavioral studies, brains were sectioned for histological confirmation of LV infection.

γ-secretase inhibition and memory consolidation

Mice were habituated and then fear conditioned in Context A as noted above. Ten minutes after training in Context A, mice were injected i.p. with either Vehicle (0.1 mL, 100% DMSO) or γ-secretase inhibitor N-[N-(3,5-Difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester (DAPT, Sigma-Aldrich) (75 mg/kg DAPT in 100% DMSO). The following day, consolidation of memory was tested in Context B as previously described.

Manipulation of Notch signaling and memory consolidation

Gain-of-function

Using stereotaxic surgery, 1ul of LV-Hes1 (Plasmid #17625 from Addgene) or a control LV-GFP was infused into the BLA as noted above (From Bregma: −1.4 A/P, −5.0 D/V, ± 3.4 M/L). Two weeks after infection, mice were habituated and fear conditioned in Context A as noted above. Consolidation of cued-fear memory was then tested 24 hours later in Context B as previously described.

Loss-of-function

Using stereotaxic surgery, guide cannulae were implanted such that infusions could be made into the BLA at −1.4 A/P, −5.0 D/V, ± 3.4 M/L from Bregma. One week after cannulation, mice were habituated and fear conditioned in Context A as noted above. Thirty minutes after conditioning, 0.5 ul of Notch inhibiting antibody or control antibody was infused into the BLA and animals were returned to the vivarium. Consolidation of cued-fear memory was then tested 24 hours later in Context B as previously described. To inhibit the Notch pathway, a previously used antibody-strategy was used. Briefly, 10 ug of Jagged-1 Fc (1277-JG: R&D systems) was incubated with anti-human Fc (i2136: Sigma) for one hour on ice before use, with the anti-human Fc being used as the control infusion. To visualize the localization of this antibody-based inhibition (Supplementary Figure 5), we conducted immunohistochemistry on fresh-frozen brain tissue that was harvested one hour after infusion of the antibodies into the BLA. Briefly, sections were fixed in 4% Paraformaldehyde for 10 minutes, washed with 1X PBS and then incubated overnight with biotinylated anti-goat antibody (Vector laboratories, 1:500), after rinsing in 1X PBS the sections were incubated for two hours with 1:1000 Streptavidin-conjugated to AlexaFluor-568, coverslipped and visualized using an epi-fluorescent microscope.

Interaction between miR-34a and Notch signaling in memory consolidation

Using stereotaxic surgery as noted above, 1ul of LV-miR-34a-sponge or LV-mmu-miR-34a-spongeSCR was infused into the basolateral amygdala (BLA) Two weeks later, mice were habituated and fear conditioned in Context A as noted above. Ten minutes later, animals were injected i.p. with either Vehicle (0.1 mL, 100% DMSO) or γ-secretase inhibitor N-[N-(3,5-Difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester (DAPT, Sigma-Aldrich) (75 mg/kg DAPT in 100% DMSO). Fifty minutes after the injection, these animals were exposed to 2 CS presentations in Context B to test for Short Term Memory (STM). Consolidation of cued-fear memory was then tested 24 hours later in Context B as previously described. Following behavioral studies, brains were sectioned for histological confirmation of LV infection.

Supplementary Material

01

Highlights.

  1. miR-34a action in Basolateral Amygdala (BLA) facilitates fear memory consolidation.

  2. miR-34a targets ligands and receptors of the Notch signaling pathway.

  3. miR-34a in BLA decreases Notch signaling to facilitate fear memory consolidation.

  4. Molecules canonically used during development have roles in adult behavior.

Acknowledgments

The authors thank Dr. Dieter Edbauer for generously providing us with cloning constructs, Xinping Huang for help with lentivirus production (Viral Vector Core, Emory University), Oskar Laur for help with cloning the miR-34a sponge and scrambled control (Custom Cloning Core Facility, Emory University), Dr. Andrew Miller and members of his laboratory for use of their luminometer for the luciferase assays, Dr. Shannon Gourley and members of her laboratory for use of the BioRad Imager for Western Blot visualization, and 3 anonymous reviewers for their constructive suggestions that have strengthened this version of the manuscript. This work was supported by funding to KJR by HHMI and Burroughs Wellcome Fund. The Viral Vector Core of the Emory Neuroscience NINDS Core Facilities is supported by grant P30NS055077. This project was partially funded by the National Center for Research Resources P51RR000165 and is currently supported by the Office of Research Infrastructure Programs/OD P51OD011132 to the Yerkes National Primate Research Center.

Footnotes

Accession number

The GEO accession number for the miRNA array is GSE59072.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Ables JL, Breunig JJ, Eisch AJ, Rakic P. Not(ch) just development: Notch signalling in the adult brain. Nature Publishing Group. 2011;12:269–283. doi: 10.1038/nrn3024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Carlson ME, Conboy IM. Regulating the Notch pathway in embryonic, adult and old stem cells. Curr Opin Pharmacol. 2007;7:303–309. doi: 10.1016/j.coph.2007.02.004. [DOI] [PubMed] [Google Scholar]
  3. Comery TA. Acute γ-Secretase Inhibition Improves Contextual Fear Conditioning in the Tg2576 Mouse Model of Alzheimer’s Disease. Journal of Neuroscience. 2005;25:8898–8902. doi: 10.1523/JNEUROSCI.2693-05.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Conboy IM. Notch-Mediated Restoration of Regenerative Potential to Aged Muscle. Science. 2003;302:1575–1577. doi: 10.1126/science.1087573. [DOI] [PubMed] [Google Scholar]
  5. Conboy L, Seymour CM, Monopoli MP, O’Sullivan NC, Murphy KJ, Regan CM. Notch signalling becomes transiently attenuated during long-term memory consolidation in adult Wistar rats. Neurobiol Learn Mem. 2007;88:342–351. doi: 10.1016/j.nlm.2007.04.006. [DOI] [PubMed] [Google Scholar]
  6. Costa RM, Honjo T, Silva AJ. Learning and memory deficits in Notch mutant mice. Curr Biol. 2003;13:1348–1354. doi: 10.1016/s0960-9822(03)00492-5. [DOI] [PubMed] [Google Scholar]
  7. Cowansage KK, Bush DEA, Josselyn SA, Klann E, Ledoux JE. Basal variability in CREB phosphorylation predicts trait-like differences in amygdala-dependent memory. Proc Natl Acad Sci USA. 2013;110:16645–16650. doi: 10.1073/pnas.1304665110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dash PK, Moore AN, Orsi SA. Blockade of γ-secretase activity within the hippocampus enhances long-term memory. Biochem Biophys Res Commun. 2005;338:777–782. doi: 10.1016/j.bbrc.2005.10.006. [DOI] [PubMed] [Google Scholar]
  9. Davis M. The role of the amygdala in fear and anxiety. Annu Rev Neurosci. 1992;15:353–375. doi: 10.1146/annurev.ne.15.030192.002033. [DOI] [PubMed] [Google Scholar]
  10. Davis M, Walker DL, Miles L, Grillon C. Phasic vs sustained fear in rats and humans: role of the extended amygdala in fear vs anxiety. Neuropsychopharmacology. 2010;35:105–135. doi: 10.1038/npp.2009.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. de Antonellis P, Medaglia C, Cusanelli E, Andolfo I, Liguori L, De Vita G, Carotenuto M, Bello A, Formiggini F, Galeone A, et al. MiR-34a targeting of Notch ligand delta-like 1 impairs CD15+/CD133+ tumor-propagating cells and supports neural differentiation in medulloblastoma. PLoS ONE. 2011;6:e24584. doi: 10.1371/journal.pone.0024584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dias BG, Ressler KJ. Parental olfactory experience influences behavior and neural structure in subsequent generations. Nat Neurosci. 2014;17:89–96. doi: 10.1038/nn.3594. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dias BG, Banerjee SB, Goodman JV, Ressler KJ. Towards new approaches to disorders of fear and anxiety. Current Opinion in Neurobiology. 2013:1–7. doi: 10.1016/j.conb.2013.01.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dudai Y. The neurobiology of consolidations, or, how stable is the engram? Annu Rev Psychol. 2004;55:51–86. doi: 10.1146/annurev.psych.55.090902.142050. [DOI] [PubMed] [Google Scholar]
  15. Edbauer D, Neilson JR, Foster KA, Wang CF, Seeburg DP, Batterton MN, Tada T, Dolan BM, Sharp PA, Sheng M. Regulation of Synaptic Structure and Function by FMRP-Associated MicroRNAs miR-125b and miR-132. Neuron. 2010;65:373–384. doi: 10.1016/j.neuron.2010.01.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fanselow MS, Poulos AM. The neuroscience of mammalian associative learning. Annu Rev Psychol. 2005;56:207–234. doi: 10.1146/annurev.psych.56.091103.070213. [DOI] [PubMed] [Google Scholar]
  17. Fitzgerald P, Seemann J, Maren S. Can fear extinction be enhanced? A review of pharmacological and behavioral findings. Brain Research Bulletin. 2013 doi: 10.1016/j.brainresbull.2013.12.007. S0361-9230, 00196-00192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Frankland PW, Josselyn SA, Anagnostaras SG, Kogan JH, Takahashi E, Silva AJ. Consolidation of CS and US representations in associative fear conditioning. Hippocampus. 2004;14:557–569. doi: 10.1002/hipo.10208. [DOI] [PubMed] [Google Scholar]
  19. Gafford GM, Guo JD, Flandreau EI, Hazra R, Rainnie DG, Ressler KJ. Cell-type specific deletion of GABA(A)α1 in corticotropin-releasing factor-containing neurons enhances anxiety and disrupts fear extinction. Proc Natl Acad Sci USA. 2012;109:16330–16335. doi: 10.1073/pnas.1119261109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gao J, Wang WY, Mao YW, Gräff J, Guan JS, Pan L, Mak G, Kim D, Su SC, Tsai LH. A novel pathway regulates memory and plasticity via SIRT1 and miR-134. Nature. 2010;466:1105–1109. doi: 10.1038/nature09271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gräff J, Mansuy IM. Epigenetic codes in cognition and behaviour. Behav Brain Res. 2008;192:70–87. doi: 10.1016/j.bbr.2008.01.021. [DOI] [PubMed] [Google Scholar]
  22. Griggs EM, Young EJ, Rumbaugh G, Miller CA. MicroRNA-182 Regulates Amygdala-Dependent Memory Formation. Journal of Neuroscience. 2013;33:1734–1740. doi: 10.1523/JNEUROSCI.2873-12.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Haramati S, Navon I, Issler O, Ezra-Nevo G, Gil S, Zwang R, Hornstein E, Chen A. microRNA as Repressors of Stress-Induced Anxiety: The Case of Amygdalar miR-34. Journal of Neuroscience. 2011;31:14191–14203. doi: 10.1523/JNEUROSCI.1673-11.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Johansen Joshua P, Cain Christopher K, Ostroff Linnaea E, Ledoux Joseph E. Molecular Mechanisms of Fear Learning and Memory. Cell. 2011;147:509–524. doi: 10.1016/j.cell.2011.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Josselyn SA, Shi C, Carlezon WA, Neve RL, Nestler EJ, Davis M. Long-term memory is facilitated by cAMP response element-binding protein overexpression in the amygdala. J Neurosci. 2001;21:2404–2412. doi: 10.1523/JNEUROSCI.21-07-02404.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kashat M, Azzouz L, Sarkar SH, Kong D, Li Y, Sarkar FH. Inactivation of AR and Notch-1 signaling by miR-34a attenuates prostate cancer aggressiveness. Am J Transl Res. 2012;4:432–442. [PMC free article] [PubMed] [Google Scholar]
  27. Konopka W, Kiryk A, Novak M, Herwerth M, Parkitna JR, Wawrzyniak M, Kowarsch A, Michaluk P, Dzwonek J, Arnsperger T, et al. MicroRNA Loss Enhances Learning and Memory in Mice. Journal of Neuroscience. 2010;30:14835–14842. doi: 10.1523/JNEUROSCI.3030-10.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Landeira-Fernandez J, DeCola JP, Kim JJ, Fanselow MS. Immediate shock deficit in fear conditioning: effects of shock manipulations. Behav Neurosci. 2006;120:873–879. doi: 10.1037/0735-7044.120.4.873. [DOI] [PubMed] [Google Scholar]
  29. Lim LP, Lau NC, Garrett-Engele P, Grimson A, Schelter JM, Castle J, Bartel DP, Linsley PS, Johnson JM. Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature. 2005;433:769–773. doi: 10.1038/nature03315. [DOI] [PubMed] [Google Scholar]
  30. Lin Q, Wei W, Coelho CM, Li X, Baker-Andresen D, Dudley K, Ratnu VS, Boskovic Z, Kobor MS, Sun YE, Bredy TW. The brain-specific microRNA miR-128b regulates the formation of fear-extinction memory. Nat Neurosci. 2011:1–3. doi: 10.1038/nn.2891. [DOI] [PubMed] [Google Scholar]
  31. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(−Delta Delta C(T)) Method. Methods. 2001;25:402–408. doi: 10.1006/meth.2001.1262. [DOI] [PubMed] [Google Scholar]
  32. Lois C, Hong EJ, Pease S, Brown EJ, Baltimore D. Germline transmission and tissue-specific expression of transgenes delivered by lentiviral vectors. Science. 2002;295:868–872. doi: 10.1126/science.1067081. [DOI] [PubMed] [Google Scholar]
  33. Maguschak KA, Ressler KJ. Wnt signaling in amygdala-dependent learning and memory. J Neurosci. 2011;31:13057–13067. doi: 10.1523/JNEUROSCI.3248-11.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Maguschak KA, Ressler KJ. A role for WNT/β-catenin signaling in the neural mechanisms of behavior. J Neuroimmune Pharmacol. 2012;7:763–773. doi: 10.1007/s11481-012-9350-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Maren S. The amygdala, synaptic plasticity, and fear memory. Ann N Y Acad Sci. 2003;985:106–113. doi: 10.1111/j.1749-6632.2003.tb07075.x. [DOI] [PubMed] [Google Scholar]
  36. Maren S, Ferrario CR, Corcoran KA, Desmond TJ, Frey KA. Protein synthesis in the amygdala, but not the auditory thalamus, is required for consolidation of Pavlovian fear conditioning in rats. Eur J Neurosci. 2003;18:3080–3088. doi: 10.1111/j.1460-9568.2003.03063.x. [DOI] [PubMed] [Google Scholar]
  37. Muiños-Gimeno M, Espinosa-Parrilla Y, Guidi M, Kagerbauer B, Sipilä T, Maron E, Pettai K, Kananen L, Navinés R, Martín-Santos R, et al. Human microRNAs miR-22, miR-138-2, miR-148a, and miR-488 Are Associated with Panic Disorder and Regulate Several Anxiety Candidate Genes and Related Pathways. BPS. 2011;69:526–533. doi: 10.1016/j.biopsych.2010.10.010. [DOI] [PubMed] [Google Scholar]
  38. Pang RTK, Leung CON, Ye TM, Liu W, Chiu PCN, Lam KKW, Lee KF, Yeung WSB. MicroRNA-34a suppresses invasion through downregulation of Notch1 and Jagged1 in cervical carcinoma and choriocarcinoma cells. Carcinogenesis. 2010;31:1037–1044. doi: 10.1093/carcin/bgq066. [DOI] [PubMed] [Google Scholar]
  39. Parsons RG, Ressler KJ. Implications of memory modulation for post-traumatic stress and fear disorders. Nat Neurosci. 2013;16:146–153. doi: 10.1038/nn.3296. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Paxinos G, Franklin K. The Mouse Brain in Stereotaxic Coordinates: Compact. 2. Elsevier Academic Press; San Diego, CA: 2004. [Google Scholar]
  41. Pierfelice T, Alberi L, Gaiano N. Notch in the Vertebrate Nervous System: An Old Dog with New Tricks. Neuron. 2011;69:840–855. doi: 10.1016/j.neuron.2011.02.031. [DOI] [PubMed] [Google Scholar]
  42. Rizzo P, Osipo C, Foreman K, Golde T, Osborne B, Miele L. Rational targeting of Notch signaling in cancer. Oncogene. 2008;27:5124–5131. doi: 10.1038/onc.2008.226. [DOI] [PubMed] [Google Scholar]
  43. Sail V, Hadden MK. Notch Pathway Modulators as Anticancer Chemotherapeutics. Annual Reports in Medicinal Chemistry. 2012;47:267–280. [Google Scholar]
  44. Sargin D, Botly LCP, Higgs G, Marsolais A, Frankland PW, Egan SE, Josselyn SA. Disrupting Jagged1-Notch signaling impairs spatial memory formation in adult mice. Neurobiol Learn Mem. 2013;103:39–49. doi: 10.1016/j.nlm.2013.03.001. [DOI] [PubMed] [Google Scholar]
  45. Sato C, Turkoz M, Dearborn JT, Wozniak DF, Kopan R, Hass MR. Loss of RBPj in postnatal excitatory neurons does not cause neurodegeneration or memory impairments in aged mice. PLoS ONE. 2012;7:e48180. doi: 10.1371/journal.pone.0048180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Sauerhöfer E, Pamplona FA, Bedenk B, Moll GH, Dawirs RR, von Hörsten S, Wotjak CT, Golub Y. Generalization of contextual fear depends on associative rather than non-associative memory components. Behav Brain Res. 2012;233:483–493. doi: 10.1016/j.bbr.2012.05.016. [DOI] [PubMed] [Google Scholar]
  47. Silva AJ, Giese KP. Plastic genes are in! Current Opinion in Neurobiology. 1994;4:413–420. doi: 10.1016/0959-4388(94)90104-x. [DOI] [PubMed] [Google Scholar]
  48. Spadaro PA, Bredy TW. Emerging role of non-coding RNA in neural plasticity, cognitive function, and neuropsychiatric disorders. Front Gene. 2012;3:1–16. doi: 10.3389/fgene.2012.00132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Steckler T, Risbrough V. Pharmacological treatment of PTSD - established and new approaches. Neuropharmacology. 2012;62:617–627. doi: 10.1016/j.neuropharm.2011.06.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Vanelzakker M, Dahlgren KM, Davis F, Dubois S, Shin L. From Pavlov to PTSD: The extinction of conditioned fear in rodents, humans, and anxiety disorders. Neurobiol Learn Mem. 2013 doi: 10.1016/j.nlm.2013.11.014. S1074-7427, 00242–00246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Wang W, Kwon EJ, Tsai LH. MicroRNAs in learning, memory, and neurological diseases. Learning & Memory. 2012;19:359–368. doi: 10.1101/lm.026492.112. [DOI] [PubMed] [Google Scholar]
  52. Zovkic IB, Guzman-Karlsson MC, Sweatt JD. Epigenetic regulation of memory formation and maintenance. Learning & Memory. 2013;20:61–74. doi: 10.1101/lm.026575.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Zovkic IB, Sweatt JD. Epigenetic Mechanisms in Learned Fear: Implications for PTSD. Neuropsychopharmacology. 2013:1–17. doi: 10.1038/npp.2012.79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zovoilis A, Agbemenyah HY, Agis-Balboa RC, Stilling RM, Edbauer D, Rao P, Farinelli L, Delalle I, Schmitt A, Falkai P, et al. microRNA-34c is a novel target to treat dementias. The EMBO Journal. 2011:1–10. doi: 10.1038/emboj.2011.327. [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

01
NIHMS616147-supplement-01.pdf (6MB, pdf)

RESOURCES