1. Neuroscience
Download icon

Noradrenergic projections from the locus coeruleus to the amygdala constrain fear memory reconsolidation

  1. Josué Haubrich  Is a corresponding author
  2. Matteo Bernabo
  3. Karim Nader
  1. Department of Psychology, McGill University, Canada
  2. Department of Neurology and Neurosurgery, McGill University, Canada
Research Article
Cite this article as: eLife 2020;9:e57010 doi: 10.7554/eLife.57010
6 figures and 5 additional files

Figures

Conceptual diagram showing differences in memory processes triggered by fear conditioning of different intensities.

(A) Auditory fear conditioning training consisting of a single tone-shock pairing (1P). The fear memory initially exists in an unstable short-term state and in order to persist as a long-term memory (LTM), it must undergo a time-dependent stabilization processes termed consolidation. Later, recall of the LTM causes memory to destabilize and become transiently unstable. Afterwards, for memory to persist it is restabilized by reconsolidation. (B) Auditory fear conditioning consisting of 10 tone-shock pairings (10P).

Fear conditioning training with 10 shocks creates memories that are resistant to undergo reconsolidation.

Here we replicated the behavioral findings reported by Wang et al., 2009. Animals were trained with either one tone-shock pairing (1P) or 10 tone-shock pairings (10P) A) One day after training a retention test was conducted. Animals in the 10P group displayed higher freezing levels than those trained with a 1P (N = 15 per group). (B) One day after training an extinction session was conducted with 20 tone presentations, followed by a retention test the next day. Unlike in the 1P group, 10P animals did not display extinction acquisition and retention and showed higher freezing levels at all time points (N = 7/8 per group). (C) One day after training a reconsolidation-blockade procedure was conducted with post-reactivation infusion of anisomycin in the BLA. The next day a retention test was conducted. Left: representation of infusion sites in the BLA. Right: the reconsolidation blockade procedure was effective in disrupting fear memory only in the 1P group (N = 9/10 per group). Graphs show the mean ± s.e.m. Individual values are represented with circles. *p<0.05. The full statistics are available in the Supplementary file 1 and individual values in Figure 2—source data 1.

Reconsolidation-resistant memories created with 10P display reduced plasticity mechanisms in comparison to memories that are reconsolidation-permissive.

(A) Animals were trained with either 1P or 10P and were tested the next day. One day later BLA samples were collected and the postsynaptic levels of GluN2B and GluA2 quantified. Left: in comparison with HC controls, GluN2B increases in the 1P group but not in the 10P group (N = 3/6 per group). Right: 1P rats displayed higher GluA2 levels than HC, and 10P pairing displayed higher GluA2 levels than all other groups (N = 6/10 per group). (B) Animals were trained in the 1P protocol and were tested the next day. BLA samples were collected either 1 hr or 1 day after test to probe for postsynaptic and extrasynaptic GluA2 expression. Left: postsynaptic GluA2 is at HC levels 1 hr after test and is increased 24 hr after test (N = 6/11 per group). Right: extrasynaptic GluA2 is increased 1 hr after test and at HC levels 24 hr after test (N = 4/8 per group). (C) Animals were trained in the 10P protocol and were tested the next day. BLA samples were collected either 1 hr or 1 day after test to probe for postsynaptic and extrasynaptic GluA2 expression. Left: postsynaptic GluA2 is increased in comparison to HC at both 1 hr and 24 hr after test (N = 4/6 per group). Right: Extrasynaptic GluA2 is at HC levels at both 1 hr and 24 hr after test (N = 3/7 per group). Graphs show the mean ± s.e.m. Individual values are represented with circles. *p<0.05. The full statistics are available in the Supplementary file 2 and individual values in Figure 3—source data 1.

The activation of β-adrenergic receptors is necessary memory to shift from a reconsolidation-permissive to a reconsolidation-resistant state.

Animals received an injection of propranolol (i.p. 10 mg/kg) or saline and 15 min later were trained in the strong training protocol. (A) One day after training rats were tested, and in the next day their BLA was collected to probe for postsynaptic GluN2B and GluA2. Left: propranolol treatment upregulated GluN2B (N = 9/10 per group). Right: propranolol treatment downregulated GluA2 (N = 5/6 per group). (B) One day after training a reconsolidation-blockade procedure was conducted with intra-BLA post-reactivation anisomycin. The next day a retention test was conducted. Left: representation of infusion sites in the BLA. Right: the reconsolidation blockade procedure was effective in disrupting fear memory only in animals receiving propranolol before training (N = 7/8 per group). Graphs show the mean ± s.e.m. Individual values are represented with circles. *p<0.05. The full statistics are available in the Supplementary file 3 and individual values in Figure 4—source data 1.

Projections from the LC to the BLA are necessary for the formation of reconsolidation resistant memories.

(A) Animals underwent 1P or 10P fear conditioning training, were sacrificed 90 min later and slices were collected to verify c-fos expression. Training resulted in greater LC c-fos expression in the 10P group than in the 1P group (N = 3 per group). (B) Animals were infused in the LC with pAAV-hM4Di-mCHerry or the control pAAV-tdTomato. Terminal projections in the BLA were silenced with local infusion of the DREADD agonist C21 (2 μg/μL, 0.5 uL/side). (C) Top: after 3 months from the viral injections, it was observed robust somatic expression in locus coeruleus neurons (right) and terminal expression in the BLA (left). Bottom: Light red represents the maximal, and dark red the minimum viral expression spread observed in the LC included in the analysis. (D) After 3 months from viral infusions, animals were infused in the BLA with C21 to block LC-BLA projections and were and trained in the 10P training protocol 5 min later. One day after training a reconsolidation-blockade procedure was conducted with intra-BLA post-reactivation anisomycin. The next day a retention test was conducted. Left: representation of infusion sites in the BLA. Right: the reconsolidation blockade procedure was effective in disrupting fear memory only in animals infused with the hM4Di virus (N = 7/9 per group). Graphs show the mean ± s.e.m. Individual values are represented with circles. *p<0.05. The full statistics are available in the Supplementary file 4 and individual values in Figure 5—source data 1.

The noradrenaline-locus coeruleus system drives memory formation towards a reconsolidation-resistant state.

(A) Fear learning engages plasticity in the amygdala that promotes long-term memory storage. Mildly aversive fear learning (1P) results in the formation of a memory where both GluN2Band GluA2 receptor subunits are upregulated. Extreme fear learning (10P) overactivates the noradrenaline-locus coeruleus systems, which prevents GluN2B upregulation and boosts GluA2 expression. (B) Upon recall, memory created with 1P fear conditioning is destabilized and undergoes reconsolidation. The memory formed under the ND-LC modulation does not.

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)