Spevack2008

Investigating auditory fear memory erasure in the basolateral amygdala Rachel Spevack* , Cyrinne Ben Mamou, Karim Nader Department of Psychology, McGill University, 1205 Avenue Docteur Penfield, Montreal, Quebec, H3A 1B1 Canada Abstract
New memories are initially fragile and need protein synthesis in order to be stabilized for long-term storage, a mechanism
called cellular consolidation. When recalled, memories are re-activated and become unstable again. They therefore need to
be re-stored through a process called reconsolidated. Behavioural studies in rats using auditory fear conditioning have
demonstrated that propranolol, a E-adrenergic receptor antagonist, administered following memory reactivation can
reduce fear expression (freezing), which has been interpreted as amnesia for the fear memory. It was recently discovered
that GluR1-containing AMPA receptors are recruited into the post-synaptic membrane of the basolateral amygdala during
auditory fear conditioning, suggesting that synaptic GluR1 increase may be a molecular correlate of long-term memory. The
present study aims to investigate what molecular mechanisms accounts for the observed amnesia following a reconsolida-
tion blockade by propranolol. Rats were trained in an auditory fear conditioning task, and fear memory reactivation was
followed by systemic propranolol administration. Rats were then euthanized and GluR1 protein levels in baso-lateral
amygdala synaptoneurosomes were quantified. We report preliminary evidence to suggest that a reconsolidation blockade
by propranolol reduces fear expression with a concomitant reduction in GluR1. Such evidence suggests that a reconsolida-
tion blockade might, at a molecular level, erase a component of the fear memory, providing support for the clinical utility of
this treatment for post-traumatic stress disorder.
Memory, the ability to acquire, store and recall learned infor- lar reactions that are thought to lead to the growth of new mation, is a fundamental feature of human experience. Ob- synapses and to the insertion of receptors into the post-sy- servations that new memories are initially sensitive to disrup- naptic membrane (for review see Milner, Squire, & Kandel, tion but strengthen over time laid the foundations for the 1998). Specifically, the α-amino-3-hydroxy-5-methylsoxazole- consolidation hypothesis (Muller & Pilzecker, 1900, as cited in 4-propionic acid receptor (AMPAr), another type of glutami- Dudai, 2004). This theory holds that memories pass through nergic receptor, is inserted into the post-synaptic membrane two qualitatively different states (Ebbinghaus, 1885 as cited and increases the neuron’s sensitivity to glutamate, increasing in Squire & Kandel, 2000). At the time of learning, a fragile the likelihood of synaptic transmission (for review see Malen- memory trace is formed in short-term memory (STM) and said to be labile as it is sensitive to disruption. By way of the AMPAr are composed of four subunits, known as GluR1 to consolidation process, a short term memory is converted into GluR4, that have different relative levels of insertion into post- a lasting and stable memory trace which exists in long-term synaptic membranes during LTP (Passafaro, Piech & Sheng, 2001; Rumpel, LeDoux, Zador, & Malinow, 2005; Yeh, Mao, Lin, The most striking evidence supporting consolidation & Gean, 2005). Both in vitro and in vivo rodent studies have de- theory comes from animal studies; amnesic agents such as monstrated increased insertion of only the GluR1-containing protein synthesis inhibitors (Flexner, Flexner, De La Haba, & AMPAr in the basolateral amygdala (BLA), the brain structure Roberts, 1965) or electro-convulsive shock (Duncan, 1949) ad- thought to be responsible for fear learning, following auditory ministered to animals shortly after learning resulted in amne- fear conditioning (Yeh et al, 2005; Rumpel et al, 2005). Moreo- sia, while the same treatment after a delay caused no memory ver, it was shown that blocking the synaptic incorporation of impairment. These results suggest that new memories must GluR1 (Rumpel, 2005) or knocking out the gene that codes undergo a time-dependent process to persist in LTM stores. for the GluR1 subunit (Humeau, Reisel, Johnson, Borchardt, Consolidation is thus defined as a stabilization process that Jensen, Gebhardt et al, 2007) impedes associative fear condi- renders a newly acquired memory stable and lasting. tioning. These findings indicate that GluR1 may provide an es- At the cellular level, consolidation occurs as a result of sy- sential contribution to the molecular mechanism of memory naptic changes following acquisition of newly learned infor- formation and maintenance, and thus may be a molecular mation. The synaptic plasticity hypothesis (Hebb, 1949) holds that the encoding of new memories results in structural modi- Previously, scientists believed that once the synaptic modi- fications of synaptic connections, causing persistent changes fications necessary for consolidation were made, the memory in synaptic strength. Long-term potentiation (LTP) is currently was permanently hardwired into the brain. Misanin, Miller & the leading model for this plasticity. LTP was first demonstra- Lewis (1968) challenged this hypothesis when they found that ted in vivo in rabbits, where it was shown that successive test 24 hours after the acquisition of a passive-avoidance task, a pulses in the neural pathways leading to the hippocampus cueing-procedure followed by electroconvulsive shock (ECS) were shown to increase the strength of active synapses (Bliss resulted in memory loss. This study was the first to suggest & Lomo, 1973). Numerous subsequent studies (Rogan, Ursula, that a consolidated memory could be susceptible to amnesic Staubli, & LeDoux, 1997) have shown that learning and LTP in- treatments. Recently, Nader, Schafe & LeDoux (2000) confir- volve similar cellular mechanisms, suggesting that LTP may be med these results using auditory fear conditioning. This robust the mechanism by which new memories stabilize over time. learning paradigm involves the pairing of a tone, the conditio- LTP is a process triggered by the activation of the excita- ned stimulus (CS), with a footshock, an aversive unconditio- tory glutaminergic N-methyl d-aspartate receptors (NMDAr). ned stimulus (US). As such, rats learn that the CS predicts the Activation of the NMDAr results in an influx of Ca2+, which footshock, and eventually fears the CS when it is presented initiates a cascade of protein synthesis dependent intracellu- alone. Nader et. al. (2000) demonstrated that post-training * Corresponding author. E-mail: rachel.spevack@mail.mcgill.ca MSURJ • Volume 3, Issue 1
hormones at the time of a traumatic event may contribute to an ‘over-consolidation’ process that renders fear memories hi-ghly resistant to extinction (Orr, Metzger, Lasko, Macklin, Peri, & Pitman, 2000). Following from this, being a b-receptor an-tagonist, propranolol was used to block the memory-enhan-cing effects of stress hormones during consolidation in both rats (McGaugh, 2000) and humans (Chamberlain et al., 2006). Pitman, Sanders, Zusman, Healy, Cheema, Lasko et al (2002) administered propranolol to humans immediately following an acute traumatic event, and found that this administration reduced subsequent PTSD symptoms. Since propranolol blocks consolidation in both rats and humans it was believed that perhaps propranolol might block memory reconsolidation. Indeed, Debiec & Ledoux (2004, 2006) used auditory fear conditioning in rats and found that administration of propranolol following reactivation blocked the expression of fear memories, suggesting that proprano- Figure 1: The dynamic nature of memory
lol blocked reconsolidation. Their study provides an animal model for the treatment of traumatic memories by way of infusion of the protein synthesis inhibitor anisomycin into reconsolidation blockade by propranolol, and suggests that the BLA immediately following memory reactivation produ- propranolol may be an effective treatment for PTSD. ced amnesia. Interestingly, anisomycin infusion in the ab- Our study aims to investigate the molecular processes sence of reactivation did not produce amnesia, suggesting that might account for the observed amnesia, as the me- the observed amnesia hinges on memory reactivation. The chanism by which reconsolidation blockade reduces fear authors concluded that if a protein synthesis inhibitor in- expression in rats is currently not clear. Although it has been fused after reactivation causes amnesia, then reactivation proposed that blocking reconsolidation actually removes must destabilize a well-consolidated memory and instigate a part of the memory (Nader, 2007), this has not yet been a second stabilization process requiring de novo protein validated at a molecular level. Furthermore, since GluR1 is synthesis in order for the memory to persist in LTM. This thought to represent a component of the memory trace,it process, termed reconsolidation, suggests that persistence is apt to examine whether a blockade of reconsolidation by of memory involves not only the storage of memories fol- propranolol can decrease this molecular memory tag. This lowing their acquisition, but also the re-storage of these would validate the hypothesis that the reduction of free- zing observed by reconsolidation blockade can be directly Reconsolidation has been demonstrated across species ranging from C. elegans to humans, and for different types Using an auditory fear conditioning paradigm, three hy- of learning, from aversive to appetitive conditioning (Nader, potheses were tested. Firstly, when propranolol is injected 2007). The memory process is therefore believed to be dyna- following memory reactivation, can it reduce freezing beha- mic in nature, where a memory cycles between being in an viour? Secondly, does fear conditioning cause a measurable active and an inactive state. As such, reconsolidation studies increase in synaptic GluR1 three days after training? Lastly, suggest that it is not the time delay following encoding that does a blockade of reconsolidation by propranolol reverse determines the durability of a memory, but its qualitative the conditioning-induced increase in GluR1? In order to test state; memories are in an active state when they are acquired these hypotheses, 12 rats were randomly assigned to four and recalled, following which they require a time-dependent experimental groups (n=3) (Table 1). The rats in the propra- stabilization process to persist in LTM (see Figure 1). nolol group (CS+P) were habituated, trained, and were gi- The discovery of reconsolidation has important theoretical ven a propranolol injection immediately following reactiva- implications, as it challenges the long-standing notion that tion. The rats in the vehicle group (CS+V) were habituated, memories are fixed when consolidated, but also introduces a trained, and given a saline injection immediately following novel clinical treatment for disorders associated with debilita- reactivation. The rats in the non-reactivated propranolol ting traumatic memories, such as post-traumatic stress disor- group (No CS+P) were habituated, trained, and given a pro- der (PTSD). Anisomycin, the amnesic agent used by Nader et. pranolol injection without memory reactivation. This group al. (2000), cannot be administered to humans, and as a result is a necessary control to determine that propranolol, in the drugs with a similar amnesic effect that can be administered absence of reactivation, does not affect memory proces- to humans have been investigated to determine the validity sing. The rats in the naïve group were habituated and admi- nistered saline. This last group provides a baseline measure Propranolol, a β-receptor antagonist, has been examined because of the involvement of endogenous stress hormones in the memory system and because propranolol can safely be administered to humans. When a rat experiences a footshock, or when a human undergoes a traumatic experience, stress Twelve adult male Sprague-Dawley rats from Charles River hormones are released (McGaugh, 2000). Also, β–adrenergic Laboratories, weighing 275-300g on arrival, were individually agonists infused into the amygdala in rats (Cahill & McGaugh, housed and maintained on a 12/12-hour light/dark cycle, with 1996) and injected in humans (Chamberlain, Muller, Blac- lights on at 7:00 a.m. All testing was performed during the li- kwell, Robbins & Sahakian, 2006) enhance memory storage. ght period. Rats were handled once a day for three consecu- Put together, these findings suggest that the release of stress March 2008 • msurj.mcgill.ca
Can fear memories be erased?
minimize generalization between the two contexts, and to eliminate any contextual conditioning, ensuring that the CS is the primary predictor of US exposure.
Training. On the fourth day, only rats in the CS+P, CS+V and No CS+P were transported to a brightly lit waiting area and remained there for five minutes before training began in Context A. Rats were then individually placed in a chamber, and after a two minute acclimatizing period, they were given three forward-presented pairings of the CS-US. The CS was a 5 KHz, 65 dB, 30 s tone that co-terminated with the US, a 1.5 Table 1: Drive efficiency of the SNOM probe as indicated by the ratio
mA , 1 s footshock delivered through the metal grid floor. The between the drive amplitude and the amplitude of the cantilever tunes inter-trial interval (ITI) between each tone-shock presenta- tion was 60 s. Rats were then returned to their home cages. Two distinct test chambers were used for this study (Med-As- Reactivation. On day five, twenty-four hours after trai- sociates). The first chamber, (context A) was used for training. ning, when the fear memory is thought to have completed It had a metal grid floor (1.5 cm bar spacing), stainless steel si- cellular consolidation (Nader et al., 2000), rats in the CS+P, dewalls, a transparent Plexiglas front wall, and was enclosed CS+V and no CS+P groups were transported to a dimly lit in a custom-built sound-attenuating isolation cubicle. During room and remained there for five minutes. Rats in the CS+P training, all lights and ventilation fans in each cubicle were on. and CS+V groups were then placed in Context B. After 120 s of The second chamber (context B) was used for habituation acclimatization, a single tone (5 KHz, 65 dB, 30 s) was played, and tone testing. These chambers were located in a second but no shock was given. Immediately following the memory room of the laboratory. Each chamber was 30 x 25 x 30 cm; reactivation, the animals in the CS+P group received an intra- two sidewalls were stainless steel and two were Plexiglas with peritoneal injection of propranolol (Sigma Aldrich, Ontario), an opaque sheet of alternating 2 cm-wide vertical black and while animals in the CS+V group received a saline injection. white stripes. The floor was a Plexiglas opaque white surface, Rats in the no CS+P group received the propranolol injection scented with peppermint before each rat was inserted. Each at this time, without undergoing the memory reactivation conditioning chamber was enclosed in a sound-attenuating session. Propranolol was dissolved in a saline solution at a isolation cubicle. The house lights were dimmed while the dose of 10 mg/ml and administered at a dose of 20 mg/kg, ventilation fans remained off. The amount of time each rat as used in Debiec & Ledoux (2004, 2006). Freezing behaviour, spent freezing during the 30 s interval preceding the CS pre- operationally defined as the cessation of all movement ex- sentation for each test, as well as the time spent freezing du- cept respiratory-related movements (LeDoux, 2000), is a spe- cies-typical fear response, and is used as a measure of fear. Freezing behaviour was scored with Freeze-View software (Actimetrics) by an experimenter blind to the experimental A Cryostat (Microm Instrumentation, Germany) was used to collect amygdala slices. A Teflon homogenizer was used to Post-reactivation short-term memory test (PR-STM). On homogenize tissue samples in a buffer to fractionate cells. day five, four hours following reactivation, rats in the CS+P, Millipore filters were used to filter the cell fractions from other CS+V and no CS+P group underwent the PR-STM test. These cellular components. A microcentrifuge was used to separate rats were transported to a dimly lit room and remained there the heavier fractions, enriched with synaptoneurosomes, for five minutes. Rats were then placed into Context A, and from lighter ones (less than 5 μm). An electrophoresis module after a 120 s acclimatization period three tones with the same (Bio-rad) was used in the present study to further separate sy- parameters as used in the reactivation trials were delivered, naptic GluR1 from other cellular proteins. A transfer module (Bio-rad) allowed a current to run through the polyacrylamide Typically, PR-STM tests are performed as an internal gel apposed to a PVDF (Millipore) membrane, and was used control in order to rule out any nonspecific effects of the to transfer proteins from the gel to the membrane. The Storm drug, in this case the effects of propranolol on the memory Laser scanner (Storm 860, Amersham Biosciences) was used abilities of the rats (Nader et al, 2000). The rats in the CS+P, to quantify GluR1 proteins in the PVDF membrane via Image CS+V, and no CS+P groups should show no significant dif- Quant software (Blot Imaging System).
ference in freezing on this test, indicating that propranolol is affecting long-term memory storage without interfering Habituation. All rats were habituated for 15 minutes on Post-reactivation long-term memory test (PR-LTM). On day three consecutive days in Context B. This was intended to ac- six, twenty hours after the PR-STM test, rats in the propra- climate the rats to the lab and the behavioural chambers, to nolol, vehicle and non-reactivated propranolol group were MSURJ • Volume 3, Issue 1
Rachel Spevack , Cyrinne Ben Mamou, Karim Nader
Figure 3: Averaged freezing data for the
three experimental groups (CS+V, CS+P,
No CS+P) in each experimental condi-tion (Reactivation, PR-STM and PR-LTM).
transported to a dimly lit room and remained there for five zed using a one-way analysis of variance (ANOVA). Significant minutes. Rats were placed into Context A, and after a 120 s interactions for both behavioural and molecular data were acclimatization period, three tones with the same parameters further analyzed using Tukey’s post-hoc test. Type one error as used in the reactivation and PR-STM trials were delivered rate was set at 0.05. Statistica version 6.0 statistical software Euthanasia. On day seven, twenty-four hours after the PR-LTM test, the animals were deeply anesthetized with urethane (1 ml/kg) and decapitated. Brains were dissected, For both the PR-STM and PR-LTM tests, the average freezing immediately frozen on dry ice, and stored at -80 ºC for later score (time spent freezing) across the three CSs was calcula- ted (CSavg). In order to distinguish and isolate the freezing induced by the CS from the freezing induced by the context, Brains were sliced at -20 ºC on a cryostat until the BLA we took the CSavg for each rat and divided it by the sum of was reached. BLA tissues were collected with a hollow nee- that rat’s pre-CS and CSavg freezing. This allowed us to isolate dle and homogenized at 4 ºC using a Teflon tissue grinder in each rat’s average freezing to the tone for each test, and to lysis buffer consisting of 10 mM Hepes/1.0 mM EDTA/2 mM factor out any contextual freezing (Figure 3). All statistics were Synaptoneurosome preparation. Homogenates were passed To determine that there were no significant differences through two 100-μm-pore nylon mesh filters, then through a between the propranolol and vehicle groups during reacti- 5-μm-pore filter. Filtered homogenates were centrifuged at vation, an independent-samples t-test was conducted. The 3600 g for 10 min at 4°C. Resultant pellets were resuspended in two groups demonstrated comparable freezing during reac- 20 μL boiling 1% SDS for 10 min and stored at -80°C. tivation, as an independent-groups t-test, t(4) = -1.589, p > Western Blot Analysis. Equal amounts of proteins (30 µg) 0.05, revealed freezing of animals in the vehicle group (M = from each sample were boiled for 10 min in SDS electropho- 87.34, SEM = 5.67) was not significantly different than ani- resis sample Laemmi buffer containing beta-mercaptoetha- mals in the propranolol group (M = 73.40, SEM = 19.64). A nol (Bio-Rad) and were run on a 8% SDS-polyacrylamide gel, two-way mixed design ANOVA, with group (CS+P, CS+V) as along with a molecular weight marker (BioRad), and transfer- between-subjects factor and test (PR-STM, PR-LTM) as within- red to PVDF membranes (Bio-rad, Hercules, CA). The mem- subjects factor revealed non-significant results (F(1, 4) = 0.15, branes were incubated overnight at 4°C in blocking solution p > 0.05). However, it was observed; on the PR-LTM test, the (0.1 % Tween 20 and 2% bovine serum albumin in TBS), and propranolol rats froze less (M = 45.48, SEM = 14.03) than the incubated for 2 hrs at room temperature with the anti-GluR1 antibody (Chemicon). After being washed in TBS, membra- To further quantify the amnesia observed in the propra- nes were incubated with the secondary fluorescent antibody nolol group, we calculated the Amnesia Index for each group (Figure 4) (Debiec, LeDoux & Nader, 2002). This was calcula- GluR1 quantification. The membranes were scanned with ted by dividing each rat’s freezing score during PR-LTM by a Storm Laser scanner and the signals quantified.
its freezing score during reactivation. A t-test, t(4) = 0.66, p > 0.05, was conducted on the averaged values and indicated no significant difference between the mean amnesia index of the vehicle group (M = 93.13, SEM = 9.33) compared to that The behavioural data was analyzed using independent-sam- of the propranolol group (M = 68.98, SEM = 35.39). However, ples t-tests or a two-way mixed analysis of variance (ANOVA), the propranolol group did freeze 24.15% less than the vehicle with testing interval as the within-subjects factor and group group, indicating that propranolol treatment following reac- as the between-subjects factor. The molecular data was analy- March 2008 • msurj.mcgill.ca
Can fear memories be erased?
would be impaired at 24 hours, but not four hours, following We divided the BLA synaptoneurosome GluR1 levels of each rat by the average naïve value and ran the statistics on this Another challenge to the observed amnesia resulting normalized data (Figure 5). A one-way ANOVA, F (3, 8) = 4.52, p from a reconsolidation blockade was advanced by Lattal & < 0.05, revealed a main effect of group. A Tukey post-hoc test Abel (2004), who suggest that the rats experience a transient revealed a significant difference between the GluR1 levels of impairment in the ability to retrieve the memory. Supporting the naïve group (M = 100, SEM = 13.64) and the vehicle group this view, they found that rats, when tested a day after recei- (M = 160.72, SEM = 8.55; p < 0.05). The Tukey post-hoc test ving anisomycin treatment following reactivation, were am- also revealed lower levels of GluR1 in the propranolol group nesic. However, when these animals were tested 21 days after (M = 125.29, SEM = 8.20) as compared to the vehicle group this manipulation, the anisomycin-infused rats froze similarly (M = 160.72, SEM = 8.55), although these results were not sta- to control rats, indicating that the fear memory spontaneously tistically significant (p> 0.05). The non-reactivated proprano- recovered over time. From these findings, they suggested the lol group showed higher levels of GluR1 (M = 136.09, SEM = fear memory remains intact following reconsolidation bloc- 15.30) than the propranolol group (M = 125.29, SEM = 8.20), kade, and the amnesia observed after one day reflects a tran- and lower levels than the vehicle group (= 160.72, SEM = 8.55), sient inability to retrieve the memory (Lattal & Abel, 2004). but not as low as the naïve group, all of which were non-signi- The debate as to the nature of the observed amnesia results from the use of a behavioural measure of memory retention; when a rat does not exhibit fear responses fol- lowing some manipulation, it is inferred that the rat is am- Although not statistically significant, the results obtained de- nesic for this memory. Since there is no valid molecular monstrate trends in the direction of our hypotheses. With re- measurement for the integrity of a memory, behaviour is gards to the first hypothesis, we demonstrated that the propra- the only way to assess memory retention. The present study, nolol group demonstrated a non-significant decrease in freezing however, provides preliminary evidence to suggest that the on the PR-LTM test, thereby partially replicating Debiec and Le- observed amnesia is directly attributable to a reduction of Doux’s (2004) finding. The second hypothesis was supported; GluR1, a molecular correlate of the memory. This therefore trained rats that received a saline injection showed a significant suggests that the amnesia following a reconsolidation bloc- increase in synaptic GluR1 levels in the BLA compared to naïve kade by propranolol may represent “true amnesia”, in that it rats, which replicates the study by Yeh et al (2005). Interestin- actually erases a component of the memory. gly, the results of the present study are, to our knowledge, the In light of the preliminary findings of the current study, first demonstration that an increase in GluR1 can be detected questions still remain as to how much of the memory is three days after training, providing further evidence that GluR1 erased. A recent study conducted by Rose & Rankin (2006) insertion in post-synaptic membranes can be viewed as a mo- investigated reconsolidation in the nematode C. elegans lecular component of the memory trace. The third hypothesis for a nonassociative learning task called habituation. The C. was partially supported; the propranolol group demonstrated elegans were repeatedly presented with a habituation sti- decreased freezing with a concomitant reduction in GluR1 mulus (a tap). Initially, the C. elegans swam in the opposite compared to trained rats, but these results were non-signifi- direction of the tap, but after repeated presentations, they cant and thus only trends can be reported. Nonetheless, since showed a decreased response to this stimulus. Twenty-four propranolol does not affect LTM in the absence of reactivation, hours after training, the habituation memory was reactiva- it appears that the effect of propranolol on the fear memory is ted by the presentation of several taps followed by the deli- not confounded by any effects it may have on memory systems very of a heat shock, which works like anisomycin to disrupt altogether. As such. these results suggest that a blockade of re- protein synthesis. When tested 24 hours after reactivation consolidation by propranolol may reduce fear responses while for the memory of the tap, these animals behaved like reducing a molecular correlate of the memory.
controls, suggesting the heat shock successfully blocked This molecular information helps clarify what molecu- reconsolidation. Interestingly, this study also investigated lar events account for the amnesia observed when reconso- GluR1 levels in the control and the heat-shock group, and lidation is blocked. Since behaviour is used as a measure of found that when reconsolidation was blocked, the C. ele- memory retention, there is some question as to the nature gans not only behaved like controls, but their GluR1 levels of the observed amnesia. Nader (2007) suggests reactivation were equivalent to that of controls. This suggested that a destabilizes a memory, and blocking its reconsolidation thus reconsolidation blockade actually re-set the GluR1 levels of interferes with its re-storage. Others suggest that the drug the trained rats to that of the untrained controls. administration following reactivation causes amnesia by da- The present study, although examining rodents using a maging the BLA. Specifically, Rudy, Biedenkapp, Moineau & fear conditioning paradigm, supports the results obtained Bolding (2006) cite studies showing that anisomycin can cau- in the Rose and Rankin (2006) study. Interestingly, Rose and se apoptosis, a form of programmed cell death. According to Rankin (2006) additionally suggest that the memory trace this view, the observed amnesia is not the result of deletion of is not only reduced, but actually erased, to the point whe- a fear memory, but rather from the destruction of the tissue re the C. elegans have no molecular trace for the learning that would normally store it. Typically, a PR-STM test is taken of this task. This stands in contrast to the data obtained in four hours after reactivation and drug infusion in order to the present study, as we demonstrate trends suggesting control for any nonspecific effects of the drug on the memory that propranolol reduced synaptic GluR1 to the level of the system altogether. This would seem to be a valid test for any trained saline group but not to the level of the naïve rats. lesion effects of a drug, and indeed with anisomycin, there are Importantly, Rose and Rankin (2006) provides additional no such observed deficits in freezing (Nader, 2000). However, evidence that the amnesia induced by a reconsolidation Rudy et al. (2006) argue that the lesion effect produced by blockade actually decreases a molecular component of the anisomycin could be delayed for several hours, so that a rat MSURJ • Volume 3, Issue 1
Rachel Spevack , Cyrinne Ben Mamou, Karim Nader
Figure 4 : Quantification of the amnesia induced by Propranolol as compared
Figure 5 : GluR1 synaptoneurosome quantification in the BLA for each experi-
mental group (Naïve, CS+P, CS+V, no CS+P) normalized to the Naïve. Beyond these theoretical implications, our study also regions of interest, as done in the Nader et. Al (2000) study supports the use of propranolol in clinical treatment for with anisomycin. This raises the possibility that propranolol PTSD. This is particularly significant because the most com- may exert nonspecific effects on the memory system. Al- mon treatment for PTSD is exposure-based psychotherapy, though we included a control group that received propra- a form of extinction involving a patient’s repeated expo- nolol without memory reactivation, propranolol may cause sure to the feared object or situation in the absence of any some long-term permanent changes in fear expression by overt danger. Although believed to attenuate the associa- peripheral sites of action , producing amnesia-like beha- ted emotional response, clinical experiments show it has a viour. Murchison, Zhang, Zhan, Lee, and Thomas (2004) poor long-term outcome (Davis, Myers, Chhatwal, Ressler, addressed such a problem; prior to fear conditioning, rats 2006). It was initially believed that extinction represented were given other β-adrenergic receptor antagonists such as “unlearning” at the synaptic level, in that it simply reversed nadolol and sotalol, which do not readily cross the blood- the plasticity associated with acquisition. Such a theory brain barrier, and found no effects on freezing. This suggests does not reflect the literature, as extinction in both rats and that the effects of propranolol are CNS-dependent, and the humans is not long-lasting (Myers & Davis, 2002). As a result, observed amnesia was not due to interference with fear me- it is currently believed that extinction is a new and distinct mory expression, but from its direct effects on the BLA. In ad- form of learning, resulting in the formation of an inhibito- dition, although in this study propranolol was administered ry association between the CS and US that competes with systemically, we detected a specific effect on synaptic GluR1 the original memory trace. This theory is more consistent levels in the BLA, indicating the behavioral effect of the drug with the literature, as the conditioned fear response often was mediated at least partially via the BLA.
returns when the animals are tested in a different context, Another limitation of the present study concerns the re-exposure to the US prior to testing reinstates the fear increase in GluR1. Although this study is the first to our memory, and the fear responses to the CS spontaneously knowledge that demonstrates that levels of GluR1 remain recover over time (Myers & Davis, 2002). elevated three days following the initial learning, this fin- A recent study conducted by Mao, Hsaio, Ya-Hsin, Gean, ding may be confounded by the test trials (reactivation, PR- & Po-Wu (2006) using a light-shock conditioning paradigm, STM, PT-LTM). It is possible that each test trial may reinstate found that extinction applied 24 hours after training redu- GluR1, helping to maintain elevated levels of GluR1 so that ced fear-potentiated freezing without influencing surface it can be detected three days after the initial learning. In or- GluR1 levels. From this, it was proposed that although ex- der to address this confound, a separate control group is re- tinction reduced fear-potentiated freezing at a behavioural quired where rats are conditioned and sacrificed three days level, it may not affect the original memory trace at a mo- later without any test trials. This would ensure that GluR1 lecular level, and this could explain why extinction training levels remain high even without the reminder trials, further is often short-lasting. In other words, this study suggests confirming that GluR1 is a molecular correlate of the me- that GluR1 might be responsible for the persistence of the memory after extinction. Interestingly, when DCS, a partial In conclusion, the preliminary results obtained in this ex- agonist for the glycine site on NMDAr was used, the rats’ periment need to be replicated with a larger sample size in fear-potentiated freezing was reduced, as was the condi- order to statistically validate the observed trends. Also, the tioning-induced increase in GluR1. From this, the authors use of central infusions of both propranolol and anisomycin suggested that extinction training with DCS may transform into the BLA would provide further evidence that a blockade the effect of light-alone trials from inhibitory learning (ex- of reconsolidation actually reduces synaptic GluR1 in the BLA. tinction) to erasure (reconsolidation blockade). The pre- This would directly correlate the observed behaviour with liminary results from our experiment further suggest that BLA synaptic GluR1, providing firmer conclusions. Nonethe- reconsolidation blockade may be an effective treatment for less, the present study does demonstrate that propranolol in- PTSD, as they indicate that a reconsolidation blockade may jection following reactivation, at a behavioural level, reduces actually decrease a portion of the emotional component of fear expression and, at a molecular level, reduces a correlate of long-term memory. These preliminary results suggest that One limitation of the present study is that propranolol a blockade of reconsolidation actually erases a component of was administered systemically and not infused into brain March 2008 • msurj.mcgill.ca
Can fear memories be erased?
tor trafficking. Annual N.Y. Academy Science, 1003, 1-11. 19. Mao, S-C. , Hsaio, Y-H. & Gean, P-W. (2006). Extinction trai- 1. Alberini, C. M. (2005). Mechanisms of memory stabiliza- ning in conjunction with a partial agonist of the glycine tion: are consolidation and reconsolidation similar or dis- site on the NMDA receptor erases memory trace. The tinct processes? TRENDS in Neuroscience, 28 (1), 51-56.
Journal of Neuroscience, 26 (35), 8892-8899.
2. American Psychiatric Association. (1994). Diagnostic and 20. McGaugh, J. L. (2000). Memory-a century of consolida- statistical manual of mental disorders (4th ed.). Washing- 21. Milner, B., Squire L. R. & Kandel, E. R. (1998). Cognitive 3. Bliss, T. V. P., & Lomo, T. (1973). Long-lasting potentiation Neuroscience and the Study of Memory. Neuron, 20(3), of synaptic transmission in the dentate area of the anaes- thetized rabbit following stimulation of the perforant 22. Misanin, J. R., Miller R. R. & Lewis, D. J. (1968). Retrograde am- path. Journal of Physiology, 232, 357-374.
nesia produced by electroconvulsive shock after reactiva- 4. Cahill, L. & McGaugh, J.L. (1996). Modulation of memory tion of a consolidated memory trace. Science, 160, 554-5.
storage. Current Opinion in Neurobiology, 6, 237-242. 23. Murchison, C. F., Zhang, X-Y., Zhang, W-P., Ouyang, M., 5. Chamberlain, S. R., Muller, U., Blackwell, A. D., Robbins, T. Lee, A. & Thomas, S. A. (2004). A distinct role for Norepi- W., Sahakian, B.J. (2006). Noradrenergic modulation of nephrine in memory retrieval. Cell, 117, 131-142. working memory and emotional memory in humans. 24. Myers, K. M. & Davis, M. (2002). Behavioral and neural analysis of extinction. Neuron, 36, 567-584. 6. Davis, M., Myers, K. M., Chhatwal, J., & Ressler, K. J. (2006). 25. Nader, K., Schafe G. E., & LeDoux, J. E. (2000). Fear memo- Pharmacological treatments that facilitate extinction of ries require protein synthesis in the amygdala for recon- fear: relevance to psychotherapy. Journal of the American solidation after retrieval. Nature, 406(6797), 722-726.
Society for Experimental Neurotherapeutics, 3, 82-96.
26. Nader, K. (2007). A single standard for memory; the case 7. Debiec, J., LeDoux , J. E. & Nader, K. (2002). Cellular and for reconsolidation. Debates in Neuroscience, In Press. systems reconsolidation in the hippocampus. Neuron, 27. Orr, S. P., Metzger, J. L., Lasko, N. B., Macklin, M. L., Peri, T., & Pitman, R. K. (2000). De novo conditioning in trau- 8. Debiec, J. & LeDoux, J. E., (2006). Noradrenergic signa- ma-exposed individuals with and without post-trauma- ling in the amygdala contributes to the reconsolidation tic stress disorder. Journal of Abnormal Psychology, 109, of fear memory. Ann. N.Y. Acad. Sci., 1071 ,521-524.
9. Debiec, J. & LeDoux, J. E., (2004). Disruption of reconso- 28. Orr, S. T., Milad, M. R., Metzger, L. J., Lasko, N. B., Gilbert- lidation but not consolidation of auditory fear conditio- son, M. W., & Pitman, R. K. (2006). Effects of beta blockade, ning by noradrenergic blockade in the amygdala. Neu- PTSD diagnosis, and explicit threat on the extinction and retention of an aversively conditioned response. Journal 10. Debiec, J., LeDoux, J.E. & Nader, K. (2002). Cellular and sys- tems reconsolidation in the hippocampus, 36(3), 527-538. 29. Passafaro, M., Piech, V. & Sheng, M. (2001). Subunit-speci- 11. Dudai, Y. (2004). The neurobiology of consolidations, or, fic temporal and spatial patterns of AMPA receptor exo- how stable is the engram? Annual Review of Psychology, cytosis in hippocampal neurons. Nature Neuroscience, 12. Duncan, C. P. (1949). The retroactive effect of electros- 30. Pitman, R. K., Sanders, K. M., Zusman, R. M., Healy, A. R., hock on learning. Journal of Comparative and Physiolo- Cheema, F., Lasko, N. B., Cahill, L., Orr, S.P. (2002). Pilot stu- dy of secondary prevention of posttraumatic stress disor- 13. Flexner, L. B., Flexner, J. B., De La Haba, G. & Roberts, der with propranolol. Biological Psychiatry, 51, 189-142 R. B. (1965). Loss of memory as related to inhibition of 31. Riccio, D. C., Millin, P. M., & Bogart, A. R. (2006). Reconso- cerebral protein synthesis. Journal of Neurochemistry, lidation: a brief history, a retrieval view, and some recent issues. Learning and Memory, 13, 536-544.
14. Hebb, D. O. (1949). The Organization of Behaviour. Wiley, 32. Rose, J. K. & Rankin, C. H. (2006). Blocking memory re- consolidation reverses memory-associated changes in 15. Humeau, Y., Reisel, D., Johnson, A.W., Borchardt, T., Jen- glutamate receptor expression. Journal Neuroscience, sen, V., Gebhardt, C., Bosch, V., Gass, P., Bannerman, D.M., Good, M.A., Hvalby, O., Sprengel, R., & Lüthi1, A. (1997). 33. Rudy, J. W., Biedenkapp, J. C., Moineau, J. & Bolding, K. A pathway-specific function for different AMPA recep- (2006). Anisomycin and the reconsolidation hypothesis. tor subunits in amygdala long-term potentiation and fear conditioning. The Journal of Neuroscience, 27(41), 34. Rumpel, S., LeDoux, J., Zador, A. & Malinow, R. (2005). Postsynaptic receptor trafficking underlying a form of 16. Lattal, M. K. & Abel, T. (2004). Behavioral impairments associative learning. Science, 308, 83-88. caused by injections of the protein synthesis inhibitory 35. Squire, L. R. & Kandel, E. R. (2000). Memory: From Mind to anisomycin after contextual retrieval reverses with time. Molecules. Scientific American Library, New York. 36. Yeh, S-H., Mao, S-C., Lin, H-C. & Gean, P-W. (2006). Synap- 17. Lewis D. J. (1979). Psychobiology of active and inactive tic expression of glutamate receptor after encoding of fear memory in the rat amygdala. Molecular Pharmaco- 18. Melenka, R.C. (2003). Synaptic plasticity and AMPA recep-

Source: http://msurj.mcgill.ca/vol3/iss1/spevack2008.pdf