Review
Pharmacological manipulation of memory reconsolidation: Towards a novel treatment of pathogenic memories

https://doi.org/10.1016/j.ejphar.2008.03.010Get rights and content

Abstract

Well-consolidated memories, when retrieved, may return to a transiently fragile state, and need to be consolidated again in order to be maintained. This process has been referred to as memory reconsolidation and presumably serves to modify or strengthen memory traces. In recent years, our understanding of the neurobiological mechanisms underlying this phenomenon has increased rapidly. Here, we will briefly review some of the pharmacological evidence, stressing a crucial role for the brain's major neurotransmitter systems, such as glutamate and noradrenaline, in memory reconsolidation. Pharmacological intervention of reconsolidation processes may have clinical relevance, especially for the treatment of psychiatric disorders that are characterized by pathological memories, including post-traumatic stress disorder and addictive behaviour.

Introduction

The traditional hypothesis of memory consolidation postulates, that during consolidation memory is initially labile after acquisition, but gradually becomes resistant to amnesic treatment. However, accumulating evidence indicates that, following retrieval, a well-consolidated memory trace may return to a labile state during which it can be modified again. For example, several studies in rodents have shown that protein synthesis inhibition immediately upon retrieval of memory impairs subsequent expression of this type of memory (Tronson and Taylor, 2007), thereby indicating that this memory updating mechanism, denoted as a reconsolidation process (similar to consolidation of memory) requires protein synthesis.

Behavioural pharmacological studies in rats and mice have provided evidence for the existence of this phenomenon, and significantly contributed to our comprehension of the neurobiological mechanisms associated with memory reconsolidation processes, by showing that certain pharmacological agents can induce retrograde amnesia when administered upon a short memory reactivation trial (Table 1). Moreover, these studies suggest that drugs interfering with memory reconsolidation processes might be of great clinical importance, particularly for the treatment of psychiatric disorders that are characterized by strong pathogenic memories, such as post-traumatic stress disorder and drug addiction.

In this article we will review how behavioural pharmacological studies in rodents have contributed to our understanding of the complex processes of memory reconsolidation. Further, we speculate on the potential therapeutic efficacy of drugs that can ‘erase’ pathogenic memories by disrupting memory reconsolidation.

Most experiments on memory reconsolidation thus far involved conditioned fear paradigms, i.e. procedures in which either a distinctive context (contextual fear conditioning) or tone (auditory fear conditioning) is paired with a foot shock, so that these contextual and/or auditory cues themselves induce a fear response, such as freezing behaviour. In memory reconsolidation experiments, rats are (after training) briefly reexposed to this cue (without the shock) for a limited period in order to reactivate the cue-shock association, immediately after which they receive an injection of a drug to disturb memory reconsolidation. The next day(s), memory retention is tested in the shock context or in the presence of the tone and the amount of freezing behaviour is used as index of conditioned fear memory.

In addition to conditioned fear paradigms, appetitive learning paradigms have also recently been employed in reconsolidation research, for example, in a conditioned place preference paradigm (in which a context is associated with a pleasant stimulus such as the effect of a drug), and in food- or drug-rewarded (instrumental) learning tasks. Accordingly, either a decreased preference for the drug-paired context or a reduction in instrumental responding upon post-retrieval drug treatment is regarded to reflect a disruption of memory reconsolidation processes.

Typically, a no-reactivation control condition (i.e. a non-contingent drug injection in the home cage) is included in reconsolidation experiments, in order to ensure that the drug effect depends on memory reactivation and thus affects molecular processes initiated by retrieval, rather than interfering with initial memory consolidation or other side-effects of the drug.

The contribution of glutamate receptors, especially the N-methyl-d-aspartate (NMDA) receptor, in learning and memory processes including long-term potentiation and consolidation has been well established. Recently, several studies have revealed that NMDA receptor activation is also critically involved in reconsolidation of memory.

Using a contextual fear paradigm, Suzuki and colleagues demonstrated that a systemic injection of the NMDA receptor antagonist [d(−)-3-(2-carboxypiperazine-4-yl)-propyl-1-phosphonic acid (CPP)] prior to brief, non-reinforced reexposure to the conditioning context, reduced conditioned freezing behaviour in mice when tested in the same context 24 h later (Suzuki et al., 2004). Furthermore, they showed that when the retrieval session was of longer duration, fear memory was extinguished in vehicle-treated animals when tested the following day. In contrast, fear memory was preserved in CPP-treated rats, indicating that NMDA receptor antagonism prevented extinction. These results indicate that a short reminder triggers a second cascade of consolidation, whereas a long reactivation session leads to the formation of a new memory that competes with the original memory, i.e. extinction. It also indicates that NMDA receptor dependent mechanisms may underlie both memory reconsolidation and extinction processes. Using the non-competitive NMDA receptor antagonist ((−)-5-methyl-10,11-dihydro-SH-dibenzo[a,d]cyclohepten-5,10-imine maleate (MK-801), Lee et al. (2006) demonstrated that NMDA receptor blockade prior to reactivation may impair reconsolidation or extinction of fear memory up to 7 days post-reactivation, depending on the duration of the reactivation session. Vice versa, d-cycloserine, a partial NMDA receptor agonist, administered either systemically or into the basolateral amygdala prior to recall, facilitated memory reconsolidation after short (i.e. 2 min) reexposure, or alternatively, blocked extinction of fear memory when reexposure was prolonged to 20 min.

Additional evidence for the role of amygdalar NMDA receptors in fear-related memory reconsolidation, was provided by Ben Mamou et al. showing that intra-amygdala infusions with d(−)-2-amino-5-phosphonovaleric acid (APV) prevented anisomycin produced amnesia of fear-related memory (Ben Mamou et al., 2006).

Besides fear-related memories, post-retrieval pharmacological manipulation of the NMDA receptor has also been found to interfere with reconsolidation of reward-related memories. For example, systemic administration of MK-801 prior to memory reactivation impaired subsequent performance on a food-rewarded spatial discrimination learning task (Przybyslawski and Sara, 1997). Moreover, an intracerebroventricular injection of APV upon retrieval disrupted the subsequent expression of an odor-reward association (Torras-Garcia et al., 2005). Importantly, the detrimental effect of APV on memory was retrieval-dependent, since rats receiving a non-contingent APV injection in the absence of memory reactivation showed perfect retention. Similar to food-related memories, reconsolidation of long-term drug-related memories also involves an NMDA receptor mediated mechanism. Thus, in conditioned place preference studies treatment with MK-801 before reexposure to the cocaine-associated context (Kelley et al., 2007) or after reexposure to the amphetamine-associated context (Sadler et al., 2007) appeared to block place preference. In the latter study, the memory-disrupting effects were only seen after four repeated treatments, consisting of memory reactivation tests followed by systemic MK-801, indicating that although long-term drug-related memories are less sensitive to interference, they can indeed be disrupted pharmacologically. Non-NMDA glutamate receptors, such as AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid)-like receptors and the metabotropic glutamate receptor mGlu5 receptor, may also be involved in memory reconsolidation processes. In this regard, it was shown that administration of the non-NMDA-type glutamate receptor antagonist 6,7-dinitroquinoxaline-2,3-dione (DNQX) during retrieval eliminated retention of non-associative memory in the worm Caenorhabditis elegans (Rose and Rankin, 2006). Additionally, in 1-day old chicks, treatment with the mGlu5 receptor antagonist MPEP (2-methyl-6-phenylethynyl pyridine hydrochloride) immediately after retrieval induced a transient impairment in passive avoidance performance (Salinska, 2006).

Noradrenaline neurotransmission plays a critical role in learning and memory processes, as evidenced by post-training increases in noradrenaline release (Tomie et al., 2004, Tronel et al., 2004) and more specifically, β-adrenoceptor involvement in long-term protestation (Gelinas and Nguyen, 2005, Straube and Frey, 2003), and consolidation of memory (McGaugh, 2002).

β-adrenergic receptor activation is also important for post-retrieval stabilization of memories, as systemic injections with the β-adrenoceptor antagonist propranolol impair expression of aversive memories in rats that received reactivation (Przybyslawski et al., 1999). Moreover, such memory-disrupting effect is still present 15 days after a single memory reactivation session followed by propranolol administration (Abrari et al., 2007) and was not reversed by a reminder shock (Abrari et al., 2008, Debiec and Ledoux, 2004).

Administration of β-adrenoceptor antagonists after memory retrieval also induced retrograde amnesia in a spatial memory task (Przybyslawski et al., 1999, Roullet and Sara, 1998) and an instrumental task measuring context-induced reinstatement of sucrose seeking behaviour (Diergaarde et al., 2006), suggesting that appetitive memories also undergo β-adrenoceptor dependent reconsolidation. Recently, it was shown by means of a self-administration paradigm, that propranolol, when administered immediately upon memory reactivation, persistently disrupts reconsolidation of Pavlovian associations between environmental conditioned stimuli and appetitive reinforcers including sucrose and cocaine (Milton et al., 2008). Finally, reactivation of drug-related memories and concomitant propranolol administration disrupted subsequent expression of cocaine (Bernardi et al., 2006) and morphine conditioned place preference (Robinson and Franklin, 2007). Collectively, these findings indicate that disrupting drug-related memories by means of β-adrenoceptor antagonism may constitute a new approach for aiding relapse prevention.

Whereas traditionally, dopamine neurotransmission has been implicated in reward and reinforcement, recent research suggests that this neurotransmitter is also critically involved in associative learning and memory. For example, medial prefrontal dopamine levels have been reported to rise during the early, but not later learning phases of an appetitive instrumental task (van der Meulen et al., 2007). Employing the same paradigm, dopamine release within the nucleus accumbens was found to be highest in rats that learned this task as compared to non-learning counterparts (Cheng and Feenstra, 2006). More specifically, the dopamine D1 receptor is required for appetitive as well as aversive memory formation, since memory impairments have been reported after post-training injections with the dopamine D1 receptor antagonist SCH 23390 (Dalley et al., 2005, Izquierdo et al., 2007). Despite the well-established role of dopamine transmission in memory consolidation, dopaminergic mechanisms involved in reconsolidation of memory have not received much attention. In this respect, it was demonstrated that cocaine treatment after memory reactivation enhanced active avoidance memory in rats (Rodriguez et al., 1993). More recently, amphetamine has been shown to enhance reconsolidation of a morphine conditioned place preference (Blaiss and Janak, 2006), whereas performance on an appetitive classical conditioning task was not affected by a post-retrieval amphetamine injection (Blaiss and Janak, 2007). Finally, it was shown that treatment with SCH23390 before retrieval impaired the subsequent expression of passive avoidance memory in chicks, indicating a role of dopamine D1 receptor involvement in memory reconsolidation (Sherry et al., 2005). However, it should be noted that since SCH23390 was administered prior to memory reactivation, this drug may have also affected memory retrieval by itself.

Taken together, since the number of studies is rather limited, no firm conclusions can as yet be drawn regarding the exact role of dopamine neurotransmission in reconsolidation of memory.

It is well documented that endogenous acetylcholine release is implicated in long-term memory consolidation (Power et al., 2003). In support of this, it was shown that intracerebroventricular injections with hemicholinium, a choline uptake inhibitor, may impair long-term expression of inhibitory avoidance memory in mice when administered immediately after training. Furthermore, inhibitory avoidance memory was impaired when hemicholinium was given after memory reactivation, but not in hemicholinium-treated mice not receiving reactivation (Boccia et al., 2004). Collectively, these data indicate hat both consolidation and reconsolidation of this type of aversive memory involve an acetylcholine-dependent mechanism. Consolidation of contextual fear memory was also shown to be under cholinergic control, as evidenced by impaired performance after post-training amygdala infusions with the muscarinic receptor antagonist scopolamine. However, post-reactivation blockade of the muscarinic receptor in the amygdala did not affect subsequent fear memory expression (Bucherelli et al., 2006). Hence, re-storage of contextual fear memory does not seem to involve activation of the muscarinergic receptor, at least not at the level of the amygdala.

In contrast, it was shown that post-retrieval systemic injections with scopolamine one week after training disrupted expression of a conditioned place preference for cocaine. These results indicate that, as opposed to contextual fear memory, post-retrieval stabilization of cocaine-associated contextual memory may indeed involve activation of muscarinergic receptors (Kelley et al., 2007). Further research, however, should clarify whether muscarinergic receptor antagonists may also disrupt drug memory reconsolidation in a drug self-administration paradigm.

Drugs acting at the GABAA receptor are widely prescribed to treat anxiety, insomnia, or epilepsy. Despite the wealth of studies showing that these drugs can induce amnesia, little is known regarding the involvement of this neurotransmitter in memory reconsolidation processes. In a recent study however, systemic injection with midazolam (a GABAA receptor agonist) after reexposure to the training context in rats, significantly reduced contextual freezing as compared to vehicle-treated control rats. This effect appeared to be long-lasting, as no spontaneous recovery of contextual fear memory was observed (Bustos et al., 2006). In another study, muscimol inactivation of the hippocampus upon memory reactivation disrupted performance on an inhibitory avoidance task (Amaral et al., 2007). The reported effect on memory was transient and additional studies are needed to verify whether the GABAA receptor constitutes another drug target for interfering with traumatic memories.

It is well established that glucocorticoid hormones may profoundly affect mnemonic functioning, including storage and retrieval of memory (Roozendaal, 2003). Thus, recent experiments suggest that reconsolidation of aversive memories can be disrupted by interfering with glucocorticoid receptor signalling, as indicated by long-term impairments of inhibitory avoidance upon post-retrieval amygdala infusions of the glucocorticoid receptor antagonist RU38486 (Tronel and Alberini, 2007). Furthermore, fear memory did not re-emerge after a reminder shock was administered, suggesting that the effect of glucocorticoid receptor antagonists on reconsolidation of fear-related memory may be persistent. Similar data were obtained in a study of Jin et al., that showed that RU486 infused in the amygdala impaired long-term auditory fear memory when memory was reactivated 10 days post-training (Jin et al., 2007). Retrograde amnesia has also been reported upon post-retrieval administration of the endogenous stress hormone corticosterone (Abrari et al., 2008, Cai et al., 2006). However, the authors suggested that rather than reflecting a disruption of memory reconsolidation, the observed reduction in freezing behaviour should be interpreted as enhanced memory extinction, which is consistent with previous findings (Yang et al., 2006).

In humans, acute administration of marijuana and its psychoactive constituents (i.e. cannabinoids) results in impaired encoding, consolidation and retrieval of memory (Ranganathan and D'Souza, 2006). In line with this, activation of the cannabinoid-1 (CB1) receptor by infusion of WIN55,212-2 into the insular cortex impaired acquisition of conditioned taste aversion memory (Kobilo et al., 2007). This report further showed that post-retrieval WIN55212-2 infusion into the insular cortex blocked expression of conditioned taste aversion memory, indicating that stimulating the brain endocannabinoid system disrupts both consolidation and reconsolidation of aversive memories. This was confirmed by a study showing that bilateral infusion of the cannabinoid CB1 receptor agonist WIN55212-2 or HU210 into the amygdala after memory retrieval, dose-dependently disrupted reconsolidation of fear memory, an effect that was reversed by co-administration of the cannabinoid CB1 receptor antagonist AM251. However, post-retrieval administration of cannabinoid CB1 receptor antagonists alone does not seem to affect reconsolidation of aversive memory, since unimpaired memory retention was observed after post-retrieval intra-amygdala AM251 infusions (Lin et al., 2006), or SR141716A either when given systemically after memory reactivation (Suzuki, et al., 2004) or when infused into the insular cortex (Kobilo et al., 2007). In contrast, Bucherelli et al. showed that, blockade of cannabinoid CB1 receptors in the amygdala following retrieval by a low dose of AM251, interferes with reconsolidation of contextual fear memory (Bucherelli et al., 2006). Given these inconsistent findings, together with the lack of studies involving appetitive learning paradigms, additional experiments are needed to address the precise role of the endocannabinoid system in post-retrieval processes.

Section snippets

Concluding remarks

Behavioural pharmacological experiments, such as those reviewed in this article, have contributed significantly to our understanding of the neurobiological mechanisms involved in memory reconsolidation. Especially the β-adrenoceptor and the NMDA receptor seem to play a key role in post-retrieval modification and stabilization of aversive as well as appetitive memories. Although the number of reconsolidation studies focusing on drugs acting at the dopamine D1-like receptor and the cannabinoid CB1

References (52)

  • McGaughJ.L.

    Memory consolidation and the amygdala: a systems perspective

    Trends Neurosci.

    (2002)
  • MilekicM.H. et al.

    Temporally graded requirement for protein synthesis following memory reactivation

    Neuron

    (2002)
  • PowerA.E. et al.

    Muscarinic cholinergic influences in memory consolidation

    Neurobiol. Learn. Mem.

    (2003)
  • Prado-AlcalaR.A. et al.

    Amygdala or hippocampus inactivation after retrieval induces temporary memory deficit

    Neurobiol. Learn. Mem.

    (2006)
  • PrzybyslawskiJ. et al.

    Reconsolidation of memory after its reactivation

    Behav. Brain Res.

    (1997)
  • RobinsonM.J. et al.

    Central but not peripheral beta-adrenergic antagonism blocks reconsolidation for a morphine place preference

    Behav. Brain Res.

    (2007)
  • RodriguezW.A. et al.

    Post-reactivation cocaine administration facilitates later acquisition of an avoidance response in rats

    Behav. Brain Res.

    (1993)
  • RoozendaalB.

    Systems mediating acute glucocorticoid effects on memory consolidation and retrieval

    Prog. Neuropsychopharmacol. Biol. Psych.

    (2003)
  • SalinskaE.

    The role of group I metabotropic glutamate receptors in memory consolidation and reconsolidation in the passive avoidance task in 1-day-old chicks

    Neurochem. Int.

    (2006)
  • SherryJ.M. et al.

    The effects of the dopamine D1 receptor antagonist SCH23390 on memory reconsolidation following reminder-activated retrieval in day-old chicks

    Neurobiol. Learn. Mem.

    (2005)
  • StraubeT. et al.

    Involvement of beta-adrenergic receptors in protein synthesis-dependent late long-term potentiation (LTP) in the dentate gyrus of freely moving rats: the critical role of the LTP induction strength

    Neuroscience

    (2003)
  • TomieA. et al.

    Pavlovian autoshaping procedures increase plasma corticosterone and levels of norepinephrine and serotonin in prefrontal cortex in rats

    Behav. Brain Res.

    (2004)
  • TronelS. et al.

    Persistent disruption of a traumatic memory by postretrieval inactivation of glucocorticoid receptors in the amygdala

    Biol. Psychiatry.

    (2007)
  • Ben MamouC. et al.

    NMDA receptors are critical for unleashing consolidated auditory fear memories

    Nat. Neurosci.

    (2006)
  • Bermudez-RattoniF. et al.

    Insular cortex is involved in consolidation of object recognition memory

    Learn. Mem.

    (2005)
  • BernardiR.E. et al.

    Postretrieval propranolol disrupts a cocaine conditioned place preference

    Neuroreport

    (2006)
  • Cited by (0)

    View full text