The role of context in the re-extinction of learned fear
Introduction
Current neurobiological models of extinction maintain that this process is: (1) dependent on activation of the NMDA receptor, and (2) mediated by a neural circuit involving the medial prefrontal cortex (mPFC), amygdala, and hippocampus. There is substantial empirical evidence for both of these claims. For example, administration of the NMDAr antagonist MK-801 has been shown to impair the extinction of learned fear (Baker and Azorlosa, 1996, Lee et al., 2006). This same finding has been reported with other NMDAr antagonists (e.g., APV: Falls, Miserendino, & Davis,1992; ifenprodil: Sotres-Bayon, Bush, & LeDoux, 2007; CPP: Santini, Muller, & Quirk, 2001). Furthermore, administration of the partial NMDAr agonist d-Cycloserine (DCS) facilitates extinction of learned fear (Ledgerwood et al., 2003, Walker et al., 2002). These effects of NMDAr antagonists and agonists on extinction are seen whether these agents are given systemically or centrally (see Myers & Davis, 2007 for an extensive review). In regard to the neural circuit mediating the extinction of learned fear, there is evidence supporting the idea that this circuit involves the mPFC, amygdala, and hippocampus (see Corcoran & Quirk, 2007 for a review). For example, extinction is impaired when an NMDAr antagonist (Goosens and Maren, 2004, Sotres-Bayon et al., 2007) or a GABAA agonist (Laurent, Marchand, & Westbrook, 2008) is infused into the lateral amygdala before extinction training. Additionally, bilateral infusions of the NMDAr antagonist CPP in the mPFC both prior to and following extinction training impairs extinction (Burgos-Robles, Vidal-Gonzalez, Santini, & Quirk, 2007). Finally, hippocampal inactivation has been shown to impair the retrieval of extinction (Corcoran, Desmond, Frey, & Maren, 2005).
Extinction is currently attracting considerable interest. From a theoretical perspective, this seemingly simple phenomenon, which was described by Pavlov (1927), has proven to be quite difficult to fully explain. Early theoretical models suggested that extinction was due to the elimination, or erasure, of the excitatory CS–US association established during the fear conditioning phase of the experiment (Rescorla & Wagner, 1972) or to the devaluation of the US representation (Rescorla, 1973). However, the currently accepted theoretical model of extinction maintains that it is due to new learning (i.e., of a context-specific inhibitory ‘CS–no US’ memory) that then competes with the original excitatory CS–US memory (Bouton, 1991). There is no doubt that this particular theoretical model captures much of the behavioral data available on extinction, especially the return of conditioned responding following a change in context (renewal), the administration of an unsignalled US (reinstatement), or the passage of time (spontaneous recovery). From a practical perspective, extinction is generally acknowledged as a model for exposure-based therapies used to treat anxiety disorders. Therefore, advances in our understanding of the process of extinction may be translated into improved treatments for anxiety disorders (Hofmann, 2007, McNally, 2007).
Recently reported findings have raised some questions about the role of NMDAr in extinction and about the neural circuit underlying extinction. These new studies all involve what can be called “re-extinction”; that is, the animals are trained to fear a CS or context, have this fear extinguished, are then re-trained to fear the same CS or context, and then have this fear re-extinguished. Current models do not make any distinctions between extinction the first time and extinction the second time. That is, NMDA receptors should be critically involved in both cases and the same neural circuit should be activated in both cases. However, recent studies have not found this to be so. For example, Kim and Richardson (2008) compared the effect of amygdala inactivation prior to initial extinction and re-extinction of fear to a noise CS. In 24-day-old rats (rats this age exhibit extinction just like adult rats) it was found that inactivation of the amygdala, by local infusion of bupivicaine (a sodium channel modulator that prevents sodium influx into nerve cells, resulting in a local anaesthetic effect), prior to extinction training impaired long-term extinction whereas inactivation of the amygdala prior to re-extinction did not. It was clear that the amygdala was inactivated in both cases though as rats in each condition exhibited markedly lower levels of fear while the bupivicaine was on-board. Similarly, Laurent et al. (2008) reported that bilateral infusions of the GABAA agonist muscimol into the basolateral amygdala (BLA) impaired extinction of learned fear to a context, but not re-extinction of fear to that same context following reconditioning. Laurent et al. (2008) also found that bilateral BLA infusions of the NMDAr antagonist DL-APV impaired extinction but not re-extinction of learned fear to a context. In addition, Langton and Richardson (2008) reported that systemic injection of the partial NMDAr agonist DCS facilitates extinction of learned fear to a light CS, but not re-extinction of fear to that same CS following a subsequent reconditioning experience. Taken together, it seems that re-extinction may not require the amygdala, and that extinction makes a transition from being an NMDAr-dependent process to an NMDAr-independent process over the course of repeated conditioning and extinction sessions.
In the present study we explored another fundamental aspect of extinction according to current theoretical models; namely, the idea that extinction is contextually bound. As mentioned earlier, changing the context between extinction training and test often leads to a return of the learned response, a phenomenon referred to as renewal. This is a very robust effect, even observed following extinction in multiple contexts (Bouton, Garcia-Gutierrez, Zilski, & Moody, 2006). However, it is also the case that the findings of NMDA receptors and the amygdala being critically involved in extinction are also quite robust effects (see Myers & Davis, 2007 for a review). Given that these latter two fundamental aspects of extinction do not seem to occur in re-extinction we examined whether the role of context would be as important in re-extinction as it is in extinction. We explored this issue in two ways. In Experiment 1, we examined renewal following extinction and re-extinction. If re-extinction is context-bound, as is extinction, then one would expect the rats tested in a different context to where re-extinction training occurred to exhibit a renewal of fear. However, given that re-extinction does not appear to depend on the amygdala or NMDA receptors, perhaps rats will also fail to exhibit renewal following re-extinction.
The second way in which we explored the role of context in re-extinction is based on the recent finding by Langton and Richardson (2008) that the transition from an NMDAr-dependent extinction process to an NMDAr-independent re-extinction process is stimulus-specific. In that experiment, rats were reconditioned either to a previously conditioned and extinguished CS or to a novel CS and then received re-extinction training. If the CS was novel, DCS enhanced re-extinction, but if the CS was previously conditioned, then DCS did not enhance re-extinction. In Experiment 2 of the present study we examined whether the transition from NMDAr-dependent extinction to NMDAr-independent re-extinction is context-specific. The question of interest was whether NMDA receptors were involved in re-extinction when it occurred in a context different from initial extinction.
Section snippets
Subjects
Experimentally naive adult male Sprague–Dawley rats obtained from the School of Psychology’s breeding colony at the University of New South Wales were used. Rats were approximately 90 days old and were housed in groups of eight in plastic boxes in a colony room maintained on a 12-h light-dark cycle (lights on at 0700) with food and water continuously available. Rats were handled for 3 days prior to the start of the experiment, and were treated in accordance with the principles of laboratory
Experiment 1
The aim of Experiment 1 was to determine whether re-extinction is context-specific. Specifically, this experiment examined whether renewal was observed following initial extinction and re-extinction using an ABA procedure. The experimental design was a 2 × 2 factorial where the first factor was extinction (initial extinction or re-extinction) and the second factor was extinction/re-extinction context (A or B). All rats were trained and tested in Context A.
Experiments 2A and 2B
As previously mentioned, recent studies have demonstrated that extinction the second time around is NMDAr-independent (Langton and Richardson, 2008, Laurent et al., 2008). However, this transition from NMDAr-dependent extinction to NMDAr-independent re-extinction is stimulus-specific (Langton & Richardson, 2008) in that the transition is only observed if rats are re-trained and re-extinguished to the same CS. The aim of this set of experiments was to examine the context-specificity of this
General discussion
The current series of experiments used two different designs to explore the context-specificity of re-extinction. Experiment 1 used an ABA renewal design to compare the context-specificity of extinction and re-extinction. As expected, renewal was observed after extinction. A novel finding was that renewal was also observed following re-extinction. Experiments 2A and 2B examined the effect of the NMDAr antagonist MK-801 on extinction and re-extinction. The results of Experiment 2A replicated
Acknowledgments
This research was supported by an Australian Postgraduate Award (JL) and an Australian Research Council Discovery Project (DP0666953; RR). Reprint requests can be sent to Julia Langton, School of Psychology, University of New South Wales, Sydney Australia, 2052 ([email protected]).
References (35)
- et al.
Extinction in multiple contexts does not necessarily make extinction less vulnerable to relapse
Behaviour Research and Therapy
(2006) - et al.
Consolidation of fear extinction requires NMDA receptor-dependent bursting in the ventromedial prefrontal cortex
Neuron
(2007) - et al.
Attenuation of the renewal effect by extinction in multiple contexts
Learning and Motivation
(1999) - et al.
Conducting exposure treatment in multiple contexts can prevent relapse
Behaviour Research and Therapy
(1998) Enhancing exposure-based therapy from a translational research perspective
Behaviour Research and Therapy
(2007)- et al.
A developmental dissociation in reinstatement of an extinguished fear response in rats
Neurobiology of Learning and Memory
(2007) Mechanisms of exposure therapy: How neuroscience can improve psychological treatments for anxiety disorders
Clinical Psychology Review
(2007)- et al.
Ventral medial prefrontal cortex and emotional perseveration: The memory for prior extinction training
Behavioural Brain Research
(2003) - et al.
A meta-analysis of d-cycloserine and the facilitation of fear extinction and exposure therapy
Biological Psychiatry
(2008) The return of fear: review and prospect
Clinical Psychology Review
(1989)