This section will discuss extinction of conditioned fear and how it is mediated by a protein called the N-methyl-D-aspartate (NMDA) receptor in the amygdala and medial prefrontal cortex. This will be followed by a review of the literature showing that a compound called D-cycloserine, which facilitates the NMDA receptor, speeding up extinction in animals and psychotherapy in people. Much of this progress can be attributed to the use of Paviovian fear conditioning as a model system. In this paradigm, an initially innocuous stimulus, the to-be conditioned stimulus (eg, a light, tone, or distinctive place) is paired with an innately aversive unconditioned stimulus (eg, a footshock in rats, a blast of air to the throat in humans) and the subject comes to exhibit a conditioned fear response to the conditioned stimulus. In rodents, fear is defined operationally as a cessation of all bodily movements except those required for respiration (freezing), an increase in the amplitude of an acoustically elicited startle response (fear-potentiated startle), an increase in blood pressure, changes in respiration, emission of ultrasonic distress calls, avoidance of the place where shock occurred, or several other possible measures, in the presence of the conditioned stimulus. In humans fear is typically measured as a change in skin conductance and increased startle when elicited in the presence of the conditioned stimulus. Unlike Pavlov's dog, which salivated when it heard the metronome, just as it did when it swallowed the dry food powder, the fear response may or may not mimic the unconditioned response to the aversive stimulus. For example, rats jump around when they are shocked, yet the conditioned fear response is just the opposite; they freeze and hold very still. Hence, fear is really a hypothetical construct that is used to describe the constellation of behaviors that are seen following fear conditioning, and these may or may not mimic what happens in the presence of the unconditioned stimulus.
Fear is a highly adaptive form of learning that prevents us from returning to a place where we were harmed (the alley where you were raped) or distraught (the airplane where we had a very bumpy flight) or contacting something that was harmful (a hot burner on a stove). Fear conditioning can be produced by a single training trial, and fear memories can last a lifetime. Normally, fear memories are suppressed by the process called fear extinction or habituation when the situation signals that these cues are no longer dangerous (eg, a soldier returning from combat) or when they are experienced over and over again in the absence of any negative consequence (eg, multiple smooth airplane flights). However, fear can become pathological if a person continues to be afraid in situations where they no longer should be afraid. For example, a soldier who is still afraid of a helicopter or the sound of a car backfiring long after he returned from service is no longer adaptive; he has a deficit in extinction or the ability to respond appropriately to safety signals (eg, as seen in post-traumatic stress disorder).
As mentioned above, extinction of fear refers to the reduction in the measured level of fear to a cue previously paired with an aversive event when that cue is presented repeatedly in the absence of the aversive event. Actually, the term extinction is used in several different ways in the literature. Extinction may refer to: (i) the experimental procedure used to produce a decrement in the fear response; (ii) the decremental effect of this procedure on the fear response, which can be measured both at the time the cue is presented in the absence of the aversive event and at a later time; or (iii) the hypothesized associative or cellular process responsible for that effect. As suggested elsewhere,we will define the experimental procedure as extinction training, the decrement in the fear response measured during extinction training as within-session extinction, and the decrement measured at some interval after extinction training as extinction retention. The term extinction will be reserved for the process underlying the loss of the fear response.
Extinction is not the same as forgetting
Although some forgetting of the original conditioned fear association may occur in extinction (see outstanding review),numerous studies show that extinction cannot fully be explained by forgetting because it requires exposure to the conditioned stimulus in the absence of the aversive event as opposed to the simple passage of time.
Extinction is generally cue-specific
Most studies show that fear extinction is cue-specific. For example, if a tone is paired with a shock and a light is paired with a shock, and then extinction training is only given to the tone, fear of the light will be undiminished. Generalization gradients of extinction are typically seen where the magnitude of extinction is greatest to the cue given during extinction training, and less so to cues along some continuum, such as a series of different auditory frequencies that received no extinction training.Generalization of extinction is negligible across cues drawn from different sensory modalities, or drawn from a single modality but differing substantially in their physical characteristics.
Extinction generally is not permanent
The decrement in conditioned fear responses during and shortly after extinction training generally is not permanent, as there are several instances in which extinguished fear responses are observed to reappear.
This refers to the reappearance of a fear response following exposure to unsignaled presentations of the unconditioned stimulus after the completion of extinction training.Unsignaled unconditioned stimulus presentations must occur within the context in which animals ultimately are tested if a return of fear is to be observed. , Thus, reinstatement seems to depend on context conditioning and is likely to involve summation of two fear-inducing tendencies, each behaviorally subthreshold when considered independently, but suprathreshold when combined: weak conditioning to context and residual conditioned fear to the extinguished stimulus (see ref 7).
Renewal refers to a reappearance of extinguished fear when animals are tested in a context different from the one in which extinction training took place. For example, when animals are first trained to fear a light in context A, then receive extinction training to the light in context B, and finally are tested for fear to the light in either context A or context B, different outcomes are obtained: animals tested in context B (the same context where extinction training took place) exhibit little fear to the light, whereas animals tested in context A exhibit robust fear to the light., A similar postextinction return of fear is observed when animals are tested in a third, novel context C following acquisition in context A and extinction in context B. , Thus, rather than learning that “now the cue is no longer paired with the shock,” the animal learns that “now, in this place, the cue is no longer paired with the shock.”
Spontaneous recovery refers to a reappearance of fear with the passage of time following extinction training in the absence of any further explicit training.
So, extinction is not a full erasure of the original fear memory but instead an active form of learning that acts to suppress or inhibit the original fear memory. This second learning process is referred to as “inhibitory” learning, as opposed to the original “excitatory” learning that occurred during pairings between the conditioned and the unconditioned stimulus. These two types of learning work at cross-purposes in terms of their tendency to stimulate or oppose, respectively, fear output, eg, refs 11-13. In other words, the conditioned stimulus emerges from extinction training with two meanings: following acquisition, the conditioned stimulus signals that an aversive event is coming, and following extinction, the conditioned stimulus signals that an aversive event will be withheld.
To account for recovery of fear following extinction, the inhibitory learning that accrues in extinction may not be expressed either because it is particularly “fragile” and subject to disruption or because it is gated by context, where “context” is defined broadly to include temporal and interoceptive cues, as well as spatial ones.That is, following extinction, appropriate behavior (no fear) is expressed within the temporal and spatial context of extinction training, whereas acquisition-appropriate behavior (fear) is expressed at other times and in other places.
Extinction may be “erased” under certain circumstances
Recently, however, new data have emerged in support of a mechanism more consistent with an “unlearning” account of extinction, in which plasticity underlying fear memory is reversed through a process known as synaptic depotentiation. Depotentiation refers to a reversal of long-term potentiation (ie, a return of potentiated synapses to baseline synaptic efficacy) when low-frequency stimulation is applied to afferent pathways shortly following induction of long-term potentiation. Evidence indicates that the biochemical and molecular mechanisms of depotentiation are opposite to those of long-term potentiation. For example, long-term potentiation is associated with membrane insertion of nonNMDA receptors.Depotentiation, by contrast, is associated with internalization of the same type of receptors (see ref 15).
Po-Wu Gean and colleagues demonstrated that depotentiation occurs in the amygdala., For example, depotentiation-inducing low-frequency stimulation of the amygdala in vivo 10 min after fear acquisition blocked the expression of conditioned fear 24 h later, an effect that could be interpreted as a mimicking of extinction. These findings are intriguing, but puzzling, because they would seem to offer no explanation of recovery of fear following extinction through reinstatement, renewal, or spontaneous recovery. Although “new learning” and “unlearning” mechanisms of extinction are often presented as mutually exclusive possibilities, it has been acknowledged that both may occur to some extent, eg, ref 2. Interestingly, depotentiation is inducible more readily at short intervals following induction of longterm potentiation and does not seem to be inducible at all at intervals greater than about 1 h (see ref 18). In rodents, extinction studies typically do not use intervals between acquisition and extinction training of less than 24 h, although biochemical processes of extinction were reported to be different when extinction training was conducted immediately following acquisition compared with 1 h or 3 h after extinction training.
To test the hypothesis that extinction training given shortly after conditioning might “erase” the original fear memory, rats were fear conditioned and then given extinction training either 10 min, 1 h, 24 h, or 3 days later.Consistent with an inhibitory learning mechanism of extinction, rats extinguished 24 or 72 h following acquisition exhibited moderate to strong reinstatement, renewal, and spontaneous recovery. By contrast, and consistent with an erasure mechanism, rats extinguished 10 min to 1 h after acquisition exhibited little or no reinstatement, renewal, or spontaneous recovery. These data support a model in which different neural mechanisms are recruited depending on the temporal delay of fear extinction. Based on these results, Dr Barbara Rolhbaum's group at Emory has been testing whether a full therapeutic dose of exposure therapy in the emergency room will lead to stronger fear extinction in traumatized individuals compared with delayed extinction, although the results are not yet fully in.
Extinction training after memory recall may also “erase” fear memories
Very similar results have been found when extinction training was carried out 10 min to 1 h after fear memory recall.Rats were trained to associate a tone with a footshock and then divided into five groups. Four groups were given a single retrieval trial presentation of the tone in the absence of a footshock. Extinction training then began either 10 min, 1 h, 6 h, or 24 h later. The fifth group was exposed to the context but did not receive memory retrieval. Twenty-four hours later, all groups were tested to see if they would show between-session extinction and then they were tested once again, 1 month later. Twentyfour hours after extinction all groups had low levels of freezing. However, 1 month later, the groups given extinction training 10 min or 1 h after recall showed no spontaneous recovery, whereas the groups extinguished 6 or 24 h later did. Very similar results were seen when relapse of extinction was measured with renewal or reinstatement. Hence, just like extinction given shortly after fear conditioning seems to block consolidation extinction given shortly after recall seems to block reconsolidation. Importantly, this work was extended in humans and extinction given shortly after recall blocked spontaneous recovery 1 year later! These are exciting results and clearly indicate that the timing of extinction either after original learning or after memory recall can have pro found effects on the durability of extinction. It remains unclear, however, as to why a 10-min interval between the first extinction trial (ie, a memory retrieval trial is identical to the first trial of extinction training) produces such a different effect than the usual intertrial interval during normal extinction training.
Role of NMDA receptors in extinction of conditioned fear in rodents
Like fear acquisition,fear extinction depends on NMDA receptors within the basolateral amygdala. Thus, intra-amygdala infusions of a compound that blocks NMDA receptors prior to extinction training dose-dependently blocked retention of extinction of fear-potentiated startle measured 1 day after extinction training. This impairment could not be attributed to an effect on NMDA receptors outside the amygdala, to damage or destruction of the amygdala, or to an impairment of sensory transmission during extinction training. Later studies showed that systemic administration of NMDA receptor antagonists prior to fear extinction training lead to dose-dependent impairments of both within-session extinction and extinction retention. - Systemic NMDA receptor antagonists also impair extinction retention or reinstatement when administered immediately after extinction training, - indicating that NMDA receptors are involved in consolidation as well as encoding of extinction memory. A similar blockade of extinction of contextual fear conditioning, and inhibitory avoidance conditioning has been reported with both systemic and localized administration of various NMDA receptor antagonists, , and additional studies have confirmed that these effects cannot be explained by state dependency. ,
Different effects of NMDA blockade in the amygdala and medial prefrontal cortex on extinction
The medial prefrontal cortex (mPFC) sends dense projections to the amygdala that terminate, in part, on inhibitory interneurons.- Hence, mPFC is in a position to inhibit the amygdala, a possible extinction mechanism, at least under some circumstances. , Electrolytic lesions or localized inactivation of the infralimbic region of mPFC impair extinction retention while having little to no effect on acquisition or within-session extinction, suggesting a role for this region specifically in consolidation and/or expression of extinction memory (see also ref 46). Single units within infralimbic cortex fire selectively to presentations of a previously fear-conditioned cue during an extinction retention test 24 h after extinction training but not during the extinction training session itself. When infralimbic cortex microstimulation was paired with presentations of a previously fear conditioned cue in nonextinguished animals, freezing to those cues was attenuated, and this effect was also seen the next day when no stimulation was given. , Collectively, these findings indicate that mPFC plays a significant role in many cases in extinction memory consolidation and expression, likely via its interactions with the amygdala. NMDA receptors within amygdala seem to be involved in the initiation of extinction, whereas in infralimbic cortex, they seem to be involved in consolidation of extinction. Microinfusions of NMDA receptor antagonists into basolateral nucleus of the amygdala prior to fear extinction training impair both within-session extinction and extinction retention. , , , , However, local infusions of NMDA 2A, 2B antagonists into basolateral amygdala block the expression of several fear-related conditioned responses, including freezing, suggesting these drugs could artifactually block extinction retention by interfering with synaptic transmission. However, infusion of ifenprodil, a drug that blocks a subtype of the NMDA receptor but does not block expression of fear conditioned responses, still blocked extinction retention. , , Immediate post-extinction training infusions into the amygdala of ifenprodil have no effect on subsequent extinction retention when extinction of fear is measured. , This suggests that NMDA receptor-dependent synaptic plasticity within amygdala is involved in encoding extinction of fear, but that the subtype of the NMDA receptor where ifenprodil acts in the amygdala is not required for consolidation of extinction, at least for conditioned fear.
In contrast, pre-extinction training infusions of NMDA receptor antagonists into mPFC have no effect on within-session extinction but generally impair later extinction retention, , ; (but see ref 27). Immediate postextinction infusions of NMDA antagonists into the infralimbic cortex do block extinction retention consistently, , - providing strong evidence that NMDA receptor-dependent synaptic plasticity within this cortical area is involved primarily in consolidation of extinction memory.
NMDA receptors do not seem to be involved the second time extinction is given
Perhaps surprisingly, when the experimental protocol involves fear acquisition and extinction followed by reacquisition and re-extinction of the same cue, systemic administration of the NMDA receptor antagonist MK801 prior to re-extinction training does not impair subsequent extinction retention. However, it does block reextinction when the extinction retention test occurs in a context different from that of initial acquisition and initial extinction,suggesting that NMDA receptor activation is required when extinction events are relatively novel but not when they are relatively familiar. On the other hand, novelty does not seem to matter for fear conditioning itself because disruption of the NMDA receptor blocks fear acquisition in both a novel and a familiar context. ,
Effects of localized infusions of NMDA receptor antagonists prior to second extinction are more complex and are reviewed elsewhere (see ref 51).
Role of D-cycloserine in fear extinction
Because blockade of the NMDA receptor impairs extinction, we wondered whether enhancing the functioning of that receptor would enhance extinction. To test this we administered a compound called D-cycloserine (DCS) either systemically or directly into the rats' amygdala prior to extinction training and then tested retention of extinction the next day.DCS does not bind to the NMDA receptor itself, but to another receptor on the NMDA protein called the glycine regulatory site. Activation of this site improves the ability of the NMDA receptor protein to flux calcium which initiates a variety of intracellular events that are critical for extinction. As predicted, when DCS was given in combination with repeated exposure to the feared stimulus without a shock, extinction retention was enhanced, when testing took place after DCS had worn off. This did not occur in control rats that received the drug alone, without extinction training. Based on these results, we concluded that the positive effects of the DCS were specifically connected with extinction and did not result from a general dampening of fear expression.
These effects have now been replicated in a large number of studies.
Systemic administration of DCS either before- or after extinction training facilitates extinction. Local infusion of DCS into the basolateral nucleus of the amygdala prior to , or after fear extinction training mimics the effects of systemic administration. Chang and Maren recently showed that although DCS infusions directly into infralimbic cortex did not facilitate extinction, these infusions did facilitate the subsequent reextinction of fear when animals were trained in a drugfree state.
DCS may normalize impaired extinction
There is a growing body of evidence to suggest that DCS reverses fear extinction deficits caused by a variety of factors including stress,- alcohol withdrawal, REM sleep deprivation, genetically modified mice that have a polymorphism in their BDNF gene, or even adolescent rats in which the medial prefrontal cortex has not yet developed fully. Perhaps consistent with these stress-related effects, DCS interacts with stress hormones: DCS blocks the extinction-impairing effect of the corticosteroid synthesis inhibitor metyrapone and enhances the extinction-facilitating effects of the synthetic glucocorticoid dexamethasone. Judo et al showed that changes in prefrontal synaptic efficacy during extinction training did not occur in adult rats exposed to early postnatal stress, and these synaptic changes and resulting deficits in extinction were restored by DCS. These observations may also be consistent with findings in clinical studies described below, that DCS facilitates exposure therapy in clinical, but not subclinical, populations, and that different types of mechanisms could be involved in the two groups of subjects.
DCS facilitates psychotherapy
Many forms of psychotherapy depend in part on extinction of fear. Patients with fear of snakes avoid snakes and do not allow themselves to extinguish this fear. However, repeated exposure to pictures of snakes, a snake in a jar, or even a live snake are extremely effective in eliminating such simple phobias and are widely used. Panic patients afraid they will have a panic attack driving over a high bridge are taken back to the bridge to show them that they will not always have a panic attack there. Patients with a contamination phobia who are forced to touch the bottom of a toilet seat, but not allowed to wash their hands, learn not only they do not die but they don't even get sick, identical to an extinction trial. Exposure to scenes of combat in people with post-traumatic stress disorder (PTSD) in the presence of a supportive therapist often leads to substantial improvement, and cognitive behavioral therapy has been found to be helpful in many PTSD patients. In all these cases of exposure-based psychotherapy extinction is the fundamental mechanism that is operating.
The finding that DCS can facilitate fear extinction in animalsand that fear extinction was so central to many types of psychotherapy suggested that DCS might also be effective in facilitating exposure therapy for fear and anxiety disorders in people. DCS had been FDA-approved for some time as an antibiotic treatment for tuberculosis at high doses. Although this effect had nothing to do with its ability to facilitate extinction it allowed us to test whether it would facilitate exposure-based psychotherapy right away.
In this study, the ability of DCS to enhance exposure therapy for acrophobia, or fear of heights, using virtual reality exposure therapy, was examined. Previous work had shown improvements on acrophobia outcome measures after seven or eight weekly virtual reality therapy sessions.Participants in the DCS study underwent a suboptimal amount of virtual reality therapy for acrophobia (only two virtual reality sessions) and were instructed to take a single dose of study medication before each session. So DCS was only taken twice: prior to each of the two sessions that were separated by average of 12 days. Similar to the rats in the preclinical work, participants receiving DCS exhibited significantly more improvement than did participants receiving placebo, measured either 1 week or 3 months later, long after the drug was out of the body ( ). At the 1-week follow-up, DCS-treated patients exhibited less subjective fear and fewer skin conductance fluctuations in the virtual reality environment.
Most importantly, outside of the virtual reality environment patients reported a decrease in overall acrophobia symptoms, increased self-reports of exposure to heights in the “real world,” and higher self-ratings of improvement. These later results are very important because they indicate that extinction of fear is not always context-specific, as seen so often in animal studies. The reason for this appears to be that humans begin to feel safe in situations they previously avoided, once they have some successful psychotherapy and avoid these situations less often. People with fear of elevators do not want to continue to walk up 20 flights of stairs once they learn the elevator will not harm them. In contrast, rats have no opportunity to continue to extinguish because they are put back in their home cage with no further exposure to the fearful conditioned stimulus. So, the several measures of relapse from extinction may be overestimated in rodent studies.
Other groups found that DCS enhanced exposure therapy for social anxiety disorder—, , obsessivecompulsive disorder— , , and panic disorder— , indicating that the DCS effect is a relatively general one. The failure of another study to see on effect in OCT) may have resulted from giving DCS 4 hours prior to exposure therapy, which may have been too early. There has been one report of a failure of DCS to facilitate exposure therapy for subclinical spider phobia, and failures of DCS to facilitate extinction of Paviovian conditioned fear in a laboratory situation in humans. , However, these negative effects may indicate that DCS is useful only in people with clinically significant, maladaptive fear—consistent, perhaps, with the preclinical data suggesting that DCS is particularly effective in stressed animals (described above).
A note of caution
DCS may not work on re-extinction
It should be recalled that NMDA antagonists block extinction the first time extinction training is carried out but not when rats are retrained and then extinguished again.The same is true for D-cycloserine.
DCS and serotonin reuptake inhibitors
DCS also failed to facilitate extinction in rats with prior administration of a serotonin reuptake inhibitor (imipramine).However, because it is likely that many patients in the positive trials of DCS in anxiety disorders were taking serotonin reuptake inhibitors, it is hard to know how important this variable is because the database is just not large enough to allow an adequate evaluation of this variable. In our own study of fear of heights we could find no relationship.
DCS shows tolerance
DCS also failed to facilitate extinction in rats given prior daily injections of DCS,consistent with several preclinical studies showing tolerance with repeated DCS treatment. Hence, we suggest spacing DCS treatments by at least a week.
DCS should not be given too far in advance of psychotherapy
As mentioned above, DCS is known to facilitate consolidation of fear extinction so it is important not to give it too early prior to psychotherapy. In fact, post-extinction training is used routinely in rodent studies and this maybe especially effective clinically. For example, if a patient had a bad session of psychotherapy it might not be useful to use DCS. But, if they have a good session then the therapist could give DCS right after therapy, which would more likely cover the consolidation period and improve compliance.In fact, we have preliminary data in rats that giving DCS prior to sleep, when many types of memory consolidate, may be especially effective (Davis, Bowser, McDevitt, and Walker, in revision).
DCS is very unstable in humid conditions and in solution
It is very important to keep it dry during compounding into lower doses from Seromycin as well as in storage.
Will DCS make patients worse?
A question that is often asked is why, if DCS is a cognitive enhancer, does it not stamp in the bad memories brought up during psychotherapy and make patients worse? DCS has been shown to facilitate retention of inhibitory avoidance and spatial learning in rats,stimulus attributes in inhibitory avoidance in rats, inhibitory avoidance in chicks or mice, , thirst-motivated maze learning in mice, object location in mice, taste aversion in rats, , delayed nonmatching- to-sample in rhesus monkeys, and acquisition of eyeblink conditioning in rabbits when trace conditioning was used. It also improves memory due to aging in mice, spatial memory in rats, and eyeblink conditioning in rabbits. In rats, DCS reverses scopolamine-induced deficits in: the T-maze and water maze, working memory, or inhibitory avoidance, or reduces deficits following brain injury or hippocampal lesions, or deficits in inhibitory avoidance in mice caused by β 25-35-amyloid peptides or convulsant drugs.
Because most, if not all, these tasks depend on the hippocampus, one might expect that DCS would facilitate hippocampally dependent declarative memory in humans. However, the literature is very inconsistent in this area. Otto et alfound no effect of DCS in improving verbal or nonverbal learning given at weekly sessions nor did D'Souza et al on several tests of verbal learning and it had no effect on a procedural task (finger tapping). On the other hand, improvement of procedural learning (sequential finger tapping) but not of declarative (word-pair) learning by DCS was found. DCS accelerated rate of learning on item-category associations, but had no beneficial effect in the object-location association task, both declarative memory tasks. There was improvement on one cognitive task (delayed thematic recall on the logical memory test) in schizophrenic patients. There was one report showing enhanced fear conditioning with DCS in humans, but the stydy desing was so complex that it is hard to know what to conclude from this study, especially because there appear to be no positive studies of DCS on classical fear conditioning in humans. Finally, no reports were found of patients getting worse on or after DCS in the six positive studies that have been published with cognitive behavioral therapy. Hence, despite the ability of DCS to facilitate learning in animal studies, for reasons that are not clear, this has not been found universally in humans, even though DCS generally has facilitated fear extinction in clinical populations. Possible reasons for this are discussed else-where.
Because excessive fear and anxiety occur in so many psychiatric disorders, research continues to investigate how the brain normally Inhibits or suppresses these emotions. Exposure-based cognitive behavioral therapy (CBT), in which patients are repeatedly exposed to anxiogenic situations in the absence of any aversive consequences, has been quite successful in treating these disorders. CBT is procedurally similar to fear extinction in animals, in which a fearful stimulus also Is exposed repeatedly without aversive events. Extinction does not erase the original fear memory but instead actively inhibits that memory. It is dependent on a protein called the NMDA receptor in brain areas such as the amygdala and medial prefrontal cortex. A medication called D-cycloserine allows the NMDA receptor to work even better and It also facilitates fear extinction, especially when extinction is compromised following stress. However, it does not work when given alone, but only in combination with extinction training. Six independent clinical trials have shown that D-cycloserine facilitates CBT in patients with phobia, obsessive-compulsive and panic disorder, and several trials are underway to tests its effects in PTSD. Continued analysis of normal and abnormal fear extinction in animals will almost surely lead to other medications to facilitate CBT.