Cueing Fear Memory during Sleep—To Extinguish or to Enhance Fear?

Fear memories can be a tremendous burden in psychiatric disorders such as phobias and posttraumatic stress disorder. Sleep is known to facilitate the consolidation, i.e., strengthening and integration of newly formed memories,^1–4 including fear memories.^5–7 Reactivation of the underlying neuronal memory traces that occur during slow wave sleep (SWS) causally contribute to this consolidation process. Importantly, this reactivation can be triggered by presenting external reactivation cues (reminders) like odors and sounds during SWS.^8–11 New findings from He and colleagues in this issue of SLEEP show that such external cueing during sleep can be used to extinguish fear memories.¹² This finding is surprising in showing that cueing memory reactivation during sleep cannot only strengthen but also weaken a memory, raising the question which mechanisms drive the effect of cueing in either one or the other direction. In fact, the study by He et al. is the fourth in a recent series of publications examining fear memory cueing during sleep with rather contrary results.^11,13,14

In the new study by He and colleagues,¹² human subjects followed a fear conditioning procedure in which a neutral tone was paired with a mild uncomfortable electrical shock. Successful conditioning was confirmed by an increase in subjects' skin conductance response (SCR) to the conditioned tone— a signal of fear. After conditioning, subjects were allowed to sleep for four hours while the conditioned tone (without electrical shock) was repeatedly presented again during SWS. This re-exposure of the conditioned tone reduced the subsequent fear response compared to two other groups of subjects who received a different tone or did not receive any tones during SWS.

These findings basically confirm previous results by Hauner and colleagues,¹³ who used a contextual fear conditioning procedure in which faces were paired with electrical shocks while subjects smelled a specific odorant. Re-exposure of the odorant during SWS after conditioning produced a striking reduction in the fear response to the faces. Inexplicably, two other scientific reports performed in rodents reported exactly the opposite, i.e., an increased fear response after cueing the fear memory during SWS. Barnes and Wilson¹¹ paired olfac-tory bulb stimulation (simulating perception of an odor) with foot shocks in rats, and Rolls and colleagues¹⁴ paired an odor with foots shocks in mice. In both studies, reapplying the conditioned stimulus (CS) during ensuing SWS distinctly increased subsequent freezing in response to the CS.

What is so exciting about extinguishing human fear by presenting cues during sleep? Obviously, the idea to use cues during sleep as an anesthesia-like state for cutting out a patient's bad memories—like tumors to be removed by the surgeon's scalpel—is highly attractive to behavioral health practitioners. Beyond its clinical relevance, the issue of whether the cued reactivation of fear memory during sleep can extinguish the respective memory, tackles a basic scientific question in this field: Can cueing of memory reactivations during sleep, beyond strengthening existing associations, induce new learning? Indeed extinction represents a complex inhibitory learning process¹⁵ that requires the evaluation of the association between the conditioned and the unconditioned stimulus, at least in the waking brain. Does the brain during sleep assess whether or not CS presentation is followed by the unconditioned stimulus? This question is related to the more general issue whether the effects of neural reactivation during sleep depend on any kind of evaluative feedback activity, including activity from structures signaling aversiveness, pleasure, reward etc. Alternatively, sleep might be considered a state in which neural reactivations form memory representations in “unsupervised” conditions, as there is normally no sensory input that could be used for shaping behavior.¹⁶ Extinguishing human fear responses by presenting conditioned stimuli during SWS, as demonstrated by He et al.¹² and Hauner et al.,¹³ tells us that new learning during sleep does indeed happen, and that this learning is obviously feedback-controlled, i.e., by the missing unconditioned stimulus (shock) during sleep. In this vein, these findings of inhibitory learning during sleep are complementary to recent demonstrations of excitatory learning, i.e., classical conditioning during sleep.^17,18

As mentioned, the picture drawn by these two studies in humans becomes somewhat blurred in light of the fact that two other studies in rodents employing an essentially similar approach (cueing of fear memory during SWS) reported the opposite result, i.e., enhanced rather than extinguished fear responses. What could explain this? Species differences might be suspected as one contributing factor. However, we consider this highly unlikely, given that beneficial effects of sleep on memory consolidation have been shown to be similar in humans and animals,¹⁹ including the spontaneous reactivation of memory representations^20–22 as well as the possibility to trigger reactivation during sleep by external reminders.^8,9,23 Other differences are more likely to play a role.

As indicated in Table 1, the four studies greatly vary with regard to the conditioning procedure and the experimental protocol. Importantly, (1) reinforcement contingencies during conditioning differed: in the human studies the CS was followed by an electrical shock in only 40% and 50% of the cases, respectively, whereas in the rodent studies the CS was always followed by foot shock. The different reinforcement schemes presumably produce memories of different strength and/or involve the recruitment of different brain structures, which ultimately might determine whether cueing during sleep extinguishes or enhances the memory. (2) Differences in aversiveness of the conditioned memories might be relevant as well, as the electrical shocks applied in the human studies appeared to be milder and, thus, less salient than the foot shocks in the rodent studies. (3) The studies differed in the frequency of cueing during sleep. It is possible that a massed presentation of the cues within 3–40 minutes (as in He et al. and Hauner et al.) leads to an extinction, whereas presenting single cues over a longer time period of 2–4 hours^11,14 strengthens the underlying memory trace. (4) Finally, the studies used different delays between cueing during sleep and testing of the fear response. Reactivation during sleep might labilize the fear memory trace, leading to reduced fear responses when tested immediately after sleep, as in He et al.¹² and Hauner et al.¹³ Yet, the trace might become re-stabilized during the ensuing 24-hour interval (including sleep) such that it comes out even stronger when tested 24 hours later.^11,14 Additional factors also differed between the studies. For instance, testing took place in the same context in the human studies but in a new context in the animal studies.

Table 1.

Overview of the experimental design and findings of the four available studies.

Whatever the mechanisms determining whether cueing during sleep produces extinction or enhancement of fears, in demonstrating that fear extinction might in fact be achieved by this procedure in humans, the findings by He et al.¹² and Hauner et al.¹³ open up entirely new perspectives on the treatment of psychiatric and behavioral disorders that originate from maladaptive memories. This perspective is appealing and should motivate research that faces the challenge to identify the mechanisms deciding when sleep cueing drives an extinction process and when the outcome turns to the opposite.

CITATION

Diekelmann S, Born J. Cueing fear memory during sleep—to extinguish or to enhance fear? SLEEP 2015;38(3):337–339.

DISCLOSURE STATEMENT

The authors have indicated no financial conflicts of interest.