Erasing Fear Memories Without Drugs


by Jim Schnabel

December 11, 2009

People’s “fear response” to a mildly frightful memory can be weakened or erased forever with a simple procedure that requires no drugs, according to a new study that could point the way toward better treatments of anxiety-related disorders and even addictive behaviors.

The finding, which confirms previous research in rats, makes use of a recent discovery that recalling a fearful memory opens up a brief window during which the memory can easily be modified before being stored away again. “The hope is that by paying attention to these memory re-storage processes, we can create [treatment] interventions that are more effective,” says Elizabeth Phelps, who led the study with fellow New York University researcher Joseph LeDoux. (LeDoux is a member of the Dana Alliance for Brain Initiatives.)

A standard neuroscience technique for creating a “fear memory,” either in lab animals or in humans, is to present an ordinary stimulus, such as a tone or visual image, just before presenting an unpleasant stimulus such as a painful electric shock. After experiencing these paired stimuli several times, a research subject forms a long-term memory of their association. Even presentation of the non-painful stimulus alone will evoke a bodily fear response involving elevated bloodstream cortisol levels, increased electrical conductance of the skin, and other easily measurable markers of stress.

This link between the first stimulus and the fear response is the fear memory, and it is believed to be stored in the amygdala, part of the brain’s limbic system. Researchers know that by presenting the first stimulus repeatedly, without any following pain, they can cause the brain to inhibit the expression of the fear memory. But this misnamed “fear extinction” effect doesn’t actually extinguish the fear memory; the memory typically re-emerges later, either on its own or under conditions of psychological stress.

Research into the mechanisms of memory storage and retrieval, however, suggests that there is a way not just to suppress but to effectively erase a fear memory.  Mainly from work in rodents, researchers have discovered that the act of recalling a fear memory apparently triggers the opening of a memory “reconsolidation window,” which begins after several minutes and lasts several hours. Within this reconsolidation window, the fear element of the memory apparently becomes unstable and can relatively easily be altered.  Researchers have hypothesized that this reconsolidation function evolved to enable animals to maintain the relevance of their memories in changing environments.

Some researchers have tried to erase fear memories simply by disrupting the reconsolidation process, using drugs that interfere with the workings of the amygdala. But in a study published May 15 in Science, LeDoux and colleagues found that drugs may not be necessary, if classic fear “extinction” techniques are used within the reconsolidation window.

The team first conditioned rats to have a fear memory linking an audio tone with an electric shock. After the fear memory had been well established, the researchers played the fear-associated tone again to reactivate the rats’ fear memories and open up the reconsolidation window. Animals exposed to a typical fear extinction regime—repeated exposure to the tone without a shock—within this window apparently lost most or all of their previous fear memories permanently. Those whose fear extinction regime occurred outside the reconsolidation window, however, soon recovered their fear memories, as in classic fear extinction experiments.

In the new paper, published in the Dec. 10 issue of Nature, Phelps, LeDoux and their colleagues report having replicated this finding in 65 human subjects. Phelps emphasizes that those subjects whose fear memories were removed during the memory reconsolidation window didn’t lose their memories of the original shock-linked stimulus (in this case a blue square); they lost only their bodily fear responses to it. In those who underwent the fear extinction regime outside the reconsolidation window, fear responses returned within 24 hours.

The researchers also found that this memory-rewrite effect worked over a remarkably long period. A year after the original experiment, they were able to re-test some of their subjects, and again they found that those who had undergone the window-extinction had no fear memory, while those who had undergone fear extinction outside the window still retained measurable fear responses.

Blocking pain and pleasure

“It’s a wonderful finding,” says Karim Nader, a researcher at McGill University who also has done prominent work in this area. “But it raises a lot of questions. For example, if the memory was very strong, as traumatic memories are, would the [technique] still be effective?”

In recently published work with rodents, Nader and his colleagues have found that a very strong fear memory may not be susceptible to reconsolidation for a period of days or weeks after the memory is formed; in humans, this period may extend much longer. The research also suggests that specific receptors on neurons in the amygdala mediate the reconsolidation process, so that drugs might be targeted to these receptors to enhance reconsolidation and make it easier to weaken fear memories.

Phelps agrees that the dynamics of the reconsolidation process aren’t yet fully understood and notes that her lab is now working on several follow-up studies. In one, she hopes to demonstrate with brain imaging technology that the reconsolidation-rewrite phenomenon uses different brain mechanisms than classical fear extinction. “There’s a lot of research to be done to figure out exactly how this works,” she says.

Nader emphasizes that the amygdala doesn’t just store “fear memories.” In principle, techniques based on memory reconsolidation could be applied as well to the “desire memories” that frequently overwhelm addicts when they see drug cues. “We’re starting to look at whether we can do this with drug addiction in humans,” he says. “Work in rodents certainly suggests that we can.”