Memories that trigger fear can last for years. In some cases they can be overwhelming, for example in post-traumatic stress disorder. A team at New York University, led by psychologist Elizabeth Phelps, now shows that fearful memories in humans can be eliminated through simple behavioural manipulation.

In studies of fearful memory processes, tones or lights are typically paired with electric shocks, and in time the tone or light alone elicits a fear response; in rats this manifests as freezing, in humans by a change in skin conductance. If at a later date study subjects are re-exposed to the tone or light, the stored memory is called up and experienced anew. The memory is then re-stored in a process called reconsolidation, which occurs over minutes to hours.

A wealth of animal studies has shown that reconsolidation can be blocked by drugs, such as protein-synthesis inhibitors — essentially wiping out memories. Although similar effects have been found for some drugs in humans, Phelps has not been able to replicate the work.

One way to lessen fearful memories without drugs is through extinction training. Here, the association with the fear memory — for example, the light or tone previously paired with the shock — is repeatedly presented without the fearful stimulus (the shock) until the fear response abates. But the effect is only temporary and the fear response can return.

An experimental disparity led Phelps and her colleagues to a clue to making extinction training more effective. Marie Monfils, then a postdoc in the New York University laboratory of Phelps' long-time collaborator Joseph LeDoux, was trying to show that drug blockade of reconsolidation is mediated by a different brain region from extinction. A procedural variation between her rat study groups troubled her, however. The drug group received a reminder of the fearful memory — a tone — a few minutes before the reconsolidation-blocking drug was administered. But the extinction group did not receive such a reminder before training began.

To remove the variable, Monfils repeated the experiment, giving both groups the reminder. Unexpectedly, the conditioned fear response was virtually eliminated in both groups. Monfils was perplexed: how could a tone block protein synthesis? Daniela Schiller, Phelps's postdoc, helped Monfils to interpret the results. “If you view reconsolidation as an adaptive mechanism, then new learning serves to update the memory, not block it,” Phelps says. In other words, extinction training was as effective as drug treatment when the training took place during the reconsolidation triggered by the reminder. “In retrospect,” Phelps says, “it makes perfect sense.”

To test the idea, the postdocs embarked on parallel studies — Monfils in rats (M.-H. Monfils et al. Science 324, 951–955; 2009) and Schiller in humans (see page 49). Schiller found that, with the new training approach, conditioned fear responses were eliminated for at least a year.

Phelps is cautiously optimistic about how the discovery might inform therapy. “I feel like we've done this light-paired-with-shock paradigm: a simple memory trace in the laboratory where the fear is, for obvious reasons, mild. I don't want people who are truly suffering to have false hopes.” Having said that, Phelps says the work shows that timing therapeutic interventions to fall within the reconsolidation window could bring about more durable positive effects.