Jol Coutu knelt on the cold cement floor of the pet supply store he managed in Montreal, his wrists bound behind him with telephone wire. He could feel the barrel of a pistol pressed against the back of his neck. “You’re lying!” the gunman screamed. “And I am going to blow your head off.”

He and another attacker had herded Coutu and a young cashier into the back room and demanded that he unlock the safe. When he told them he did not have the key, they became enraged. They ripped out all the wires of the fax and telephones in his office and tossed the contents of his desk drawers. Now they were getting ready to execute him. “Go ahead and pop him,” he heard one of them tell the other. “Blow his head off.’”

Coutu had just enough time to turn to his co-worker and ask her to tell his girlfriend that he loved her before the gunman smacked him with the butt of the weapon and sent him sprawling. Lying facedown, Coutu watched his blood pooling around him and waited for the coup de grce. Then, suddenly, the front door of the store slammed. The assailants had fled, empty-handed.

Yet Coutu’s ordeal was just beginning. For years he would be tormented by violent nightmares, panic attacks brought on by the mere hint of aggression around him, and severe depression—signs of post-traumatic stress disorder (PTSD). His girlfriend moved out. He retreated from his friends and left his job.

Then, one day in 2009, he saw an advertisement in a local newspaper for a trial of an experimental therapy run by Alain Brunet, a McGill University psychiatrist. Brunet suggested something radical: he wanted to erase portions of Coutu’s memory.

For decades scientists believed that long-term memories were immutable—unstable for a few hours and then etched into the brain for good. Research now suggests that recalling a memory causes it to revert temporarily to an insecure state, in which the recollection can be added to, modified, even erased. “Memory is more dynamic, more fluid and malleable than we thought,” says neuroscientist Daniela Schiller of Mount Sinai School of Medicine.

That idea, brought to the fore about a decade ago, has opened up a new controversial research area exploring the possibility of deleting, or at least muting, parts of human memory with drugs or targeted therapies. Some experts have found that a drug used to treat high blood pressure works to unseat recollections; others are testing novel biochemical means or behavioral interventions to interfere with unwanted remembrances. [For more on psychological forgetting strategies, see “Trying to Forget,” by Ingrid Wickelgren.]

Although scientists and ethicists worry that such drugs might be abused or have unsettling side effects, these treatments could also liberate individuals from experiences that haunt them—including a traumatic event, such as that experienced by Coutu—and the emotions that linger, such as the agony from the death of a loved one or the crippling apprehension from a car accident or sports injury. “Imagine a high jumper who fell during the Olympics,” says neuroscientist Karim Nader of McGill. “They may have a lot of anxiety associated with jumping, and it could severely affect their future performance. If we can make these drugs work, you could help them, too—or anyone with anxiety that is proving a problem.”

Window of Vulnerability
To create, or consolidate, stable long-term memories, the brain must synthesize specific proteins in the hours after events occur. Those proteins are part of a cascade of chemical processes that remodel some of the tiny junctions, or synapses, between brain cells to make these cells communicate more efficiently. The construction process often includes the production of more synapses, which further facilitates neuronal chatter.

A decade ago most memory researchers believed these synaptic connections were extremely stable and resistant to degradation. They might fade with time, but they could not be changed or erased. Yet Nader, as a 33-year-old postdoctoral student at New York University back in 1999, was new enough to question that dogma. After attending a speech on memory delivered by Nobel laureate Eric R. Kandel, a Columbia University neuroscientist, Nader wondered exactly what happens when we recall an event. To do so, it seemed to him, you would have to take the memory out of storage. What if you added new information or blocked the chemical processes needed to put that memory back?

To find out, Nader and his colleagues created the kind of searing emotional memory that should have been permanent and immutable. He placed a rat in a cage and played a tone while delivering a shock through the metal floor. Soon all Nader had to do was play the sound, and the rat would freeze in terror. The two stimuli, convention held, had been permanently connected.

Fourteen days later the researchers played the tone and simultaneously injected a drug that blocks protein synthesis into the rat’s amygdala, an emotion hub in the brain with an important role in establishing emotionally rich memories. Nader’s intent was to see if the drug would interfere with the memory’s return to storage. The strategy worked. In subsequent trials, the animal no longer froze at the sound. It had forgotten the meaning of the tone and therefore had been liberated from its trauma.

The experiment provided powerful support for a theory called reconsolidation that was first floated back in the 1960s but largely abandoned because of lack of evidence. It holds that reminding a person or animal of something makes that memory temporarily unstable. During a brief window before the memory is “reconsolidated,” it is susceptible to perturbation. “We used to think the memories we had were pictures of the original event. Now we know that it is the last version of the memory because each time we retrieve it, it changes a little bit,” Schiller says.

Shutting Off the Alarm
Nader’s findings were a revelation to Brunet. The Montreal-based psychiatrist had already been experimenting with ways to prevent the initial consolidation of traumatic memories as a preventive measure against PTSD. Brunet, along with Roger K. Pitman of Harvard University, had drawn his inspiration from a series of groundbreaking experiments conducted by James L. McGaugh and his colleagues at the University of California, Irvine, in the 1990s. McGaugh had demonstrated that a drug called propranolol, a so-called beta blocker used to treat high blood pressure and anxiety, could also weaken new memories.

Propranolol interferes with a key signaling agent that normally augments memory formation in response to an emotional event. [For more on emotional memory, see “A Feeling for the Past,” by Ingfei Chen.] Anytime we get emotionally aroused, the adrenal gland releases stress hormones, which trigger the release of a chemical in the brain called norepinephrine. This neurotransmitter binds to receptors in the amygdala, which in turn discharges a flood of chemicals that signal the rest of the brain to encode the memory. Propranolol binds to, and blocks, those receptors. McGaugh showed he could inhibit typical memory formation by administering propranolol, which he thinks interferes with the action of norepinephrine—thus preventing the memory-boosting signal from ever going out.

Pitman and Brunet immediately recognized the potential for treating patients who had been exposed to trauma—triggering what Brunet calls “pathological remembrances.” In 2002 and 2003 teams led first by Pitman and then by Brunet administered propranolol to trauma victims who came through emergency rooms in Boston and in Lille, France. Both research groups demonstrated that administering the drug was far more effective at reducing the likelihood the participants would develop PTSD than a placebo was.

Brunet and Pitman were both excited by the effects of the drug. Yet the limitations of the therapy were clear. The procedure would help patients only during the brief window before the long-term memory had consolidated, within hours of the initial event. By definition, PTSD does not set in until at least six months later.

Nader’s findings offered new hope. They showed that established memories could be made labile again just by taking them out of storage. So, in 2005, Nader, Brunet and Pitman joined forces to test whether propranolol might also be able to tweak older memories. The researchers asked 19 patients suffering from chronic PTSD to recall their trauma. They gave half of them propranolol and the other half a sugar pill. A week later Brunet monitored the physiological response of the patients as they listened to an audio account of their event. Those who had received the beta blocker still retained a memory of the factual details but were significantly less aroused than those given the dummy drug. A few theories attempt to explain propranolol’s action. As with the initial trauma, recalling an agonizing memory releases stress hormones, which may well be involved in reconsolidating the memory afterward. One possibility is that propranolol blunts the action of norepinephrine then, too. Alternatively, the drug might be inhibiting the protein synthesis needed to put emotional memories back into storage.

Either way the initial evidence for propranolol’s effects, published in 2008, led to the larger study for which Coutu volunteered. Once a week for six weeks, Coutu took propranolol and read a one-page description of the armed robbery out loud. The task was so unsettling that Coutu considered dropping out after just one session. That night he was terrorized by nightmares. Halfway through the fifth session, however, something remarkable happened. “I will never forget it,” he says. “I’m reading the one page, and it’s like I have no more attachment to the story—like I am reading a novel or watching a movie. I started to smile, I was so excited.”

Brunet published the results of his study, which included 66 patients in Boston, France and Montreal, this past August. On average, physiological symptoms of fear such as a racing heart and sweating diminished by 50 percent for the 40 PTSD patients who took propranolol, compared with a 7 percent decrease for the 26 patients who did not take the drug. After the experiment, Brunet claims, roughly three in four propranolol patients were so improved they no longer met the criteria for PTSD.

New Knowledge
Yet propranolol may not offer a foolproof way to forget. Neuroscientist Elizabeth A. Phelps of N.Y.U. and her colleagues spent several years attempting to erase fear memories in mentally healthy people using propranolol. They found that the drug could only temporarily expunge a learned association between a visual stimulus (a colored square) and a shock. The fear later returned as if the therapy had never been applied at all.

Phelps believes the propranolol failed because her subjects still knew that the colored squares were associated with the shocks—and that this conscious memory generated a fear response even after the emotional record of the initial event was erased. Exactly why the drug worked for trauma victims is unclear, but Brunet says their memories are starkly different from the associations Phelps studied. “We are dealing with people with PTSD, and her subjects are dealing with a very simple task using squares and triangles,” he says.

In a study published in 2010 Phelps and Schiller demonstrated a method Phelps believes eliminates the potential interference from overt knowledge. Instead of simply sending subjects home after the visual reminder of the shock, the researchers added an experience designed to modify both the conscious and emotional aspects of the memory.

After showing volunteers a picture of the square, Schiller and Phelps waited for a variable period, then delivered “extinction training,” a kind of behavioral therapy intended to overwrite the dreaded association with one that is benign. In this case, the researchers exposed the volunteers to images of the colored squares, but this time they did not deliver a shock so that these individuals would think of the images as “safe” again. The timing of this extinction training was key. Previous research held that the initial reminder, the square, would spark chemical processes that would render the memory of the shock temporarily vulnerable to modification or erasure while the memory was being reconsolidated. This so-called reconsolidation window would close once those processes were complete.

Some volunteers viewed the square 10 minutes before receiving this extinction training, a time point within the reconsolidation window. Others saw the square six hours before the extinction training—safely outside that window. A third group did not see the square prior to extinction training.

All three groups returned to the lab on a third night and were presented with pictures of the squares as researchers monitored their fear response. The response virtually disappeared in those who had received the extinction training during the reconsolidation window, whereas it returned for those who had not, providing evidence that human memories are malleable during this window and can be blunted without drugs. In fact, altering memory with new information in this manner might be especially effective because it adds to conscious knowledge, rather than just altering an instinctive fear memory—a strategy that might not work in the long run. (Other forms of behavioral therapy, such as memory suppression, may also work most effectively during recall, within the reconsolidation window.)

Chemical Intervention
If you asked one of Phelps and Schiller’s volunteers what happened the first night of the trial, they would very likely be able to tell you about the shocks, even if they no longer linked them with the squares. The same holds for the propranolol-taking trauma victims. But what if we could erase those memory traces altogether?

Neuroscientist Todd C. Sacktor of S.U.N.Y. Downstate Medical Center in Brooklyn is developing a compound that would do just that. In 1990 Sacktor and his colleagues discovered an enzyme known as protein kinase M-zeta (PKMzeta) they suspected might play a role in long-term memory. Not only was the enzyme present in the appropriate regions of the brain, but it also had chemical properties that scientists thought were ideally suited to supporting the maintenance of such neural traces.

In 2006 Sacktor’s team confirmed its hunch. The researchers trained a rat to avoid an area of a room where it received an electric shock. Then they waited a day and injected a drug that inhibited PKMzeta into the hippocampus, where the memory was presumably stored. When they put the rat back in the room, it could not remember what area to avoid. Blocking the actions of PKMzeta had wiped out the rat’s memory of the event, proving the enzyme had a role in maintaining the memory. This past March, Sacktor and his colleagues reported the same effect with a mutation that crippled PKMzeta. They also did the reverse and enhanced memory in rats with a genetic manipulation that caused the animals to produce additional copies of the enzyme.

Meanwhile Sacktor’s team had figured out how the enzyme worked. It catalyzes a reaction that enables the transport of key proteins to the synapses. These proteins respond to the neurotransmitter glutamate, allowing a neuron to detect the firing of a neighboring cell by its resulting release of glutamate. The upshot is effective information transfer.

A drug that shuts down PKMzeta, however, is like a “nuclear bomb,” Nader says; it obliterates all memory, not just the recollections you want to detonate. Yet Sacktor may have found a way around this problem. Every time a memory is pulled out of storage, he believes, the brain breaks down the PKMzeta connected to that memory. To put the memory back, he posits, the brain must create the enzyme anew. Sacktor has developed a drug that, in unpublished experiments, blocks the synthesis of new PKMzeta in rats for about two hours. In theory, then, a person could selectively shut out troublesome memories by recalling them, making them active and then taking this drug, which would stop the brain from restocking them.

If the drug works as Sacktor imagines—a big “if” at the moment—it promises to be more powerful than propranolol. “With propranolol, the problem is trying to get that very potent and effective erasure,” Sacktor says. “That is not the issue with PKMzeta.”

Bioethicists such as Paul Root Wolpe of Emory University worry about such strong medicine for the mind. “Memory is such a crucial part of what makes us who we are that we have to be extremely cautious about changing or erasing [memories],” Wolpe says. “To what degree will we use this technology in ways that threaten selfhood and personality?” He also frets that people with sinister motives could abuse a potion that makes others forget—enabling an intelligence officer to get away with torture, say, or a parent with the abuse of a child.

Yet the terror Coutu, and others like him, endured is arguably not a critical part of who he is. For Coutu, the memory of his attackers seemed to accomplish quite the opposite: it cracked his sense of self. Only calibrating that recollection, in fact, could enable him to reassemble the person he had been—and the delightfully ordinary life he once led.