Health for Life: The Science of Forgetting

Karim Nader sounds giddy as he recalls the day in November 2001 when he stood before hundreds of experts in the science of memory and presented a radical theory. At 34, with one research paper on the topic to his name, Nader was a newcomer to the field. He was so nervous, he considered ditching the San Diego conference and fleeing to Mexico: "I thought, Tijuana is only 20 minutes away. I can go there and surf for the rest of my life."

Instead, the young scientist composed himself, walked through an intriguing rat experiment and presented his stunning conclusion. Long-term memories, Nader proposed, aren't fixed in a permanent form once they're filed away in the brain, as researchers had long believed. When a memory is recalled, it returns to an unstable state, like ice melting to water. As such, it can be altered and then it is stored again. The original memory? No longer there. The notion challenged decades of dogma and rattled seasoned scientists who believed that old memories could never be changed. Nader's nerves were so fried by the end, he says, "I couldn't believe I was still alive."

Memory isn't like the heart. You can't count its vessels or hold it in your hands. Our recollections are molecular enigmas, vast and elusive. The scientists who study them have the patience to examine neurons and enzymes, talk about "intrahippocampal infusions" and spend a lot of time hanging around rats. One of their colossal endeavors is figuring out how to enhance memory—and we could all benefit from that. (Car keys? Password?) But some of the most intriguing advances involve the science of forgetting—how and why we lose memories—and now researchers are raising the stakes with a mind-boggling question: can certain memories be intentionally targeted and changed, maybe even eradicated? In their quest for an answer, scientists are transforming what we know about how our brains process the images and sounds and feelings we encounter. One day, the research might lead to innovative treatments for conditions like posttraumatic stress disorder and addiction.

Some basic facts about memory are clear. Your short-term memory is like the RAM on a computer: it records the information in front of you right now. Some of what you experience seems to evaporate—like words that go missing when you turn off your computer without hitting SAVE. But other short-term memories go through a molecular process called consolidation: they're downloaded onto the hard drive. These long-term memories, filled with past loves and losses and fears, stay dormant until you call them up. Here's where the brain is truly ingenious: when events are supercritical or meaningful or scary—a first kiss, a baby's birth, a bike accident—stress hormones alert the amygdala, the brain's emotional control center, which then ramps up the memory-processing machinery, etching that particular event more deeply. "There's a punch to it," says James McGaugh, the University of California, Irvine, scientist who discovered this. These memories are coated with emotional glue. They stick.

Thinking back on the superhappy sticky stuff is fun. We can luxuriate in years past. But the unwanted terrible memories of rapes and robberies and combat can disrupt people's lives and, in some cases, lead to lifelong struggles with PTSD. Some 8 million Americans have the disorder in a given year and the military is at even greater risk. Antidepressants help relieve the symptoms, but fearful recollections often persist. Many patients receive cognitive behavioral therapy, which encourages them to confront their experiences in a safe way. A car-accident victim will talk through the details during repeated therapy sessions, then visit the place where the crash took place. The treatment doesn't get rid of the old memory; instead, patients form a new, competing memory of the event—I can stand at this corner and not get hit—that isn't nearly as toxic. The approach has helped countless patients, but it isn't wildly successful. Overall, symptoms improve by only about 50 percent, and that drops to one third over the long run, says Dr. Roger Pitman, professor of psychiatry at Harvard Medical School. When people see a "cue"—a reminder of the event—it can all come rushing back.

Nader wasn't thinking about PTSD when he launched his research. He became fascinated by the mechanics of memory in the late 1990s after watching Dr. Eric Kandel, a Nobel Prize winner in the field, give a lecture about how the brain processes information. A spark went off in Nader's head: what if memories were consolidated not just once, as researchers believed, but every time they were recalled? What if an original memory could be changed and "reconsolidated"? Joseph LeDoux, of New York University, remembers the moment his postdoc came up with a simple rat experiment to test the idea. "That's crazy, it'll never work," he told Nader.

Using a technique called fear conditioning, Nader gave rodents a foot shock at the same time that he played a sound. The rats now formed a memory: tone equals terror. The next day, Nader prompted the rats to recall the memory by turning the sound on again; as expected, the animals froze in fear. For a memory to be consolidated from short term to long term—from unstable to stable—it must undergo a chemical process called protein synthesis. Nader believed the same might hold true when memories were reconsolidated. Under the reconsolidation theory, some memories can be modified by new information, either intentionally or naturally after they're recalled; that may be why people who witness a crime will testify about what they heard in a news report, rather than what they saw, says LeDoux.

Nader injected a drug that stops protein synthesis into the rat's amygdala. These inhibitors have been used for decades in memory research, but Nader was testing the technique in a new way. His hypothesis was that if he could block reconsolidation of the rat's original fearful memory, it would no longer freeze. When he played the tone and gave the drug a day later and even 14 days later, the rats were indeed blasé about the whole ordeal. That memory linking sound and fear, which was supposed to be immutable, was now defunct. "I went, 'holy cow'," Nader says. " 'This is insane!' "

Researchers first proposed reconsolidation in the late 1960s, but it never caught on. Now science was showing it might be real, and Pitman, who sees the ravages of PTSD in his patients, wanted to get into the act. Already, Pitman had conducted a study in accident victims to see if propranolol, a drug that reduces anxiety, could stop long-term traumatic memories from forming in the first place. Propranolol blocks the action of adrenaline—a stress hormone known to strengthen an emotionally significant memory—and McGaugh had previously shown that the drug could weaken memories in rats. But the approach had two major drawbacks in humans: not everybody who goes through a wrenching ordeal develops PTSD, and nobody wants to medicate unnecessarily. And because memories can consolidate from short term to long term in a matter of hours, patients might get the drug too late. Reconsolidation offered a new line of attack. "If it's true that reactivating memory returns it to an unstable state, here we have a golden second chance," says Pitman. "The implications for PTSD are huge."

Pitman teamed up with Nader, now a neuroscientist at McGill University, and Alain Brunet, a McGill psychologist. The team tested a single dose of propranolol in patients with PTSD. Participants did do better, but most still reported feeling upset by their trauma. Now researchers are giving multiple doses to see if they can get an even stronger effect. "The idea is that over time, we will chip away at this memory," says Pitman. Joël Coutu, 43, signed up for a trial conducted by Brunet. Twelve years ago, two men robbed the pet-food store where Coutu was manager, put a gun to his head and threatened to kill him. Coutu played dead after one of the criminals bashed his head with the weapon. For years, he had nightmares and flashbacks. He broke up with his girlfriend; he quit a job he loved as president of a bird club in Montreal. Usually upbeat, Coutu became depressed. "Inside, I was dying," he says.

On his first visit with Brunet, Coutu wrote down the details of his trauma. Once a week for six weeks after that, he received propranolol, then read the disturbing account. At first, it was agony. "I realized all these emotions I thought were gone weren't," he says. But the fifth time, Coutu noticed a distinct change. "I felt like smiling. All of a sudden, it wasn't me anymore." At his last session, doctors tested Coutu's physiological reactions—his heartbeat, his palm sweat, his facial muscles—to his script and to neutral stories, like a beach scene. His response was similar. "It feels like there's been water poured on the fire," he says. Brunet's early data are compelling: participants' symptoms dropped by 50 percent, and 70 to 80 percent no longer meet the full criteria for PTSD.

What's so fascinating about this research is how it plays on the geography of memory. We often think of memory as one entity wrapped neatly in a bow. But our remembrance of a single experience lands like confetti in the brain, scattered into different locations. The dry facts about what happened—I was walking home, a man assaulted me—appear to lodge in the hippocampus. But the emotional trauma of that same event—the anger at the man, the horror of the moment—seem to be housed in the amygdala. When the memory is recalled, both parts emerge together, like the sound and images in a movie, says Brunet. This is critical to the science of forgetting: researchers believe they may be able to target the fear part of the memory but leave the details of what happened intact. "People cherish their memories, even their bad memories," says Brunet. "They don't want them to be erased, they want to recall them with less pain."

Will memory eradication become all the rage? "I get calls from patients weekly wanting to erase their traumatic memories," says LeDoux. "One guy wants to get rid of the memory of his ex-wife." That will almost certainly never happen. But improving treatment for PTSD and other conditions (researchers are targeting the "appetitive" memory in addicts) is a laudable goal. Not every memory can be reconsolidated, however, and no one's sure that propranolol will be the best therapy. Pitman is studying other medications, too, including a painkiller, a nausea drug and RU-486 (the abortion pill).

Is all of this ethical? Is it real? Kandel says science must always move forward, but, he adds, "removing memory gets into dangerous territory. We have to think about it very carefully." McGaugh, citing conflicting data, doesn't even buy the theory of reconsolidation, though he admits to being in the minority.

Science is filled with fits of successes and legions of failures. Our understanding of human memory, as baffling as the universe, is still emerging. And that, says Yadin Dudai, of Israel's Weizmann Institute of Science, is the fun of it. "There will be enough work for my students' students," he says. "We contribute a little, we incite debates, we experiment to refute or resolve the issues. We go on. We don't know the entire story, but we know more and more." Neuron by wondrous neuron.

Join the Discussion