It was the day before Christmas, and the normally busy MIT laboratory on Vassar Street in Cambridge was quiet. But creatures were definitely stirring, including a mouse that would soon be world famous.
Steve Ramirez, a 24-year-old doctoral student at the time, placed the mouse in a small metal box with a black plastic floor. Instead of curiously sniffing around, though, the animal instantly froze in terror, recalling the experience of receiving a foot shock in that same box. It was a textbook fear response, and if anything, the mouse’s posture was more rigid than Ramirez had expected. Its memory of the trauma must have been quite vivid.
Which was amazing, because the memory was bogus: The mouse had never received an electric shock in that box. Rather, it was reacting to a false memory that Ramirez and his MIT colleague Xu Liu had planted in its brain.
“Merry Freaking Christmas,” read the subject line of the email Ramirez shot off to Liu, who was spending the 2012 holiday in Yosemite National Park.
The observation culminated more than two years of a long-shot research effort and supported an extraordinary hypothesis: Not only was it possible to identify brain cells involved in the encoding of a single memory, but those specific cells could be manipulated to create a whole new “memory” of an event that never happened.
“It’s a fantastic feat,” says Howard Eichenbaum, a leading memory researcher and director of the Center for Neuroscience at Boston University, where Ramirez did his undergraduate work. “It’s a real breakthrough that shows the power of these techniques to address fundamental questions about how the brain works.”
The prospect of tinkering precisely with memory has tantalized scientists for years. “A lot of people had been thinking along these lines,” says Sheena Josselyn, a senior neuroscientist at the Hospital for Sick Children in Toronto, who studies the cellular underpinnings of memory, “but they never dreamed that these experiments would actually work. No one ever thought that you could actually, really do this.”
Except Ramirez and Liu. Their work has launched a new era in memory research and could someday lead to new treatments for medical and psychiatric afflictions such as depression, post-traumatic stress disorder and Alzheimer’s disease. “The sky is really the limit now,” says Josselyn.
Though the work so far has been done on lab mice, the duo’s discoveries open a deeper line of thought into human nature. If memories can be manipulated at will, what does it mean to have a past? If we can erase a bad memory, or create a good one, how do we develop a true sense of self? “Memory is identity,” the British author Julian Barnes writes in his memoir Nothing to Be Frightened Of. “You are what you have done; what you have done is in your memory; what you remember defines who you are.”
“I was always amazed by the level of control that science can have over the world,” says Ramirez, who collected rocks as a kid and remembers being astounded that there actually were ways to figure out how old rocks were. “The example is kind of banal by now,” he says, “but as a species we put somebody on the moon. And we figured out for the most part how to eradicate things like smallpox, things that you can’t see, whose existence you have to infer from indirect measurements, until your microscopes get good enough.”
What Ramirez, now 26, and Liu, 36, have been able to see and control are the flickering clusters of neurons, known as engrams, where individual memories are stored. Joining forces in late 2010, a few months after Ramirez began his graduate work at MIT, the two men devised an elaborate new method for exploring living brains in action, a system that combines classic molecular biology and the emerging field of optogenetics, in which lasers are deployed to stimulate cells genetically engineered to be sensitive to light.
Armed with state-of-the-art tools, and backed by MIT’s Susumu Tonegawa, a Nobel laureate for his work in immunology whose lab they were a part of, Ramirez and Liu embarked on a quest that resulted in two landmark studies published 16 months apart, back-to-back blasts of brilliance that advanced our understanding of memory at the cellular level. Ramirez describes the discoveries, as he does almost everything, with exuberance: “The first paper was like catching lightning in a bottle, and the second paper was like lightning striking the same place twice.”
In the first study, published in Nature in March 2012, Ramirez and Liu identified, labeled and then reactivated a small cluster of cells encoding a mouse’s fear memory, in this case a memory of an environment where the mouse had received a foot shock. The feat provides strong evidence for the long-held theory that memories are encoded in engrams. Most previous attempts involved tracking either the chemical or the electrical activity of brain cells during memory formation. Ramirez and Liu rejected those methods as too inexact. Instead, they assembled a customized set of techniques to render mouse brain cells in their target area (a part of the hippocampus called the dentate gyrus) sensitive to light.
Working with a specialized breed of genetically engineered lab mice, the team injected the dentate gyrus with a biochemical cocktail that included a gene for a light-sensitive protein, channelrhodopsin-2. Active dentate gyrus cells—those participating in memory formation—would produce the protein, thus becoming light-sensitive themselves. The idea was that after the memory had been encoded, it could be reactivated by zapping those cells with a laser.
To do that, Ramirez and Liu surgically implanted thin filaments from the laser through the skulls of the mice and into the dentate gyrus. Reactivating the memory—and its associated fear response—was the only way to prove they had actually identified and labeled an engram. The researchers sacrificed the animals after the experiment and examined the brain tissues under a microscope to confirm the existence of the engrams; cells involved in a specific memory glowed green after treatment with chemicals that reacted with channelrhodopsin-2.
When Ramirez and Liu looked at the treated neurons through the microscope, “it was like a starry night,” says Liu, “where you can see individual stars.” Though these active cells were just one part of a widely distributed foot shock engram, reactivating them was enough to trigger a fear response.
The next step was to manipulate a specific engram to create a false memory, an elegant experiment detailed in Ramirez and Liu’s second paper, published in Science in July 2013. They prepared the mouse, injecting the biochemical cocktail into the dentate gyrus. Next, they put the mouse in a box without shocking it. As the animal spent 12 minutes exploring, a memory of this benign experience was encoded as an engram. The following day, the mouse was placed in a different box, where its memory of the first (safe) box was triggered by shooting the laser into the dentate gyrus. At that exact moment, the mouse received a foot shock. On the third day, the mouse was returned to the safe box—and immediately froze in fear. It had never received a foot shock there, but its false memory, created by the researchers in another box, caused it to behave as if it had.
There was no chance that the mouse could have mistaken one box for another: They were different shapes and colors and had different scents. Ramirez and Liu also used multiple control groups—ruling out the possibility that the flash of the laser itself and not the engram activation caused the fear reaction the next day, for example. They had indeed created a memory.
The announcement generated a news media frenzy. “Scientists Trace Memories of Things That Never Happened,” read the headline in the New York Times. Ramirez and Liu woke up in the middle of the night to do live interviews on European radio. Liu’s parents, back in China, read about his achievements online. The public fascination with the role of false memory in criminal trials (the tall, dark-haired bank robber the eyewitness saw was actually short and bald) helped drive the story. But no doubt the science fiction overtones made it especially fascinating. To many it seemed to confirm familiar (and frightening) ideas from movies like Inception and Eternal Sunshine of the Spotless Mind. Nothing is as it seems; reality is but a dream; who you gonna trust, me or your lying eyes?
To neuroscientists, Ramirez and Liu’s discovery was downright dashing. “For me, what made them successful was their fearlessness,” says Josselyn. “You could imagine all the things that could go wrong, but these guys went in there, they got the best tools, they applied the best sort of mind power.” Eichenbaum agrees that the young scientists went “out on a limb” and took a major risk with their careers. “They could have spent three years and ended up with nothing to show for it,” he says.
Spend a little time around Ramirez and Liu, and you quickly sense their upbeat attitude. They come from different worlds—Liu was born and raised in Shanghai, the son of a chemical engineer father and a mother who worked for the railroad, and Ramirez’s parents fled the civil war in El Salvador in the 1980s and settled in Everett, Massachusetts—but their well-matched personalities are no accident. In the fall of 2010, as Liu was interviewing potential partners to explore the mysteries of memory with him, he at first concentrated on scientific expertise. But as time passed he put a different attribute at the top of his wish list—happiness. “If you are going to collaborate with people, you want to collaborate with happy people,” says Liu. “And Steve’s one of the happiest guys I’ve ever seen.” He is also a speed talker who squeezes a whole lot of words into every breath. “He cannot stop talking,” Liu jokes. “Otherwise he will die.”
When Ramirez was young, he often went to the Harvard animal locomotion lab with his father, who started out cleaning cages there and sweeping floors and later became the head animal technician. During visits to the lab Ramirez saw llamas, ostriches and other creatures, and “people doing cool things with animals, taking measurements and stuff.” He figures he “inadvertently absorbed something” that helped push him toward science.
But it was the brain that closed the deal. “Whether it was a sonnet, or getting somebody to the moon or figuring out the biological molecules of life, all of that was the product of brains, of neural activity,” says Ramirez, describing how his broad interests—in Shakespeare, engineering, biology and more—finally led him to neuroscience. “Why not study that which produced everything?”
Liu also demonstrated a scientific bent early in life. And while he is certainly not the first scientist who spent his childhood collecting bugs, Liu’s dedication was distinctive. He raised families of centipedes, had lots of shiny beetles, and kept locusts in tiny cages. He usually fed the locusts edamame but discovered that hot peppers caused an interesting reaction. “They would sing even more,” he says. After studying biology as an undergraduate at Fudan University in Shanghai, Liu received his doctorate from Baylor College of Medicine studying memory in the fruit fly.
As a teenager he’d dabbled in science fiction and wrote a novel called The Challenge. It was about a future in which athletes no longer competed directly against each other but, rather, submitted to various objective measurements of performance or physiology: speed, strength, lung capacity and so on. The hero wants to return to real competition and restore the unmeasurable factors of luck and chance.
One day this past spring, as Liu was listing the many things that could have gone wrong in his work with Ramirez—they could have been beaten to the discovery by a rival team, they could have picked the wrong part of the brain to zero in on—he said he was convinced that luck had played a role in their success. If so, I said, then his work as an adult had delivered on the theme of his boyhood novel. “That’s amazing,” he said after a long silence. “I never made that connection between the book and this work, but I think you are right.”
More than two dozen labs around the world have projects going that build on Ramirez and Liu’s research. Eichenbaum, for instance, is interested in reproducing a larger experience, a memory that occurs over time, like navigating a maze.
At a time when the treatments for many serious mental illnesses are lacking, the potential clinical applications of memory modification are enticing. “This is kind of crazy,” says Josselyn, whose work centers on Alzheimer’s disease and other memory-related disorders, “but maybe somebody with Alzheimer’s... maybe we can figure out a treatment to just go in and do what these guys did in their papers, and sort of activate these cells artificially, boost the activation and have the memories recalled better.”
In another theoretical application, PTSD might be eased by repeatedly reactivating a bad memory to show that the memory itself is not harmful, or by erasing the traumatic components of a specific bad memory, or by replacing it with a positive one. Building on Ramirez and Liu’s work, others in the Tonegawa lab did exactly that in male mice earlier this year, converting a negative memory of a foot shock into a positive memory of an encounter with a female mouse.
Ramirez, who is finishing his PhD at MIT, and Liu, who is headed to Northwestern University to start his own lab, have recently taken on another big memory question: Can we intervene in a depressed state in an animal by reactivating positive memories? The answer appears to be yes. They are studying mouse models of anhedonia, or loss of interest in pleasure, a symptom of depression. Experimental mice subjected to stress until they no longer seek pleasure (such as a sip of sugar water) recover their interest when engrams for pleasant experiences are reactivated. The success rate so far is 80 percent.
“Because the proof of principle is there that we can artificially reactivate memories and create false memories in animals,” Ramirez says, “the only leap left between there and humans is just technological innovation.”
What about the ethical concerns of memory manipulation? Patricia Churchland, a professor at UC San Diego and author of Touching a Nerve: The Self as Brain, says therapy of this sort won’t be as profound a change as it seems. Human memories, inexact and labile to begin with, have long been the target of intervention, from cognitive-behavioral therapy to electroshock to medication. Treating conditions like depression at the engram level “is continuous with what we are already doing,” says Churchland, a leading philosopher of neuroscience.
Ramirez believes that memory surgery is inevitable, though there are a great many questions to address. How could it be done safely? Noninvasively? Ethically? How would patients be selected? As painful as heartbreak usually is, most of us also recognize that it’s a natural, even healthy, part of life. A high-school boy who just broke up with his girlfriend might not be a good candidate for memory surgery. But people with dementia or severe depression—would it be inhumane not to ease their suffering if an effective, safe memory intervention were possible?
The inroads that Ramirez and Liu have made into the mechanics of memory are opening a wide new world of possibilities that are profound, frightening, astonishing—and urgent. “We need to start the conversation yesterday about what we are going to do when this does happen,” Ramirez says, “so that we are ready and know how to handle it.”