“The underlying grip of movies is that they program us to have experiences. They create events in our heads…”
- Jeffrey Zacks, Flicker: Your Brain On Film. Forthcoming from Oxford University Press
If you’re interested in science, you probably heard last week’s news that memories were successfully implanted in the brains of mice. In a study led by Steve Ramirez of MIT, published in Science, and covered by nearly every major news outlet including us, a mouse’s behavior could be manipulated by implanting a past experience directly into its brain.
This is the stuff of The Matrix or Eternal Sunshine of the Spotless Mind: decoding experiences from neural patterns and creating false ones that feel real. But the story behind the science has been largely skipped. Where did the idea come from? Hollywood, as it turns out.
Steve Ramirez is a PhD student in the lab of Nobel laureate Susumu Tonegawa. He’s also the college roommate of Co.Labs’ News Hacker Gabe Stein; and, like me before I became a journalist, he studies the brain basis of memory. I spent three years in labs at Princeton, Kyoto, and St. Louis studying how memory happens in humans. Like Steve’s lab, the ones where I worked were searching for memories in the brain—the “temporary constellations,” as Harvard psychologist Dan Schacter calls them, that light up when a person lives an event and then echo back when the event is recalled. So I was eager to pick the brain of the world’s first-ever “memory inceptor” about where he gets his ideas.
“I thought memory sounded interesting,” Ramirez tells me, when asked how he first got involved in neuroscience research as an undergrad at Boston University. “I knew that if you got brain damage in certain parts of the brain, you’d lose certain parts of your memory. And there are movies like Eternal Sunshine of the Spotless Mind, Memento, and 50 First Dates that I thought were interesting. But I didn’t know what was their real neuroscience basis. So I said, ‘Hollywood has asked these questions before; here’s a lab that studies memory. Maybe I can get better insight into these movies, and at the same time get some fundamental insight into how memory works.’"
Here’s how it worked. The experimenters used flashes of light directed into the brain of the mouse to convey the neural patterns representing a dwelling place--let’s call it Box A-- where the mouse had been before. They did this while administering a shock to the mouse in a new place. That shock created a negative experience, while the light imprinted the old place on the mouse's brain-- associating a new, bad memory with the sensation of being in Box A.
Afterward, the mouse was found to freeze in fear when he walked into Box A, where he was never shocked, but had had the place’s neural signature shot into his brain.
Implanting the memories required surgically implanting an apparatus in the mouse’s brain. “You can eavesdrop on the brain when memories are being formed,” says Ramirez. “I fell in love with it.”
Here he explains how he went from this original inspiration to what he has nicknamed his “Inception experiment.”
Where’d you get the idea to create false memories?
We began touching on these ideas mainly because all of us are huge fans of movies like Inception, the ideas behind movies like Total Recall and Eternal Sunshine or Memento. For me personally, looking to Hollywood is a great source of questions.
I feel that Hollywood is a repository of all these fantastic ideas, because nobody in Hollywood is limited. Chris Nolan [director of Memento and Inception], for example, is not limited by a budget anymore. He can ask whatever question he wants, and turn it into art, through film. So I look at [filmmaking] as unfiltered creativity, and wonder: Can this have a basis in reality? Or how could this have a basis in reality? When [scientists] look at Hollywood, the goal is: filter these questions through the lens of science, and re-pose that question in a way that could be experimentally tractable, that can be framed as a scientifically tractable problem. I love it.
So you were interested in memory. You’d been recording from mouse brains, at BU. Did that drive you to the lab you’re working in at MIT?
The Tonegawa lab is the one I thought would let me ask the overarching questions I’d like to devote my career to. Big questions like: Is it possible to find memories in the brain, at the level of individual brain cells? Could you tinker with those brain-cells and therefore tinker with the memory? There are few labs in the world where I thought I’d be able to ask those questions.
How did you decide to try to implant memories in mice?
Three years ago, Xu [Liu, Steve's research partner, a postdoc in the Tonegawa lab] and I were asking questions about identifying memories in the brain, and playing with them. So last year we focused work on this big question: Is it possible 1) to find the underlying neurons representing a memory, and 2) can we trick those neurons to respond to light?
Last year, we were able to do just that [in this paper published in Nature]: We were able to find the memory, and trick the memory to respond to just pulses of light--by shooting light into the brain and activating the corresponding brain cells.
What do you imagine for Eternal Sunshine? You’d try to isolate a memory in the brain, like you guys have done, and then delete it?
Yep. One thing Eternal Sunshine got wrong was localizing memories. There’s a scene with Elijah Wood, where they’re going into the brain, and [saying] ‘There’s a memory right here, it’s at point A in the brain’, and boom, they delete it. But in reality, memories are distributed throughout the brain. There’s no one XYZ coordinate and bam, that’s the memory. Different aspects of memories are processed in different parts of the brain—high and low, front and back.
So, there’s the memory of Kate Winslet, and then there’s the awful underlying, visceral feelings that Jim Carey has when he recalls Kate Winslet [because of their break-up]: the emotional undertones that color in that memory. Now, the emotional undertones and the memory of Kate Winslet itself are largely mediated by separate brain systems. So you can imagine going into the brain, finding the brain cells that represent that dark feeling of a break-up, and inactivating only those. So the memory of Kate Winslet is still intact, but the emotional feeling is gone. You could imagine this might one day be useful for treating people with PTSD.
If mice had a Hollywood, things like Eternal Sunshine would totally be possible right now. I think that’s a possibility in mice, for sure.
You’ve talked about your collaboration with Xu Liu. How did the genesis of the false memory experiment come about? Did one person come up one day and say ‘Hey, why don’t we try this specific thing?’ or was it more collaborative and gradual?
When I joined the lab, the ideas of isolating and trying to reactivate memories were there already. Xu was the postdoc who was ballsy enough to take that on and say ‘Can we make this work?’ When I joined the lab, everything kind of reached escape velocity. I teamed up with Xu and we started asking ‘How can we do this?’
With the Inception Project, our team had grown in size. There were a couple of other postdocs, a couple grad students who were just as committed as we were to asking these questions, and they also shared the air of scientific camaraderie and this hugely ambitious goal of trying to play with single memories in the brain.
We sat down, about a dozen of us, in a conference room for hours saying ‘Well what about this experiment?’ ‘What if we labeled this memory?’ and it was just this unbelievably fruitful dialogue. Sort of like: You have an idea, and give it your all. The Hollywood influence was undoubtedly there. And then it was just sitting down and getting our hands dirty. These ideas were really born through those meetings.
So, in the movie version of this discovery, you and Xu will be main characters. You’ve mentioned ten others. How big is the Tonegawa lab?
Unbelievably humongous. Right now it’s around fifty-something people. That’s one of the reasons I joined. This is fifty different perspectives, fifty different criticisms, fifty different contributors, intellectually or hands-on, to a given project. That’s in the service of trying to make your science as watertight as possible.
So imagine when you’re presenting your ideas at lab meeting: fifty [MIT scientists] are sitting there saying ‘Hey, this is good. Maybe you could improve on A, B, and C.’ Or: ‘Hey, you need to do this control if you’re going to claim A, B, and C.’ It's great to shave off the fat of a project and get to the substance very fast.
You know, everybody says that a lab is like a family; and I’ve realized how true that actually is. For me, coming in day in and day out, science feels more like a hobby and less like a 9-to-5 job. The team that we have of about a dozen people, they’re the kind of people who are crazy-serious about science, and they love these lofty ideas, but on Friday once it’s 5:00 it’s obviously time to go have a beer and decompress.
Tonegawa is very famous, seventy-three years old, and Japanese. There are cultural differences between the way Japanese science is often done and the way American labs tend to work. Is Tonegawa’s default mentality more like the Japanese top-down style?
No. He has 100% embraced the sort of ‘Western’ mentality. He dislikes the Japanese or traditional [hierarchical] ‘Eastern’ style of running labs, because he likes [a lab] to be this egalitarian environment where you’re basically judged by the quality of your ideas, regardless of whether you’re a postdoc or a grad student or a professor: he cares about your ideas, the substance behind them.
You can still tell there is this child-like wonder when he’s talking about an unanswered question. For a guy in his mid-seventies, it’s pretty amazing. He doesn’t care who posed the question or who claims to have answered it: he just wants to know the truth behind the question. All that matters is the answer to that question, because that’s how science progresses. His mentality is: It’s an even playing field and you are judged by the quality of your ideas, and your execution of those ideas.
Let’s talk about reconsolidation: the idea of a memory being rendered flexible when it's recalled, and modified. Did this concept play into planning the experiment?
Absolutely. We update our memories all the time. Every time you recall a memory, it is reconstructed, reconsolidated; it's constantly updated. It's interesting because this, ironically, means that your most accurate memories are those you never recall.
Our idea with the false memory project was to find a memory in the brain that represents an entirely neutral environment, and then update that with aversive emotional information: to change that memory from a neutral environment to an aversive one by artificially associating the two. This really speaks to just how flexible this memory process is when it's being recalled.
When I read the press about your study, and even the title of your paper, I questioned the language: You say you've induced a "false memory," but this seems more like an implanted experience. I don't doubt that a perceptual, spatial experience flashed into the mouse's mind's eye. But do you believe this feels like recalling a moment from the past, or more like a hallucination or a dream?
What we're doing is very artificial, so I wouldn't be surprised if it had elements of both [a memory and a hallucination]. You’re right, we just don’t know.
You mention the headlines and tabloids. One common misconception that a lot of media read into the paper is that they approached [the experiment] as if we made a memory from scratch, de novo--like we wrote a memory into the brain. But it wasn't like that. We found a memory, then associated it with an emotional stimulus. To update a neutral environment and incorporate aversive information into it, that's what we're calling a 'false memory.' Whereas writing a memory from scratch? There are some experiments that could possibly get at that, but it's definitely the next step, not what we've done yet.