Skip to main content
Move the World.
Brains Store Memories in a Temporary "Cache" (and We Can Read It)
Mouse neurons grown in a lab (brightened and colorized).Credit: Kelsey Tyssowski (CC BY-ND)

The first dance at my wedding lasted exactly four minutes and 52 seconds, but I'll probably remember it for decades. Neuroscientists still don't entirely understand this: How was my brain able to translate this less-than-five-minute experience into a lifelong memory? Part of the puzzle is that there's a gap between experience and memory: our experiences are fleeting, but it takes hours to form a long-term memory.

In recent work published in the journal Neuronmy colleagues and I figured out how the brain keeps temporary molecular records of transient experiences. Our finding not only helps to explain how the brain bridges the gap between experience and memory. It also allows us to read the brain's short-term records, raising the possibility that we may one day be able to infer a person's, or at least a laboratory mouse's, past experience – what they saw, thought, felt – just by looking at the molecules in their brain.

Electrical pulses carry signals along
the branches of neurons.
Electrical pulses carry signals along the branches of neurons. Credit: Santiago Ramón y Cajal (CC BY)

Electrifying experience

To uncover how the brain keeps track of an animal’s experience, we started by asking how the brain records its electrical activity. Every experience you have, from chatting with a friend to smelling french fries, corresponds to its own unique pattern of electrical activity in the nervous system and brain. These activity patterns are defined by which neurons are active and in what way they’re active.

For example, say you’re at the gym lifting weights. Which neurons are active is fairly straightforward: If you’re lifting with your right arm, different neurons will be active than if you’re lifting with your left arm because different neurons are connected to the muscles of each arm.

The way in which a neuron is active, on the other hand, encompasses an infinite number of possibilities. Neuronal activity consists of pulses of electricity that can occur in pretty much any pattern over time that you can imagine. Electrical activity can vary in duration, or whether the pulses occur in clumps or steadily. In this case, lifting a heavier weight will lead to more pulses per minute.

So, it’s a combination of which neurons are active and how frequently they’re pulsing that makes your experience of lifting a 10-pound weight with your right hand different from that of lifting a 5-pound weight with your left hand.

Activated neurons activate genes

In our experiments, we couldn't test every possible pattern of electrical activity, so we focused just on the way neurons record how long they are active.

We predicted they'd keep these records by turning on genes. All the cells in your body have pretty much the same genes encoded in their DNA. But different genes turn on depending on the type of cell and what it's encountered in its life. Which genes are activated in a particular cell are what makes it different from other cells.

For about 30 years, researchers have known that neurons turn on certain genes when they're electrically active. When a gene in a neuron is turned on, the cell sends a molecular Xerox machine to that gene's place in the DNA. The molecular Xerox makes lots of copies of the gene in the form of new molecules. These new molecules aren't made of DNA, but rather the closely-related RNA. These RNA molecules remain in the cell for hours to days and serve as the brain's record of which neurons were active.

But we wanted to know whether the genes that are on in neurons can record not just that they've been at all active but also the way they've been active. That is, do neurons that are activated differently – for longer or shorter time periods, for instance – turn on different genes?

We thought they would: Long-term memories are stored in physical changes to the neurons themselves, and the type of change is determined by the pattern of electrical activity the neuron experiences. So we predicted that the brain would need to keep track not only of which neurons were active, but also the way those neurons were active in order to make those lasting changes.

Researchers activated mouse neurons
growing in a dish.
Researchers activated mouse neurons growing in a dish. Credit: Kelsey Tyssowski (CC BY-ND)

In our experiments, we activated mouse neurons growing in a dish by exposing them to a chemical that turned them on. As long as the chemical was there, the neurons were active, allowing us to keep them turned on for various lengths of time.

We found that, indeed, neurons in a dish that are activated for different lengths of time turn on different genes. And this genetic record-keeping is unexpectedly simple: The longer neurons are active, the more genes they turn on.

This turned out to be true not only in neurons growing in a dish, but also in the brains of living mice. By exposing mice to bright lights, we were able to activate the neurons in the vision center of their brains for as long as the lights were on. The longer the lights shone, the more different genes turned on, their RNA copies building up in the cell. This means that the set of molecules found in a briefly active neuron is different from that found in a neuron that was active for a long time.

That this simple record-keeping was present in the brains of living mice suggests it's likely also in the brains of humans.

Each neuron contains a metaphorical
machine that translates its electrical
activity into molecular records.
Each neuron contains a metaphorical machine that translates its electrical activity into molecular records. Credit: Anastasia Nizhnik and Kelsey Tyssowski (CC BY-ND)

Temporary records of breaking news

Our work only explains how neurons keep track of how long they were active, but we think neurons may well keep track of all aspects of their activity in the same way. But why would the brain keep this molecular record of an animal's experiences?

I think of these molecular records as being like a newspaper. The brain writes an article about each experience by turning on a specific set of genes in a specific set of neurons. These articles – in the form of RNA molecules – will remain around for hours to days. But just as days-old newspapers are usually tossed out, the copies of the activated genes are not how the brain stories memories for decades.

Instead, the brain reads its temporary newspaper-like records to write its history books: long-term memories. When your brain stores a memory of an experience, it physically changes the connections between the neurons that the experience activated. Those changes can last a lifetime – like my wedding memories. Our group thinks the genes that are on in a neuron probably tell it what kinds of changes to make, like the articles in a newspaper tell scholars what to write in history books.

Reading records

My colleagues and I thought that if the brain is able to read these molecular records when writing its long-term memories, we should be able to read them, too. Like any reliable record, the genes that turned on in response to short versus long activity were predictable. They were actually so predictable that we were able to figure out if a group of neurons had been activated for a long time or a short time just by looking at which genes they had turned on.

So far we can only read how long a neuron's been active, but if we could fully read the brain's records, we might be able to infer someone's experience of their day just by looking at the RNA molecules present in their brain. We could look at the genes that are on in your neurons and figure out that in your workout this morning you lifted a 5-, not 10-, pound weight with your right, not left, hand. And we could tell that you were daydreaming about your date tonight while you did it.

Unfortunately for aspiring mind-readers who are willing to put aside any ethical qualms, it's not actually possible to look at the molecules present in the brain of a living person and probably won't be in the foreseeable future. Furthermore, we don't entirely understand which brain activity patterns correspond to which experiences. So even if we could read these records fluently, we wouldn't be able to infer experience.

Instead, we hope that understanding the brain's record-keeping will provide an easier way to measure brain activity in lab animals for researchers trying to figure out the correspondence between experience and brain activity. Current technologies are somewhat inefficient and can only measure activity in real time, so reading the brain's genetic records could make these experiments more feasible.

So while molecular mind-reading in humans stays relegated to science fiction for now, our work begins to allow scientists to read the records in the brains of lab mice. It's a step toward understanding how the brain converts experience to electrical activity to memory.

·   ·   ·

Kelsey Tyssowski is a graduate student in Biomedical Science at Harvard University. This article was first published on The Conversation.

The Conversation

More About

Innovation
How the Technology Behind Deepfakes Can Help Us Create a Better World
How the Technology Behind Deepfakes Can Help Us Create a Better World
Innovation
How the Technology Behind Deepfakes Can Help Us Create a Better World
Deepfakes have ignited fierce media criticism and call into question the public’s ability to discern fact from fiction. But the…
By Jackie Snow

Deepfakes have ignited fierce media criticism and call into question the public’s ability to discern fact from fiction. But the technology behind Deepfakes, called GANs, has enormous potential to drive innovation beyond fake social media videos. Read more to find out some of the amazing things being done with this technology.

Innovation
An Address for Everywhere on Earth
An Address for Everywhere on Earth
Watch Now
Innovation
An Address for Everywhere on Earth
Can three simple words change how we find each other?
Watch Now

We take addresses for granted - but billions of people and places don’t have them, and it’s a big problem. Whether it’s voting, disaster relief, or pinpointing a spot on festival grounds, not having an address makes things that should be simple difficult. Enter Chris Sheldrick, who coordinated events in the music industry where he was frustrated by address-related problems. He created What3Words, a method of dividing the entire world…

Dispatches
How Redesigning Labs Can Demystify Genetic Science
How Redesigning Labs Can Demystify Genetic Science
Dispatches
How Redesigning Labs Can Demystify Genetic Science
"Scientists work in high-security buildings that are banned to the public and then wonder why they are misunderstood."
By Brook Muller

"Scientists work in high-security buildings that are banned to the public and then wonder why they are misunderstood."

Dispatches
Neuroscience Has a Low-Tech Answer for a Good Night's Sleep
Neuroscience Has a Low-Tech Answer for a Good Night's Sleep
Dispatches
Neuroscience Has a Low-Tech Answer for a Good Night's Sleep
Neuroscientists say that we may be ignoring a basic fact that could defuse the "screen-time wars" between parents and kids.
By Adriana Galván

Neuroscientists say that we may be ignoring a basic fact that could defuse the "screen-time wars" between parents and kids.

Sponsored
Saving Lives with AI
Saving Lives with AI
Watch Now
Sponsored
Saving Lives with AI
Artificial intelligence can find hidden patterns in patient’s vital signs - and stop emergencies before they happen.
Watch Now

As hospitals collect more and more data, analyzing it is a challenge and an opportunity. Montefiore Medical Center of the Albert Einstein College of Medicine is a case study in how using artificial intelligence in hospitals can help improve outcomes. They’re working with Intel’s Healthcare AI team to develop machine learning algorithms that can see patterns within it. The result, which they call the Patient Centered Analytic Learning Machine or…

Sponsored
Can Science Make People Live Healthier for Longer?
Can Science Make People Live Healthier for Longer?
Watch Now
Sponsored
Can Science Make People Live Healthier for Longer?
An MIT researcher has turned 30 years of aging research into something you can use right now.
Watch Now

Most of medical science focuses on combating disease and managing the impact of aging. But one MIT researcher wants to tackle aging head on. Through decades of research, Dr. Leonard Guarente has uncovered a basic mechanism to regulate aging and co-founded Elysium to turn his research into a product. Elysium’s mission is to help people live healthier for longer. Freethink is proud to present this story in partnership with Elysium.…

Dispatches
Zika Could Be a "Smart Missile" for Brain Cancer
Zika Could Be a "Smart Missile" for Brain Cancer
Dispatches
Zika Could Be a "Smart Missile" for Brain Cancer
Zika can devastate fetal brains; scientists want to turn it against brain tumors instead.
By Dan Bier

Zika can devastate fetal brains; scientists want to turn it against brain tumors instead.

Dispatches
The Gut Microbiome Affects Brain Structure
The Gut Microbiome Affects Brain Structure
Dispatches
The Gut Microbiome Affects Brain Structure
What happens in your gut in childhood can change how your brain works later in life.
By Dan Bier

What happens in your gut in childhood can change how your brain works later in life.

The New Space Race
Who Owns the Moon?
Who Owns the Moon?
The New Space Race
Who Owns the Moon?
Throughout history, different organizations, governments, and even individuals have attempted to establish rules for, and ownership of, outer space.
By Mike Riggs

Throughout history, different organizations, governments, and even individuals have attempted to establish rules for, and ownership of, outer space.

What to Expect at the First Cyborg Olympics
What to Expect at the First Cyborg Olympics
What to Expect at the First Cyborg Olympics
The event will seek to answer one of the most interesting technology questions of the early 21st century: How close…
By Mike Riggs

The event will seek to answer one of the most interesting technology questions of the early 21st century: How close are we to integrating humans with machines?