The Rajasethupathy lab's recent research reveals a cascade of molecular timers that determine whether short-term impressions consolidate into long-term memory, with implications for memory-related diseases. (Credit: Matthew Septimus)
One of the brain's most essential tasks is deciding which experiences to store as memory, and which to forget. For decades, scientists assumed this process was governed by simple on-off switches in the brain. But a new study in Nature overturns that view, showing instead that the persistence of memory is regulated by a cascade of molecular timers that unfold across multiple brain regions.
The findings, from Priya Rajasethupathy's lab at Rockefeller, further reveal that an unexpected node-the thalamus-plays a central role in shepherding memories from short to long-term, via gene programs that progressively stabilize each memory. These discoveries not only offer a fresh framework for how the brain sorts and preserves memory, but also open the door to potential new strategies for treating memory disorders.
We sat down with Rajasethupathy to discuss how her team uncovered these hidden timers, why memory may be more malleable than we thought, and what this work could mean for Alzheimer's disease and beyond.
How have neuroscientists historically thought about memory formation?
For a long time, when scientists were trying to think about how memories persist, the major model was that there are transistor-like memory molecules. Something with an on-off switch where, if you flip that switch, the memory is tagged to be "on" and it stays that way forever. Such switches are attractive because they provide a mechanism for fleeting experiences to be coded into lasting changes in the brain. Maybe these switches were long-lived enzymes/proteins, or modifications to DNA, or mechanical changes in synaptic or cellular structure, but the consistent idea in the field was the existence of some kind of enduring switch that tells the brain that we want to maintain this memory long-term.
So what's really happening?
One of the key insights of our study is that we shouldn't be thinking of the conversion of short- to long- term memory as one-and-done. Instead, it's a progressive process. The brain sets a timer for, let's say, a few minutes. If the memory continues to be important, the brain then promotes that memory and turns on a second timer that can last, say, a few hours. If that memory is still important, the brain turns on yet another timer that can last days, then weeks, and so on.
In this view, everything we experience can be formed as a memory, but we have mechanisms in place to rapidly forget-unless a memory is promoted onto one of these timers. In our present work, we identified three such timers that essentially allow memories to last longer and longer.
Do these timers also explain how we forget things?
Yes, this was an important aspect of our study. The strength of the switch model is that it can explain memory persistence. But such a model, while good at storage, is bad at forgetting. If memory involved turning on a persistent switch, it would be difficult to forget something in the future that had already been committed to long-term memory. A cascade of timers offers an explanation for how the brain retains the flexibility to promote or demote memories over time.
Let's say you experienced something emotionally rewarding, or traumatic, and you kept thinking about it. Over time, as you thought about it, that memory would get promoted through several of these timers. But then say a month or a year goes by, and you haven't thought about the event. Since each timer is set for a longer duration, eventually you might reach a checkpoint that the memory does not pass- thus, time and experience can be integrated to continuously sculpt the persistence of a memory.
Where does all of this happen?
These molecular timers span the brain-not acting in isolation. For decades, the field has thought of short-term memory as forming in the hippocampus and long-term memory as being represented in cortex. In a series of recent studies, my lab has identified the thalamus as a key link between these areas, helping to assign value to memories and rout them for longer term stabilization. In the present study, it was surprising to see three separate timers acting in separate circuits across the brain-spanning the hippocampus, thalamus, and cortex to progressively prolong a memory. Such a process allows for integrating an initial system of fast memory acquisition (but fast decay), with subsequent systems for slower acquisition but longer retention.
We don't believe there are only three timers, but it provides a framework for why multiple circuits across the brain, each acting on different time-scales, are recruited to support continuous memory stabilization.
What are the implications for diseases impacting memory?
In many cases of memory loss, like Alzheimer's, the hippocampus-where that initial timer first forms and stores a short-term memory-is getting damaged. Now that we know there is robustness and redundancy built into this system with multiple timers in multiple brain regions, we can begin to ask: what if we could bypass damaged areas by turning on molecules that can route memories to healthier circuits? In other words, our brains are built to compensate-can we leverage brain redundancy to improve cognitive resilience?
Where will these findings lead us next?
For many decades, there has been a large focus on studying the hippocampus to understand memory. If you read the patient accounts, you can see why there was so much interest. Patients with lesions of their hippocampus simply couldn't form new memories. The patients would come in, talk to their physician, and a short time later, ask "who are you?" The role of the hippocampus in memory formation was so striking that it attracted a lot attention. But the flip side is that we know almost nothing about what happens to memory beyond the hippocampus. An area that my lab is just starting to make some exciting progress on.
By systematically exploring all of the output pathways of the hippocampus, we have been able to identify the thalamus as a key center that sorts, routes, and maintains memories in longer term repositories. How it does this is a major set of next steps for us.
