Flexible neural modules could help explain why our brains are able to take on so many functions, with little difficulty, like when we go on a grocery run.
Image Credit: MIT News; iStock
(CC BY-NC-ND 3.0)
Scientific Frontline: Extended "At a Glance" Summary: Flexible Neural Modules
The Core Concept: The brain uses flexible, multipurpose neural circuits—specifically in the prefrontal cortex—that can be repurposed to store different types of information, such as sensory inputs or action plans, rather than relying on separate, dedicated neurons for every distinct task.
Key Distinction/Mechanism: While some areas, like the parietal cortex, may store only specific types of data (e.g., sensory memories like a tone), these flexible modules in the prefrontal cortex act like "cognitive Legos." The same cluster of neurons can maintain sensory information in working memory during one phase of a task and then switch to holding a motor/action plan in a later phase.
Major Frameworks/Components:
- Working Memory: The system responsible for temporarily holding and manipulating information needed for complex cognitive tasks.
- Prefrontal Cortex: The brain region involved in executive functions, such as planning, decision-making, and hosting these flexible memory modules.
- Parietal Cortex: The brain region that processes sensory information and plans movement, shown in this study to handle more specialized, single-purpose memory storage.
- Compositionality: The concept that complex behaviors are built by combining and reusing smaller, discrete cognitive processes.
Branch of Science: Neuroscience, Cognitive Science, and Computational Neuroscience.
Future Application: Understanding how to manipulate or inhibit these modules could provide deeper insights into cognitive flexibility, learning, and the fundamental mechanisms of decision-making. This could potentially inform treatments for cognitive disorders or influence the development of more efficient, brain-inspired artificial intelligence and neural network architectures.
Why It Matters: This discovery fundamentally shifts the understanding of how the brain manages an infinite variety of tasks with a finite number of neurons. By proving that neural circuits can be flexibly reused, it explains how humans and animals can learn new tasks and adapt to complex, everyday situations without needing to build entirely new neural pathways for every action.
As we move through everyday life, our brains engage in a huge variety of cognitive tasks. For example, during a grocery run, we might have to recall the items for a recipe, remember where the clerk said the flour was located, and count out money to pay.
Scientists have long theorized that the brain contains modules, or clusters of neurons, that perform the same computation across many different types of tasks. This type of modularity could help explain why our brains can take on so many functions with little difficulty.
In a new study of mice, MIT neuroscientists have found the first evidence for the existence of these flexible modules. They identified neurons in the prefrontal cortex that can be used to store either a sensory input or an action plan in working memory.
“We found that the brain doesn’t dedicate a separate group of neurons for every type of information. Instead, it uses the same populations of neurons to perform the same computation on different kinds of information, which means the same subset of neurons can hold both an action and a sensory stimulus in working memory,” says Yuma Osako, an MIT postdoc and the lead author of the new study.
The discovery supports the theory that reusable circuits allow the brain to mix and match components to generate a rich variety of behaviors, the researchers say.
Mriganka Sur, the Newton Professor of Neuroscience at MIT’s Picower Institute for Learning and Memory, and Timothy Buschman, PhD ’08, a professor at the Princeton Neuroscience Institute, are the senior authors of the paper, which appears today in Nature Neuroscience. MIT graduate student Greggory Heller and postdoc Sofie Ahrlund-Richter are also authors of the study.
Cognitive Building Blocks
Dating back to his time as a graduate student at MIT, Buschman has been interested in understanding how the brain can perform so many different kinds of behaviors.
“One of the solutions that’s always been proposed has been this idea of compositionality—that you can take pieces of cognition that perform part of a task and reuse them in another task,” he says.
In a study published last year, Buschman’s lab at Princeton showed that when animals perform a task such as categorizing objects based on their shape or color, they assemble neural circuits that perform different pieces of the task. Just like “cognitive Legos,” these building blocks can be flexibly combined to generate new behaviors.
Osako, who joined Sur’s lab several years ago, was also interested in studying cognitive flexibility. He and Sur teamed up with Buschman to explore a related question: whether individual neural circuits can be repurposed to perform different functions.
“Our everyday lives require us to temporarily hold many different kinds of information. One big question is how the brain can represent an unlimited variability of information using only a finite number of neurons,” Osako says.
To get at that question, the researchers trained mice on a task in which they had to determine whether two sensory stimuli (high- or low-pitched tones) were the same and respond accordingly.
The researchers recorded electrical impulses from the brain while the mice performed this task, focusing on the prefrontal cortex, which is involved in executive functions such as planning and decision-making, and the parietal cortex, which processes sensory information and plans movement.
After measuring electrical activity from thousands of neurons, the researchers performed computational analyses that allowed them to identify groups of neurons that encode specific pieces of information.
They focused on two time periods—the time between the first and second tones, when the animals were holding a memory of the first tone, and the time between the second tone and the point when they had to decide on an action. During that second period, the animals held their decision and action plan in their working memory.
Within the parietal cortex, the researchers found that neurons appeared to exclusively store memory of the tone. However, in the prefrontal cortex, they identified a cluster of neurons that could switch between the two types of memory. During the first period, they stored a memory of the first tone, but during the second, they were responsible for remembering the plan of action.
Reusing these clusters for different purposes allows the animals to flexibly store different types of information, the researchers say.
“When mice do tasks that test whether memory computations can be reused, the answer is they are. There are subspaces of functional activity in the prefrontal cortex that can be the substrate of mixing and matching toward flexible cognition,” Sur says.
Computational Flexibility
The new findings offer support for the idea that the same computational circuits can be used for different purposes, Buschman says.
“The main result from this study is that there’s a circuit in the brain that maintains items in working memory, and you can put either sensory or motor information into it and flexibly reuse it depending on what your current task is,” he says. “This means you do not have to build an entire new circuit for holding information in mind every time you want to learn a new task.”
The researchers now plan to study whether inhibiting these modules during different parts of the task affects the animals’ behaviors, which could offer additional evidence that the flexible modules they identified participate in a variety of functions.
Funding: The research was funded by the National Institutes of Health, a MURI grant, the Picower Institute Innovation Fund, the Japan Society for the Promotion of Science Overseas Research Fellowships, and the Uehara Memorial Foundation Postdoctoral Fellowship.
Published in journal: Nature Neuroscience
Authors: Yuma Osako, Greggory R. Heller, Sofie Ährlund-Richter, Timothy J. Buschman, and Mriganka Sur
Source/Credit: Massachusetts Institute of Technology | Anne Trafton
Edited by: Scientific Frontline
Reference Number: ns081726_01