In mice, MIT finds brain circuits that switch tasks

A mouse hears a tone, waits, hears another, and must decide whether the two match. During that short sequence, researchers found that a group of neurons in the prefrontal cortex could switch what it was holding in working memory: first the sensory memory of the tone, then the animal’s plan for action.

The team, led by MIT postdoctoral researcher Yuma Osako with senior authors Mriganka Sur of MIT and Timothy Buschman of Princeton, recorded electrical activity from thousands of neurons while the mice performed the task. Computational analyses separated the periods when the animals remembered the first sound from the period when they prepared their response.

The contrast between brain regions was telling. In the parietal cortex, neurons appeared to store only the tone memory. In the prefrontal cortex, the researchers identified a cluster that could be repurposed for both kinds of information. The finding is the first evidence, according to the researchers, for these flexible modules in the mouse brain.

The mechanism gives substance to an old idea in neuroscience: that cognition may be built from reusable components, much like “cognitive Legos.” Instead of constructing an entirely new circuit whenever an animal learns a new task, the brain could place different information into the same computational structure and combine it with other modules.

So what changes, concretely? The result points to one way the brain can manage many forms of information with a limited number of neurons—holding a sound in mind in one moment and a movement plan in the next. It is still a laboratory result in mice, not a demonstration of human cognition. The researchers now plan to inhibit the modules during different stages of the task to see whether the animals’ behavior is affected.