Two sets of neurons in the cortex help mice assess their performance on motor tasks.
If at first you don’t succeed, try again – a different way.
Whether it’s hitting a baseball or flipping a pancake, learning any coordinated movement involves messing up and then adjusting future attempts accordingly. Now, researchers have identified two distinct groups of neurons in the mouse brain cortex that fire separately when a mouse either succeeds or fails at a motor task, they report June 25 in Neuron.
These signals seem to tell the mouse “that performance was good or bad, regardless of the reward,” says Adam Hantman, a group leader at HHMI’s Janelia Research Campus.
Hantman led the study alongside Jackie Schiller of Technion-Israel Institute of Technology. Schiller and members of her lab spent a year working at Janelia through its Visiting Scientist Program, then continued collaborating with Hantman after they returned to Israel.
The motor cortex helps animals plan and execute movements, and Hantman and Schiller wanted to know how it monitored the outcome of those actions. Their team trained mice to reach out and grab a food pellet and then put it in their mouth. The task requires sophisticated motor control, and mice succeed only about half the time, so it’s a useful tool for studying learning, Hantman says.
While monitoring brain activity of mice during the grasping task, the researchers noticed distinct activity patterns in the cortex after each round. If a mouse successfully got the pellet, one set of neurons in the cortex lit up. If the mouse fumbled, a different set of neurons in the same area lit up.
The neurons appeared to be responding to the performance itself, rather than any reward that the mice got from eating the food, Hantman notes. The neurons showed the same response when mice reached for and grabbed an inedible plastic pellet. And they didn’t fire when mice got a food reward without having to reach for it.
These success- and failure-marking neurons may talk to other group of neurons in the cortex that send signals to the spinal cord and ultimately direct movement. Their responses seem to cue these downstream neurons to execute the next task differently, based on poor past performance, Hantman says.
“Understanding what cues guide these neurons to respond, how they develop during learning, and how they are modified when we adapt to new conditions will help us understand what happens in the brain when we learn via a reinforcement process,” Schiller adds.
Shahar Levy, Maria Lavzin, Hadas Benisty, Amir Ghanayim, Uri Dubin, Shay Achvat, Zohar Brosh, Fadi Aeed, Brett D. Mensh, Yitzhak Schiller, Ron Meir, Omri Barak, Ronen Talmon, Adam W. Hantman, and Jackie Schiller. “Cell-Type-Specific Outcome Representation in the Primary Motor Cortex.” Neuron. Published online June 25, 2020. doi:10.1016/j.neuron.2020.06.006