After scurrying down a track to find a tasty food pellet, a rat, like any sensible creature, pauses to eat. Now, neuroscientists have found that while the rat chews, its brain appears to replay a memory of the path that led to the bounty. Surprisingly, the memory plays in reverse, which may strengthen it so the rat can later retrace his steps to this mysterious food-giving place.
Neuroscientists David Foster and Matthew Wilson at The Massachusetts Institute of Technology in Cambridge recorded the electrical activity of neurons in the hippocampus of four rats while each animal ran back and forth on a straight track. In particular, the pair was interested in the activity of hippocampal neurons called "place cells," each of which becomes associated with a particular place and fires rapid bursts of electrical pulses when the rat is in that location. By recording the activity of more than 100 neurons simultaneously in each rat, Foster and Wilson identified place cells corresponding to each part of the track. When a rat ran down the track, its place cells always fired in the same sequence as the animal passed through the neurons' favorite spots on the track.
But the place cells didn't stop chattering when the rat stopped moving. When a rat paused at the end of the track to eat a snack that popped up from a dispenser, its place cells fired in reverse order, as if replaying the rat's steps from the food back down the track, Foster and Williams report online 12 February in Nature. Foster notes that previous work has shown hippocampal place cell activity that suggests forward replay during sleep. Whereas the forward replay during sleep may help solidify memories of the day's activities, the reverse replay during wakefulness likely reinforces the behavior that led to the food reward, Foster speculates.
"It's a fascinating and exciting paper," says Robert Stickgold, a neuroscientist at nearby Harvard University. Very little is known about how animals learn the path to a goal, Stickgold says. "I think this is probably the first physiological model for ... a mechanism in the brain that lets you backtrack from the goal to the steps that led you there." To put the idea to the test, Stickgold says the researchers should determine whether the replay still happens when there's no food reward.