Neurons pull together as a brain learns

When a cartoon character gets an idea, you know it. A lightbulb goes on over Wile E. Coyote’s head, or a ding sounds as Goofy puts two and two together.

While the lightbulb and sound effects are the stuff of cartoons, scientists can, in a way, watch learning in action. In a new study, a learning task in rats was linked to increases in activity patterns in groups of brain cells. The results might help scientists pin down what learning looks like at the nerve cell level, and give us a clue about how memories are made.

Different areas of the brain communicate with each other, transferring information from one area to another for processing and interpretation. Brain cell meets brain cell at connections called synapses. But to transfer information between areas often takes more than one neuron firing a lonely signal.

It takes cortical oscillations — networks of brain cells sending electrical signals in concert — over and over again for a message to transmit from one brain area to another. Changes in electrical fields increase the probability that neurons in a population will fire. These cortical oscillations are like a large crowd chanting. Not all voices may be yelling at once, some people may be ahead or behind, some may even be whispering, but you still hear an overwhelming “USA! USA!”

Cortical oscillations can occur within a single brain area, or they can extend from one area to another. “The oscillation tells you what the other brain area is likely to ‘see’ when it gets that input,” explains Leslie Kay, a neuroscientist at the University of Chicago. Once the receiving area ‘sees’ the incoming oscillation, it may synchronize its own population firing, joining in the chant. “A synchronized pattern of oscillations in two separate brain regions serves to communicate between the two regions,” says Kei Igarashi, a neuroscientist at the Norwegian University of Science and Technology in Trondheim.

Cortical oscillations are found all over the brain. They play a role in everything from motor coordination to seizures to sleep. They are also thought to be associated with learning and memory.

To examine the role of these oscillations in learning, Igarashi and colleagues trained rats to perform a very simple task. If the rat placed its nose into a hole, it received an odor sample. The odor corresponded to a place where they could find a treat. So if they smelled, say, pine, a sugar pellet would be available in the left cup. If they smelled banana, it was on the right. Greedy rats will master this task pretty quickly; after three weeks, they went to the right cup 85 percent of the time.

Then the researchers implanted electrodes into the rats’ entorhinal cortex and the hippocampus, two areas important in learning and memory. The electrodes allowed the scientists to watch the electrical activity in those areas as the rats behaved. As the rats sampled the odor cues, the brain cell populations in the two brain areas showed matching electrical oscillations.

It turns out these oscillations develop as the rats learn their odor task, the researchers report June 5 in Nature. The authors performed another experiment in a different group of rats, placing the electrodes before the animals learned the task. The authors were then able to observe the oscillations coming together and becoming more coherent as the rat’s performance improved. When the rats made a mistake, the oscillations were off, less coherent than in perfectly performed trials. If the odors were changed, say, to rose and peach instead of pine and banana, the rat’s performance, and the oscillations, dropped. But as the rat re-learned the task, the oscillations came cycling back, this time associated with the new scents.

Igarashi and his colleagues were even able to localize the oscillation to particular neurons in the hippocampus and entorhinal cortex. Some of the cells in each area responded specifically to one odor and not the other. The selective response was absent when the animals made a mistake, and was associated with the development of the oscillations between the two brain areas.

The results, Igarashi says, suggest that the entorhinal cortex and hippocampus are more closely coordinated and more efficient communicators after learning. As the authors watch the oscillations grow, they are watching learning take place. And by watching established oscillations in trained animals, the scientists can observe a link to a memory in action.

But the oscillations aren’t necessarily “memories” themselves. While the firing patterns are associated with successful task performance, they may not be the cause or the result of the action. Oscillations could serve to make the connections between brain cells in the two areas stronger, or they could be a side effect of other activities. To determine whether the oscillations themselves are memories, scientists will need to find a way to artificially eliminate the brain cell crowd chorus without affecting other functions, a tricky prospect.

And scientists still don’t know how these large firing patterns begin, how neurons begin to chant together. “It’s kind of magical when it occurs,” Kay notes. “The oscillations are enormous and they just kind of appear. It’s something we’re all working on.”

But this study is a good first step. “The authors really nailed the phenomenon,” Kay says. “It supports the idea that oscillations are part of the learning process, and gets us closer to teasing out how exactly they are involved.” And the next time you visualize the learning process, forgo the lightbulbs or bells. Instead, think of growing, oscillating pulses of electricity. When they come together, you might just get it.