The mice that are part of the Allen Brain Observatory have a difficult task. A series of images flashes in front of their eyes, and the animals must signal when that image changes identity by licking a tiny waterspout dangling in front of them.

“The mouse is effectively playing a video game. It’s seeing images on a screen and making decisions, but instead of clicking a button on a video game controller, it’s licking a little spout,” said Marina Garrett, Ph.D., Assistant Investigator at the Allen Institute for Brain Science, a division of the Allen Institute. “The images are repeated in a sequence that will go something like flower, flower, flower, bear. That’s what they have to detect, the change from flower to bear, and they have to respond very quickly.”

Before the mice are put to the test, the research team trains them on the task. And before that training, the animals were genetically engineered so that certain neurons will glow under a fluorescent microscope when those cells kick into action.

Garrett and her colleagues want to understand which neurons are active, and when they’re active, as the animals learn to perceive the world around them. Recently, they found that a class of neurons, known as VIP interneurons, switch on when the animals see unfamiliar images. The researchers published a study describing the details of these novelty-detecting neurons in the journal eLife Wednesday.

The mice are trained on one series of images (say, the flower-and-bear sequence) and then are tested both with those familiar images and with a second set of new-to-them photos. They have to lick the spout right after the image changes, whether it’s a familiar or a new series, or they don’t get the reward (a drink of water). It’s an artificial set-up — mice didn’t evolve to recognize photos in a video-game-playing context, of course — but the researchers are trying to understand the basic principles of how the brain perceives our environment, and how our neurons’ activity changes during behavior and learning.

“Detection of novelty is critical for an animal’s survival,” said Shawn Olsen, Ph.D., Associate Investigator at the Allen Institute for Brain Science, who led the study along with Garrett. “The most important thing to be aware of in your environment is when something is unexpected.”

From novelty to learning

VIP interneurons are a type of inhibitory neuron, the general class of neurons whose job it is to suppress other neurons’ activity. They sit in the outermost shell of the brain known as the cortex. But VIP interneurons silence other inhibitory neurons, so the cumulative result is more brain activity, not less — the neural equivalent of a double negative.

It’s not completely clear what the end result of all this activity is in the brain, but the researchers hypothesize that VIP interneurons may be gatekeepers of novelty or general importance, pointing other neurons toward the most important thing to focus on.

Think about the brainpower difference between a task you do every day — say, brushing your teeth — and something you’re trying for the first time, like a new hobby. Your brain doesn’t exactly ignore the rote task, but it doesn’t take as much mental effort as something completely novel.

“We think this difference has to do with learning,” Garrett said. “When something’s familiar, you know what to expect. You can rely on your internal predictions and past knowledge. When something’s new, you need to pay more attention to it to learn more about it.”

To test that hypothesis, the researchers are now asking how the VIP interneurons’ activity changes over time. With repeated exposure, at some point the new images will become familiar. What happens in the brain at that tipping point? And do VIP interneurons play a similar role in other parts of the brain?

The current study is the first stage in a much larger Allen Institute project to study how different kinds of visual neurons in different sections of the brain react when mice are performing this task. The larger project should generate enough data to allow the researchers to address these questions more comprehensively and to understand how VIP neurons work together with other cell types, they said.

As for whether human VIP interneurons also recognize novelty, that’s impossible to test in the same way — researchers obviously can’t genetically engineer a human to study their neurons under a microscope. But there’s reason to think the general concept might apply to more than just mice seeing photos, Olsen said.

“I would not be surprised if you saw a similar phenomenon in the mouse auditory system or the part of the brain that senses touch,” he said. “To me, this result is less about mouse vision than it is about the core circuitry of the cortex, which we know is conserved between mice and humans.”