How does the brain decide what memories are good and bad? A new mouse study offers clues

SAN DIEGO — Your first kiss. The blissful cool of ice cream. That time you smashed your finger while shutting the car door. Life is full of memories that we instinctively and instantly label as good or bad. In a new study, scientists took a key step toward unraveling how our brain assigns positive or negative emotions to our experiences.

Researchers at the Salk Institute working in mice found that a signaling molecule called neurotensin profoundly shaped whether the animals associated cues in their environment with positive or negative stimuli. Eliminating neurotensin in certain brain cells enhanced fear-based responses and blunted reward-based learning. And driving up levels of neurotensin had the opposite effect.

It’s too soon to say whether the findings, published on Wednesday in the journal Nature, will hold up in people. But if so, researchers are hopeful this line of research could lead to new ways of understanding — and maybe even treating — conditions such as anxiety, depression, and post-traumatic stress disorder.

“How do you put the fear aside for a moment and learn about something positive? And what switches you into that brain state? All of those things were mysterious before,” said Kay Tye, the study’s senior author. “It’s just completely reshaped the way that I think about the brain.”

The new findings build on years of work by her team. In 2015, they discovered two distinct circuits of neurons involved in positive and negative memory associations. These neurons live in the amygdala, an almond-shaped region buried deep within the brain that plays an outsized role in our behavior and how we process emotion. Scientists found that electrical signals among one group of neurons strengthened as lab mice learned to associate a certain tone with sweet sips of sugary water. A different circuit responded when the animals were taught that another sound preceded a quick and mild electrical zap.

You can think of these circuits as two separate train tracks, says Tye. But that study left a key question unanswered: What’s the switch that controls which track that train of thought ends up on?

That’s what researchers set out to learn in the recent study. And they already had a clue. When they compared neurons in the two circuits they’d previously identified, they found that these cells activated certain genes at different levels. One of them coded for a receptor that latches onto neurotensin, a known chemical messenger, or neurotransmitter, that regulates neuron activity.

To understand how neurotensin worked, researchers started tinkering with the neurotransmitter. They used the gene-editor CRISPR to knock out the gene that codes for neurotensin in a specific population of neurons in the amygdala. Doing so made it harder for mice to associate sound with sugar-laden water. But knocking out neurotensin enhanced negative, fear-fueled responses, and the animals learned to dart or freeze much more quickly when they heard a tone that preceded an electrical jolt.

Researchers then tried the opposite experiment — increasing neurotensin levels. To do so, they stimulated neurotensin-producing neurons engineered to activate in response to certain types of light, an approach known as optogenetics. That blunted fear-based responses and promoted reward-driven behavior. Mice with higher neurotensin levels ran more readily to a spout of sugar water each time they heard a tone that presaged the sweet reward.

“I was so surprised to see that,” Tye said. “I’m so used to it being a blurry mess. To see something that was the ‘switch operator’ and fit this model so perfectly was really exciting.”

Researchers now plan to study neurotensin’s role in conditions where positive reward circuits are overactive, such as addiction, as well as those in which negative circuits dominate, including depression and PTSD. Postdoctoral researcher Hao Li, who spearheaded many of the paper’s key experiments, plans to pursue these questions at Northwestern University, where he’ll be starting his own lab.

Li and Tye are hopeful that scientists could one day tweak neurotensin signals to help those living with these common conditions, perhaps with molecules that target the neurotensin receptor. About 6% of Americans will have PTSD at some point in their lives, and roughly 30% of U.S. adults will have an anxiety disorder. And while it’s too soon to know if therapeutic tweaks are possible, Vincent Costa, a neuroscientist at Oregon Health and Science University, says the new findings are encouraging.

“There’s a strong case to be made that the effects you’re seeing with neurotensin are large enough that they matter and potentially have consequences where, if you were to develop drugs around neurotensin receptors, you might expect to have therapeutic implications,” said Costa, who was not involved in the study.

The amygdala, often dubbed the “lizard brain,” is one of the most ancient and conserved brain regions across the animal kingdom. Tye says that makes it likely the neural circuits they’ve identified in mice also exist in people, though it’ll take additional work to confirm that. And she sees her work as part of a shift away from considering dopamine as the key neurotransmitter that encodes reward signals.

“We need to rewrite textbooks,” she said. “It’ll take a little while, but I fully expect this to be neuroscience gospel.”