Mice that wouldn’t behave reshaped Costa’s learning quest

how animals – Kauê M. Costa’s path through neuroscience pivoted on a simple, frustrating glitch: amphetamines calmed some of his genetically engineered mice, but not others. The difference—sex-based learning and dopamine circuits—pushed the University of Alabama at Birmingh

On a bench in Germany, Kauê M. Costa was watching mice instead of listening to the theory he’d gone looking for. He was a Ph.D. student trying to understand how amphetamines cause hyperactivity, and the animals in front of him weren’t matching the clean story he expected.

The mice were genetically engineered to help researchers study dopamine in the brain. Yet they were unusually hyperactive. Then the amphetamines came in as a test of the mechanism—calming the females but leaving the males hyperactive. What should have been a straightforward biological signal turned into a stubborn. confusing divergence: the mice were learning differently based on sex. and that mismatch started scrambling his results.

Costa, 36, would later take that disruption and turn it into a focus. Born in Amazonian Brazil. he is now an assistant professor at the University of Alabama at Birmingham. studying how animals learn. In his work, the question isn’t only what dopamine does when learning happens. It’s what principles decide what gets learned in the first place—what information the brain treats as worth storing. and what can be thrown out.

To chase that, Costa is applying newer tools to neuroscience’s classic behavioral tests. His lab looks at food cues, mazes, and lever pressing. Across those tasks, he is trying to understand how the brain chooses which information matters, and how that selection shapes what animals retain.

“ We are constantly bombarded with, let’s say, a practically infinite amount of information,” Costa says. “What are the principles that determine what we learn and how we learn it?”

His experiments have also shaped a broader picture of learning that goes beyond the immediate reward or punishment. His work suggests that dopamine is involved not just in helping us learn what is good and what is bad. Dopamine also plays a role in a more complex type of learning that helps infer what’s next based on past events.

The arc of his career has kept a similar theme: the most revealing answers often arrive when a result doesn’t cooperate. Costa could have left the U.S. after finishing his postdoctoral fellowship. but he stayed—despite “federal funding chaos. ” he said the experience of doing research in the U.S. is still “incomparable.” He teaches his students to follow their interests and their own minds. a philosophy he now ties back to the way his own path—through unexpected mouse behavior—pulled him toward questions that neuroscience had been trying to answer for years.

For Costa, the adventure hasn’t stopped at the lab bench. Even as he once thought he would explore the Amazon. he says the brain is more complex—and thus more appealing. uncharted territory. His work sits inside an editorially independent effort called “The Young American Scientists,” produced with financial support from Regeneron.

Kauê M. Costa neuroscience dopamine amphetamines hyperactivity learning mice University of Alabama at Birmingham behavioral tests food cues mazes lever pressing dopamine and learning