Animals frequently display complex behaviors, and the neural signals underlying those actions are an intricate and often unintelligible mix of electric pulses. Rui Costa believes that beneath the complexity lies a set of simple principles, like the laws that underlie the intricacy of the universe.

“The universe is very complex, but Einstein’s equations describe it in a very simple manner. The brain has more synapses than stars in the universe, so the potential paths of information are immense,” Costa says. His lofty goal: to reveal these principles and understand how the brain generates the behavior of organisms.

Growing up in a rural area of Portugal, Costa spent a lot of time watching the surrounding wildlife as they hunted for food and evaded predators. Those years infused him with a curiosity about how such complex behaviors were possible. That inquisitiveness he felt as a child now drives his research on how animals move, behave, and learn ways to interact with their environment.

Using a combination of techniques, Costa investigates the functions of different types of neurons in the brain. With arrays of electrodes implanted in the brains of mice, he’s able to record the activity of individual neurons as the mice move and learn. And by engineering mice so that certain neurons can be stimulated with light through an optical fiber—a technique known as optogenetics—he is gaining a better understanding of which neurons are important for certain activities and how dysfunctional ones might be corrected.

For an organism to act in its environment, it must be able not just to access inborn behaviors but also to learn new ways of doing things. This process of learning is what Costa wants to understand. He has been making inroads into how the brain learns since his Ph.D. research at the University of California, Los Angeles. There, under the guidance of his advisor Alcino Silva, he studied learning by observing what happens in learning disorders. He studied mouse models for neurofibromatosis-1 (NF1), a disorder characterized by benign, soft, neural tumors and learning disabilities. The research led him to find a mechanism that not only explained certain learning deficits but also suggested they could be reversed—overturning the established theory that the learning disabilities associated with NF1 were developmental and therefore permanent. It turned out that certain mutations in the Nf1 gene led to more inhibition of cells in the brain. Drugs reversing this inhibition are already on the market, approved for use in other diseases, and are now being tested in Nf1 patients.

As a postdoc in the lab of neuroscientist Miguel Nicolelis at Duke University, Costa delved deeper into the mechanisms of learning. He explored how changes in cortico–basal ganglia circuits, groups of cells in one area of the brain, play a role in learning and retaining skills. He also found that dopamine, a neurotransmitter that carries signals between brain cells, is essential to produce new patterns of brain activity and for an organism to carry out new physical actions. The discovery has implications for understanding some of the motor deficits of Parkinson’s disease.

Since setting up his own lab in 2006, first at the U.S. National Institutes of Health and then at the Champalimaud Center for the Unknown in Portugal, Costa has continued to research how animals learn new actions, showing that different cortico–basal ganglia circuits are responsible for different aspects of action learning. He found, for example, that changes in particular circuits are critical for animals to organize their physical actions into discrete sequences of movements. And that a separate group of cells signals the start and the stop of each of these sequences.

Already, Costa is beginning to link some of these findings with diseases in which the ability to learn or shape precise motor sequences is impaired, such as Parkinson’s and Huntington’s. “We think that we’re relatively closer to understanding why it is that if you have these diseases you can do things you learned a long time ago but have difficulty initiating something new, or terminating it, and can’t learn novel sequences,” Costa says.

It’s a start to uncovering the laws that dictate how the brain generates behavior and learns novel actions. “Now we’re at the stage where we can test to see how the brain works and what different cells are doing. It’s much more refreshing than to have arguments based on how well thought out your theory is,” Costa says. “I think it will completely change how we see behavior and neuroscience.”