The brain’s so-called “reward” system plays a large role in human behavior. Certain activities can trigger the release of dopamine, one of the chemical messengers involved in this system, which produces positive, reinforcing feedback within the brain. People learn to repeat those dopamine-producing activities, though this process can be problematic, as in the cases of substance abuse or gambling addiction.

Now, a new study from Penn team of neuroscientists and neurosurgeons has shown that electrically stimulating a region of the brain that features dopamine-containing neurons can influence this learning process. With the help of Parkinson’s disease patients undergoing neurosurgery, the researchers were able to investigate the workings of a deep brain structure known as the substantia nigra. The team found that stimulating this region may alter learning by biasing individuals to repeat physical actions that resulted in reward.

The study was conducted by lead author Ashwin Ramayya, a graduate student in the Department of Neuroscience in the Perelman School of Medicine, and senior authors Gordon Baltuch, a professor of neurosurgery at Penn Medicine, and Michael Kahana, a professor of psychology in Penn Arts & Sciences.

“Stimulating the substantia nigra as participants received a reward led them to repeat the action that preceded the reward,” says Kahana, “suggesting that this brain region plays an important role in modulating action-based associative learning.”

As with some of Kahana’s previous research, this study recruited participants who were already undergoing brain surgery: deep brain stimulation treatment for Parkinson’s disease. During a portion of the procedure where the patients are awake, they played a computer game during which they had to choose between pairs of objects that carried different reward rates, like choosing between rigged slot machines in a casino.

The objects were displayed on a computer screen and participants made selections by pressing buttons on hand-held controllers. When they received a reward, they were shown a green screen and heard a sound of a cash register, as they might in a casino.

Participants were not told which objects were more likely to yield a reward, but that their task was to figure out which ones were “good” options based on trial and error. 

When stimulation was provided in the substantia nigra following reward, participants tended to again press the button that resulted in a reward. This was the case even when the good object was no longer associated with that button press, resulting in poorer performance on the game when stimulation was given, compared to when stimulation was not given.

“While we’ve suspected, based on previous studies in animal models, that these dopaminergic neurons in the substainia nigra play an important role in reward learning, this is the first study to demonstrate in humans that electrical stimulation near these neurons can modify the learning process,” says Baltuch. “This result also has possible clinical implications through modulating pathological reward-based learning, for conditions such as substance abuse or problem gambling, or enhancing the rehabilitation process in patients with neurological deficits.”