University of California, San Francisco
Dr. Brainard is also a professor of physiology and psychiatry at the University of California, San Francisco.
Neural and Genetic Mechanisms of Motor Skill Learning
Michael Brainard was deep into the details of how gateways within cell membranes, called ion channels, open and close. But the Stanford University graduate student began to feel he was becoming too specialized in the wrong research area. "I felt increasingly divorced from what really interested me: how experience shapes behavior to give rise to our unique abilities and differences," he says. "So I pulled the plug and switched to a lab studying learning and memory."
Brainard soon found himself crawling through barns to collect barn owls. Owls are masters at locating sounds, he explains. "By studying an animal that's expert at a behavior, we hoped to generalize to similar behaviors in humans," he says.
In his PhD work with Eric Knudsen, Brainard was able to fool the owls and change their perception of the source of a sound by outfitting them with special goggles that bent the light; objects appeared to be where they weren't. The altered perception caused their brains to adapt, shifting the pattern of neurons reacting to the sounds. He also found that brains of young owls could readily adjust, but those of older birds could not. "This form of learning has a critical period in owls, just as for many forms of learning in humans," he says.
After getting his PhD, Brainard needed an animal he could breed and study in larger numbers to pursue bigger questions about how learning in early life shapes the function of the nervous system. He found the ideal subject in Bengalese finches. These birds, which sing complex songs of courtship, learn their signature melody by memorizing, then mimicking the song of their fathers. They warble the tune over and over until they get it right. The process is very similar to how human babies learn language. Infants refine their initially "babbling" vocalizations over years of trial-and-error adjustments. Birds do this much faster—in about a month. And what the birds learn during this critical period stays with them. Just as those who immigrate to the United States as a teenager might still have a strong accent from their native language at age 70, many adult birds can't change their songs.
How does this learning occur? Brainard's research at the University of California, San Francisco (UCSF), combining behavioral and neurophysiological techniques shows the importance of a cluster of interconnected regions deep in the brain, the basal ganglia. The basal ganglia provide corrective feedback when a bird hits a wrong note.
In an added twist, the basal ganglia also seem to send a signal that causes the bird to sing a slightly wrong note so that it can hear the error and learn to make a correction. A male bird singing alone makes subtle mistakes, but when he sings to a female, the song gets more precise.
"We hypothesize that the variability when singing alone reflects practice, while the precision when singing to a female reflects performance," Brainard says. Consistent with this, when signals from the basal ganglia are blocked, much of the variation in the practice singing goes away. The learning process may be very similar in humans, he suggests, in whom such skills as hitting a tennis ball or throwing darts improve through extensive practice with small variations. The brain gets feedback from the results of those variations and adjusts to do better.
When Brainard blocked the feedback signal from basal ganglia, as expected, the bird couldn't learn. But when he unblocked the signal, the bird immediately hit many more correct notes. The basal ganglia are "covertly" learning from subtle variations in the song that are not generated by that region, he says. When feedback is restored, the bird's song reflects that jump in learning. "The basal ganglia not only generate variation, but also monitor variations from elsewhere," Brainard concludes.
Of course, it's not that simple. In his lab at UCSF, Brainard uses computerized systems to precisely control the sounds his birds hear. Yet even with identical experiences, individual birds vary in their ability to learn. That suggests inherited differences in learning ability, Brainard says, an idea reinforced as he watches his 8-year-old fraternal twins, who, he says, act and learn very differently.
Brainard is now breeding distinct species of songbirds to create hybrids with greater genetic variation so he can probe the link between genes and learning. He plans to search for variations in genes that affect things like the patterns of song or the brain's dopamine reward circuits involved in reinforcement learning. The long-term goal, he says, is to "figure out why some individuals can learn a new accent while others can't, or why some learn well and others poorly. Ultimately, we believe such knowledge will contribute to our ability to repair and enhance brain function in cases of injury or disease.