Mark Frye wants to unravel the neuron-by-neuron code for sight, smell, and motor control. That's why he sometimes glues flies to sticks before placing them in small flight arenas he's constructed in his lab at the University of California, Los Angeles (UCLA). He then turns on a light show, pumps in some appetizing smells, and watches.
"It's like virtual reality for insects," says Frye. "If they turn to the left, the scene on the viewscreen rotates to the right. For some reason, fruit flies really love to play these video games."
Frye began building insect arenas in graduate school as a way to overcome the limitations of the existing techniques in insect neurobiology—mainly implanting electrodes into the brains of immobile animals. The arenas, in contrast, generate quantitative data on more natural behaviors as the insects actively control their own sensory experience. "My postdoc adviser Michael Dickinson introduced me to a whole new level of technology, and I've cooked up some new variations. My lab looks less like a neurobiology lab and more like a high-tech flea circus," he says.
A fly's brain is about 200,000 times smaller than a human brain. And yet, the insects "can do amazing things, things we can't yet build robots to do," says Frye. "A fly can appear out of nowhere following the smell of your dinner, zoom around your kitchen evading your flyswatter, and touch down perfectly on the edge of your wine glass. And it can do all of this with a brain the size of a coarse grain of sand."
Figuring out exactly how that tiny biological computer controls all of those behaviors could help us understand how own brain works as well as help us build faster computers, more capable robots, even artificial eyes. Already, an engineer at the University of Maryland, Sean Humbert, has harnessed Frye's discoveries about how flies control their flight. He has built an autonomous hovercraft, which moves around and avoids crashing by deploying the same motion-control equations flies use—algorithms that Frye helped to work out with his experiments.
Frye spends much of his time trying to understand how the brain combines and interprets input from different senses, including sight, smell, and a form of mechanical feedback (a sense called proprioception). As a graduate student, Frye discovered that hawk moths need both sight and proprioception from their wings to steer correctly. "When they're blown off course, they sense that with their eyes, but to make a correction they also need to detect forces on their wings using special sensors," says Frye. "The hawk moths need both. Without the wing sensors, visual signals alone can't produce the proper steering maneuvers."
This finding highlights a general phenomenon all humans know by instinct—our senses are wired together. "But we have almost no idea how those circuits actually work," Frye says.
To explore that in his lab at UCLA, Frye built a fly arena that streams appetizing odors, like apple cider vinegar, toward a fly that is free to steer in a magnetic field. He then manipulates the visual background and studies how the shifting visual scene impacts the fly's ability to track the odor stream. "We want to know, how do different sensory modalities interact to allow a hungry fly to stay in an odor plume?" Already, Frye has discovered that scent and sight combine to push flies toward rewards much more aggressively than either sense alone. He is pursuing the underlying neural circuits that make these behavioral computations.
Frye also studies social behavior in insects. For this, he puts a tube that holds 100 walking flies into a hallway light arena. As the light show flashes, the flies all "stampede" to the same spot. Frye thinks a chemical alarm call might be partly responsible for the groupthink.
In any case, Frye is convinced the tiny fly brain holds enough mysteries to keep him busy for years. "Here's this teeny brain that can outperform any supercomputer on the planet," he says. "That's incredible."