Latest Issue of Science News

5/2/15 Cover

News

Ten thousand neurons linked to behaviors in fly

New view of brain reveals functions of specific cells

By
3:43pm, March 27, 2014

THERE IS A SEASON  A larval fly will turn and then turn again when researchers activate two neurons in its brain (yellow, top).

Magazine issue: Vol. 185 No. 8, May 3, 2014
Sponsor Message

View the video

By scrutinizing the twists, turns, wiggles and squirms of 37,780 fruit fly larvae, neuroscientists have created an unprecedented view of how brain cells create behavior. The results, published March 27 in Science, draw direct connections between neurons and specific movements.

“Understanding how neural activity gives rise to behavior is the most important question in neuroscience,” says neuroscientist Kay Tye of MIT, who was not involved in the research. The new study provides a way for scientists to start answering that question, she says. “I think this is a really important approach that’s going to be very influential.”

Scientists led by Marta Zlatic of the Howard Hughes Medical Institute’s Janelia Farm Research Campus in Ashburn, Va., took advantage of an existing set of specially mutated flies. In each animal, small groups of neurons, usually between 2 and 15 cells, were engineered to respond to blue light. By activating handfuls of neurons with light and analyzing videos of the resulting behaviors, the researchers systematically explored most of the 10,000 neurons in Drosophila melanogaster larvae’s brain. 

“It’s like saturation bombing,” says neuroscientist George Augustine of the Center for Functional Connectomics in Seoul, South Korea. “They’re marching through pretty much all the neurons in the nervous system of this simple little creature and finding out what all of them do. That’s dramatic. That’s profound.”

Because much of the experimental work was automated, it took only several months to test 37,780 larvae. “The real challenge,” Zlatic says, “was dealing with the data.”

The team developed a mathematical approach to look for patterns of behavior elicited by activating small groups of neurons. This computational approach revealed behaviors that would have been impossible to identify otherwise, says neuroscientist Aravinthan Samuel of Harvard University. “If you just look at tons of videos, it’s very hard to see structure,” he says. “The human eye just isn’t able to handle that much data.”

But mathematics certainly can. By tracing larval movement in thousands of videos, the algorithm neatly described 29 distinct sequences of behaviors, the team reports. These behaviors included wiggles that help a larva escape from a threat, left turns followed by right turns, and backwards crawling.

The blue light failed to trigger some characteristic fly behaviors, such as a particular sort of predator-escaping roll, Zlatic says. That action is usually observed in moist environments, unlike the dry surface the larvae were tested on.

With the 29 behaviors in hand, scientists then used mathematics to look for neuron groups that seemed to bias the fly toward each behavior. The relationship between neuron group and behavior is not one to one, the team found. For example, activating a particular pair of neurons in the bottom part of the larval brain caused animals to turn three times. But the same behavior also resulted from activating a different pair of neurons, the team found. On average, each behavior could be elicited by 30 to 40 groups of neurons, Zlatic says.

And some neuron groups could elicit multiple behaviors across animals or sometimes even in a single animal.

Stimulating a single group of neurons in different animals occasionally resulted in different behaviors. That difference may be due to a number of things, Zlatic says: “It could be previous experience; it could be developmental differences; it could be somehow the personality of animals; different states that the animals find themselves in at the time of neuron activation.”

Stimulating the same neurons in one animal would occasionally result in different behaviors, the team found. The results mean that the neuron-to-behavior link isn’t black-and-white but rather probabilistic: Overall, certain neurons bias an animal toward a particular behavior.

The results have implications for more complex brains. “I don’t think anybody cares about how fly maggots move,” Augustine says. But the same principles in this simple organism are probably at work in more sophisticated brains, he says. 

Zlatic and her colleagues are currently studying a handful of these neurons and behaviors in more detail to understand how the neurons communicate with one another. The team ultimately plans on overlaying their results on a forthcoming map of all the physical connections between neurons in the Drosophila larval brain, offering scientists a more powerful way to study how all of the neurons work together to control behaviors. 



BODY AND BRAIN  A computational analysis of Drosophila larvae’s behavior resulted in videos like this and revealed 29 distinct behaviors, such as crawling and turning. The scientists were able to link each behavior to specific sets of nerve cells.
Credit: J. Vogelstein et al., Science/AAAS; adapted by Ashley Yeager

More from this issue of Science News