A Neural Circuit That Helps Flies Stay on Course

Navigating an environment is a key skill for many organisms. For years, scientists have wondered how the brain pieces together environmental inputs, compares them to an animal’s desired destination, and guides the organism to navigate successfully. But exploring these questions in the complex mammalian brain has proved challenging. 

In two studies published in Nature, researchers from Harvard University and the Rockefeller University peered into the tiny brains of fruit flies to uncover the neural circuit that makes those computations. Both research teams described how neurons called PFL3 integrate information from the insect’s current location with information about its target destination. While the Harvard University scientists reported how another population of brain cells, anti-goal neurons, helps the fly correct its path when it strayed far from its target, the Rockefeller University team described a set of neurons that encodes information about the insect’s goal.^1,2 By combining experimental and modeling approaches, these studies reveal a detailed neural circuit that is essential for Drosophila melanogaster’s navigation and provide insights into fundamental principles that may guide navigation in more complex brains. 

“These behaviors that they’re looking at, goal-directed steering, are universal to animals that navigate,” said Daniel Turner-Evans, a neuroscientist at the University of California, Santa Cruz, who was not involved in the studies. “It’s just beautiful to see how these behaviors unfold across these different layers and different neurons in the brain, and how you can create these really nice conceptual and quantitative models that really match the anatomy and the biology.”

In insects, the central complex, a highly conserved conglomerate of brain structures, is key to integrating a variety of sensory inputs and guiding locomotion during navigation. While some neurons within this center act as an internal compass that represents the fly’s heading direction relative to a landmark, others can directly influence an insect’s body steering.^3,4 By reconstructing the neuronal connections in the brains of different insects, scientists identified the specific lines of communication established between different cell types.^5,6 For instance, compass neurons communicate with PFL3 neurons.⁷ Two sets of PFL3 cells sit on each side of the fly’s brain, with neurons in each hemisphere sending projections to the steering center located on the opposite side. These previous findings suggested that PFL3 cells may allow an insect to directly compare its direction with its goal and adjust its trajectory to align the two.      

In the latest Nature papers, the researchers sought to experimentally test this idea while also investigating this navigation circuit more closely. Both teams recorded the neuronal activity of tethered flies as they walked in a floating sphere. “It basically acts as a treadmill. The fly can run, but because it is a ball, the fly can also turn,” explained Elena Westeinde, a graduate student in Rachel Wilson’s lab at Harvard University and coauthor of one of the studies. In both papers, the researchers placed the spherical treadmill in a virtual reality environment where they presented the fly with a bright bar, a stimulus known to attract the insect. The fly walked in a straight line for extended periods of time, which allowed the researchers to infer the insect’s heading direction each time they changed the position of the visual cue. 

In both studies, the researchers found that PFL3 neurons are involved in steering the fly’s body when it deviates from its intended path. When the fly went left of its goal, for instance, PFL3 cells located on the right side of the brain fired to make the insect correct its course. The opposite was true for the PFL3 neurons on the left side. 

Since almost all the inputs that PFL3 cells receive are shared with another group of neurons called PFL2, Westeinde and her colleagues investigated the role of these cells in fly’s navigation. They found that PFL2 neurons fired robustly when the fly was facing the opposite direction of its goal. When the team stimulated these cells, the insect increased its turning speed. “It’s like, ‘no, you’re really wrong! You just got to turn, you’ll just get back closer,’” explained Westeinde.

At the Rockefeller University, Gaby Maimon’s team decided to look upstream of PFL3 to find the cells that encode information about the goal. As the fly walked in the spherical treadmill, the researchers rotated the ball by some degrees and observed how specific neurons fired. They found that a group of cells called FC2 did not change its activity when the fly turned to a different direction. To confirm that these cells encoded a representation of the insect’s goal, Peter Mussells Pires, a postdoctoral researcher in Maimon’s lab, and his colleagues optogenetically stimulated the FC2 neurons. “[When] Peter stimulated these FC2 goal-communicating neurons, [he] could get the flies to walk in different directions,” explained Maimon. 

While Maimon’s work showed that the FC2 neurons read out the goal location, Turner-Evans remains curious about how the insect’s brain establishes that goal. Exploring this question is one of Maimon’s next steps. Turner-Evans also believes that exploring how these steering signals feed into the complex motor network and are then transformed into movements is another important next step. 

Even though the fly brain is much simpler than the human brain, its extensive characterization by researchers over the years has turned it into a powerful system for understanding how the brain processes information. “In the fly, we can uncover these sorts of fundamental principles that seem to be universal at some level,” said Turner-Evans. ”I wouldn’t be surprised if some of the conclusions from these studies as well, looking at how these different angular directions and vectors are used to steer an animal for navigation, proved to be true [in] mammals.”