New research sheds light on how the brain of a fruit fly constructs a precise internal compass to navigate its environment using just a handful of neurons, enhancing scientists’ comprehension of the capabilities of small neural networks.
Neuroscientists faced a significant challenge.
For many years, there has been a prevailing theory about how an animal’s brain keeps track of its spatial position without external signals—similar to how we can identify our location with closed eyes.
This theory, which drew on brain recording studies of rodents, posited that neuronal networks known as ring attractor networks maintain an internal compass to monitor one’s position in the environment. It was believed that a precise internal compass necessitated a sizable network with numerous neurons, while a smaller network would lead to inaccuracies and drift.
However, researchers identified an internal compass in the minuscule fruit fly.
“The fly’s compass is astonishingly accurate, but it relies on a remarkably small network, contrary to earlier assumptions,” says Ann Hermundstad, who leads a group at Janelia. “This indicated a clear gap in our understanding of brain compasses.”
Now, research spearheaded by Marcella Noorman, a postdoctoral researcher in Hermundstad’s lab at HHMI’s Janelia Research Campus, clarifies this mystery. The new findings reveal how it’s feasible to create a highly accurate internal compass within a small network, like that of the fruit fly.
This research reshapes the perspective of neuroscientists regarding various brain functions, including memory, navigation, and decision-making.
“This truly broadens our understanding of what small neural networks can achieve,” Noorman notes. “They are capable of performing more intricate computations than previously recognized.”
Creating a ring attractor
Upon joining Janelia in 2019, Noorman tackled the issue that had perplexed Hermundstad and others: How does the small brain of the fruit fly generate a reliable internal compass?
Noorman initially aimed to demonstrate that creating a ring attractor with a small number of neurons was impossible; she believed that additional components—such as different cell types and more complex biophysical characteristics of the cells—were necessary for success. To investigate this, she removed all the “extra components” from existing models to determine if a ring attractor could still be formed with what remained. She anticipated this would prove impossible.
However, Noorman found it difficult to confirm her hypothesis. That’s when she opted for a different strategy.
“I had to shift my thinking and consider that maybe a ring attractor *can* be generated with a small network,” she explains, “and then identify the specific requirements for that network to function.”
By altering her premise, Noorman uncovered that it is indeed possible to create a ring attractor with just four neurons, provided that their connections are precisely adjusted. She collaborated with fellow researchers at Janelia to experimentally validate this new theory, discovering physiological evidence that supports the notion that the fly brain can produce a ring attractor.
“Smaller networks and brains can perform complex computations far beyond what we previously thought possible,” Noorman asserts. “However, achieving this requires much more precise connections among neurons compared to larger brains where many neurons can accomplish the same computation.”
“Thus, there’s a trade-off between the number of neurons used for a computation and the precision of their connections,” she explains.
Looking ahead, the researchers plan to investigate whether the “extra components” could enhance the robustness of the ring attractor network and if the foundational computation could serve as a fundamental element for more intricate calculations in larger networks with various variables. Further experiments could also aid researchers in understanding how neuronal connections in the network are modified and how sensory signals might influence the network’s representation of spatial orientation.
For Noorman, a mathematician turned neuroscientist, the journey of translating biology into a solvable mathematical problem has been both challenging and enjoyable.
“The head direction system of the fly has been my first encounter with neural activity, making it a fascinating experience to figure out and comprehend how it operates,” she shares.
