How Fruit Flies Turn an Internal Compass into Steering Commands
Summary: A recent study reveals how fruit flies translate an internal sense of direction into corrective steering actions. Researchers identified three distinct neuron groups that link the brain’s compass region to steering circuits, enabling rapid course corrections when a fly is displaced from its intended path.
These results sharpen our understanding of navigation in simpler nervous systems and offer a foundation for future work on how internal goals and spatial maps are transformed into behavior in more complex animals, including mammals.
By examining the connections and activity between compass cells and locomotor circuits, the study uncovers basic principles about how internal cognitive states—like a remembered heading—are converted into concrete motor commands.
Key Facts:
- Three neuron populations (two that nudge turns to the left or right, and one that drives larger directional adjustments) translate head-direction signals into steering commands.
- The study maps functional connections between the brain’s head-direction “ring attractor” and downstream steering centers, clarifying how directional estimates guide movement.
- Insights from fruit flies point to general principles of neural computation for goal-directed behavior that may apply across species.
Source: Harvard
Our sense of direction acts like an internal compass, helping animals maintain or change course. Yet until now, the direct pathway from that internal heading signal to the motor actions animals make has been poorly understood.
A team led by researchers at Harvard Medical School addressed this gap by studying fruit flies during navigation. Their experiments, published on Feb. 7 in Nature, clarify how compass signals are relayed and transformed into steering commands.
The researchers studied flies that were displaced while walking in a fixed direction within a virtual environment. Using the fruit fly connectome as a guide, they built a computational model to identify neurons that might bridge head-direction circuits and locomotor centers. They then recorded neural activity and manipulated these cells while monitoring the animals’ navigation.
When the virtual world was rotated to put flies off course, the animals quickly corrected their heading. The team traced these corrections to three neuron groups: two that provide leftward or rightward steering nudges, and a third that produces stronger steering when the fly is far from its goal. Together, these groups compare the fly’s current heading with a stored goal heading and convert the discrepancy into appropriate motor commands.
“Until now, the link between an internal sense of direction and the animal’s actions was unclear,” said senior author Rachel Wilson, Joseph B. Martin Professor of Basic Research in Neurobiology at Harvard Medical School. “These results give a concrete account of how a cognitive variable—directional error—produces real-time, guided behavior.”
How the circuit performs corrections
Head-direction cells form a ring-like map of heading, making fruit flies a particularly tractable model. The new work shows that three downstream populations each receive a shifted copy of that heading signal, with the reference frames offset by roughly 120 degrees. Each population evaluates how well its shifted heading matches an internal goal heading through a nonlinear comparison. The combined outputs generate left or right turning commands and modulate steering vigor.
Specifically, one population (PFL3R) becomes active when the fly is left of its goal and drives rightward turns, while a complementary population (PFL3L) does the opposite. A third population (PFL2) increases steering strength when the directional error is large, adaptively balancing speed and accuracy during corrections.
The authors liken these populations to sentries posted at different bearings around a target, each signaling the adjustment needed to bring the animal back on course.
The study complements a second paper published the same day by another team, together offering a more complete picture of how head-direction signals are transformed into motor commands in Drosophila.
Implications and next steps
Beyond mapping connections, the results point to where and how navigational goals are represented in the brain and suggest mechanisms by which stored intentions can be held in a latent form and later acted upon. Wilson and colleagues plan to probe whether similar arrangements—distinct pathways for fine versus coarse adjustments—exist across other brain systems and species.
“By understanding a compact neural system in detail, we can form targeted hypotheses about analogous processes in larger brains,” Wilson said. “This work moves us toward general principles of how goals and spatial maps generate behavior.”
Authorship, funding, disclosures
Additional authors on the paper include Elena Westeinde, Emily Kellogg, Paul Dawson, Jenny Lu, Lydia Hamburg, Benjamin Midler, and Shaul Druckmann. The research was supported by the National Institutes of Health (U19NS104655).
About this neuroscience research news
Author: Dennis Nealon
Source: Harvard
Contact: Dennis Nealon – Harvard
Image: Image credit: Neuroscience News
Original Research: Open access. “Transforming a head direction signal into a goal-oriented steering command” by Rachel Wilson et al., published in Nature.
Abstract
Transforming a head direction signal into a goal-oriented steering command
To navigate, animals continuously estimate their heading and correct deviations from a goal. Directional estimates arise from ring attractor networks in the head-direction system, but it has been unclear how that sense of direction is converted into action.
Connectome analysis in Drosophila identified three cell populations—PFL3R, PFL3L and PFL2—that link head-direction neurons to locomotor circuits. Using imaging, electrophysiology and targeted stimulation during navigation, the study shows that each population receives a shifted version of the head-direction signal, with reference frames approximately 120° apart. Each population compares its shifted heading to a common goal vector via nonlinear computations, and their combined outputs generate steering commands.
PFL3R activity drives rightward turns when the fly is left of its goal; PFL3L drives leftward turns when the fly is right of its goal. PFL2 increases steering intensity as directional error grows, adjusting the balance between speed and accuracy. Together, these results explain how a spatial map in the brain can be integrated with an internal goal to produce body-centric motor commands.