Summary: Researchers compare human habits—like a vinyl collector favoring a particular record store—to a widespread decision-making strategy called operant matching, in which animals distribute choices according to expected rewards. New experiments show fruit flies use the same strategy, and the work pinpoints synaptic mechanisms in the fly brain that implement it.
A recent study at HHMI’s Janelia Research Campus demonstrates that fruit flies make value-based decisions using reward expectations. By tracing this behavior to specific synapses in the mushroom body, the researchers provide a mechanistic link between theoretical models of operant matching and neural circuitry. The findings clarify how simple synaptic rules can produce sophisticated decision patterns and suggest broader relevance for understanding decision-making in larger brains and in conditions such as addiction.
Key facts:
- Operant matching is a common strategy across species in which choices are allocated in proportion to expected rewards.
- Janelia researchers showed fruit flies learn to choose between odor cues in proportion to the probability each cue is paired with reward.
- The behavior can be explained by synaptic plasticity in the mushroom body that incorporates reward expectation signals.
Source: HHMI
Everyday choices and foraging behavior
Imagine a record collector who usually finds the best albums at one shop, so he visits that shop most often, but occasionally goes elsewhere and sometimes finds a rare find. He is implicitly weighing the likelihood of reward from each option and allocating his time accordingly. Neuroscientists call this pattern matching behavior, and it appears across many animals that must choose where to forage or what cue to follow to find food.
Glenn Turner, a group leader at Janelia, and colleagues asked how brains implement this intuitive rule. Although operant matching has been documented in pigeons, rodents and humans, the neural mechanism behind it remained unproven. Theoretical work predicted that simple synaptic plasticity rules, modulated by reward expectation, could produce matching. The Janelia team set out to test that prediction directly in an experimentally tractable animal.
Led by graduate scholar Adithya Rajagopalan with contributions from James Fitzgerald, Ran Darshan and Karen Hibbard, the team developed a controlled foraging task for individual fruit flies. In a Y-shaped arena, a fly sampled one arm while two odors were presented from the other arms. Each odor was associated with a reward delivered probabilistically by activating sugar-sensing neurons: one odor produced reward more often (for example, 80% of trials) and the other less often (for example, 20%).
Over repeated trials, flies distributed their choices in proportion to the reward probabilities: they chose the richer odor around 80% of the time and the poorer odor around 20% of the time. This behavioral pattern matched the predictions of operant matching and demonstrated that even a small insect uses expectations about reward probability to guide choice.
From behavior to synapses
Because the fruit fly brain and its mushroom body circuitry are well mapped, the group could test where and how value is represented. They traced the matching behavior to synapses within the mushroom body, a structure central to learning and memory in flies. Building a neural model of the mushroom body, the team showed that biologically realistic synaptic plasticity rules—those that strengthen or weaken connections based on differences between expected and received reward—can generate matching behavior.
Crucially, the experiments showed that plasticity must incorporate reward expectation signals, not merely stimulus or reward presence. When the researchers bypassed the neural representation of expectation using optogenetic manipulations, matching behavior disappeared, supporting the idea that expectation signals are essential to how synapses adapt to produce value-based choice.
These results bridge theory and experiment by revealing synapse-level mechanisms that implement an elegant theoretical explanation for matching. The work demonstrates that a relatively simple plasticity rule, operating within a compact circuit, can yield complex economic-like decision-making.
“We found that flies are using expectation to assign value to their world,” Turner says. “It nicely connects back to theoretical work and explains a widespread phenomenon.” Fitzgerald adds that the study is an example of mechanistic cognitive neuroscience: using small animals to precisely tie behavior to neural changes.
Understanding how expectation-driven synaptic plasticity supports decision-making in flies provides a foundation for exploring similar principles in larger brains. Since decision-making processes break down in disorders like addiction, identifying basic rules in simpler systems may inform how those processes evolve and malfunction across species.
About this neuroscience research news
Author: Nanci Bompey
Source: HHMI
Contact: Nanci Bompey – HHMI
Image: The image is credited to Neuroscience News
Original research: Open access. “Reward expectations direct learning and drive operant matching in Drosophila” by James Fitzgerald et al., PNAS.
Abstract
Reward expectations direct learning and drive operant matching in Drosophila
Foraging animals must use decision-making strategies that adapt to fluctuating reward availability. Many species distribute choices in proportion to rewards obtained from each option, consistent with Herrnstein’s operant matching law. Theoretical work proposes that operant matching can arise naturally from simple synaptic plasticity rules applied within relevant neural circuits, but direct mapping of matching behavior onto synaptic mechanisms has been lacking.
In this study, operant matching was identified in Drosophila and shown to depend on synaptic plasticity in the mushroom body that incorporates reward expectation. Using a dynamic foraging paradigm in which individual flies learned odor–reward probabilities, and by constructing a model of mushroom body circuitry with biologically realistic plasticity rules, the authors demonstrated that plasticity incorporating reward expectations explains sequential choice behavior. Optogenetic manipulations that bypassed the representation of reward expectation abolished matching, indicating that the plasticity rule must specifically include reward expectation signals. These results reveal synapse-level mechanisms of operant matching and underscore the role of expectation in value-based learning.