Summary: Researchers have identified a population of neurons in the orbitofrontal cortex that become especially active when outcomes are uncertain. These “uncertainty neurons,” observed in both rats and humans, appear to help the brain balance flexibility and precision in decision-making and support learning in unpredictable environments. The discovery could inform treatments for conditions characterized by rigid thought patterns, including anxiety disorders, PTSD, and addiction.
When these neurons were temporarily inactivated in rats, their learning performance declined, indicating a critical role for this cell group in tracking changing conditions and adjusting behavior. The work sheds light on how the frontal cortex supports adaptive strategies when rewards are inconsistent or probabilistic.
Key Facts:
- Uncertainty Neurons: A subset of orbitofrontal cortex neurons shows peak activation during decisions with uncertain outcomes.
- Adaptive Role: These neurons promote flexible reward learning, helping animals and people update choices when contingencies shift.
- Clinical Potential: Targeting these cells might improve cognitive flexibility in disorders where adapting to uncertainty is impaired, such as anxiety, PTSD, and substance use disorders.
Source: UCLA
Newly identified neurons in the front of the brain act like an internal signal for uncertainty — prompting learning and flexibility when outcomes are unpredictable.
The UCLA team discovered these specialized neurons in rats and notes that humans possess comparable cells. The findings point to a neural mechanism that encourages exploration and adjustment when the environment does not offer perfectly reliable feedback.
“If you already know with certainty what will happen, you have little need to learn or change your behavior,” said Alicia Izquierdo, professor of behavioral neuroscience at UCLA and senior author on the study published in Nature Communications. “But real life is rarely certain. We found neurons in the orbital part of the frontal cortex that are tuned to uncertainty, and they appear essential for learning under those conditions.”
The orbitofrontal cortex (OFC), located above the eyes in both humans and rodents, processes emotions, taste and smell, and the rewarding aspects of choices. Prior research has implicated the OFC in flexible reward learning — the ability to discover and update which actions lead to positive outcomes when those relationships are not fixed.
In flexible reward learning, rewards reinforce correct choices, but the learner must infer which choices are correct because reward probabilities change or are noisy. This drives behavior: the desire for rewards pushes learners to persist through mistakes until they identify the pattern that produces the best outcomes.
To isolate the neurons involved, UCLA doctoral student and first author Juan Luis Romero-Sosa and colleagues focused on pyramidal cells in the OFC while rats performed touchscreen tasks to earn food rewards. They used a calcium indicator to visualize neuronal activity and chemogenetic tools — synthetic receptors activated by a designer drug — to switch those neurons off transiently. A miniature lens and camera implanted over the OFC recorded activity as the animals learned.
Tasks began with guaranteed rewards for correct responses, then progressed to harder conditions where reward probabilities became uncertain. At the most challenging stage, one option yielded a reward 70% of the time while the alternative rewarded only 30% of the time. Certain OFC cells became more active as uncertainty increased, signaling engagement specifically under ambiguous conditions.
When the researchers broadly inactivated the OFC, rats’ learning suffered: they failed to maintain preference for the higher-probability choice and showed fewer adaptive strategies, such as repeating choices that had just been rewarded (win-stay). In contrast, neurons in secondary motor cortex (M2) signaled choices more reliably under certainty, not uncertainty, highlighting functional differences across frontal regions.
“There’s a balance between becoming an expert at a particular strategy and staying adaptable,” Izquierdo explained. “If you want flexibility, you sacrifice some precision, and vice versa. Our results show a frontally distributed system where OFC neurons bias behavior toward adaptability when outcomes are unclear.”
Because rats and humans share evolutionary pressures to balance expertise with adaptability, the team suspects many species possess neurons tuned to uncertainty. If similar mechanisms are confirmed in humans, targeted interventions might be developed to improve cognitive flexibility in people who struggle to adapt — for example, patients with anxiety, PTSD, or certain dementias.
This research was supported by two National Institutes of Health grants that funded equipment, reagents, animal care, and trainee support for the project.
Key Questions Answered:
A: They identified neurons in the orbitofrontal cortex that become most active under outcome uncertainty and appear to guide flexible learning and decision-making.
A: These neurons help animals and people adapt to changing environments by tracking which choices are likely to produce the best outcomes even when results are unpredictable.
A: Difficulty adjusting to uncertainty is a hallmark of conditions like anxiety and PTSD. Understanding and potentially targeting these neurons could improve cognitive flexibility and resilience.
About this neuroscience and learning research news
Author: Holly Ober
Source: UCLA
Contact: Holly Ober – UCLA
Image: The image is credited to Neuroscience News
Original Research: Open access. “Neural coding of choice and outcome are modulated by uncertainty in orbitofrontal but not secondary motor cortex” by Alicia Izquierdo et al., published in Nature Communications.
Abstract
Neural coding of choice and outcome are modulated by uncertainty in orbitofrontal but not secondary motor cortex
Orbitofrontal cortex (OFC) and secondary motor cortex (M2) both contribute to flexible reward learning, but how they are differentially engaged under varying levels of uncertainty has been unclear. The authors recorded calcium signals from individual neurons in rat OFC and M2 during learning with reward probabilities that became progressively more uncertain.
Choice predictions decoded from M2 neurons were accurate across all certainty levels, while OFC neurons provided more accurate decoding under greater uncertainty. In OFC, decoding accuracy for choice and outcome was associated with behavioral strategies such as Win-Stay and Lose-Shift; these associations were not observed in M2. Chemogenetic inhibition of OFC impaired learning across schedules, whereas M2 supported learning primarily when reward contingencies were most certain.
These results indicate that OFC neurons preferentially encode choices and outcomes that promote adaptive strategies under uncertainty, revealing functional heterogeneity within frontal cortex that supports flexible learning.