How does the brain adapt to uncertainty?
Does an unexpected event signal a lasting change, or is it merely a coincidence? The brain must constantly resolve this dilemma: revising its expectations quickly enough to adapt to changes, without mistaking every unexpected event for a lasting upheaval. A study led by researchers from INCIA and Institut de Génomique Fonctionnelle (Montpellier), published on 13 July in the journal PNAS, highlights a key brain circuit involved in this ability to adapt.
To understand how the brain distinguishes between these two situations, the scientists first observed rats faced with two levers associated with different probabilities of receiving a reward. When the reward probabilities changed, sometimes abruptly, the rats adjusted their learning rate according to the nature of the uncertainty. When rewards were delivered at random, they gradually revised their expectations; by contrast, when there was a genuine change in the rules, they accelerated their learning to adapt more quickly.
By measuring the release of noradrenaline in the orbitofrontal cortex in real time, the scientists also found that this neurotransmitter peaked precisely when the rules were reversed. Using modelling, the researchers predicted that if noradrenaline does indeed signal a change in the environment, then blocking its release should impair adaptation to a rule reversal. The scientists then tested this hypothesis. The result: rats deprived of this neurotransmitter struggled to adapt following a rule reversal, while continuing to perform well during the stable phase.
Combining behavioural observation, modelling, neurochemical measurement and causal manipulation, this study sheds light on the biological basis of behavioural flexibility. Although this mechanism has yet to be demonstrated in humans, disruptions to it could contribute to several neuropsychiatric disorders.
Article
Orbitofrontal noradrenaline supports adaptive learning-rate adjustment in probabilistic reversal learning. H. Plat, C. Chevallier, A. Piccin, A. R. Marchand, J. Naudé and É. Coutureau. PNAS, 13 July 2026.
DOI : 10.1073/pnas.2536535123
Contact
Last update 16/07/26