Animals organize reward seeking around aversive events. An abundance of research shows that foot shock, as well as a shock-associated cue, can elicit freezing and suppress reward seeking. Yet, there is evidence that experience can flip the effect of foot shock to facilitate reward seeking. Here we examine cue suppression, foot shock suppression and foot shock facilitation of reward seeking in a single behavioural setting. Male Long Evans rats received fear discrimination consisting of danger, uncertainty and safety cues. Discrimination took place over a baseline of rewarded nose poking. With limited experience, all cues and foot shock strongly suppressed reward seeking. With continued experience, suppression became specific to shock-associated cues and foot shock facilitated reward seeking. Our results provide a means of assessing positive properties of foot shock, and may provide insight into maladaptive behavior around aversive events.

Animals organize reward seeking around aversive events. An abundance of research shows that foot shock, as well as a shock-associated cue, can elicit freezing and suppress reward seeking. Yet, there is evidence that experience can flip the effect of foot shock to facilitate reward seeking. Here we examine cue suppression, foot shock suppression and foot shock facilitation of reward seeking in a single behavioural setting. Male Long Evans rats received fear discrimination consisting of danger, uncertainty and safety cues. Discrimination took place over a baseline of rewarded nose poking. With limited experience, all cues and foot shock strongly suppressed reward seeking. With continued experience, suppression became specific to shock-associated cues and foot shock facilitated reward seeking. Our results provide a means of assessing positive properties of foot shock, and may provide insight into maladaptive behavior around aversive events.

When faced with potential threat we must estimate its probability, respond advantageously, and leverage experience to update future estimates. Threat estimates are the proposed domain of the forebrain, while behaviour is elicited by the brainstem. Yet, the brainstem is also a source of prediction error, a learning signal to acquire and update threat estimates. Neuropixels probes allowed us to record single-unit activity across a 21-region brainstem axis during probabilistic fear discrimination. Against a backdrop of widespread threat probability and behaviour signaling, a dorsally-based brainstem network rapidly signaled threat probability. Remapping of neuronal function following shock outcome gave rise to brainstem networks signaling prediction error on multiple times-cales. The results reveal construction of threat probability, behaviour, and prediction error along a single brainstem axis.