Differential fear conditioning generates prefrontal neural ensembles of safety signals.

Fear discrimination is critical for survival, while fear generalization is effective for avoiding dangerous situations. Overgeneralized fear is a typical symptom of anxiety disorders, including generalized anxiety disorder and posttraumatic stress disorder (PTSD). Previous research demonstrated that fear discrimination learning is mediated by prefrontal mechanisms. While the prelimbic (PL) and infralimbic (IL) subdivisions of the medial prefrontal cortex (mPFC) are recognized for their excitatory and inhibitory effects on the fear circuit, respectively, the mechanisms driving fear discrimination are unidentified. To obtain insight into the mechanisms underlying context-specific fear discrimination, we investigated prefrontal neuronal ensembles representing distinct experiences associated with learning to disambiguate between dangerous and similar, but not identical, harmless stimuli. Here, we show distinct quantitative activation differences in response to conditioned and generalized fear experiences, as well as modulation of the neuronal ensembles associated with successful acquisition of context-safety contingencies. These findings suggest that prefrontal neuronal ensembles patterns code functional context-danger and context-safety relationships. The PL subdivision of the mPFC monitors context-danger associations to conditioned fear, whereas differential conditioning sparks additional ensembles associated with the inhibition of generalized fear in both the PL and IL subdivisions of the mPFC. Our data suggest that fear discrimination learning is associated with the modulation of prefrontal subpopulations in a subregion- and experience-specific fashion, and the learning of appropriate responses to conditioned and initially generalized fear experiences is driven by gradual updating and rebalancing of the prefrontal memory representations.