Transfer to a naturalistic setting restructures fear responses in laboratory mice

Appropriate classification of a novel stimulus as threatening or benign depends on a repertoire of prior environmental experiences involving challenge, risk, and opportunity^1,2. Without this library, individuals may classify harmless stimuli as dangerous - a hallmark of generalized anxiety^1,2. In humans, insufficient exposure to uncertainty or manageable risks is associated with heightened anxiety and maladaptive fear generalization and is theorized to contribute to rising rates of anxiety in children^3,4,5. Although animals in natural environments accumulate a wide range of experiences that allow them to calibrate threat assessment, most behavioral studies of anxiety rely on laboratory animals housed in static, impoverished conditions. In this laboratory context, the widely used elevated plus maze (a measure of anxiety) induces a persistent fear response in mice after a single exposure. Here we show that transferring adult mice from the lab to a large field enclosure mimicking natural mouse environments was sufficient to block the development of this fear response and restore baseline levels of anxiety behavior. A canonical rodent anxiety phenotype is thus environmentally contingent and rapidly reversible, underscoring the risks of inferring general behavioral principles from impoverished housing conditions.