The sinking platform test: a novel paradigm to measure persistence in animal models.

Persistence is the propensity to maintain goal-directed actions despite adversities. While this temperamental trait is crucial to mitigate depression risk, its neurobiological foundations remain elusive. Developing behavioral tasks to capture persistence in animal models is crucial for understanding its molecular underpinnings. Here, we introduce the Sinking Platform Test (SPT), a novel high-throughput paradigm to measure persistence. Mice were trained to exit a water-filled tank by ascending onto a platform above water level. Throughout the training, mice were also occasionally exposed to "failure trials," during which an operator would submerge a platform right after the mouse climbed onto it, requiring the mouse to reach and ascend a newly introduced platform. Following training, mice were subjected to a 5-min test exclusively consisting of failure trials. Male and female mice exhibited comparable persistence, measured by the number of climbed platforms during the test. Furthermore, this index was increased by chronic administration of fluoxetine or imipramine; conversely, it was reduced by acute and chronic haloperidol. Notably, six weeks of social isolation reduced SPT performance, and this effect was rescued by imipramine treatment over the last two weeks. A 4-week regimen of voluntary wheel running also improved persistence in socially isolated mice. Finally, comparing transcriptomic profiles of the prefrontal cortex of mice with high and low SPT performance revealed significant enrichment of immediate-early genes known to shape susceptibility for chronic stress. These findings highlight the potential of SPT as a promising method to uncover the biological mechanisms of persistence and evaluate novel interventions to enhance this response.

Arc regulates a second-guessing cognitive bias during naturalistic foraging through effects on discrete behavior modules.

Foraging in animals relies on innate decision-making heuristics that can result in suboptimal cognitive biases in some contexts. The mechanisms underlying these biases are not well understood, but likely involve strong genetic effects. To explore this, we studied fasted mice using a naturalistic foraging paradigm and discovered an innate cognitive bias called "second-guessing." This involves repeatedly investigating an empty former food patch instead of consuming available food, which hinders the mice from maximizing feeding benefits. The synaptic plasticity gene Arc is revealed to play a role in this bias, as Arc-deficient mice did not exhibit second-guessing and consumed more food. In addition, unsupervised machine learning decompositions of foraging identified specific behavior sequences, or "modules", that are affected by Arc. These findings highlight the genetic basis of cognitive biases in decision making, show links between behavior modules and cognitive bias, and provide insight into the ethological roles of Arc in naturalistic foraging.