Delivering a new generation of translational animal models for depression research

Early animal models of depression focused on developing methods that could predict treatment efficacy and were validated based on pharmacological responses to known antidepressants. As our understanding of major depressive disorder (MDD) and the pharmacology of antidepressants progressed, so did the need for better animal models. This need was met with the development of new disease models, such as the chronic mild stress model, and behavioural readouts such as the sucrose preference test, which more closely aligned with risk factors and symptoms seen in patients. These approaches have supported huge advances in the understanding of how stress affects the brain and impacts on reward-related behaviours. However, there remain significant challenges when trying to model complex psychiatric symptoms and disorders in non-human animals. In this perspective article, a brief history of animal models of depression and associated readouts is discussed with specific reference to the important contributions from Paul Willner. The main discussion then focuses on translational validity and approaches that may support delivering this objective. This is illustrated with the example of the affective bias test and reward learning assays, which have been developed to recapitulate in animals the neuropsychological impairments observed in MDD and modulation by antidepressants.