Comparison of the predictive validity of the mirror chamber and elevated plus maze tests in mice

The mirror chamber (MC) is a putative test of anxiety-like behavior in mice, and is increasingly popular. Nonetheless, it is unclear whether the observed behaviors rely on the presence of mirrored panels. If so, it is unclear whether the behaviors are sensitive to clinically effective anxiolytics, and how the test compares to the elevated plus maze (EPM) in terms of predictive validity. The present studies assessed anxiety-like behaviors in different mouse strains in the MC using mirrored and non-mirrored panels, under variable lighting conditions. We also assessed the pharmacological validity of the MC and EPM tests, and the locomotor properties of active test compounds. Seven mouse strains exhibited different levels of anxiety-like behaviors in the MC, and differential sensitivity to panel and light conditions. DBA/2J mice appeared most sensitive to the mirrored, versus black or white, panels and were therefore used in pharmacological MC studies. The mGlu5 receptor antagonist MPEP significantly decreased anxiety-like behaviors, similar to an intermediate dose of the benzodiazepine diazepam. The benzodiazepines chlordiazepoxide and alprazolam and the 5HT(1A) partial agonist buspirone had no effects on anxiety-like behaviors in the MC. None of the MC effects of active test compounds were attributable to non-specific/locomotor effects. The antidepressants fluoxetine and venlafaxine increased anxiety-like behaviors in the MC. By contrast, the anxiolytic-like effects of chlordiazepoxide, diazepam and MPEP were revealed in the EPM in C57Bl6/J mice. In conclusion, the EPM test exhibits superior predictive validity compared to the MC test, despite the sensitivity of the MC to mouse strain differences.