The anxious mouse: implications for preclinical research and animal welfare

Anxiety is an essential emotion that is highly conserved during evolution and is present in animals and humans. Although anxiety is a biological adaptive response, anxiety disorders in humans are common and affect about 10-17% of the world population. To gain more insight in the underlying neurobiological mechanisms of pathological anxiety, animal models are often used. However, commonly used animal models of anxiety are based on the single exposure to for example novelty and thus measure normal i.e. adaptive anxiety. A biological appropriate animal model for pathological anxiety would be of high use for both, translational research on human anxiety disorders and for the investigation of animal welfare aspects, since welfare might be compromised in animals suffering from non-adaptive anxiety. The main aim of the present studies was to evaluate behavioural habituation as an indicator of non-adaptive, i.e. pathological anxiety in mice. Two mouse strains BALB/c and 129P3 mice were characterized for their ability to habituate to an initially novel environment. Behavioural performance as well as stress hormones and the neural activity marker c-Fos were investigated. After behavioural testing for their ability to habituate, it was clear that initially highly anxious BALB/c mice were characterized by rapid habituation i.e. adaptive anxiety. In contrast, 129P3 mice showed a profound lack of habituation which indicated non-adaptive anxiety. This characteristic appeared to be gender-independent and sensitive to pre-exposure to mild stress. The lack of habituation in 129P3 mice was accompanied by a lower c-Fos expression in the prelimbic cortex, a brain area primarily involved in cognition. We hypothesized, that this might reflect a deficit in the integration of cognitive and emotional information, a process needed to adapt over time. Interestingly, the metabotropic glutamate 5 antagonist (mGlu5R) MPEP improved habituation and, in parallel, increased c-Fos expression in the prelimbic area. At the same time c-Fos expression decreased in brain areas known to regulate emotions, such as the amygdala, suggesting that increased glutamatergic neurotransmission might play a role in the behavioural profile of 129P3 mice. We concluded from strain-specific expression of the neuromodulator corticotropin-releasing factor (CRFR1), that CRFR1 mRNA (down) regulation in the prelimbic cortex is associated with habituation and that the non-adaptive phenotype of 129P3 mice is primarily a cognitive dysfunction, a theory which has also been suggested for human anxiety disorders. Thus 129P3 mice not only seem to be a promising translational model for human anxiety disorders. The fact that these mice show a possibly pathological phenotype suggests that their welfare may be compromised under standard housing conditions as applied in our studies. For 129P3 animals, strain-specific demands with respect to habituation processes, such as transfer to a new home cage, can be assumed. Characterizing the emotional phenotype, not only in 129P3 mice but also in other laboratory animals, would lead to a better understanding of potentially confounding factors on laboratory animal welfare