Effects of Cohousing Mice and Rats on Stress Levels, and the Attractiveness of Dyadic Social Interaction in C57BL/6J and CD1 Mice as Well as Sprague Dawley Rats.

Rats, including those of the Sprague Dawley strain, may kill mice. Because of this muricidal behavior, it is standard practice in many research animal housing facilities to separate mice from rats (i.e., the predators) to minimize stress for the mice. We tested the effect of cohousing on the stress levels of mice from either the C57BL/6J (BL6) or the CD1 strain and Sprague Dawley rats (SD rat) by quantifying their fecal corticosterone and metabolites (FCM) concentration. We also investigated cohousing impacts a behavioral assay, i.e., conditioned place preference for intragenus (i.e., mouse-mouse or rat-rat) dyadic social interaction (DSI CPP) that was shown be sensitive to social factors, especially to handling by humans. We found that the two delivery batches of BL6 mice or SD rats, respectively, had different stress levels at delivery that were statistically significant for the BL6 mice. Even so, the BL6 mice cohoused with rats had significantly increased FCM concentrations, indicative of higher stress levels, as compared to (1) BL6 mice housed alone or (2) BL6 mice at delivery. In contrast to their elevated stress levels, the attractiveness of contextual cues associated with mouse-mouse social interaction (DSI CPP) even increased in rat-cohoused BL6 mice, albeit non-significantly. Thus, cohousing BL6 mice and rats did not impair a behavioral assay in BL6 mice that was proven to be sensitive to handling stress by humans in our laboratory. SD rats cohoused with BL6- or CD1 mice, and CD1 mice cohoused with SD rats, showed DSI CPP that was not different from our previously published data on SD rats and BL6 mice of the Jackson- or NIH substrain obtained in the absence of cohousing. CD1 mice cohoused with rats did not show an increased FCM concentration compared to delivery. Our findings suggest that the effect of cohousing rats and mice under the conditions described above on their stress levels as opposed to their behavior might be less clearcut than generally assumed and might be overriden by conditions that cannot be controlled, i.e., different deliveries. Our findings can help to use research animal housing resources, which are usually limited, more efficiently.