The phenotype of genetically modified animals is thought to result from an interaction of gene manipulation with the genetic background and environmental factors.

To test the behavioral and drug responses of Penk1^−/− mice on different genetic backgrounds.

Congenic C57BL/6J and DBA/2J mouse strains with a targeted deletion of the Penk1 gene were generated. Behavior and drug effects were tested in models of pain and anxiety.

Penk1^−/− mice showed exaggerated responses to painful or threatening environmental stimuli, but the expressivity of the mutant phenotype was strongly dependent on the behavioral paradigm and on the genetic background. For example, elevated levels of anxiety were readily detectable in C57BL/6J-Penk1^−/− mice in the light–dark and startle response tests, but not in the social interaction test. In contrast, we found elevated levels of anxiety in DBA/2J-Penk1^−/− mice only in the zero-maze and social interaction tests. In some cases, the idiosyncratic behavior masked the appearance of the knockout gene effect. The activity of the anxiogenic drug, m-chlorophenylpiperazine, but not the anxiolytic drug diazepam, was strain and genotype dependent. Mice with the Penk1 mutation on the DBA/2J, but not on other genetic backgrounds, showed an increased opioid-dependent stress-induced analgesia.

(1) The behavioral effects of the Penk1 gene deletion persists on different genetic backgrounds, but its detection sometimes requires the use of different behavioral paradigms. (2) The behavior of the background strain should be considered in the analysis of knockout mice to avoid floor and ceiling effects, which may mask the phenotype.