Abstract
Combining strains to generate mutant mice may obscure conclusions regarding the targeted gene. Specifically, cocaine may have reduced reinforcing effects in 129 substrains compared to the C57BL/6 strain, commonly used for ES cells and breeding, respectively. We tested the hypothesis that reinforcing effects of cocaine differ between the C57BL/6J strain and two substrains of 129, 129X1/SvJ and 129S6/SvEvTac. To assess and reduce performance differences, operant responding was established with liquid food as a reinforcer and evaluated under fixed and progressive ratio schedules. Dose–effect functions for intravenous cocaine self-administration were then determined under both schedules. Finally, reinforced and nonreinforced manipulanda were reversed to assess acquisition of self-administration using a previously nonreinforced response. Relative to C57BL/6J mice, 129X1/SvJ mice showed decreased reinforcing effects of low-magnitude food and cocaine reinforcers. Dose–effect functions for cocaine self-administration were comparable between C57BL/6J and 129S6/SvEvTac mice, despite delayed acquisition of operant behaviors and rightward shifts in the food concentration–effect functions in 129S6/SvEvTac mice. A high cocaine dose clearly served as a positive reinforcer in all three strains in a reversal procedure. Relative to C57BL/6J mice, the reinforcing effects of cocaine were diminished in 129X1/SvJ mice, but only for low cocaine doses, and a similar profile was observed with food reinforcement. 129S6/SvEvTac mice required more extensive operant training than C57BL/6J mice did, but after acquisition, reinforcing effects of cocaine were similar in the two strains. We suggest that comparable phenotypes observed in gene-targeting studies may result from genetic background, whereas more profound or qualitatively different phenotypes may be more confidently attributed to targeted mutations.