a Representative example photographs of dorsal surface of the coat in BTBD3 WT (left), HT (middle), and KO (right) mice. b Increased incidence of barbering behavior was identified in Btbd3 HT and KO mice in a chi-square analysis (n = 81 female/90 male WT, 150 female/165 male HT, 83 female/78 male KO mice). Across genotypes c, chronic fluoxetine reduced onset of barbering over time in a cohort of initially 100% non-barbers, whereas desipramine did not significantly alter barbering behavior at any time point in a Kaplan–Meier survival curve analysis (n = 15 female/13 male WT, 16 female/15 male HT, and 19 female/12 male KO for vehicle treatment, 19 female/13 male WT, 13 female/12 male HT, and 14 female/16 male KO for desipramine treatment, and 14 female/14 male WT, 16 female/17 male HT, and 19 female/15 male KO mice in the fluoxetine treatment group). Drops indicate time points when mice began to barber a cage mate, with the cumulative fraction of non-barbers indicated on the Y-axis. Within Btbd3 WT mice d, fluoxetine completely prevented onset of barbering behavior. Within Btbd3 HT mice e, fluoxetine-reduced barbering behavior. Within Btbd3 KO mice f, barbering was unaffected by fluoxetine treatment. g
Btbd3 KO mice showed reduced accuracy on the PLT (n = 13 WT, 16 HT, 17 KO mice, all male), which was driven by an increased frequency of lose-shift responses on the target port h and an increased frequency of win-stay responses on the non-target port i. j Breakpoint was unaffected by Btbd3 genotype in the PRBP task in an ANOVA. k
Btbd3 KO mice exhibited excessive wheel-running behavior during the dark cycle (n = 10 WT, 9 HT, and 10 KO mice of each sex). Results are expressed as mean values ± SEM, except for panels depicting categorical data, which does not have error. *p < .05 vs. WT or vehicle groups in a chi-square analysis b, log-rank analysis c–f, or an ANOVA with post hoc tests g–k. WT wild-type, HT heterozygous, KO knockout, PLT probabilistic learning task, PRBP progressive ratio breakpoint task