Rationale and objectivesCognitive deficits, including an impaired ability to shift perceptual attentional set, belong to the core features of schizophrenia, are associated with prefrontal cortical dysfunctions, and may involve glutamate NMDA receptors. Although phencyclidine disturbs cognitive flexibility, little is known about the effects of ketamine and other NMDA antagonists that differ in receptor subunit selectivity, particularly in the mouse species.MethodsAt different times following the administration of ketamine, the NMDA NR2B-subtype specific antagonist Ro 25-6981, or the atypical antipsychotic sertindole, male C57Bl/6J mice were investigated in a modified version of attentional set-shifting task (ASST).ResultsSpecific extra-dimensional shift (EDS) deficit was observed in all control mice. As revealed by the increased number of trials, time and errors to reach criterion, ketamine at 10 or 20 mg/kg given 50 min prior to sessions, but not at 10 mg/kg given 3 or 24 h prior to sessions, further worsened the EDS performance. Sertindole (2.5 mg/kg) prevented ketamine-induced cognitive inflexibility, although it did not affect ASST performance when given alone. In contrast to ketamine, Ro 25-6981 at 10 but not 3 mg/kg, reduced the number of trials and errors to criterion, suggesting a facilitation of cognitive flexibility. Finally, as revealed by the number of trials and time to criterion measures, Ro 25-6981 (10 mg/kg) administration to ketamine (10 mg/kg)-pretreated mice inhibited ketamine-induced cognitive inflexibility.ConclusionThe present study provides an improved and reliable mouse ASST protocol and confirms and extends previous findings demonstrating that NR2B subunit-selective antagonists improve cognitive processes.