Discrimination of motor and sensorimotor effects of phencyclidine and MK-801: Involvement of GluN2C-containing NMDA receptors in psychosis-like models.

Non-competitive NMDA receptor antagonists like PCP and MK-801 produce schizophrenia-like effects, but the specific receptor subunits and brain regions involved are unclear. Using GluN2C knockout mice, the authors found that deleting the GluN2C subunit reduced stereotyped behaviors such as circling, rearing, and ataxia signs induced by PCP and MK-801, indicating better motor coordination. However, other motor effects and sensorimotor gating deficits (pre-pulse inhibition) remained unchanged. PCP and MK-801 activated c-fos in thalamo-cortical networks but reduced it in the cerebellum, with differences between genotypes matching motor coordination changes. Resting-state fMRI showed enhanced cortico-thalamic-cerebellar connectivity in knockout mice that was less disrupted by MK-801. Thus, GluN2C-containing NMDA receptors in cerebellar circuits mediate some motor incoordination effects but not sensorimotor gating deficits.

Study at a glance

Abstract

Non-competitive NMDA receptor (NMDA-R) antagonists like ketamine, phencyclidine (PCP) and MK-801 are routinely used as pharmacological models of schizophrenia. However, the NMDA-R subtypes, neuronal types (e.g., GABA vs. glutamatergic neurons) and brain regions involved in psychotomimetic actions are not fully understood. PCP activates thalamo-cortical circuits after NMDA-R blockade in reticular thalamic GABAergic neurons. GluN2C subunits are densely expressed in thalamus and cerebellum. Therefore, we examined their involvement in the behavioral and functional effects elicited by PCP and MK-801 using GluN2C knockout (GluN2CKO) and wild-type mice, under the working hypothesis that psychotomimetic effects should be attenuated in mutant mice. PCP and MK-801 induced a disorganized and meandered hyperlocomotion in both genotypes. Interestingly, stereotyped behaviors like circling/rotation, rearings and ataxia signs were dramatically reduced in GluN2CKO mice, indicating a better motor coordination in absence of GluN2C subunits. In contrast, other motor or sensorimotor (pre-pulse inhibition of the startle response) aspects of the behavioral syndrome remained unaltered by GluN2C deletion. PCP and MK-801 evoked a general pattern of c-fos activation in mouse brain (including thalamo-cortical networks) but not in the cerebellum, where they markedly reduced c-fos expression, with significant genotype differences paralleling those in motor coordination. Finally, resting-state fMRI showed an enhanced cortico-thalamic-cerebellar connectivity in GluN2CKO mice, less affected by MK-801 than controls. Hence, the GluN2C subunit allows the dissection of the behavioral alterations induced by PCP and MK-801, showing that some motor effects (in particular, motor incoordination), but not deficits in sensorimotor gating, likely depend on GluN2C-containing NMDA-R blockade in cerebellar circuits.