Essential Role of NMDA Receptor Channel ε4 Subunit (GluN2D) in the Effects of Phencyclidine, but Not Methamphetamine

Phencyclidine (PCP), a noncompetitive N-methyl-D-aspartate (NMDA) receptor antagonist, increases locomotor activity in rodents and causes schizophrenia-like symptoms in humans. Although activation of the dopamine (DA) pathway is hypothesized to mediate these effects of PCP, the precise mechanisms by which PCP induces its effects remain to be elucidated. The present study investigated the effect of PCP on extracellular levels of DA (DAex) in the striatum and prefrontal cortex (PFC) using in vivo microdialysis in mice lacking the NMDA receptor channel ε1 or ε4 subunit (GluRε1 [GluN2A] or GluRε4 [GluN2D]) and locomotor activity. PCP significantly increased DAex in wildtype and GluRε1 knockout mice, but not in GluRε4 knockout mice, in the striatum and PFC. Acute and repeated administration of PCP did not increase locomotor activity in GluRε4 knockout mice. The present results suggest that PCP enhances dopaminergic transmission and increases locomotor activity by acting at GluRε4.

The NMDA receptor channel subunit family is composed of seven subunits-GluRf (GluN1), GluRe1-4 (GluN2A-D), and GluRx1, 2 (GluN3A, B)-which are all products of separate genes [25]. In the rodent and human brains, GluRe1 and GluRe2 are predominant subunits expressed in the forebrain. GluRe3 is expressed largely in cerebellar granule cells and selectively in several other brain regions. GluRe4 is expressed in the diencephalon and midbrain and is more prominent during early development [26]. Highly active NMDA receptor channels are produced when the GluRf subunit is expressed together with one of the four GluRe subunits in Xenopus oocytes and mammalian cells [27][28][29][30]. Four GluRe subunits are major determinants of the functional properties of NMDA receptor channels [31]. Noncompetitive NMDA receptor antagonists (i.e., PCP, ketamine, and SKF-10,047) block the four GluRe/GluRf channels to similar extents in Xenopus oocytes [32]. Gene-targeting techniques provide an efficient method for clarifying the distinct functions of these NMDA receptor channel subunits. GluRe1 knockout mice display increased locomotor activity, whereas GluRe4 knockout mice exhibit reduced locomotor activity in a novel environment [33][34][35][36]. GluRe3 knockout mice show few apparent deficits [37][38][39]. Investigating the physiological functions of GluRf or GluRe2 knockout mice, in contrast, is nearly impossible because these two mutants die shortly after birth [40][41][42].
To clarify the contributions of NMDA receptor channel subunits in the PCP-induced increases in extracellular levels of dopamine (DA ex ) and locomotor responses, we investigated the effects of METH and PCP on DA ex in the striatum and prefrontal cortex (PFC) using in vivo microdialysis and measuring locomotor activity in GluRe1 knockout (GluRe1 2/2 ) and GluRe4 knockout (GluRe4 2/2 ) mice.

Discussion
The present study showed that PCP-induced increases in DA ex in the striatum and PFC and locomotor activity were absent in GluRe4 2/2 , but present in GluRe1 2/2 , mice, indicating that GluRe4 plays an important role in PCP-increased DA ex and locomotor activity. Phencyclidine exerts psychotomimetic effects, whereas another NMDA receptor antagonist, MK-801, exerts no clear psychotomimetic effects in humans [43]. Interestingly, whereas MK-801 suppresses GluRe3/GluRf1 and GluRe4/ GluRf1 channels more weakly than GluRe1/GluRf1 and GluRe2/GluRf1 channels, PCP blocks the four GluRe/GluRf channels to similar extents in Xenopus oocytes [32]. The absence of psychotomimetic effects of MK-801 may be attributable to its weak ability of blocking the GluRe4/GluRf1 channel.
Systemic administration of PCP reportedly increases DA ex in the striatum and PFC [44][45][46][47][48][49]. Similarly, PCP (3 mg/kg) increased DA ex in wildtype and GluRe1 2/2 mice in the present study. However, PCP failed to increase DA ex in the striatum and PFC in GluRe4 2/2 mice. Phencyclidine is known to be a DA reuptake blocker and a noncompetitive NMDA antagonist [9][10][11]. It inhibits DA uptake by binding to the DAT at doses approximately 10-fold greater than those at which it binds to NMDA receptor channels [1]. Phencyclidine at the low dose used in the present study appears to have few effects on the DAT. Furthermore, no PCP-induced increases in DA ex in GluRe4 2/2 mice that possess an intact DAT gene indicates that PCP increases DA ex not via DAT inhibition but via blockade of NMDA receptor channels. The present results support the hypothesis that GluRe4 is an important determinant of increased DA ex induced by PCP. Acute administration of METH increased DA ex in the striatum and PFC in wildtype, GluRe1 2/2 , and GluRe4 2/2 mice. No differences in DA ex increases were found between genotypes. The similar DA ex increases among these mice in response to acute METH challenge suggest that increased DA ex occurs independently of GluRe1 2/2 and GluRe4 2/2 .
Locomotor activity in a novel environment is reportedly high in GluRe1 2/2 mice [34,36] and low in GluRe4 2/2 mice [33,35]. Consistent with these findings, increased locomotor activity in GluRe1 2/2 mice and reduced locomotor activity in GluRe4 2/2 mice were observed in the present study. GluRe1 2/2 mice did not habituate during the 180 min period compared with wildtype mice. Interestingly, acute METH administration decreased locomotor activity in GluRe1 2/2 mice. Hyperactivity and a paradoxical response to METH suggest that GluRe1 2/2 mice may be an animal model of attention-deficit/hyperactivity disorder.
Psychostimulants, such as METH and PCP, increase locomotor activity [2,3,12,13]. In GluRe4 2/2 mice, acute METH administration increased locomotor activity, but PCP did not. Acute PCP administration increased locomotor activity in wildtype and GluRe1 2/2 mice, but not in GluRe4 2/2 mice. The absence of locomotor-stimulating effects of PCP in GluRe4 2/2 mice indicates that locomotor responses to PCP require the GluRe4 subunit.
Repeated administration of PCP produces sensitization to its locomotor-stimulating effects in wildtype mice. In GluRe4 2/2 mice, locomotor activity did not increase after repeated PCP treatment. Acute PCP did not increase locomotor activity, and repeated PCP did not produce sensitization to the locomotorstimulating effects of PCP in GluRe4 2/2 mice. The GluRe4 subunit appears to be necessary for behavioral sensitization to occur in response to repeated PCP administration. A previous study demonstrated that acute PCP treatment increased locomotor activity in wildtype and GluRe1 2/2 mice. Chronic PCP treatment at a low dose (3 mg/kg/day) for 7 days produced sensitization to the locomotor-stimulating effects of PCP in wildtype mice, but not  in GluRe1 2/2 mice [50]. The present study confirmed that repeated PCP administration (3 mg/kg/day) did not produce sensitization during Session 8 in GluRe1 2/2 mice. Repeated METH administration produced behavioral sensitization in wildtype, GluRe1 2/2 , and GluRe4 2/2 mice. The development of sensitization in GluRe1 2/2 and GluRe4 2/2 mice was delayed compared with wildtype mice. The noncompetitive NMDA receptor antagonist MK-801 has been shown to block the development of behavioral sensitization to AMPH and METH [51][52][53][54]. Molecular and cellular adaptive changes during chronic drug exposure are hypothesized to lead to the development of sensitization. Our findings support the hypothesis that adaptive changes through NMDA receptor channels play a role in the development of locomotor sensitization to METH.
Schizophrenia is a disease that has been hypothesized to be associated with hyperfunction of the dopaminergic neuronal system and dysfunction of glutamatergic transmission [55,56]. Administration of PCP to normal humans induces symptoms similar to those of schizophrenia [57]. This finding has been replicated over the years, and PCP has been shown to exacerbate the primary symptoms of schizophrenic patients [56]. Phencyclidine-treated animals have been used as an animal model of schizophrenia, and the amelioration of hyperlocomotion in these animals has been used as a screening test to assess the efficacy of antipsychotic drugs [58,59]. GluRe4 immunoreactivity and protein expression increase in the frontal cortex following repeated PCP treatment, whereas GluRe1 immunoreactivity and protein expression are not altered in rats [60]. Furthermore, polymorphisms of several genes known to interact with NMDA receptor channels are related to altered risk for schizophrenia, and psychotic patients display changes in the levels of mRNA encoding NMDA receptors [61]. Interestingly, Makino et al. reported that the GluRe4 gene locus is a possible genomic region that contributes to schizophrenia susceptibility in a Japanese population [62]. In the present study, we first demonstrated that deletion of GluRe4 abolished PCP-induced hyperlocomotion and potentiated the increases in DA ex in mice. Our data and previous findings suggest that GluRe4 might be a potential target for antipsychotic drug development.
Although NMDA receptor channels are highly expressed in adult brains, adult GluRe4 expression is very limited [26]. GluRe4 is expressed in the substantia nigra compacta (SNc), subthalamic nucleus, globus pallidus, and ventral pallidum in adult rats [63]. Jones and Gibb reported that functional GluRe2 and GluRe4 subunits form somatic NMDA receptors, possibly as triheteromeric receptors, whereas no somatic GluRe1 subunits are present in SNc dopaminergic neurons in rats aged postnatal day 14 [64]. A small subset of NMDA receptor channels (i.e., channels containing GluRe4) may be implicated in the effects of PCP on DA ex and locomotor activity. This possibility is consistent with the lack of psychotic effects of ifenprodil, a selective blocker of NMDA receptor channels containing GluRe2, which is highly expressed in adult brains. Additionally, GluRe4 is highly expressed in the brain during development [26], suggesting that GluRe4 knockout during the developmental stage may alter neuronal function in the adult brain. Although the expression of the genes related to dopaminergic signaling pathways are not altered in GluRe4 2/2 mice during adulthood (see Table S1), other developmental changes may alter the effects of PCP in GluRe4 2/2 mice. Further studies of synapses, neurons, and neuronal networks regulated by GluRe4 and developmental changes in neuronal function in GluRe4 2/2 mice may lead to a better understanding of the mechanisms underlying PCP-induced psychosis and schizophrenia.

Ethics statement
The experimental procedures and housing conditions were approved by the Institutional Animal Care and Use Committee (Animal Experimentation Ethics Committee of Tokyo Institute of Psychiatry, Approval ID: , and all animal were cared for and treated humanely in accordance with our institutional animal experimentation guidelines.

Animals
Wildtype and GluRe1 2/2 or GluRe4 2/2 mouse littermates from crosses of heterozygous/heterozygous GluRe1 or GluRe4 knockout mice, respectively, on a C57BL/6 genetic background [33,65] served as subjects. Naive adult mice were housed in an animal facility maintained at 2262uC and 5565% relative humidity under a 12 h/12 h light/dark cycle with lights on at 8:00 am and off at 8:00 pm. Food and water were available ad libitum. In the behavioral experiments, 13-to 23-week-old male mice were used. In the microdialysis experiments, 10-to 24-weekold male and female mice were used.

Microdialysis and analytical procedures
Twenty-four hours after implantation, the dialysis experiments were performed in freely moving animals. Ringer's solution (145 mM NaCl, 3 mM KCl, 1.26 mM CaCl 2 , and 1 mM MgCl 2 , pH 6.5) was perfused at a constant flow rate of 1 ml/min. Perfusates were directly injected into the high-performance liquid chromatography system every 10 min using an autoinjector (EAS-20; Eicom). Dialysate DA was separated using a reverse-phase ODS column (PP-ODS; Eicom) and detected with a graphite electrode (HTEC-500; Eicom). The mobile phase consisted of 0.1 M phosphate buffer (pH 5.5) containing 500 mg/l sodium decanesulfonate, 50 mg/l EDTA, and 1% methanol. Perfusion was initiated 180 min prior to the collection of baseline samples. Baseline levels of DA ex were obtained from the average concentrations of three consecutive samples when they were stable. The DA detection limit of the assay was 0.3 fmol/sample with a signal-to-noise ratio of 2.

Locomotor activity measurements
Each mouse were exposed to an illuminated chamber (30640625 cm) at an ambient temperature of 2262uC, and locomotor activity was measured with Supermex (Muromachi Kikai, Tokyo, Japan), a sensor monitor mounted above the chamber. In this system, a sensor detects the radiated body heat of an animal [67]. This measurement system can detect changes in heat across multiple zones of the chamber and count all horizontal movements. All counts were automatically summed and recorded every 5 min. After a 180 min habituation period, METH or PCP was administered subcutaneously (s.c.), and locomotor activity was monitored continuously for 180 min.

Drugs
Drugs were dissolved in saline and administered s.c. in a volume of 10 ml/kg. In the microdialysis experiment, saline, METH (1 mg/kg), or PCP (3 mg/kg) was administered after establishing a stable baseline, and the dialysate was continuously collected for 180 min. In the acute behavioral experiments, saline, freshly prepared METH (1 mg/kg; Dainippon Sumitomo Pharma, Osaka, Japan), or PCP (3 mg/kg; Shionogi Pharmaceutical Co. Ltd., Osaka, Japan) was administered. In the repeated behavioral experiments, METH (1 mg/kg) or PCP (3 mg/kg) was administered repeatedly at 2 or 3 day intervals for a total of seven injections. One week after withdrawal, METH or PCP challenge injections were administered as described above.

Statistical analysis
DA ex responses to drugs are expressed as a percentage of baseline. The AUC of DA ex during the 180 min period after drug administration was calculated as the effects of the drugs. Areaunder-the-curve values of all groups were analyzed using two-way ANOVA. Individual post hoc comparisons were performed with Fisher's PLSD test. The responses to acute administration were analyzed using Student's t-test, one-way ANOVA, or two-way ANOVA. To evaluate behavioral sensitization, the response to drugs in Session 8 was compared with the response to the first drug injection (Session 1) in the same animal using a paired t-test or mixed-design ANOVA. Values of p,0.05 were considered statistically significant. Data were analyzed using Statview J5.0 software (SAS Institute, Cary, NC, USA).