Reversal of Phencyclidine Effects by a Group II Metabotropic Glutamate Receptor Agonist in Rats

See allHide authors and affiliations

Science  28 Aug 1998:
Vol. 281, Issue 5381, pp. 1349-1352
DOI: 10.1126/science.281.5381.1349


Glutamatergic abnormalities have been associated with several psychiatric disorders, including schizophrenia and addiction. Group II metabotropic glutamate receptors were targeted to normalize glutamatergic disruptions associated with an animal model of schizophrenia, the phencyclidine model. An agonist of this group of receptors, at a dose that was without effects on spontaneous activity and corticolimbic dopamine neurotransmission, attenuated the disruptive effects of phencyclidine on working memory, stereotypy, locomotion, and cortical glutamate efflux. This behavioral reversal occurred in spite of sustained dopamine hyperactivity. Thus, targeting this group of receptors may present a nondopaminergic therapeutic strategy for treatment of psychiatric disorders.

Several lines of evidence suggest that glutamatergic mechanisms contribute to the pathophysiology of schizophrenia (1–3). For example, phencyclidine (PCP) and other antagonists ofN-methyl-d-aspartate (NMDA) receptors have psychotomimetic properties in healthy individuals (4) and exacerbate preexisting symptoms of schizophrenia (5). However, drugs that target ionotropic glutamate receptors are not considered therapeutically useful because of the ubiquitous involvement of these receptors in mediating fast synaptic transmission throughout the central nervous system. Metabotropic glutamate receptors (mGluRs) on the other hand may provide important pharmacotherapeutic targets for psychiatric disorders associated with increased or decreased glutamatergic neurotransmission. These receptors modulate synaptic neurotransmission, and the heterogeneous localization of at least eight subtypes of mGluRs (mGluR1 to mGluR8) with distinct functional properties suggests that glutamatergic neurotransmission may be modulated in an anatomically and functionally distinct manner (6). The subtypes of mGluRs are currently classified within three groups (I to III) on the basis of sequence homology and pharmacology. The group II family of mGluRs, which consists of mGluR2 and mGluR3, is primarily distributed in forebrain regions (6, 7). Stimulation of this group of mGluRs mediates presynaptic depression and decreases evoked release of glutamate, suggesting that these receptors regulate activated glutamate release by presynaptic mechanisms (8–10).

PCP and other psychotomimetic NMDA antagonists increase glutamate efflux (11, 12) and may produce their cognitive and locomotor effects, at least in part, by potentiating glutamatergic neurotransmission at non-NMDA receptors (2, 11). In rodents (11,13) and humans (14), antagonists of non-NMDA receptors or pretreatments that attenuate glutamate release reduce mnemonic and other behavioral effects of NMDA receptor antagonists. Hence, reduction of presynaptic glutamatergic activity by targeting group II mGluRs may present an approach for reversing those behavioral effects of PCP associated with increased glutamatergic activity.

Animals were treated with the systemically active and highly selective agonist of group II mGluR (+)-2-aminobicyclo- [3.1.0]-hexane-2,6,-dicarboxylate monohydrate (LY354740) (15) or vehicle before they received PCP (16). The highest dose of LY354740 used [10 mg of LY354740 per kilogram of body weight intraperitoneal (ip) injection] has been shown to not affect basal glutamate efflux but to normalize depolarization-induced activation of glutamate release (9). This dose also does not affect neuromuscular coordination, spontaneous locomotor activity, or learning and memory as assessed by passive avoidance responding (15).

The increase in glutamate efflux in the prefrontal cortex produced by PCP was abolished in animals that were pretreated with the mGluR agonist LY354740 (Fig. 1). In contrast, activation of dopamine release by PCP in the prefrontal cortex or the nucleus accumbens was not reduced by this pretreatment (Fig. 2).

Figure 1

Effect of pretreatment with the group II mGluR agonist LY354740 (10 mg/kg ip injection) (•; n = 6) or vehicle (○; n = 7) on stimulation of glutamate efflux by PCP (5 mg/kg ip injection) in the (A) prefrontal cortex and (B) nucleus accumbens of freely moving rats. PCP injection after vehicle (water) produced a significantly greater effect on extracellular glutamate concentrations in the prefrontal cortex (P < 0.05, f = 2.2) than in animals pretreated with LY354740. Similar results were observed in the nucleus accumbens; however, because of a greater degree of variability, the difference in the two groups did not reach statistical significance.

Figure 2

Effect of pretreatment with the group II mGluR agonist LY354740 (10 mg/kg ip injection) (•) or vehicle (○) on stimulation of dopamine efflux by PCP (5 mg/kg ip injection) in the (A) prefrontal cortex or (B) nucleus accumbens of freely moving rats. Animals pretreated with LY354740 exhibited a significant increase in dopamine release in the nucleus accumbens (P < 0.001, f= 6.4) and prefrontal cortex (P < 0.001,f = 7.2). •, n = 6 (A) and 8 (B); ○, n = 7 (A) and 8 (B). (C) The lack of effect of LY354740 (10 mg/kg) by itself on dopamine release in both regions is shown. Prefrontal cortex, •; nucleus accumbens, ○.

Stimulation of group II mGluRs nearly abolished PCP-induced locomotor activity and significantly reduced the total stereotypy score, primarily because of a decrease in head rolling, a hallmark of behavioral activation by NMDA receptor antagonists (Fig. 3). This behavioral reversal at a dose of LY354740 that is without an effect on spontaneous locomotion (15) is distinct from similar effects reported with monoamine receptor antagonists, including the antipsychotic drugs haloperidol and clozapine, in that this latter class of drugs reduces locomotor activation by PCP at doses that, by themselves, produce marked locomotor inhibition (17).

Figure 3

Pretreatment with the group II mGluR agonist LY354740 (10 mg/kg ip injection) [•; n = 11 (A) and 6 (B)] significantly reduced the (A) locomotor-activating effects of PCP (P < 0.001,f = 8.9) and (B) total stereotypy score (P < 0.05, f = 1.7) as compared with vehicle (water)–pretreated rats [○; n= 10 (A) and 5 (B)]. (C) The effect of this pretreatment on oral stereotypy and head rolling scores totaled for the 2-hour period after PCP injection is shown. A lower dose of LY354740 (1 mg/kg ip injection) was without a significant effect on the locomotor activations produced by PCP (25).

The effect of LY354740 was also assessed on PCP-induced impairment of discrete-trial delayed alternation task. This is a rodent working memory paradigm that involves continuous changing of strategy and thus is not prone to overtraining as in ordinary T-maze delayed alternation tasks (12). Pretreatment with LY354740 reduced the deficits induced by 5 mg of PCP per kilogram of body weight and abolished the deficits induced by a PCP dose of 1 mg/kg (Fig. 4).

Figure 4

Effect of vehicle (water), PCP (1 or 5 mg/kg ip injection), and PCP + LY354740 (10 mg/kg ip injection) on discrete-trial delayed alternation performance in rats. Animals were trained until their performance of percentage of correct choice (total of 10 choice runs) at (A) 10- and (B) 40-s delays was stabilized for at least five consecutive days (C). A “baseline performance” at each delay was defined for individual animals as the average of correct choices on three consecutive days before drug injection. Water (H2O) or LY354740 (10 mg/kg) was injected immediately before PCP and 50 to 60 min before testing. Treatment with both doses of PCP impaired the performance at 10- (○) and 40-s (▴) delays. Pretreatment with LY354740 significantly reversed the impairment of a PCP dose of 1 mg/kg at both delays and a PCP dose of 5 mg/kg at the 40-s delay (*, P < 0.05, as compared with water-treated group; n = 5 to 9).

The reduction in PCP-induced glutamate efflux by the mGluR agonist suggests that this drug ameliorates the behavioral effects of PCP by attenuating presynaptic glutamatergic activity, a mechanism consistent with reports that non-NMDA receptor antagonists and glutamate release inhibitors reduce the behavioral effects of PCP or ketamine (11, 13, 14). However, because in vivo selectivity of LY354740 and the source of the activated extracellular glutamate by PCP is presently unknown (11, 12), other mechanisms involving glial cells or postsynaptic interactions between mGluRs and NMDA receptors cannot be ruled out.

Although the psychotomimetic and cognitive effects of PCP are often attributed to the effects of this drug on dopamine neurotransmission (3), attenuation of the behavioral effects of PCP by the mGluR agonist was not accompanied by a reduction in dopamine release. This observation is consistent with findings indicating that dopaminergic neurotransmission is dissociated from locomotor and other behavioral effects of NMDA receptor antagonists (12,18) and questions a direct role for dopamine in the psychotomimetic effects of PCP.

Although extrapolation of data from rodent models to complex human syndromes such as acute PCP intoxication or schizophrenia may be problematic, the behavioral measures in the present study are generally considered relevant to the clinical phenomenology. Locomotor activity in the rodent may be a useful indicator of the propensity of a drug to elicit or exacerbate psychosis in humans (3, 19), and cognitive tasks with a working memory component may model the frontal lobe dysfunctionality and working memory deficits associated with schizophrenia (19, 20). Furthermore, stereotyped tendencies have been reported in schizophrenics, and this behavior may be of importance to schizophrenic symptomatology, not merely in the context of movement, but also relating to organization of thought and perseveration (21). The present observation that LY354740 reduced the head-rolling stereotypy in the rodents may be of particular importance because repetitive head rolling is commonly reported in individuals suffering from PCP psychosis (22). Thus, our finding that stimulation of group II mGluRs reduces all the above behavioral effects of PCP suggests that targeting these receptors may ameliorate symptoms of acute PCP psychosis in humans and, assuming that PCP psychosis is a valid model of schizophrenia, may be effective in treating related symptoms of schizophrenia. Of note, the putative peptide neurotransmitterN-acetyl-aspartylglutamate, whose levels have been reported to be abnormal in schizophrenic brains (23), is an endogenous agonist for group II mGluRs (24).

The important functional characteristic of this group of mGluR agonists, which has relevance for their clinical utility, is that, unlike dopamine antagonists or drugs that target the ionotropic glutamate receptors, a significant amelioration of the cognitive and motoric effects of PCP was observed at a dose that was without a measurable effect on animals' spontaneous activity or on dopaminergic neurotransmission. This observation is consistent with an autoregulatory function for these receptors whereby they inhibit glutamate release under conditions of high agonist availability. Thus, targeting mGluR receptors may offer exquisite functional selectivity by specifically modulating those presynaptic sites where glutamatergic neurotransmission is abnormally regulated.

  • * To whom correspondence should be addressed. E-mail: bita.moghaddam{at}


View Abstract

Navigate This Article