Abstract
The objective of this study was to characterize the behavior induced by the N-methyl-d-aspartate receptor antagonist MK-801 (dizocilpine maleate) in rats as a model of psychosis. The temporal profile, dose dependence, age, and sex differences of the behavior are described. A gas chromatographic method for the analysis of MK-801 in plasma and brain was developed. Female rats showed 4 to 10 times more MK-801-induced behavior and displayed around 25 times higher serum and brain concentrations of MK-801 than male rats. Twenty-one neuroactive compounds, including a number of excitatory amino acid-active substances, were tested for the effect on MK-801-induced behavior. Neuroleptics blocked MK-801-induced behavior in a dose-dependent manner that correlated to their antipsychotic potency in humans. Adenosine receptor agonists and anN-methyl-d-aspartate receptor-associated glycine site antagonist showed putative antipsychotic effects. In conclusion, MK-801-induced behavior represents a rat excitatory amino acid hypofunction model of psychosis that appears to be of clinical relevance and may be of value in the search for new antipsychotic agents.
A large number of different animal models for the study of psychotic disorders, such as schizophrenia, have been described (Lyon, 1991). However, no models have yet been demonstrated to show all signs of the disorders. Most widely used are pharmacological models based on different neurochemical pathophysiological theories, e.g., the administration of dopamine agonists, serotonin agonists, opioids, and anticholinergics. These models each show some, but not all, of the changes in animal behavior that are considered to be related to psychosis (e.g., continuous exploratory behavior, stereotypies, postural imbalance, aggression).
In addition to the dopaminergic (Carlsson, 1988) and other hypotheses of psychotic disorders, a dysfunction in the main excitatory neurotransmitter system of the brain, the excitatory amino acids (EAAs), has been suggested in psychoses (Kim et al., 1980; Deutsch et al., 1989; Carlsson and Carlsson, 1990; Javitt and Zukin, 1991; Squires and Saedrup, 1991; Moghaddam, 1994; Olney and Farber, 1995). This is mainly based on the schizophrenia-like psychotomimetic action of phencyclidine in humans (Snyder, 1980; Rosse et al., 1994), which can be attributed to noncompetitive blockade of theN-methyl-d-aspartate (NMDA) type of EAA receptor (see Javitt and Zukin, 1991). In rodents, phencyclidine and the highly selective NMDA antagonist (5R,10S)-(+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine hydrogen maleate (dizocilpine maleate, MK-801; Wong et al., 1986), induce a behavior with increased locomotion, stereotypies, and ataxia (Sturgeon et al., 1979; Koek et al., 1988; Contreras, 1990; Tiedtke et al., 1990; Hoffman, 1992; Ginski and Witkin, 1994). Antipsychotic agents in clinical use (neuroleptics) antagonize both phencyclidine- and MK-801-induced behavior (Sturgeon et al., 1981; Freed et al., 1984; Tiedtke et al., 1990; Hoffman, 1992), indicating that NMDA antagonist-induced behavior may be used as a complementary model of psychosis in the search for new and better antipsychotic agents. The aim of this study was to further characterize MK-801-induced behavior in rats as a putative model of psychosis.
Materials and Methods
Animals
About 1100 Sprague-Dawley rats (B&K Universal, Sollentuna, Sweden) were used. They were housed in groups of five in a temperature-controlled room (22°C) with the light on from 6:00 AM until 6:00 PM. The rats had free access to food and water. Each rat was experimentally naive and was tested only once. In the behavioral experiments at different ages, the rats were age-determined by their body weight and a standard correlation paradigm (B&K Universal, Sollentuna, Sweden); 10 days of age: 12 to 20 g, both males and females; 20 days: 50 g, both males and females; 40 days: males 200 g, females 150 g; 60 days: males 360 g, females 230 g; 80 days: males 450 g, females 260 g. When the 21 neuroactive compounds were tested for their effect on MK-801-induced behavior, only adult female rats weighing 200 to 250 g (i.e., around 60 days of age) were used (n = 686), except for (αS, 5S)-α-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid (AT-125; acivicin), which was also tested in males.
Drugs
MK-801 was a gift from Dr. Karl A. Rudolphi at Hoechst AG (Wiesbaden, Germany). Acivicin was generously supplied by Upjohn (Partille, Sweden) and risperidone (RISP) by Janssen Pharma (Göteborg, Sweden). The commercially available solutions of remoxipride (REM; Roxiam, Astra Arcus, Södertälje, Sweden), diazepam (DIAZ; Apozepam, A.L, Nacka, Sweden), chlorpromazine (CHLOR; Hibernal, Rhône-Poulenc, Birkeröd, Denmark), perphenazine (PERPH; Trilafon, Schering, Kenilworth, NJ), haloperidol (HAL; Haldol, Janssen Pharmaceutica, Beerse, Belgium), and theophylline (THEO; Theofyllin, Draco, Lund, Sweden) were used. The chemicalsN6-cyclohexyladenosine (CHA), 1-(4-aminophenyl)-4-methyl-7,8-methylenedioxy-5H-2,3-benzodiazepine hydrochloride (GYKI; GYKI 52466 hydrochloride), clozapine (CLOZ), 1,2,3,4-tetrahydro-6-nitro-2,3-dioxo-benzo[f]quinoxaline-7-sulfonamide disodium (NBQX), 3-bicyclo[2.2.1]hept-5-en-2-yl-6-chloro-3,4-dihydro-2H-1,2,4-benzothiadiazine-7-sulfonamide 1,1-dioxide (cyclothiazide), (2-hydroxypropyl-β-cyclodextrin,d-cycloserine (DCS),R(+)-3-amino-1-hydroxy-2-pyrrolidinone (HA-966), α-(4-hydroxyphenyl)-β-(4-benzylpiperidin-1-yl) β-methylethanol tartrate (IFEN; ifenprodil tartrate), R(−)N6-(2-phenylisopropyl)adenosine (R-PIA), 8-cyclopentyl-1,3-dipropylxanthine (DPCPX), 1-(6-amino-9H-purin-9-yl)-1-deoxy-N-ethyl-β-l-ribofuranuronamide (5′-N-ethylcarboxamido adenosine) (NECA), 3,7-dimethyl-1-propargylxanthine (DMPX), and 2-p-(2-carboxyethyl)phenethylamino-5′-N-ethylcarboxamido adenosine hydrochloride (CGS-21680) were obtained from Research Biochemicals International (Natick, MA).
The following substances were dissolved in physiological saline on the day of the experiment: MK-801, acivicin, CGS-21680, DCS, DMPX, GYKI, HA-966, NECA, CHA, and R-PIA. CLOZ was dissolved in a minimal volume of HCl (0.1 M), diluted with saline, and pH-adjusted with NaOH. Cyclothiazide was suspended in 2-hydroxypropyl-β-cyclodextrin (45% w/v). DPCPX was dissolved in Tween 80 after 60 min at 75°C and diluted with warm water to obtain a Tween 80 concentration of 20% (w/v) or less, thereby avoiding behavioral effects of Tween 80 (Castro et al., 1995). IFEN was dissolved in saline by adding a minimal amount of tartaric acid. RISP could also be dissolved by adding tartaric acid to water. NBQX was dissolved in a minimal amount of NaOH (1 M), diluted with water, and pH-adjusted with HCl (0.1 M). All drugs were administered i.p.
Procedure for Behavioral Experiments
On the day of the experiment, 10 rats were placed in individual standard clear plastic cages (20 × 26 × 14 cm) with a wire mesh top and a thin layer of wood chips as bedding. For the 10-day-old rats, a cage of half the size was used. The rats were allowed 1 h of accommodation in the cages before the onset of the experiment. Then MK-801 (0.05–3.0 mg/kg) or an equal volume of saline was administered i.p., followed by a behavioral observation and rating procedure that started 15 min after the injection and continued for another 60 min (i.e., 15–75 min postinjection). The rating procedure included observation of each of the 10 rats during 30 s every 5 min during the 60-min observation period, which resulted in 13 observations per rat per experiment. A total score for each behavior was obtained by summing the individual ratings from the 13 observation periods. Three types of behavior were rated: locomotion, stereotyped sniffing, and ataxia (Table 1). For locomotor activity, the rating scale described by Sturgeon and coworkers (1979) was used, and stereotyped sniffing behavior was rated according to the rating scale described by Hoffman (1992). For ataxia rating, a simplified rating scale was developed (Table 1) based on empirical findings in pilot studies. Ataxia was rated during locomotion and thus rats that had been treated with any neuroactive compound and were lying still during the observation period presented a special problem. Therefore, the righting reflex was tested in all animals after the experiments. The righting reflex was tested directly after the behavioral experiment by placing the rat on its back. The way the rat returned to normal posture was used to evaluate the righting reflex. Normally the rat immediately returned to normal position, but when the return occurred more slowly it was denoted as “impairment of the righting reflex”. The interventional compounds that enhanced MK-801-induced behavior were also tested for effect on locomotor activity.
When the influence of an interventional neuroactive compound on MK-801-induced behavior was tested, the drug (or an equal volume of saline) was administered at different times in relation to the MK-801 injection. The timing was based on previous experiments presented in the literature and the pharmacokinetics of the drug to obtain a maximal brain concentration during behavioral testing. The following drugs were administered i.p. 15 min before MK-801: NBQX, DIAZ, IFEN, NECA, THEO,R-PIA, DPCPX, DMPX, and CGS-21680. The following drugs were administered 30 min before MK-801: GYKI, DCS, HA-966, RISP, PERPH, CLOZ, CHLOR, HAL, CHA, and cyclothiazide. REM and acivicin were administered 60 and 120 min, respectively, before MK-801. When testing the interventional compounds, the dose of MK-801 was adjusted to obtain optimal experimental conditions. Lower doses of MK-801 (0.05–0.1 mg/kg) caused minor behavioral activation that was suitable for detecting whether an interventional drug increased the behavior. A higher dose of MK-801 (0.2 mg/kg), resulting in major behavioral activation, made it easier to detect any reduction in behavior caused by the interventional drug. For each interventional drug that was tested, at least three doses were used and the effect compared with separate saline- and MK-801-treated groups.
After the behavioral experiments, the rats that had been given RISP, PERPH, CLOZ, CHLOR, or HAL, and some rats that had received MK-801 only, received 60 mg/kg of methohexital (Brietal, Eli Lilly, Stockholm, Sweden). The heart was exposed and blood (around 5 ml) was sampled with a syringe from the right atrium. The blood was centrifuged and the serum was stored at −80°C until it was analyzed. In addition, the brains of some animals that had only received MK-801 were carefully removed and stored at −80°C.
Determination of Concentrations of Neuroleptics in Serum
Neuroleptics were determined by standard high pressure liquid chromatography or gas chromatographic procedures. PERPH, CLOZ, and HAL were analyzed at the Psychiatric Research Unit, Sahlgrenska University Hospital. Unfortunately, we were not able to determine the serum concentration of REM during the experimental period. CHLOR was analyzed at the Laboratory of Neurochemistry, Lund University Hospital, and RISP at the Department of Clinical Pharmacology, Lund University Hospital.
Determination of MK-801 Concentration in Serum and Brain
Serum and brain tissue from animals that had only received MK-801 were analyzed for the concentration of MK-801. The brain dissection resulted in the following six brain regions: parietal cortex (PC), frontal cortex (FC), hypothalamus (HY), striatum posterior (SP), striatum anterior (SA), and hippocampus (HI). The dissection procedure was performed according to Glowinski and Iversen (1966) with some modifications: A transverse section was made through the cerebrum at the level of the optic chiasma and two regions were dissected from the anterior (FC and SA) and four regions from the posterior part of the cerebrum (PC, HY, SP, and HI). The reproducibility of the dissection procedure is presented in Table 2. The following analysis of MK-801 was based on a chromatographic technique that has been described previously (Hucker et al., 1983; Vezzani et al., 1989).
Sample Preparation Step I: Brain.
The brain regions were homogenized according to Vezzani and coworkers (1989). The brain homogenate was transferred to a centrifuge tube with the aid of two portions of acetone:1 M formic acid (85:15, v/v) so that the final volume was 6 ml/g brain tissue. The homogenate was centrifuged (3,000g, 15 min) and the supernatant was transferred to an extraction tube. If the original brain sample weight was greater than 70 mg, an appropriate aliquot of the supernatant was taken for extraction, the volume being made up to 6 ml with the acid acetone. An internal standard solution (25 μl), containing 6.25 pmol methadone in i-propanol, was added.
Sample Preparation Step I: Serum.
A volume of 0.50 ml of serum sample or serum calibrator was added to a glass homogenizer and homogenized with 3 ml acetone:1 M formic acid (85:15, v/v) and transferred to a centrifuge tube. The homogenizer was rinsed with an additional portion of acid acetone that was added to the first in the centrifuge tube. After mixing, the homegenate was centrifuged (3,000g, 10 min) and the supernatant was transferred to an extraction tube. An internal standard solution (25 μl), containing 6.25 pmol methadone in i-propanol, was added.
Sample Preparation Step II: Brain and Serum.
The supernatant from serum or brain homogenization was washed by shaking for 5 min with 4 ml hexane: chloroform (9:1, v/v), the phases were separated by centrifugation and the upper hexane layer aspired to waste. The washing procedure was repeated with a fresh portion of hexane chloroform. The acid water phase was transferred to a new extraction tube, made alkaline with 1 ml of 1 M NaOH, and extracted for 15 min with 8 ml hexane:isoamyl alcohol (100:1, v/v). The phases were separated by centrifugation and the upper hexane layer was transferred to a new extraction tube. Two milliliters of 0.1 M HCl was added and the phases were mixed by inversion for 5 min. After centrifugation, the acid water phase was transferred to a test tube (100 × 12 mm, i.d.), 0.2 ml 2 M NaOH in 3 M NaCl was added and the resulting alkaline water phase extracted with 0.5 ml hexane:isoamyl alcohol (100:1, v/v) by inversion for 5 min, followed by brief centrifugation to separate the phases. The hexane layer was transferred to a small test tube (60 × 6 mm, i.d.) with a tapered bottom and was mixed briefly with 30 μl formic acid:methanol:water (7:30:30, by volume). After the phases had separated, the hexane layer was aspired to waste and the remaining acid was evaporated to dryness under reduced pressure in a desiccator containing solid NaOH. The dried extract was dissolved in 10 μl of a mixture of heptane:isopropanol:strong ammonia (10:3:0.005, by volume), and 4 μl of this solution was then injected into the gas chromatograph. A Varian model 3300 gas chromatograph (Psychiatric Research Unit, Sahlgrenska University Hospital) was used, equipped with an Rtx-1 nitrogen-sensitive detector and a 15 m × 0.53 mm i.d. fused silica column (film thickness 1 μl). The column temperature was held at 110°C for 1 min, increased by 10°C/min to 210°C, then increased by 40°C/min to 295°C, which was maintained for 2 min. The injection temperature was 210°C and the detector temperature was 280°C. Nitrogen was used as the carrier gas (35 ml/min). Hydrogen and air flow were 4 and 200 ml/min, respectively. Homogenized human and rat serum standards containing MK-801 were used for standardization of serum and brain samples. For male rat samples, the standard concentrations used were 0, 2, 5, 10, 20, and 50 nM, and for females 0, 10, 20, 50, 100, and 300 nM. Standard curves of chromatographic peak height ratios (MK-801/methadone) as a function of drug concentration were prepared for each run and used for quantitation of samples.
Extraction yields of MK-801 and methadone, 10 pmol of each, added to human serum homogenate supernatant, were determined (mean ± S.D.): MK-801 63 ± 8% (n = 6), methadone 73 ± 6% (n = 6). Two internal quality control samples were analyzed in each analysis run with a coefficient of variation for MK-801 of 7.0% [MK-801 mean 14.8 nM (n = 54) when analyzing rat serum with MK-801 added to contain 15 nM].
The analysis procedure described by Schwartz and Wasterlain (1993)using gas chromatographic separation on a HP-5 fused silica column with a nitrogen/phosphorus-sensitive detector after a single step extraction was considered and tried by us. Because we could not achieve sufficiently clean extracts, we had to use a more elaborate extraction procedure. Schwartz and Wasterlain (1993) reported 96 ± 2% extraction yields of MK-801 from both serum and brain, indicating that losses in the homogenization steps are insignificant. In our hands, the serum concentrations of MK-801 in treated male rats were considerably lower compared with Schwartz and Wasterlain (1993), and we had to increase the sample volume 10-fold to be able to measure serum concentrations in male rats. We had difficulties in obtaining the 5-propyl derivative of MK-801 and instead we used methadone as the internal standard.
Statistics
The data are presented as mean ± S.E.M. The nonparametric two-tailed Mann-Whitney U test or Student’s ttest was used to evaluate the statistical significance of differences between groups of rats. Comparison of i.p. doses versus plasma and brain concentrations was made by linear regression analysis.
Results
Behavior in Saline-Treated Controls.
Both male and female adult rats that received saline only (n = 111) were typically asleep or lying still in one corner of the cage during the entire 60-min observation period, with one or two periods of exploratory behavior, and consequently had very low scores for locomotion and stereotypy, and no ataxia (Fig.1).
Temporal Profile of MK-801-Induced Behavior.
Injection (i.p.) of MK-801 caused a significant behavioral activation, consisting of increased locomotion, stereotyped sniffing and, at higher doses, ataxia. This behavior was dose-dependent and varied with age and sex. As an example of the temporal development of behavior after injection, Fig. 1 shows the behavior in 60-day-old female rats treated with 0.2 mg/kg of MK-801 (n = 30) compared to saline-treated controls (n = 13). The first signs of locomotion and stereotyped sniffing were observed 20 to 25 min after MK-801 administration, whereas ataxia was first detected 5 to 15 min later. Locomotor activity was maximal after 35 min and thereafter displayed a minor decline over the experimental period. Stereotyped sniffing was maximal after 30 min and persisted at this level throughout the experiment. In contrast, the observed degree of ataxia continued to increase until 75 min after MK-801 injection. Female rats receiving a dose of MK-801 that induced maximal activity typically displayed periods of intermittent movements within half or the whole cage (locomotion score 2 or 4) combined with intense sniffing (stereotypy score 2) and a tendency to fall or fallings (ataxia scores 1 or 2). Male rats showed the first signs of locomotion and stereotyped sniffing at the same time after MK-801 injection as females. The behavior in males was much less uniform as compared with female rats and typically consisted of bursts of behavioral activation (locomotion score 1–2 and stereotypy score 1–2) with a duration of around 5 to 10 min.
Age and Sex Differences in MK-801-Induced Behavior.
The behavioral response to MK-801 varied with the age of the rats, and females responded more than males to MK-801 at most ages. Figure2 shows the behavior in male (n = 42) and female (n = 39) rats of different ages that were all given 0.2 mg/kg of MK-801. In male rats, maximal MK-801-induced locomotion and stereotypy was seen at the age of 20 days. At all other ages, males displayed only minor behavioral activation, approximately 2 to 3 times more than saline-treated controls at this dose. Ataxia was not seen after 0.2 mg/kg of MK-801 in males, except in the 10-day-old rats, which displayed massive ataxia with loss of the righting reflex, and to a lesser extent 20-day-old rats. In contrast, females showed a major behavioral activation after 0.2 mg/kg of MK-801 at all ages, except for the 10-day-old rats, where mostly ataxia was seen. All three types of behavior were prominent in females after MK-801 administration. There were no statistically significant sex differences in behavior in the 10- and 20-day-old groups, but in 40- to 80-day-old rats, females showed about 4, 2, and >10 times more locomotion, stereotyped sniffing, and ataxia than males, respectively. It should be mentioned that the 10-day-old rats were difficult to evaluate as they responded with loss of the righting reflex to MK-801, which was not the case in older rats.
Dose-Dependence of MK-801-Induced Behavior.
The effect of different doses of MK-801 was investigated in both male (n = 41) and female (n = 53) 60-day-old rats (Fig. 3). In males, a dose of 0.2 mg/kg was required to obtain significant locomotion and stereotyped sniffing, and no ataxia was observed in male rats after doses of MK-801 up to 0.5 mg/kg. Maximal behavioral activation in male rats was seen after 1.0 mg/kg of MK-801 with regard to locomotion and stereotyped sniffing, whereas 3.0 mg/kg caused extensive ataxia without locomotion (data not shown). In contrast, female rats displayed significant locomotion and stereotyped sniffing after 0.05 mg/kg of MK-801, with minor ataxia at 0.1 mg/kg and major ataxia at higher doses (Fig. 3). In females, locomotion was maximal at 0.1 and 0.2 mg/kg, stereotyped sniffing at 0.1 to 0.5 mg/kg, and ataxia at 0.2 to 1.0 mg/kg of MK-801. At 0.5 and 1.0 mg/kg, locomotion was reduced because the female rats more frequently showed ataxia score 3, i.e., they were almost unable to move.
Concentration of MK-801 in Serum and Brain.
In a separate set of experiments, low- (0.2 mg/kg) and high- (1.0 mg/kg) dose MK-801-induced behavior was evaluated in adult male (n= 29) and female (n = 29) rats (Fig.4). In females, low-dose MK-801 caused marked behavioral activation whereas high-dose MK-801 mainly caused severe ataxia. In males, low-dose MK-801 induced a minimal behavioral activation, whereas high-dose MK-801 caused an activation similar to that in the female low-dose group but with 28% less locomotion.
The blood and brain samples were collected consecutively after the behavioral experiments, from about five animals per h. Thus, the analysis reflected blood and brain concentrations from about 3 to 5 h after the i.p. injection of MK-801. However, no significant changes in MK-801 concentration were seen during this 2-h period within each experiment.
Female and male rats that received the same dose of MK-801 differed most dramatically with respect to MK-801 concentrations in blood (Fig.4) and brain (Table 3). In serum, the concentration of MK-801 in males was 1.9 ± 0.24 and 8.0 ± 0.68 nM in low- and high-dose animals, respectively. In contrast, the MK-801 concentration in female serum was 27 and 26 times higher in the low- (50.8 ± 1.82 nM) and high- (205 ± 8.2 nM) dose groups, respectively. The serum levels of MK-801 in low-dose-treated male rats were close to the detection limit of the chromatographic technique whereas the serum levels in the other treatment groups could easily be detected. In addition, brain levels of MK-801 could be detected in all groups (Fig. 5). The 5-fold increase in dosage of MK-801, from low- to high-dose MK-801, resulted in a 4-fold increase in serum concentration in both male and female rats.
Male and female rats showed the same major difference with regard to brain MK-801 concentrations as they did in serum (Table 3). In both low- and high-dose-treated rats, females showed around 25 times higher levels of MK-801 than the respective male group in all six brain regions. When comparing low- and high-dose treatment within the same sex, a 5-fold increase in the dosage of MK-801 resulted in a 3- to 4-fold increase of the MK-801 levels in all brain regions in both male and female rats. Regardless of sex or dosage of MK-801, brain regions differed with regard to MK-801 levels. The highest levels were detected in PC, FC, and HI, whereas HY, SP, and SA displayed lower levels (Table3).
Effect of Neuroleptics on MK-801-Induced Behavior.
All tested neuroleptics (CHLOR, n = 26; HAL, n = 20; PERPH, n = 27; RISP, n = 36; REM,n = 47; CLOZ, n = 30) blocked MK-801-induced behavior in a dose-related fashion (Fig.6). The serum concentrations of the neuroleptics 2 to 3 h after i.p. injection correlated well to the administered doses (Table 4). The neuroleptics inhibited all three types of behavior, except REM (and to a minor extent also RISP), which marginally affected stereotyped sniffing. The inhibitory effects of the neuroleptics were seen throughout the observation period.
Effect of Adenosinergic Drugs on MK-801-Induced Behavior.
Four adenosine receptor agonists were tested (Fig.7). The adenosine analog NECA (n = 36), an agonist with almost equal activity at A1 and A2 adenosine receptors, reduced MK-801-induced locomotion and stereotypy by 60 and 54%, respectively. This was achieved at the highest dose (2.0 mg/kg), at which NECA also caused ataxia with impairment of the righting reflex. Treatment with the predominant A1 adenosine receptor agonist, CHA (n = 36), inhibited all three types of behavior, with an intermediate effect at 0.1 and 0.5 mg/kg and total blockade at 2.0 mg/kg. All three doses of CHA induced some loss of muscle tone, but the righting reflex did not differ from that of saline-treated controls. Another predominant A1 adenosine receptor agonist, R-PIA (n = 30), caused a 46, 38, and 63% reduction in MK-801-induced locomotion, stereotypy, and ataxia, respectively, at a dose that did not cause obvious ataxia (0.25 mg/kg). A 10-fold higher dose caused a further behavioral reduction but with ataxia and loss of the righting reflex. The highly selective A2 adenosine receptor agonist, CGS-21680 (n = 30), reduced MK-801-induced locomotion, stereotypy, and ataxia by 53, 63, and 58%, respectively. CGS-21680 induced no ataxia and did not affect the righting reflex.
Three adenosine receptor antagonists were also tested for effect on MK-801-induced behavior (Fig. 7). The unspecific adenosine receptor antagonist, THEO (n = 47), did not influence MK-801-induced behavior at the doses of 1.0 and 5.0 mg/kg, but 20 mg/kg of THEO significantly enhanced locomotion 2.2-fold and stereotyped sniffing 1.5-fold. Female adult rats treated with 20 mg/kg of THEO only developed a similar pattern of behavior to rats treated with 0.05 mg/kg of MK-801 (increased locomotion and stereotyped sniffing, without ataxia). Rats that were treated with THEO (20 mg/kg) and MK-801 (0.05 mg/kg) displayed a 36% higher degree of locomotion (not statistically significant) as compared with rats treated with THEO (20 mg/kg) only. Neither the A1 adenosine receptor antagonist DPCPX (n = 25) nor the A2 adenosine receptor antagonist DMPX (n = 29) influenced MK-801-induced behavior. Whereas DMPX (5.0 mg/kg) showed a behavioral activation with increased locomotion and stereotyped sniffing, DPCPX had no such effect.
Effect of EAA-Modulating Drugs on MK-801-Induced Behavior.
Figure 8 shows the effect of various EAA-modulating substances on MK-801-induced behavior. An agonist at the glycine modulatory site of the NMDA receptor, DCS (n = 40), significantly increased MK-801-induced locomotion at 10 mg/kg but had no effect on stereotypy and ataxia. Rats that were treated with DCS only (10 mg/kg) did not differ from saline-treated controls. In contrast, HA-966 (n = 29), an antagonist at the glycine-modulatory site, reduced MK-801-induced locomotion, stereotypy, and ataxia by 76, 80, and 81%, respectively. At the dose required (10 mg/kg), however, HA-966 induced impairment of the righting reflex. The antagonist at the polyamine site of the NMDA receptor, IFEN (n = 36), enhanced MK-801-induced locomotion by 39% in a dose of 1.0 mg/kg, with minor effect on stereotypy and ataxia. IFEN, like DCS, had no obvious effect on spontaneous behavior. The competitive α-amino-3-hydroxy-5-methyl-isoxazole-4-proprionic acid (AMPA)/kainate receptor antagonist, NBQX (n = 32), caused no statistically significant changes in MK-801-induced behavior. However, 20 mg/kg of NBQX increased (by 35%) whereas 40 mg/kg of NBQX decreased (by 42%) MK-801-induced locomotion. The noncompetitive AMPA/kainate receptor antagonist, GYKI (n = 34), increased MK-801-induced stereotyped sniffing, ataxia, and locomotion (by 90%) significantly at 1.0, 5.0, and 10 mg/kg, respectively. GYKI alone (10 mg/kg) did not induce any locomotion but caused a minor impairment in the righting reflex in about half of the animals. Finally, cyclothiazide (n = 30; 0.5–5.0 mg/kg), a benzothiadiazide that inhibits desensitization of AMPA receptors, and thereby potentiates AMPA agonist responses, did not alter MK-801-induced behavior (data not shown).
DIAZ.
The benzodiazepine, DIAZ, inhibited MK-801-induced behavior at 5.0 mg/kg (Fig. 9), but at this dose DIAZ caused a considerable slowing of the righting reflex.
Acivicin.
Administration of the γ-glutamyltransferase (γ-GT) inhibitor, acivicin, to adult male rats (n = 52) in doses of 5.0 or 50 mg/kg, followed by MK-801 2 h later, did not result in any different behavior compared with rats that received MK-801 (0.2 mg/kg) only (Fig. 9). However, 100 mg/kg of acivicin, in combination with MK-801, produced a marked increase in locomotion (4.1-fold) and stereotyped sniffing (by 80%, not statistically significant). Typically, the duration of periods with locomotion score 4 combined with sniffing score 2 was prolonged in high-dose acivicin-treated rats. Rats that received acivicin (100 mg/kg) only did not differ from saline-treated controls. In these experiments, ataxia was never observed. In additional experiments in adult female rats (n = 34), no potentiating effect of acivicin (5.0–100 mg/kg) on MK-801-induced behavior (0.05–0.2 mg/kg) could be found (data not shown).
Discussion
Characterization of the Model.
This study showed that female rats were considerably more sensitive to MK-801 than males, probably due to the pharmacokinetic differences. Sex differences in amphetamine-, phencyclidine-, and MK-801-induced behavior have been reported previously. The increased amphetamine-elicited rotational behavior that is seen in females is probably due both to a slower metabolism of amphetamine in females and a modulation of neurotransmission by gonadal steroid hormones (Becker et al., 1982). Female rats are also more sensitive to phencyclidine, due to a lower efficiency of the hepatic metabolizing system, resulting in higher plasma and brain concentrations of phencyclidine (Nabeshima et al., 1984). MK-801 is also metabolized in the liver and the presently reported pharmacokinetic differences between male and female rats may thus in a similar way be explained by a low capacity for MK-801 metabolism in the liver of females. Sex differences have also been reported for MK-801 with regard to morphine- and stress-induced analgesia (Lipa and Kavaliers, 1990), MK-801-induced behavior in adult rats (Criswell et al., 1993), MK-801-induced reduction of prolactin in plasma (Wagner et al., 1993), and MK-801-induced neurodegenerative changes in corticolimbic regions of the rat brain (Olney and Farber, 1995). The present study adds to this by showing the marked differences in MK-801-induced behavior between male and female rats.
Previous studies have shown that the rise in serum and brain concentrations is almost linear in doses up to 4 mg/kg of MK-801 (Vezzani et al., 1989). The concentration of MK-801 is maximal in serum and brain 10 and 30 min after i.p. administration, respectively (Vezzani et al., 1989; Schwartz and Wasterlain, 1993). The brain acts as a sink for MK-801 and from 30 min after injection brain concentrations are around 10-fold higher than in serum (Vezzani et al., 1989; Schwartz and Wasterlain, 1993). The elimination of MK-801 in brain parallels that in serum, with aT1/2 of 2 h. In the present study, serum and brain were sampled consecutively after the behavioral experiment, i.e., 3 to 5 h after the i.p. administration of MK-801, when the serum and brain region concentrations were stable enough to allow comparison of MK-801 concentrations in serum and brain with the degree of behavioral activation in the same animal.
The six evaluated brain regions displayed significant differences with regard to MK-801 accumulation, to a greater extent than what was reported by Vezzani and coworkers (1989). The differences were almost identical in both male and female rats and regardless of MK-801 dosage. The finding may be explained by the observed regional differences in binding of NMDA receptor blockers (Bresink et al., 1995; Porter and Greenamyre, 1995), NMDA receptor subunit mRNA distribution (Watanabe et al., 1993), or other factors such as different gray/white matter ratios. The differential accumulation of MK-801 should be studied further as it may have major implications for the pharmacological effects of NMDA receptor antagonists within the brain.
EAAs participate in the development of the central nervous system, and markers for EAA neurotransmission, such as the NMDA receptor agonist site, are transiently overexpressed during early development, most often around 20 days of age in rats (McDonald and Johnston, 1990). We are not aware of any previous report on MK-801-induced behavior at different ages in the rat. However, the MK-801-induced neuropathological changes described by Olney and coworkers (see Olney and Farber, 1995) are not evident until days 30 to 40, i.e., until the onset of puberty. In our study, rats of both sexes and of 10 to 80 days of age were given 0.2 mg/kg of MK-801 for comparison. Both female and male rats showed massive ataxia at 10 days of age, which made behavioral evaluation difficult. To our surprise, there were no differences between females and males at 20 days of age with regard to the three types of behavior. Male rats showed pronounced locomotor activity at 20 days after 0.2 mg/kg of MK-801, a degree of locomotion that was never observed in postpubertal male rats at any dose. Thus, maximal locomotion in male rats coincided with the period when NMDA receptor density is highest (McDonald and Johnston, 1990). At all postpubertal ages (40, 60, and 80 days), females displayed 3- to 5-fold more locomotion and stereotypy than males, whereas ataxia was almost absent in males. The much lower degree of MK-801-induced behavioral activation in males postpubertally correlated to low MK-801 serum and brain concentrations as compared to female rats, which is most likely due to developmental changes in the MK-801-metabolizing capacity of the liver. In addition, MK-801-induced behavior may be influenced by androgen modulation of neurotransmission (Kus et al., 1995).
Neuroleptics.
A model for central glutamate and dopamine interactions has been suggested based on a cortico-striato-thalamo-cortical feedback loop (Carlsson and Carlsson, 1990). Briefly, the thalamus, which is controlled by the striatum, serves as a filter for sensory inputs to the cortex. Activation of the dopaminergic input, or reduced activation of the glutamatergic input, to the striatum will open the filter and cause increased wakefulness and locomotion or even psychotic symptoms. Thus, behavior due to opening of the thalamic filter by NMDA antagonists may be counteracted by dopamine antagonists due to interactions in the striatum.
Several studies have shown that neuroleptics with dopamine antagonistic properties reduce NMDA antagonist-induced behavior (Sturgeon et al., 1981; Freed et al., 1984; Tiedtke et al., 1990; Hoffman, 1992), although cataleptogenic doses are often required (Sturgeon et al., 1981; Ögren and Goldstein, 1994). For example, HAL and CLOZ potently inhibit MK-801-induced locomotion and stereotypies (Tiedtke et al., 1990; Hoffman, 1992). In this study, we showed that three classical (CHLOR, HAL, PERPH) and three more novel (RISP, CLOZ, REM) neuroleptics inhibited MK-801-induced locomotion in i.p. doses that correlated (r = 0.906, p < .05) to the doses that are used clinically in Sweden for the treatment of psychosis. Except for REM (and to a minor extent RISP), which only marginally reduced stereotypy, neuroleptics blocked locomotion, stereotyped sniffing, and ataxia to the same extent. Lack of effect on stereotypy has been shown for the atypical neuroleptics CLOZ (Hoffman, 1992) and REM (Ögren and Goldstein, 1994) and may correlate to the lower degree of extrapyramidal side effects in humans (Hoffman, 1992). However, in our hands CLOZ did not show differential effects on locomotion and stereotypy. The reason for this is unclear but speculatively it may be dose-related as e.g., Hoffman (1992) found differential effects at intermediate doses only.
Adenosine.
In 1982, Browne and Welch found that the discriminative properties of phencyclidine were antagonized by adenosine analogs, and they suggested adenosinergic drugs in the treatment of phencyclidine-induced psychosis. The present study supports the findings of Browne and Welch (1982) as MK-801-induced behavior was inhibited by adenosine receptor agonists (NECA, CHA,R-PIA, and CGS-21680). The more selective adenosine A1 agonists, CHA and R-PIA, appeared to be somewhat more efficient than A2 active agonists. Thus, neuroleptics and adenosine agonists have the same ability to block MK-801-induced behavior. These two groups of substances share, in fact, many behavioral properties, e.g., adenosine agonists reduce amphetamine-induced hyperactivity (Heffner et al., 1989).
The adenosine A1 and A2 antagonist, THEO, increased MK-801-induced locomotion and stereotypy. However, similar behavior was observed in rats treated with THEO only, in agreement with the psychostimulant nature of methylxanthines (White et al., 1976; Fuxe et al., 1993). Caffeine, another methylxanthine, has been reported to worsen psychopathology in schizophrenic patients (Lucas et al., 1990). In addition, acute psychotic reactions have been described after high-dose administration of THEO to patients with no previous psychiatric diagnosis (Mansheim, 1989). The stimulatory effect on locomotion seems to be A2 receptor-mediated as the A2 antagonist, DMPX, but not the A1 antagonist, DPCPX, caused behavioral activation. Thus, adenosine antagonist-induced behavior in rats may offer another alternative experimental model of psychosis, additional to the models with dopamine agonists and NMDA antagonists.
We can only speculate upon the mechanism behind the present putative antipsychotic action of adenosine agonists in this NMDA antagonist model. The adenosine neuromodulatory system is inhibitory in action via A1 receptors, with the ability to decrease the release of various transmitters, e.g., EAAs (Fredholm and Dunwiddie, 1988) and thereby decrease NMDA receptor-mediated postsynaptic calcium currents (Schubert, 1988). Both a postsynaptic inhibition of NMDA receptor activation and a reduced release of EAAs would reduce NMDA receptor-operated channel opening. This may, in turn, cause less MK-801 binding within the channel due to the use-dependence of noncompetitive NMDA antagonists (Kemp et al., 1991). However, the relevance of use dependence in vivo is unclear (Davies et al., 1988; see below). Other mechanisms may also be valid. For example, A2 receptor agonists have been suggested to have antipsychotic actions (Fuxe et al., 1993) via inhibited dopamine receptor activation (Ferré et al., 1991).
EAAs.
The psychotomimetic effects of the NMDA receptor antagonist phencyclidine (Snyder, 1980; Javitt and Zukin, 1991; Rosse et al., 1994) have been verified by both noncompetitive antagonists, such as ketamine (Krystal et al., 1994), dextrorphan (Albers et al., 1995), and MK-801 (Troupin et al., 1986), and competitive NMDA antagonists (Kristensen et al., 1992; Grotta et al., 1995). The finding that both competitive and noncompetitive NMDA antagonists are psychotomimetics is interesting from a pathophysiological point of view. The nature of the proposed EAA dysfunction in psychosis is as yet unknown, but it is generally believed to be an EAA hypofunction. Thus, a pharmacotherapy capable of increasing EAA activity has been suggested to be antipsychotic. Therefore, the NMDA receptor-associated glycine site agonists, glycine and DCS, have been tried as adjuvant therapy to conventional neuroleptic treatment in chronic schizophrenic patients in several studies but the results have differed. Glycine has been reported both to be antipsychotic (Waziri, 1988; Javitt et al., 1994) and to have no effect or even worsen psychosis (Rosse et al., 1989). Adjuvant therapy with DCS caused clinical improvement in one study (Goff et al., 1995) and aggravation of psychotic symptoms in another (Cascella et al., 1994). In addition, the administration of high-dose DCS to humans, without ongoing neuroleptic treatment, induces a range of neuropsychiatric symptoms including psychotic reactions (Simeon et al., 1970). In experimental studies, glycine site agonists have also been reported to both antagonize (Toth and Lajtha, 1986; Contreras, 1990; Javitt et al., 1997) and enhance (Kretschmer et al., 1992) NMDA antagonist-induced behavior in rodents. In our hands, the glycine site agonist DCS potentiated MK-801-induced behavior. In addition, the glycine site antagonist HA-966 blocked MK-801-induced behavior, which has been described previously in both intact (Bristow et al., 1993) and monoamine-depleted (Carlsson et al., 1994) rodents. Thus, although the literature is inconsistent, data suggest that agents active at the glycine site may be considered as putative antipsychotics and should be tested clinically.
The mechanism behind the minor potentiation of MK-801-induced locomotion and stereotypy caused by IFEN is unclear but may be related to a direct effect on the NMDA receptor complex. IFEN is usually classified as an antagonist at the polyamine stimulatory site of the NMDA receptor. However, the effect of IFEN on MK-801 binding inside the channel is complicated by the fact that IFEN is also an agonist at a polyamine inhibitory site (Marvizón and Baudry, 1994). The effect of IFEN on MK-801 binding is thus dose-dependent but mainly inhibitory (Marvizón and Baudry, 1994). IFEN caused no motor activation per se in our study but has been shown to induce circling behavior when administered i.c.v. (Murata and Kawasaki, 1993).
Three agents related to AMPA/kainate receptors were tested; NBQX, GYKI, and cyclothiazide. Blockade of AMPA receptor desensitization by cyclothiazide (Larson et al., 1994) did not influence MK-801-induced behavior. The two AMPA/kainate antagonists NBQX and GYKI showed opposite effects, with GYKI potently enhancing and NBQX reducing (although not significantly) MK-801-induced behavior. This may be related to the competitive and noncompetitive nature of GYKI and NBQX antagonism at AMPA/kainate receptors, or to the fact that NBQX reduces spontaneous locomotion in rats more than GYKI (Danysz et al., 1994).
The drug acivicin markedly potentiated MK-801-induced locomotion. The potentiating effect of acivicin was, however, only observed in male rats, which may be due to the more pronounced behavioral activation in females, making it difficult to detect any potentiation. Acivicin was introduced as a glutamine antagonist and has been used as an antitumor agent in the central nervous system (Taylor et al., 1991). Acivicin irreversibly inhibits the activity of γ-GT in a number of tissues including brain (Rambabu et al., 1986). This enzyme metabolizes glutathione and may participate in cellular uptake of amino acids (Meister and Tate, 1976). The doses used in the present study are likely to have produced different degrees of γ-GT inhibition during the experimental session (Rambabu et al., 1986; Peacock et al., 1994), although other mechanisms of action for acivicin cannot be ruled out. It has been suggested that γ-GT may participate in glutamate uptake, and in the hippocampal slice model acivicin influences the extracellular concentrations of various amino acids and dipeptides (Li et al., 1996) that may potentiate glutamate-induced NMDA receptor activation, and acivicin may thus mediate activation of NMDA receptor-operated channels. Acivicin has been reported to cause sedation, ataxia, hallucinations, and personality changes in humans (Taylor et al., 1991).
A number of neuroactive compounds not directly active at the EAA system have been reported to enhance MK-801-induced behavior in rodents. For instance, in monoamine-depleted mice, MK-801-induced locomotion is increased by administration of the α adrenergic agonist, clonidine, and the cholinergic antagonist, atropine (Carlsson and Carlsson, 1989). A similar enhancement of MK-801-induced locomotion is produced by agents that increase extracellular catecholamines (amphetamine and cocaine), an opioid receptor agonist (morphine), the muscarinic antagonist scopolamine, and the adenosine antagonist caffeine (Kuribara et al., 1992). All these classes of compounds can cause disorientation and hallucinations as side effects when given to humans. To our knowledge, it is not known if acivicin modulates any of these neurotransmitter systems. Whether acivicin interacts with the metabolism of MK-801 and influences serum and brain concentrations of MK-801 remains to be studied.
Finally, the benzodiazepine DIAZ blocked MK-801-induced behavior, but a high dose that also reduced muscle tone was needed. This is in agreement with the finding that benzodiazepines block the neuropathological changes induced by MK-801 in rodents (Olney and Farber, 1995) and appear to suppress the psychotomimetic effects of ketamine in humans (Korttila and Levänen, 1978).
Modulation of MK-801-Induced Behavior and Clinical Implications.
As we have discussed, the type of EAA dysfunction involved in schizophrenia and other types of psychoses is unknown. Speculatively, the model of MK-801-induced behavior in rats represents a model where the EAA dysfunction in psychosis would be due to a malfunctioning of NMDA receptors. In vitro, MK-801 has very low affinity to the inactivated state of the receptor, but when the receptor is activated by agonists and coagonists, MK-801 binds to its site within the NMDA receptor-associated channel with high affinity (Kemp et al., 1991). Thus, factors influencing NMDA receptor activation may also influence MK-801 binding within the channel, and thereby modify MK-801-induced behavior. At least two major issues must be addressed with regard to this theory. Firstly, the significance, in vivo, of the use dependence of MK-801 binding to the NMDA channel is debated (Davies et al., 1988). However, both glutamate and glycine increase the binding of MK-801 to the NMDA receptor-associated ion channel (Ransom and Stec, 1988), and MK-801-induced behavior in rats is, in fact, potentiated by injection of l-glutamate into the nucleus accumbens (Raffa et al., 1989). Secondly, depending on whether an EAA hypofunction is mimicked by a competitive or a noncompetitive NMDA antagonist, these two types of hypofunction should theoretically be affected in opposite directions if the glutamatergic activity is increased. This has been shown in a study by Kretschmer and coworkers (1992), in which DCS potentiated the behavior in rats induced by MK-801 but reduced the behavior induced by a competitive antagonist. The behavioral differences between noncompetitive and competitive NMDA antagonists in rodents have also been suggested to be due to the phenomenon of use dependence (Carlsson, 1993).
Table 5 shows the effects of all agents in this study that were tested for effect on MK-801-induced behavior. Inhibition or reduction of MK-801-induced behavior was seen with neuroleptics, adenosine agonists, a glycine-site antagonist, and a benzodiazepine. These agents have many central effects, but they all reduce the activity at EAA receptors, either by reducing the release of EAAs or by antagonizing the NMDA receptor. In contrast, an adenosine antagonist, a glycine site agonist, and acivicin (which speculatively increases extracellular NMDA receptor-potentiating agents) all potentiated MK-801-induced behavior. IFEN does not fit this pattern. However, the effect of IFEN on behavior was small, and the pharmacology of the polyamine sites at the NMDA receptor is complex (Marvizón and Baudry, 1994). Thus, agents that increase EAA activity (i.e., increase the probability for EAA receptor-mediated neuronal activity) potentiate MK-801-induced behavior and are generally capable of inducing psychosis-like reactions in humans. Agents that reduce EAA activity also reduce MK-801-induced behavior, and the agents that thus far have been tested clinically (neuroleptics, benzodiazepines) are capable of reducing psychotic symptoms. The present study suggests that a pharmacological strategy aiming at activating NMDA receptors may, in addition to the risk of excitotoxicity, in fact exacerbate psychotic symptoms in humans. Speculatively, a new antipsychotic agent should instead reduce EAA activity, and adenosine agonists and glycine-site antagonists seem promising in this respect.
Conclusions
The aim of the present study was to characterize MK-801-induced behavior in rats as a putative model of psychosis. The following are the main findings of the study: 1) MK-801 induces reproducible behavior in rats, with locomotion, stereotyped sniffing, and ataxia, which can easily be determined by the use of rating scales; 2) female rats show a major behavioral activation after considerably lower doses of MK-801 than males, and the behavior in females is more easily rated; 3) the behavioral sex differences are probably due to lower metabolizing capacity of MK-801 in the female rat liver, which results in around 25 times higher serum and brain concentrations in females; 4) neuroleptics inhibit MK-801-induced behavior in a dose-dependent manner that correlates to their antipsychotic potency in humans; 5) adenosine receptor agonists and an NMDA receptor-associated glycine site antagonist show putative antipsychotic effects by mimicking the inhibitory effects of neuroleptics; and 6) MK-801-induced behavior represents a rat EAA hypofunction model of psychosis that appears to be of clinical relevance and may be of value in the search for new antipsychotic agents.
Acknowledgment
We thank Ulrika Hallin for expert technical assistance.
Footnotes
-
Send reprint requests to: Dr. Peter Andiné, Institute of Clinical Neuroscience, Department of Psychiatry, Sahlgrenska University Hospital, SE-413 45 Göteborg, Sweden. E-mail:peter.andine{at}sahlgrenska.se
-
↵1 This work was supported by the H. Lundbeck Psychosis Foundation, the Tore Nilsson Foundation, the Lars Hierta Foundation, the Medical Faculty of Göteborg University, the Åke Wiberg Foundation, the Swedish Society of Medicine, the Royal Society of Arts and Sciences in Göteborg, the Åhlén Foundation, the Swedish Care and Treatment of Psychoses Committee, the Adlerbertska Foundation, the Sahlgrenska University Hospital Foundations, and the Swedish Medical Research Council (11643 and 11840).
- Abbreviations:
- EAA
- excitatory amino acid
- NMDA
- N-methyl-d-aspartate
- AMPA
- α-amino-3-hydroxy-5-methyl-isoxazole-4-proprionic acid
- MK-801
- (5R,10S)-(+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine hydrogen maleate (dizocilpine maleate)
- acivicin
- (αS, 5S)-α-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid (AT-125)
- GYKI
- 1-(4-aminophenyl)-4-methyl-7,8-methylenedioxy-5H-2,3-benzodiazepine hydrochloride (GYKI 52466 hydrochloride)
- NBQX
- 1,2,3,4-tetrahydro-6-nitro-2,3-dioxo-benzo[f]quinoxaline-7-sulfonamide disodium
- DIAZ
- diazepam
- DCS
- d-cycloserine
- HA-966
- R(+)-3-amino-1-hydroxy-2-pyrrolidinone
- IFEN
- α-(4-hydroxyphenyl)-β-(4-benzylpiperidin-1-yl) β-methylethanol tartrate (ifenprodil tartrate)
- RISP
- risperidone
- PERPH
- perphenazine
- CLOZ
- clozapine
- CHLOR
- chlorpromazine
- HAL
- haloperidol
- REM
- remoxipride
- NECA
- 1-(6-amino-9H-purin-9-yl)-1-deoxy-N-ethyl-β-l-ribofuranuronamide (5′-N-ethylcarboxamido adenosine)
- THEO
- theophylline
- R-PIA
- R(−)N6-(2-phenylisopropyl)adenosine
- DPCPX
- 8-cyclopentyl-1,3-dipropylxanthine
- CHA
- N6-cyclohexyladenosine
- DMPX
- 3,7-dimethyl-1-propargylxanthine
- CGS-21680
- 2-p-(2-carboxyethyl)phenethylamino-5′-N-ethylcarboxamido adenosine hydrochloride
- cyclothiazide
- 3-bicyclo[2.2.1]hept-5-en-2-yl-6-chloro-3,4-dihydro-2H-1,2,4-benzothiadiazine-7-sulfonamide 1,1-dioxide
- PC
- parietal cortex
- FC
- frontal cortex
- HY
- hypothalamus
- SP
- striatum posterior
- SA
- striatum anterior
- HI
- hippocampus
- γ-GT
- γ-glutamyltransferase
- Received January 22, 1999.
- Accepted April 11, 1999.
- The American Society for Pharmacology and Experimental Therapeutics