Abstract
Multiple studies indicate that N-methyl-d-aspartate (NMDA) receptor hypofunction underlies some of the deficits associated with schizophrenia. One approach for improving NMDA receptor function is to enhance occupancy of the glycine modulatory site on the NMDA receptor by increasing the availability of the endogenous coagonists d-serine. Here, we characterized a novel d-amino acid oxidase (DAAO) inhibitor, compound 8 [4H-thieno [3,2-b]pyrrole-5-carboxylic acid] and compared it with d-serine. Compound 8 is a moderately potent inhibitor of human (IC50, 145 nM) and rat (IC50, 114 nM) DAAO in vitro. In rats, compound 8 (200 mg/kg) decreased kidney DAAO activity by ∼96% and brain DAAO activity by ∼80%. This marked decrease in DAAO activity resulted in a significant (p < 0.001) elevation in both plasma (220% of control) and cerebrospinal fluid (CSF; 175% of control) d-serine concentration. However, compound 8 failed to significantly influence amphetamine-induced psychomotor activity, nucleus accumbens dopamine release, or an MK-801 (dizocilpine maleate)-induced deficit in novel object recognition in rats. In contrast, high doses of d-serine attenuated both amphetamine-induced psychomotor activity and dopamine release and also improved performance in novel object recognition. Behaviorally efficacious doses of d-serine (1280 mg/kg) increased CSF levels of d-serine 40-fold above that achieved by the maximal dose of compound 8. These findings demonstrate that pharmacological inhibition of DAAO significantly increases d-serine concentration in the periphery and central nervous system. However, acute inhibition of DAAO appears not to be sufficient to increase d-serine to concentrations required to produce antipsychotic and cognitive enhancing effects similar to those observed after administration of high doses of exogenous d-serine.
Schizophrenia has traditionally been associated with hyperfunction of the mesolimbic dopamine system, primarily because existing antipsychotics produce their effects by antagonizing the dopamine D2 receptor (Kapur and Mamo, 2003). However, accumulating evidence suggests that the glutamate system is also important. In particular, it is believed that N-methyl-d-aspartic acid (NMDA) receptor hypofunction underlies at least some of the behavioral and neurobiological deficits observed in the disease. In support of this idea, NMDA receptor antagonists such as ketamine and phencyclidine produced psychotomimetic symptoms and cognitive deficits in normal subjects, which closely resembled those found in patients with schizophrenia (Javitt, 1987; Krystal et al., 1994; Lahti et al., 2001). Furthermore, NMDA receptor antagonists reinstated schizophrenia-like symptoms in remitted patients and worsened psychosis in antipsychotic-free patients (Lahti et al., 2001).
Because of the apparent relationship between decreased function of the NMDA receptor and schizophrenia, researchers have examined whether increasing NMDA receptor function ameliorates the behavioral and neurobiological symptoms associated with the disease. One approach for enhancing NMDA receptor function is to increase occupancy of the NMDA/glycine regulatory site. The NMDA receptor complex is unique in that it requires the presence of both agonist (glutamate) and coagonist (glycine, d-alanine, and/or d-serine) for the ligand-gated ion channel to open (Johnson and Ascher, 1987; McBain et al., 1989; Kemp and Leeson, 1993; Mothet et al., 2000). Notably, that d-serine and other endogenous coagonists of the NMDA receptor are decreased in both the serum and cerebrospinal fluid (CSF) of schizophrenics (Hashimoto et al., 2003, 2005). Furthermore, several small clinical trials showed that coadministration of atypical antipsychotics with glycine, d-serine, or d-alanine improved positive, negative, and cognitive symptoms, relative to when these antipsychotics were given alone (Heresco-Levy et al., 1996, 1999, 2002; Tsai et al., 1998, 2006). Recent studies have also demonstrated that the glycine transporter (GlyT1) inhibitor sarcosine, which presumably increased extracellular glycine by preventing reuptake, improved psychosis and cognitive deficits in schizophrenic patients (Tsai et al., 2004; Lane et al., 2008).
Another strategy for indirectly increasing NMDA receptor function is by preventing the metabolic breakdown of d-serine by the enzyme d-amino acid oxidase (DAAO) (Konno and Yasumura, 1983, 1992). This approach has several potential advantages over d-serine administration, including decreasing the likelihood of d-serine to induce nephrotoxicity, which is dependent on DAAO-mediated d-serine metabolism (Krug et al., 2007). Mice lacking DAAO activity show elevated brain levels of d-serine, enhanced hippocampal long-term potentiation, and improved spatial memory on the Morris water maze (Maekawa et al., 2005; Almond et al., 2006). Furthermore, genetic studies reveal an association between schizophrenia and single-nucleotide polymorphisms of the gene coding for DAAO (Chumakov et al., 2002; Schumacher et al., 2004), suggesting that activity of this enzyme is involved in at least some of the symptoms observed in schizophrenia. Here, we compared the effects of a novel DAAO inhibitor, compound 8 (Fig. 1) (Sparey et al., 2008), with d-serine on their ability to elevate d-serine in the plasma, CSF, and cortex. We also determined their influence on amphetamine-induced psychomotor activity, dopamine efflux in the nucleus accumbens shell, and novel object recognition.
Materials and Methods
Animals. All experimental protocols described in this study were approved by the Merck and Co., Inc. Institutional Animal Care and Use Committee and conducted in accordance with the Guide for Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, 1996). Adult male Wistar rats were used in all in vivo experiments and animal body weights are listed separately for each experiment. Animals were maintained on a 12-h light/dark cycle (lights on 7:00 AM), and temperature and relative humidity were maintained at 22 to 24°C and 50 to 55%, respectively. All experiments were conducted using a randomized design.
Drugs. Compound 8 was synthesized in house as previously reported (Sparey et al., 2008). d-Serine, MK-801, and d-amphetamine were purchased from Sigma-Aldrich (St. Louis, MO). Compound 8 was dissolved in sterile water, whereas d-serine, MK-801, and d-amphetamine were administered in saline. Routes of administration and drug concentration are indicated separately for each experiment. Compound 8 was dosed as a sodium salt, and all other compounds were dosed as HCl salts.
In Vitro DAAO Activity Assay. Cell-based enzyme activity was evaluated essentially as described by Brandish et al. (2006). In brief, enzyme activity was evaluated in Chinese hamster ovary (CHO) cells transiently transfected with human or rat DAAO in pcDNA 3.1+. Enzyme activity was determined by monitoring hydrogen peroxide production, a by-product of the DAAO reaction, using Amplex Red reagent (Invitrogen, Carlsbad, CA) together with horseradish peroxidase to produce the red fluorescent product resorufin. Fluorescence was detected at an excitation of 544 nm and an emission of 590 nm with a Tecan Safire2.
Evaluation of ex Vivo DAAO Inhibition, Pharmacokinetic Parameters, and Plasma d-Serine Levels after Compound 8 Administration. The effects of compound 8 (10-200 mg/kg i.p. in a volume of 2 ml/kg) on DAAO activity and plasma d-serine levels were evaluated in male Wistar rats (220-250 g). Animals were decapitated 1, 4, 6, and 8 h after drug administration, then CSF, blood, kidneys, cerebellum, and cortex were collected for analysis. Brain tissues and plasma samples were collected to analyze the pharmacokinetic properties of compound 8 and to determine plasma and tissue concentration of d-serine by HPLC-ECD following the procedures detailed below. Kidneys and cerebellum were flash frozen in liquid nitrogen for analysis of DAAO enzyme activity using the methods listed below.
Ex Vivo DAAO Activity Assay. Determination of DAAO activity was completed according to methods adapted from D'Aniello et al. (1993) utilizing the Pyruvate Assay Kit (BioVision, Mountain View, CA) to detect α-keto acid production. Rat kidneys or cerebellum were homogenized over ice with a Polytron homogenizer in Pyruvate Assay Buffer (0.25g tissue/ml) containing protease inhibitors (Complete, EDTA-free; Roche Diagnostics, Indianapolis, IN). Homogenate was clarified by centrifugation at 30,000g for 30 min at 4°C, and 100 μl of supernatant was mixed with 100 μl of 0.1 M d-alanine (in 0.1 M Tris-HCl, pH 8.2) or with 100 μl of 0.1 M Tris-HCl, pH 8.2, and then incubated at 37°C for 30 min. Pyruvate produced from the oxidation of d-alanine by DAAO was measured using the Pyruvate Assay Kit from BioVison. Fifty microliters of the d-alanine conversion reaction was added to 50 μl of Pyruvate Detection Mix and incubated at room temperature for 30 min. A standard curve covering a range of 10 to 0.1 nmol/well was created as a control. Absorbance was measured at 570 nm, and results were calculated relative to standard curve. Pyruvate production was normalized to total protein as determined using the Coomassie Plus Protein Assay kit (Pierce Chemical, Rockford, IL).
Evaluation of d-Serine Concentration by HPLC-ECD. Plasma, brain, and CSF concentrations of d-serine were evaluated in male Wistar rats (220-250 g). Animals were injected with compound 8 (10-200 mg/kg i.p.) or d-serine (3-1280 mg/kg s.c.), and CSF was sampled from the cisterna magna 1, 4, 6, or 8 h after injection. Immediately after CSF sampling, animals were decapitated, and the prefrontal cortex was dissected and stored at -80°C until HPLC analysis.
d-Serine, glycine, and d-alanine were simultaneously measured in plasma, CSF, and brain tissue essentially using the method of Hashimoto et al. (1992). Samples were deproteinized by adding 3 volumes of methanol. Precolumn derivatization was then carried out by ali-quoting 60 μl of derivatizing reagent (1 mg of o-phthaldialdehyde and 2 mg of N-isobutloxycarbonyl-l-cysteine in 0.1 ml of methanol and 0.9 ml of 0.1 M sodium borate buffer, pH 10) in 60 μl of sample, standard, or blank solution in the sample vial. The solution was then mixed thoroughly and allowed to react for 5 min at room temperature. The derivatization mixture (20 μl) was then injected into the HPLC using a model 542 autosampler (ESA Inc., Chelmsford, MA). Separation occurred on a C18 reversed phase column (Symmetry, 4.6 × 150 mm, 3.5 μm; Waters, Milford, MA), and analytes were detected by an Coulochem III controller (ESA Inc.) using a model 5011A microdialysis analytical cell (E1 = +400 mV, E2 = +600 mV). A gradient program of two mobile phases was used with mobile phase A consisting of 0.1 M sodium phosphate, pH 6.75, and 15% (v/v) methanol and B consisting of 0.1 M sodium phosphate, pH 6.75, with 44% methanol and 2.4% tetrahydrofuran. The flow rate of mobile phase was 0.8 ml/min. l-Homoserine was added to samples as an internal standard.
Amphetamine-Induced Psychomotor Activity. Two groups of male Wistar rats (n = 39 and 40) weighing 225 to 250 g were received on site and housed 10 per cage for 7 days with food and water available ad libitum. One group of rats was used to examine the influence of d-serine on amphetamine-induced psychomotor activity, and the other group was used to test the effects of compound 8.
To examine the influence of d-serine, animals were given either vehicle (saline) or d-serine (320, 640, or 1280 mg/kg; 5 ml/kg s.c.) 30 min before being placed in locomotor activity monitors (43.2 × 43.2 cm; MED Associates, St. Albans, VT; each test cage was housed in a sound-attenuating chamber). To characterize the effects of DAAO inhibition, animals were given vehicle (water) or compound 8 (100 or 200 mg/kg i.p. in a volume of 2 ml/kg) 5 h 30 min before being placed in locomotor activity monitors. This timepoint was chosen because the ability of compound 8 to increase CSF d-serine was maximal from 6 to 8 h. After being placed in the locomotor activity monitors, all animals were given 30 min to habituate. After habituation, animals were given a subcutaneous injection of vehicle (saline) or d-amphetamine (1.5 mg/kg) and placed back into the locomotor activity monitors for an additional 90 min. Total distance traveled was used to assess the influence of d-serine and compound 8 during habituation and after amphetamine treatment.
Novel Object Recognition. Two groups of male Wistar rats weighing 225 to 250 g were housed in identical conditions as the experiment described above. One group of rats was used to examine the effects of d-serine on novel object recognition, and the other group was used to examine the effects of compound 8 (n = 80 for both experiments).
On the day before the first exposure, animals were brought to the dimly lit test room and allowed to habituate to the room for 30 min. After habituation, animals were placed in the test cages for 90 s with no objects in the cages. After 90 s had elapsed, animals were put back into their home cages.
On the following day, animals were given vehicle, d-serine (320, 640, or 1280 mg/kg; 5 ml/kg s.c.) or compound 8 (100 or 200 mg/kg; 2 ml/kg i.p.) and placed back into their home cages. Thirty minutes or 5.5 h later, animals were brought to the dimly lit room, given an intraperitoneal injection of MK-801 (0.3 mg/kg), and allowed to habituate to the test room for 30 min. After habituation, animals were put in the test cages, and two identical objects (sample objects) were placed on opposite ends of the cage. Animals were allowed to freely explore these objects for 90 s and then were brought back to their home cages.
Twenty-four hours after the initial exposure animals were brought back to the testing room and allowed to habituate for 30 min. After habituation, two objects were placed at opposite ends of the test cage, one being identical to the object the animals had previously been exposed to (familiar object; cleaned between first and second exposures) and the other being an object they had never been exposed to (novel object). The objects serving as familiar and novel were counterbalanced across treatment conditions. Animals were allowed to freely explore the objects for 2 min before being returned to their home cage.
Each animal's behavior was video recorded (WV-CP240EX; Panasonic Corporation of North America, Secaucus, NJ), and analysis of the videos was performed by a scorer blind to the treatment conditions. The animal was considered to be exploring an object when its nose was directed toward the object and was less than 1 cm away from the object. Animals were excluded if they did not explore both objects during the first exposure or if exploration of both objects during the second exposure was less than 6 s. Recognition was evidenced if the animal explored the novel object to a greater extent than the familiar object. In addition, Clever Sys., Inc (Reston, VA) Top Scan was used to quantify locomotor activity during the first and second exposures.
In Vivo Microdialysis. Adult male Wistar rats (250-300 g) were implanted unilaterally with a microdialysis probe guide under ketamine/xylazine anesthesia in the nucleus accumbens (NAc) shell (anteroposterior, +1.7 mm; mediolateral, +1.0 mm; dorsoventral, -6.0 mm; relative to bregma), and the probe guide was anchored to the skull with bone screws and dental acrylic. Rats were allowed to recover for at least 2 days before experiments.
Microdialysis experiments were performed in freely moving rats tethered to a liquid swivel system (CMA 120; CMA, Chelmsford, MA). Concentric dialysis probes (CMA12, 2-mm dialysis tip) were inserted into guides and connected to syringe pumps with FEP tubing. Dialysis probes were perfused using artificial cerebral spinal fluid (145 mM NaCl, 2.7 mM KCl, 1.0 mM MgCl2, and CaCl2, pH 7.4) at a flow rate of 1 μl/min. Baseline levels of analyte were allowed to stabilize for 2 h before sample collection. Fractions were collected every 20 min by hand and stored on ice until HPLC-ECD analysis. Probe placement was confirmed by infusing cresyl violet solution into the brain via the dialysis probes. Animals with probes not correctly positioned in the NAc shell were excluded from data analysis.
Concentrations of monoamines were determined by HPLC coupled to electrochemical detection. Dialysate samples (20-μl volume) were injected into the HPLC using a model 542 autosampler (ESA Inc.) and separated on a C18 reversed phase column (MD150; ESA Inc.). Monoamines were detected by an ESA Coulochem III controller using a model 5014B microdialysis analytical cell (E1, -150 mV; E2, +220 mV). A commercially available mobile phase MD-TM (ESA Inc.) was used at a rate of 0.4 ml/min under isocratic conditions, and 3,4 dihydroxybenylamine was added to samples as an internal standard.
Statistics. The effect of compound 8 on d-serine concentration was analyzed using a two-way ANOVA to compare time and treatment effects followed by post hoc tests using a Bonferroni correction factor for individual group comparisons. Microdialysis data were analyzed using a two-way mixed ANOVA, with time as a repeated measures factor and treatment as a between-subjects factor. Post hoc analyses were made using the Bonferroni correction factor. All other neurochemical measures were analyzed using one-way ANOVA and Tukey's multiple comparisons tests. To examine the effects of compound 8 or d-serine on amphetamine-induced psychomotor activity as a function of time, two-way mixed ANOVA with a repeated measures factor (time) and between-subjects factor (group) was used. In addition, summed data during habituation and after amphetamine treatment were analyzed with one-way ANOVA, and group differences were followed by post hoc comparisons (Fischer's protected least significant difference). To examine the influence of d-serine and compound 8 on novel object recognition, one-way ANOVA was used to examine whether there were between-group differences in percentage exploration of the novel object, relative to the familiar object [(novel object exploration/(novel object + familiar object exploration) × 100)], and group differences were followed by post hoc comparisons (Fischer's protected least significant difference). In addition, one-sample Student's t tests were performed for each group to determine whether exploration of the novel object occurred at greater than chance levels (>50%), indicating recognition. Finally, to examine whether treatment influenced total object exploration during the first or second exposure, one-way ANOVA was performed on each of these variables.
Results
Inhibition of DAAO Activity by Compound 8 in Vitro. The ability of compound 8 to inhibit DAAO activity was examined in CHO cells transiently expressing either the human or rat DAAO gene. Compound 8 was found to inhibit human DAAO with an IC50 of 144.6 ± 16.1 nM (n = 8). A similar potency was exhibited against rat DAAO with compound 8 having an IC50 of 113.9 ± 5.5 nM (n = 8) (Fig. 2). In addition, compound 8 (30 μM) did not have measurable activity against a panel of over 150 receptors, enzymes, and ion channels (Panlabs Screen; MDS Sciex, Concord, ON, Canada). All assays in this panel were run in duplicate with semiquantitative data, such as estimated IC50 and Ki, reported for each biochemical (enzyme target) or binding assay (receptors and channels). Therefore, compound 8 has a minimum selectivity of 200-fold over other biologically relevant proteins.
Effects of Compound 8 on Kidney DAAO Activity and Peripheral d-Serine Concentrations. In rats, basal activity of DAAO in the kidney was determined to be 22.7 nmol/mg protein/h. Administration of compound 8 (10-200 mg/kg i.p.) produced both time- and dose-dependent inhibition of kidney DAAO activity (Fig. 3A). Main effects were identified for time (F3,80 = 20.14, p < 0.0001) and group (F5,80 = 308.38, p < 0.0001), and there was also a group × time interaction (F15,80 = 3.11, p < 0.0002). The greatest inhibition of kidney DAAO activity (∼96% inhibition) was observed 1 h after compound 8 administration. Significant inhibition (∼87%) of kidney DAAO activity was still observed 8 h after compound 8 administration (Fig. 3A).
Plasma d-serine levels were significantly elevated after administration of compound 8 (group, F5,80 = 22.46, p < 0.0001; time, F3,80 = 13.98, p < 0.0001; group × time interaction, F15,80 = 2.03, p = 0.02). Peak levels of plasma d-serine, 220 ± 7% of control (4.0 ± 0.3 μM), were observed 8 h after administration of 200 mg/kg compound 8 (Fig. 3B). Plasma d-serine concentrations ranged from 1.2 to 2 μMin vehicle-treated animals. Post hoc comparisons revealed significant elevations in d-serine from 4 to 8 h postinjection in groups treated with 10 to 200 mg/kg compound 8 (Fig. 3B). Administration of compound 8 did not alter circulating levels of glycine or d-alanine, amino acids with high affinity for the NMDA coagonist binding site (data not shown).
Effects of Compound 8 on Cerebellar DAAO Activity and Central d-Serine Concentration. Injection of compound 8 (10-200 mg/kg i.p.) significantly inhibited DAAO activity in the cerebellum [Fig. 4A; main effects of time (F3,80 = 18.39, p < 0.0001), group (F5,80 = 29.12 p < 0.0001), and group × time interaction (F15,80 = 2.09, p = 0.01]. The maximum inhibition of DAAO activity (∼80% inhibition) was observed 1 h after compound 8 (200 mg/kg i.p.) administration. The inhibitory effects of compound 8 on cerebellar DAAO were long-lasting, and ∼60% inhibition was observed 8 h after injection of 200 mg/kg compound 8 (Fig. 4A). Basal activity of cerebellar DAAO in these assays was 2.8 nmol/mg protein/h.
Figure 4B reveals that administration of compound 8 significantly increased CSF levels of d-serine [main effects of time (F3,80 = 16.21, p < 0.0001), group (F5,80 = 35.24 p < 0.0001), and group × time interaction (F15,80 = 3.97, p < 0.0001)]. Post hoc comparisons revealed significant elevations of CSF in d-serine concentration in groups treated with 50, 100, and 200 mg/kg compound 8. CSF d-serine levels increased to 175 ± 5% of control (4.8 ± 0.2 μM) 8 h after compound 8 administration (Fig. 4B). In contrast, compound 8 failed to influence d-serine concentration in rat cortical tissue as revealed by a lack of a main effect of treatment group (F5,80 = 1.40, p = 0.23) or time (F3,80 = 0.72, p = 0.54) (Fig. 5C). The exposure of compound 8 in the brain and plasma were also determined for all treatment groups and values are shown in Table 1. Brain concentration of compound 8 peaked at 164.6 ± 7.1 μM, 1 h after intraperitoneal administration in animals treated with 200 mg/kg of this compound. In this group, concentrations of compound 8 were steadily decreased to 47.4 ± 10.1 μM 8 h after injection.
Comparison of the Effects of Compound 8 and d-Serine on Peripheral and Central d-Serine Concentration. Both compound 8 (main effect of group, F5,80 = 22.46, p < 0.0001) and d-serine (main effect of group, F7,24 = 56.8, p < 0.0001) significantly increased circulating levels of d-serine. Injection of compound 8 (10-200 mg/kg i.p.) dose-dependently increased plasma d-serine to a level 2.2-fold greater than basal concentrations (Fig. 3B). Systemic injection of d-serine produced markedly greater increases in plasma d-serine than those induced by compound 8. Specifically, a behaviorally efficacious dose of d-serine (1280 mg/kg. s.c.) increased plasma d-serine concentration 500-fold over basal levels (Fig. 5A). Administration of 3 mg/kg d-serine produced elevations in plasma d-serine that were similar to the maximum effect of compound 8 (2.2-fold elevation above vehicle).
Figure 5B shows the effects of d-serine administration on CSF d-serine levels. d-Serine administration produced significant elevations in CSF d-serine (main effect of group, F7,24 = 109.1, p < 0.001). Similar to effects on plasma d-serine, systemic injection of d-serine produced a much larger increase in CSF d-serine (70-fold elevation) than the greatest increase achieved by compound 8 (1.75-fold elevation). The effects of d-serine and compound 8 on cortical d-serine concentration were also examined (Fig. 5C). d-Serine (320-1280 mg/kg) significantly elevated d-serine in the prefrontal cortex (main effect of group, F7,24 = 36.1, p < 0.001), whereas compound 8 did not alter d-serine concentration in the prefrontal cortex (main effect of group, F5,80 = 1.40).
The Influence of Compound 8 and d-Serine on Amphetamine-Induced Psychomotor Activity.Figure 6 reveals the influence of compound 8 on locomotor activity during habituation and after amphetamine administration. During habituation, locomotor activity decreased as a function of dose (F2,30 = 8.05, p = 0.002; Fig. 6, left) because 200 mg/kg compound 8 reduced spontaneous activity. Figure 6 (middle and right) reveals the influence of compound 8 on amphetamine-induced psychomotor activity. Amphetamine increased distance traveled in all groups, relative to vehicle, which resulted in a main effect of group (F3,27 = 8.56, p < 0.001), time (F29,783 = 9.19, p < 0.001), and a group × time interaction (F87,783 = 2.37, p < 0.001). Subsequent ANOVA comparing the vehicle-treated group with each of the other groups receiving amphetamine further indicated a significant effect of amphetamine in all groups (main effect of group and time × group, p < 0.001). Subsequent ANOVA comparing the vehicle amphetamine-treated group with each of the other groups receiving amphetamine revealed that none of the doses of compound 8 influenced the response to amphetamine (main effect of group and group × time, p > 0.17). When collapsed across the entire 90 min after amphetamine treatment, there was a main effect of group (F3,30 = 8.56, p < 0.001), which was because amphetamine increased psychomotor activity in all groups, regardless of compound 8 treatment (p > 0.012). Furthermore, compound 8 did not influence the amphetamine response when just the first 60 min were collapsed (p > 0.11).
Figure 7 reveals the influence of d-serine on locomotor activity during habituation and after amphetamine administration. d-Serine had no effect on spontaneous activity (F3,38 = 0.98, p = 0.41). However, in contrast to compound 8, which did not significantly influence the response to amphetamine, d-serine significantly reduced amphetamine-induced psychomotor activity (Fig. 7, middle and right). Thus, when all groups were included in the analysis, there was a main effect of group (F4,34 = 8.69, p < 0.001), time (F29,986 = 16.60, p < 0.001), and a group × time interaction (F116,986 = 2.43, p < 0.001). Follow-up ANOVA revealed that amphetamine increased psychomotor activity when given after saline [main effect of group (F1,13 = 42.42, p < 0.001), group × time (F29,377 = 4.89, p < 0.001)] or after any of the doses of d-serine tested (main effect of group, p < 0.002; time × group, p < 0.001). However, the highest dose of d-serine tested (1280 mg/kg) decreased amphetamine-induced psychomotor activity relative to when vehicle was given before amphetamine. The decrease in the response to amphetamine was most pronounced during the 1st h after treatment, as revealed by a significant group × time interaction (F29,406 = 2.56, p < 0.001). When distance traveled was collapsed across the 1st h after amphetamine treatment, there was a main effect of group (F4,34 = 9.33, p < 0.001), which was not only because amphetamine increased psychomotor activity in all groups (p < 0.022) but also because the highest dose of d-serine tested decreased the response to amphetamine (p < 0.009).
Effects of Compound 8 and d-Serine on Amphetamine-Evoked Dopamine Efflux. Injection of amphetamine (1.5 mg/kg s.c.) significantly increased dopamine (DA) efflux from the NAc shell over time (F13,56 = 10.61, p < 0.0001), with a maximal increase ∼500% above basal concentration occurring 40 min after administration for both vehicle- and compound 8-treated groups (Fig. 8A). Pretreatment with compound 8 prior to (6 h) amphetamine did not alter baseline DA release and failed to attenuate the marked increase of DA efflux from the NAc shell (main effect of group, F1,56 = 0.01, p = 0.9; group × time interaction, F13,56 = 0.19, p = 0.9). In contrast, systemic administration of d-serine (1280 mg/kg) given 1 h before amphetamine challenge significantly attenuated amphetamine-stimulated DA efflux (Fig. 8B). A two-way ANOVA revealed significant effects of group (F1,193 = 8.40, p < 0.001), time (F13,193 = 12.19, p < 0.001), and group × time interaction (F13,193 = 1.98, p = 0.02). Bonferroni post hoc tests indicated significant differences between d-serine- and saline-pretreated animals 20 (p < 0.05) and 40 (p < 0.01) min after amphetamine injection (Fig. 8B). Microdialysis data were reported as percentages of base-line, not corrected for in vitro recovery.
The Effects of Compound 8 and d-Serine on Novel Object Recognition.Figure 9A and Table 2 depict the influence of compound 8 on novel object recognition. There was a main effect of group (F3,66 = 7.06, p < 0.001), which was because MK-801 disrupted novel object recognition. However, compound 8 had no effect on the MK-801 impairment (p > 0.3). In addition, none of the compound 8-treated groups demonstrated recognition memory relative to chance (p > 0.7), whereas the V-V-treated group did (p < 0.001). Finally, there was no effect of group on exploration during the first exposure (F3,66 = 1.54, p = 0.21), indicating that differences in first exposure exploration did not mask an effect of treatment.
Figure 9B and Table 2 reveal the effect of d-serine on novel object recognition. There was a main effect of group on percentage exploration of the novel object (F4,68 = 6.03, p < 0.001). Post hoc tests revealed that MK-801 impaired novel object recognition, as reflected by a significant difference between the V-V and V-MK-801 groups (p < 0.001), but that 1280 mg/kg d-serine reversed this effect, as revealed by a significant difference between the V-MK-801 and 1280 mg MK-801 groups (p = 0.001). The vehicle group and the group receiving the highest dose of d-serine (1280 mg/kg) also explored the novel object at greater than chance levels (p < 0.001 and p = 0.001, respectively), but none of the other groups differed from chance. Finally, there were no differences in object exploration during the first exposure (Table 2).
Discussion
Several lines of evidence suggest that NMDA receptor hypofunction plays a primary role in the underlying pathophysiology of schizophrenia (Moghaddam, 2003; Heresco-Levy, 2005). Therefore, restoring NMDA receptor function by increasing the availability of d-serine for the NMDA/glycine modulatory site may ameliorate some of the behavioral and neurobiological deficits associated with this disorder. High doses of d-serine given to humans have been shown to improve some symptoms of schizophrenia (Tsai et al., 1998). Because of the potential for nephrotoxicity with exogenous d-serine administration (for review, see Krug et al., 2007), we have considered alternative approaches for increasing d-serine concentrations. Here, we characterized a novel DAAO inhibitor, compound 8, and compared the behavioral and neurochemical effects of this compound with d-serine.
Compound 8 was found to be an inhibitor of DAAO in both in vitro and in vivo assays. Thus, compound 8 significantly decreased kidney DAAO activity by 96% and inhibited cerebellar DAAO by approximately 80%. Significant elevations in plasma and CSF d-serine were observed from 4 to 8 h after compound 8 administration in male Wistar rats. Therefore, further behavior and neurochemical studies were performed 6 h after compound 8 administration to ensure enough time for this compound to elevate d-serine concentration.
Our observation of increased plasma d-serine is consistent with those of Hashimoto et al. (1993), who reported an elevation in plasma d-serine in mutant mice lacking DAAO activity (ddY/DAAO mice). The magnitude of the effect reported by Hashimoto et al. (1993) (5-fold elevation in plasma d-serine) appears to be greater than we observed, a difference that might be due to the duration of DAAO inhibition produced by compound 8 or compensatory changes in mutant mice. Adage et al. (2008) reported recently that administration of the DAAO inhibitor AS057278 failed to elevate plasma d-serine above their detection limit (5 μM) in Sprague-Dawley rats, further confirming that the elevations in plasma d-serine that occur from acute pharmacological inhibition of DAAO are quite modest (Adage et al., 2008).
Compound 8 significantly increased plasma and CSF d-serine; however, this compound did not significantly alter d-serine concentrations in cortical tissue. These findings are in contrast to Adage et al. (2008), who reported that administration of the DAAO inhibitor AS057278 significantly increased relative levels of d-serine in the cortex and midbrain. The discrepancy in our findings may stem from the way in which d-serine concentration in these tissues was expressed. Although we report absolute concentration of d-serine in brain regions, Adage et al. (2008) reported relative d-serine as a function of total serine, a value that can be heavily influenced by fluctuations in l-serine concentrations. Furthermore, Adage et al. (2008) did not report the magnitude and duration of in vivo DAAO inhibition in either peripheral or brain tissues produced by AS057278. Thus, differences in brain d-serine data may result from varying levels of peripheral and/or central DAAO inhibition.
In line with the observed modest effects of compound 8 on CNS d-serine levels, we report that compound 8 did not influence amphetamine-induced psychomotor activity, DA release in the NAc shell, or novel object recognition. The lack of efficacy was not due to insensitivity of these assays to engagement of the NMDA/glycine modulatory site because we found that high doses of d-serine decreased the behavioral and neurochemical responses to amphetamine and improved novel object recognition. These findings are consistent with clinical evidence showing that agonists at the NMDA/glycine site can improve both psychosis and the cognitive deficits observed in schizophrenia. The lack of efficacy of compound 8 appears to be related to the modest changes in d-serine achieved by this compound. Thus, the dose of d-serine necessary to see an effect in these assays (1280 mg/kg) increased CSF d-serine levels ∼40 times over that produced by the highest dose of compound 8 tested, a dose that maximally inhibited kidney DAAO activity and decreased brain DAAO activity by 80%. That compound 8 did not influence the behavioral response to amphetamine is consistent with the findings of Adage et al., (2008), who reported that acute treatment with the DAAO inhibitor AS057278 did not influence phencyclidine-induced hyperlocomotion. Compound 8 did not influence novel object recognition, a finding that contrasts somewhat with previous studies reporting that mice lacking DAAO activity show improved performance in the Morris water maze, enhanced hippocampal long-term potentiation, and increased NMDA receptor function (Maekawa et al., 2005; Almond et al., 2006). It is currently unclear whether these discrepancies stem from procedural differences, including the cognitive domains tested (the water maze depending heavily on the hippocampus and recognition memory on the perirhinal cortex), the species used, or pro-longed inhibition and/or compensatory changes inherent in using mutant mice as tools.
Our findings suggest that acute DAAO inhibition does not elevate basal d-serine to levels that produce behavioral effects in the assays we employed. It is worth noting that because of the slow clearance of d-serine from the rat cortex (Hashimoto, 2002), it may be possible to achieve a more robust increase in brain d-serine after chronic inhibition of DAAO using a repeated dosing paradigm. Furthermore, complete pharmacological inhibition of brain DAAO activity may produce greater neurobiological and behavioral effects than were achievable with compound 8. However, it is unknown whether complete inhibition of DAAO activity can be achieved in vivo. In addition, it seems unlikely that a slight increase in DAAO inhibition above and beyond what we observed (96% in kidney and 80% in brain) would produce an accumulation of d-serine to the levels observed in animals treated with behaviorally efficacious doses of d-serine.
In contrast to DAAO, other mechanisms that regulate glycine site function might serve as more appropriate targets for drugs aimed at treating schizophrenia. For example, GlyT1 inhibitors, which increase engagement of the NMDA/glycine modulatory site by increasing CNS glycine levels, reportedly have clinical efficacy (Tsai et al., 2004), attenuate amphetamine-induced psychomotor activity, and improve certain aspects of cognition (Kinney et al., 2003; Depoortère et al., 2005). We speculate that the differences in efficacy between these mechanisms may stem from dissimilarities in CNS localization of GlyT1 and DAAO. GlyT1 is heavily expressed in forebrain regions rich in NMDA receptors (Smith et al., 1992), whereas DAAO is expressed primarily in the hindbrain and cerebellum (Verrall et al., 2007). Thus, the high density of GlyT1 in the cortex and the close proximity of GlyT1 to the synapse (Cubelos et al., 2005) may produce a greater increase in NMDA receptor transmission resulting from GlyT1 inhibition, relative to DAAO inhibition. In addition, extracellular levels of d-serine are also regulated by reuptake through the alanine-serine-cysteine (Asc-1) transporter (Rutter et al., 2007). Thus, Asc-1 activity may lessen the influence of DAAO inhibition in the brain by reducing d-serine availability to NMDA receptors.
In summary, our results suggest that increasing d-serine concentrations in the CNS, and presumably increasing NMDA receptor function, produces both antipsychotic and cognitive enhancing effects. In this respect, our findings support the theory that reversing NMDA receptor hypofunction has utility for treating schizophrenia. However, the acute pharmacological inhibition of DAAO produced by compound 8 might not be a feasible mechanism for increasing basal d-serine to concentrations that produce antipsychotic-like effects or improve recognition memory. We speculate that the concentration of d-serine required for enhancing activity of NMDA receptors and having therapeutic utility is much greater than can be achieved by DAAO inhibition alone.
Footnotes
-
This work was supported by Merck and Co., Inc.
-
S.M.S. and J.M.U. contributed equally to this work.
-
doi:10.1124/jpet.108.147884.
-
ABBREVIATIONS: NMDA, N-methyl-d-aspartate; CSF, cerebrospinal fluid; GlyT1, glycine transporter 1; DAAO, d-amino acid oxidase; compound 8, 4H-thieno [3,2-b] pyrrole-5-carboxylic acid; MK-801, dizocilpine maleate; CHO, Chinese hamster ovary; HPLC, high-performance liquid chromatography; NAc, nucleus accumbens; ANOVA, analysis of variance; DA, dopamine; CNS, central nervous system; ECD, electrochemical detection; AS057278, 5-methylpyrazole-3-carboxylic acid.
- Received October 23, 2008.
- Accepted December 15, 2008.
- The American Society for Pharmacology and Experimental Therapeutics