Makoto Bannai, PhD, Institute of Life Sciences, Ajinomoto Co., Inc., 1-1 Suzuki-cho, Kawasaki-ku, Kawasaki-shi, Kanagawa 210-8681, Japan. Email: firstname.lastname@example.org
Aim: Glycine, one of the non-essential amino acids, has been reported to be effective in reducing negative symptoms of schizophrenia. Recently, we found that glycine improves subjective sleep quality in humans. The aim of this study was to investigate the effects of oral glycine administration on endogenous 5-hydroxytryptamine (5-HT) and dopamine in the prefrontal cortex (PFC) of living rats.
Methods: Microdialysis probes were inserted stereotaxically into the rat prefrontal cortex. Cortical levels of 5-HT and dopamine were measured following oral administration of 1 or 2 g/kg glycine, 2 g/kg d-serine, or 2 g/kg L-serine.
Results: Both glycine and d-serine significantly increased extracellular 5-HT levels for 10 min, whereas dopamine levels remained unchanged. L-serine, in contrast, had no significant effects on 5-HT levels.
Conclusions: It is possible that the increase in 5-HT in response to glycine and d-serine was mediated by N-methyl-D-aspartate receptors. The transient increase in 5-HT in the PFC might be associated with the alleviation of negative symptoms in patients with schizophrenia and the amelioration of sleep quality in patients with insomnia.
GLYCINE, ONE OF the non-essential amino acids, functions as an inhibitory neurotransmitter in the central nervous system, especially in the spinal cord and brain stem. When ionotropic glycine receptors are activated, chloride enters the neuron, which causes an inhibitory postsynaptic potential. In contrast to its inhibitory role in the spinal cord, glycine acts as a co-agonist for channel opening of the N-methyl-D-aspartate (NMDA) subtype of ionotropic glutamate receptors, which are excitatory.1,2
We recently reported that oral administration of glycine improves sleep quality in humans, correlating with polysomnographic changes.3,4 Darracq et al.5 reported that glycine may tonically inhibit noradrenergic locus coeruleus neurons throughout the sleep–wake cycle. Not only the noradrenergic systems but also the serotonergic and dopaminergic systems have been implicated in the regulation of the sleep–wake cycle.6 However, the inhibitory or excitatory effects of oral administration of glycine on serotonergic and dopaminergic systems in the brain remain unknown.
On the other hand, oral administration of glycine or d-serine is effective in reducing negative symptoms of schizophrenia.7 The NMDA receptor hypofunction hypothesis of schizophrenia suggests that increasing NMDA receptor function via pharmacological manipulation could provide a potential new strategy for the management of schizophrenia.7–12 Negative symptoms of schizophrenia have classically been attributed to downregulation of the mesocortical dopaminergic pathways.13 In addition, serotonin (5-hydroxytryptamine; 5-HT) 2 (5-HT2) receptors in the prefrontal cortex (PFC) have been proposed to be responsible for the expression of negative symptoms in schizophrenia.14,15
The neurochemical basis for the finding that orally ingested glycine reduces negative symptoms in schizophrenia has not yet been clarified. Considering that glycine improves sleep quality and negative symptoms of schizophrenia, it would be interesting to investigate the acute effects of oral glycine administration on dopamine and 5-HT dynamics in the brain. In the present study, we investigated the effects of oral glycine administration on extracellular 5-HT and dopamine in the PFC of the rat by in vivo microdialysis in order to examine a possible part of mechanisms underlying the effectiveness of orally ingested glycine on sleep disorders and negative symptoms of schizophrenia.
Male Wistar rats (n = 73, 300–500 g) were housed individually in acrylic cages (30 × 30 × 37 h cm). Rodent chow (CRF-1, Oriental Yeast, Tokyo, Japan) and tap water were available without restriction. Distilled water (1 mL/100 g bw) was administered around noon every day by a gavage-feeding needle to acclimatize the animals to the oral administration procedure. The vivarium was maintained at 23 ± 1°C, 60% humidity, and a 12-h light/dark photocycle (lights on at 07.00 hours). The experiments were conducted in strict observance of the ‘Guide for the Care of Laboratory Animals’.16 The Institutional Animal Care and Use Committee (IACUC) of Ajinomoto approved all procedures.
Rats were anesthetized with sodium pentobarbital (50 mg/kg, i.p., Dainippon Seiyaku, Osaka, Japan) and stereotaxically implanted with a guide cannula (AG-8; Eicom, Kyoto, Japan) aimed 2 mm above the PFC. The coordinates were AP +3.0 mm, ML +0.8 mm from the bregma and DV −2.7 mm from the skull surface.17 A dummy cannula (AD-8; Eicom, Kyoto, Japan) was inserted and fixed with a cap nut (AC-1; Eicom, Kyoto, Japan) to prevent infections. Rats were given at least 7 days to recover from the surgery.
In vivo microdialysis
Rats were briefly anesthetized using isoflurane inhalation (Dainippon Seiyaku, Osaka, Japan), and an A-I-8–02 microdialysis probe (2 mm membrane length, Eicom, Kyoto, Japan) was inserted and fixed to the guide cannula with a cap nut. Tubing was connected from the probe to the swivel (TCS2-23; Tsumura, Tokyo, Japan) and syringe pump (ESP-64, Eicom, Kyoto, Japan). The swivel arm (TSB-23; Eicom, Kyoto, Japan) and dual channel swivel allowed free 360° movement of the tethered rat. The probe was perfused with perfusion fluid T1 (in mM, 147 NaCl, 4 KCl, 2.3 CaCl2; CMA microdialysis, Solna, Sweden) at a flow rate of 0.9 µL/min (in the case of d- and L-serine, 2.0 µL/min).
Nine-microliter samples were directly injected to high-speed liquid chromatography (HPLC) every 10 min through an autoinjector (EAS-20, Eicom, Kyoto, Japan). In the case of d- and L-Serine, 20-µL samples were collected every 10 min using a microfraction collector (EFC-82, Eicom, Kyoto, Japan) and 10-µL samples were injected to HPLC by an autosampling injector (M500, Eicom, Kyoto, Japan). Samples were collected for 60 min before oral administration and for 180 min after administration of drugs or vehicle. After microdialysis, the rats were deeply anesthetized with sodium pentobarbital and then intracardially perfused with 0.9% saline and 4% paraformaldehyde. To verify the position of the probe, the brains were sliced on a frozen cryostat (CM3050; Leica, Wetzlar, Germany) into 50-µm sections, and the sections were stained with cresyl violet. Data from rats in which the probe was placed outside the PFC were excluded from the analysis.
HPLC-electrochemical detector analysis
Both 5-HT and dopamine were analyzed by HPLC with an electrochemical detector (ECD) system (HITEC-500; Eicom, Kyoto, Japan). Nine-microliter samples were automatically injected into an octadecyl silane (ODS) column (PP-ODS, 4.6 mm × 30 mm; Eicom, Kyoto, Japan). Detection was performed with ECD set at 400 mV against a reference electrode. The signals were detected and analyzed using an interface (EPC-300, Eicom, Kyoto, Japan) and Power Chrom software (Ver.2.2.3; eDAQ, Australia), respectively. The mobile phase (0.1 M sodium phosphate buffer, 50 mg/L Na2EDTA, 500 mg/L sodium 1-decanesulfonate, 1% MeOH, pH 6.0) was pumped at a flow rate of 500 µL/min. The retention times of the peak of 5-HT and dopamine were approximately 4.6 min and 1.8 min, respectively.
Glycine concentration in the cerebrospinal fluid
After oral administration of 2 g/kg of glycine, rats (n = 39) were anesthetized with sodium pentobarbital and positioned in a stereotaxic frame with their heads flexed downward at 45°. Approximately 100 µL of cerebrospinal fluid was collected from the cisternae of each rat at each time-point (pre-administration, 30 min, 1 h, 2 h, 4 h, 8 h, and 24 h after administration). The glycine concentration in the cerebrospinal fluid (CSF) was measured using the protocol for measuring plasma amino acids suggested by Noguchi et al.18
Glycine and L-Serine were obtained from Ajinomoto (Kawasaki, Japan), and d-serine was obtained from Sigma-Aldrich (Poole, UK). Glycine and L- and d-serine were dissolved in distilled water. All oral administrations were performed with a gavage needle in an injection volume of 1 mL/100 g bodyweight.
Baseline levels for 5-HT and dopamine were analyzed with one-way anova, which revealed no statistical differences between the glycine and vehicle groups (described in the Results section). Thereafter, 5-HT and dopamine levels were expressed as a percentage of the baseline for each animal. Both 5-HT and dopamine levels were analyzed by a two-way repeated measures anova, followed by Tukey's post hoc test with contrasts. Glycine concentrations in the CSF were analyzed by one-way anova on ranks, and Dunn's method was performed to compare pre-administration with later time-points.
The probe was correctly positioned in five (vehicle group; one rat was excluded), five (1 g/kg glycine administered group; two rats were excluded), six (2 g/kg glycine administered group; one rat was excluded), five (2 g/kg d-serine administered group; four rats were excluded), and five (2 g/kg L-serine administered group; no rats were excluded) rats.
Effects of glycine on extracellular 5-HT concentrations in the PFC
Basal levels of 5-HT were 2.25 ± 0.23, 2.27 ± 0.34, and 1.53 ± 0.24 fmol/sample (9 µL) in the vehicle, 1 g/kg, and 2 g/kg glycine groups, respectively. One-way anova revealed no significant differences among the groups. Two-way repeated measures anova revealed that the 5-HT levels in the PFC did not change among groups (Fgroup[2, 296] = 0.896, P = 0.432), but the interaction between groups and times changed significantly (Fgroup × time[48, 296] = 1.475, P = 0.029; Fig. 1). Post hoc analysis demonstrated that 5-HT levels were significantly higher for 10 min after administration in the 1 g/kg glycine administration group (Fig. 1A), whereas 5-HT levels were significantly higher for 20–30 min after administration in the 2 g/kg group (Fig. 1B). Basal levels of 5-HT were 1.30 ± 0.10 and 1.52 ± 0.13 fmol/sample (10 µL) in the d-serine and L-serine group, respectively. Two-way repeated measures anova revealed no significant changes among groups or in the interaction between groups and times (Fgroup[1, 146] = 4.229, P = 0.074; Fgroup × time[19, 146] = 1.180, P = 0.282) due to d-serine. However, post hoc analysis showed a significant increase of 5-HT at 0–10, 60–70, and 100–110 min after administration (Fig. 2A); L-serine, in contrast, did not induce significant changes (Fgroup[1, 137] = 0.717, P = 0.421; Fgroup × time[19, 137] = 0.559, P = 0.929; Fig. 3A).
Effects of glycine on extracellular dopamine concentrations in the PFC
Basal levels of dopamine were 3.70 ± 0.32, 4.30 ± 0.30, and 3.36 ± 0.28 fmol/sample (9 uL) in the vehicle, 1 g/kg, and 2 g/kg glycine groups, respectively. Statistical analyses of dopamine baselines were not significant among the groups by one-way anova. Two-way repeated measures anova showed that dopamine levels in the PFC did not change significantly after glycine administration (Fgroup[2, 296] = 0.148, P = 0.864; Fgroup × time[48, 296] = 1.117, 0.287: Fig. 1C). Two-way repeated measures anova revealed no significant changes among groups (Fgroup[1, 144] = 0.118, P = 0.740) or in interactions between groups and times (Fgroup × time[19, 146] = 1.159, P = 0.300) based on d-serine administration. However, post hoc analysis showed d-serine significantly increased dopamine levels for 0–10 min after administration (Fig. 2A). L-serine did not yield significant changes (Fgroup[1, 137] = 0.820, P = 0.391; Fgroup × time[19, 137] = 0.894, P = 0.591; Fig. 3A).
Concentration of glycine in the CSF after oral administration
Changes in glycine levels in the CSF after glycine administration are shown in Figure 4. Tmax and Cmax were 52.72 (±4.42) µM and 30 min, respectively.
In the present study, oral administration of glycine increased the extracellular concentration of 5-HT, but not that of dopamine, in the PFC of living rats. To the best of our knowledge, this is the first report to demonstrate that glycine increases 5-HT in the PFC. Based on our hypothesis, it is possible that the effects of glycine on sleep disorders and negative symptoms of schizophrenia may be associated with an increase in 5-HT in the PFC.
In humans, oral administration of glycine increases plasma glycine levels, with the peak occurring approximately 40 min after oral administration and then gradually decreasing.19 Our previous study showed that systemic administration of 1.6 g/kg of glycine caused a significant increase in glycine in the prefrontal cortex for up to 210 min.20 As shown in Figure 4, glycine concentration was highest at 30 min and the maximum concentrations of glycine (Cmax) in the CSF of rats after oral administration of 2 g/kg of glycine were 52.72 (±4.42) µM. Considering that the ED50 of glycine to NMDA receptors is 0.2–1 µM21,22 and the ED50 of glycine to glycine receptors is 90–100 µM,23 the Cmax of glycine after oral treatment in the current study did not reach the ED50 of glycine to glycine receptors. These findings suggest that oral treatment with 2 g/kg of glycine might mainly act on NMDA receptors and not on glycine receptors. Additionally, d-serine, an NMDA receptor agonist that acts on the glycine site, significantly increased 5-HT and the effect was longer than that of glycine. The reasons for the longer effects were unclear. However, because d-serine is a specific agonist to NMDA receptors, the effects on NMDA receptors were stronger than those induced by glycine. Our previous study indicated that selective glycine site antagonist treatment of NMDA receptors blocks the effects of systemic administration of 1.6 g/kg glycine on prepulse inhibition deficits. This study also suggested that glycine directly involves the glycine site of NMDA receptors accessed through the blood–brain barrier.20 On the other hand, L-serine did not have a significant effect on the 5-HT and dopamine concentration in the PFC, suggesting a specific effect of glycine.
Using primary cultures, Becquet et al.24,25 reported that glycine decreases 5-HT release in rat rhombencephalic raphe cells. Their research also indicated that L-glutamate and NMDA stimulate 5-HT release in raphe primary cultures. Our in vivo data for glycine were inconsistent with their in vitro results. One possible explanation for this inconsistency is as follows: in primary culture raphe cells, glycine acts as an inhibitory neurotransmitter through ionotropic glycine receptors. In contrast to the inhibitory role of glycine in culture cells, 2 g/kg of oral glycine treatment acts as a co-agonist of the excitatory NMDA receptors in the living brain because of the maximum concentration in the CSF.
Perfusion of NMDA (10 and 50 µM) through a microdialysis probe significantly increases the extracellular concentration of 5-HT and 5-HIAA in the lateral parabrachial nucleus,26 and the NMDA receptor antagonist memantine significantly decreases 5-HT in the dorsal hippocampus.27 The findings that NMDA and glycine have the same effect on 5-HT in microdialysis studies suggest that glycine increases 5-HT through NMDA receptors. Oral treatment of glycine at a dose of 1–2 g/kg may act as a co-agonist of the NMDA receptors in the living brain. The PFC receives robust serotonergic innervation from the raphe nuclei.28 Activation of NMDA receptor channels facilitates the regulation of spikes by 5-HT receptors in PFC pyramidal neurons.29 Following the agonistic effect on the NMDA receptor glycine site in the living brain, 5-HT may be acutely released by activation of NMDA receptors. Our results suggest that oral glycine treatment increases 5-HT through NMDA receptor-mediated modulation in PFC neurons.
On the other hand, our results revealed no effect of glycine on dopamine in living PFC. Hernandes et al.30 reported that 300 µM of glycine stimulates the release of labeled acetylcholine but not dopamine or glutamate from superfused, dispersed rat striatal cells via glycine receptors. However, Bennett and Gronier31 reported that glycine (100 µM) increases dopamine release in striatum blocks via NMDA receptors. The concentrations of glycine and tissue preparations are different between these two reports, i.e. acetylcholine release increases at 300 µM of glycine from dispersed cells and dopamine increases at 100 µM of glycine from tissue blocks. Considering the ED50 of glycine, 300 and 100 µM of glycine theoretically act on both NMDA and glycine receptors. However, strychnine attenuates acetylcholine release and an NMDA receptor antagonist attenuates dopamine release. These results suggest that the localization of the glycine and NMDA receptors is different. In our study, the Cmax of the glycine concentration in the cerebrospinal fluid was approximately 50 µM after oral administration of 2 g/kg of glycine, suggesting that the concentrations of glycine in this experiment might not be enough to increase dopamine release. Interestingly, 2 g/kg of d-serine significantly increased dopamine in the PFC, suggesting that NMDA receptors partially modulate dopamine concentrations. On the other hand, the effects of NMDA receptor antagonists, such as ketamine and phencyclidine, on extracellular dopamine in the PFC have been studied in an animal model of schizophrenia. Acute administration of phencyclidine or ketamine increases extracellular glutamate, dopamine, and 5-HT in the PFC.32,33 However, subchronic administration of phencyclidine decreases dopamine,34 suggesting a change in glutamatergic neural plasticity in response to NMDA receptor antagonists. Glycine is used chronically in clinical cases, and further study of the effects of chronic glycine is thus necessary.
In the present study, 5-HT increased for 10–20 min after glycine administration. The duration of this increase was very short compared with that of a selective serotonin reuptake inhibitor, such as citalopram or venlafaxine, which increase brain 5-HT for more than 180 min in the rat frontal cortex.35 Following the transient increase in 5-HT after glycine treatment, compensatory mechanisms may function to normalize the extracellular concentrations of 5-HT. Portas et al.36 reported a change in endogenous extracellular 5-HT in the dorsal raphe nucleus and frontal cortex during the sleep–wake cycle. Additionally, the 5-HT2 receptor antagonist ritanserin substantially increases non-REM sleep in humans37 and rats.38,39 Furthermore, 5-HT2A knockout mice demonstrate less non-REM sleep than do wild mice.40 Kantor et al.39 concluded that serotonin increases electroencephalography desynchronization and produces an increase in vigilance level and motor activity. These findings suggest a contribution of serotonergic regulation to the sleep–wake cycle. However, a description of the comprehensive relationship between 5 and HT and sleep is not likely possible based on this evidence because receptor subtypes of 5-HT are located in various regions of the brain and the distributions of the drugs differ. Further studies investigating the regulation of sleep by serotonin, especially in the PFC, are necessary.
Both serotonin and NMDA signaling in the PFC are implicated in mental disorders, including schizophrenia, sleep disorder, epilepsy, depression, and anxiety disorders. Breier41 proposed a hypothesis of cortical–subcortical imbalance, with an increase in subcortical 5-HT function responsible for positive symptoms and a decrease in prefrontal 5-HT function responsible for negative symptoms in schizophrenia. Postmortem brain tissue analysis, CSF studies, and pharmacological challenges suggest a deficit in 5-HT function in the cortex of patients with schizophrenia.42,43 One of the possible explanations for the effects of glycine on negative symptoms of schizophrenia is that glycine treatment may act not only directly to improve NMDA hypofunction, but also indirectly in the transient compensation for 5-HT hypofunction in PFC.
In the near future, advanced research studies investigating the glycine–serotonin interaction in the PFC should be performed to examine the pathophysiology of mental disorders, including schizophrenia and insomnia.
The authors would like to thank E. Furuhata, H. Kumeya and R. Nishijima for technical support and Animal Support Kobe Co., Ltd. for animal care. The authors also would like to thank Prof K. Nakayama from the Department of Psychiatry, Jikei University School of Medicine, for critical comments on this study.