The present address of Karen R. Aubrey is Laboratoire de Neurobiologie Moleculaire et Cellulaire, (CNRS UMR 8544), Ecole Normale Superieure, Paris, France.
N-Arachidonyl-glycine inhibits the glycine transporter, GLYT2a
Article first published online: 26 JUL 2006
Journal of Neurochemistry
Volume 99, Issue 3, pages 781–786, November 2006
How to Cite
Wiles, A. L., Pearlman, R.-J., Rosvall, M., Aubrey, K. R. and Vandenberg, R. J. (2006), N-Arachidonyl-glycine inhibits the glycine transporter, GLYT2a. Journal of Neurochemistry, 99: 781–786. doi: 10.1111/j.1471-4159.2006.04107.x
- Issue published online: 8 AUG 2006
- Article first published online: 26 JUL 2006
- Received May 14, 2006; accepted June 27, 2006.
- glycine transporter;
N-arachidonyl-glycine is one of a series of N-arachidonyl-amino acids that are derived from arachidonic acid. N-arachidonyl-glycine is produced in a wide range of tissues with greatest abundance in the spinal cord. Here we report that N-arachidonyl-glycine is a reversible and non-competitive inhibitor of glycine transport by GLYT2a, but has little effect on glycine transport by GLYT1b or γ-amino butyric acid transport by GAT1. It has previously been reported that the activity of GLYT2a is down-regulated by protein kinase C and therefore we investigated whether the actions of N-arachidonyl-glycine on GLYT2a are mediated by second messenger systems that lead to the activation of protein kinase C. However, the protein kinase C inhibitor, staurosporine, had no effect on the actions of N-arachidonyl-glycine on GLYT2a. Thus, the actions of N-arachidonyl-glycine are likely to be mediated by a direct interaction with the transporter. We have further defined the pharmacophore by investigating the actions of other N-arachidonyl amino acids as well as the closely related compounds arachidonic acid, anandamide and R1-methanandamide. Arachidonic acid, anandamide and R1-methanandamide have no effect on glycine transport, but N-arachidonyl-l-alanine has similar efficacy at GLYT2a to N-arachidonyl-glycine, and N-arachidonyl-γ-amino butyric acid is less efficacious. These observations define a novel recognition site for the N-arachidonyl amino acids.
γ-amino butyric acid transporter subtype 1
glycine transporter subtype 1
glycine transporter subtype 1b
glycine transporter subtype 2
glycine transporter subtype 2a
N-arachidonyl amino acid
N-arachidonyl-γ-amino butyric acid
fatty acid amino hydrolase
oocyte transcription vector
Arachidonic acid is a precursor for a large number of lipid molecules that exert a range of potent biological actions. These lipids include various inflammatory mediators, such as the prostaglandins, and also analgesic compounds, such as the endocannabinoids. The N-arachidonyl-amino acids (NAAAs) are a recently characterized family of arachidonic acid derivatives that include N-arachidonyl-glycine (NA-Gly), N-arachidonyl-l-alanine (NA-L-Ala) and N-arachidonyl-γ-aminobutyric acid (NA-GABA; Huang et al. 2001; Fig. 1). Two of these compounds, NA-Gly and NA-GABA, are analgesic in behavioural models of pain in rodents (Burstein 1999; Huang et al. 2001). However, the biochemical mechanisms for the analgesic effects of these agents are unknown. Arachidonic acid and its various derivatives mediate their biological actions through a number of mechanisms and include: activation of G-protein-coupled receptors (e.g. Sheskin et al. 1997), modulation of ion channels (e.g. Chemin et al. 2001) and transporters (e.g. Barbour et al. 1989), inhibition of enzymes (e.g. Grazia Cascio et al. 2004); and influencing lipid membrane fluidity which may indirectly modulate the activity of many membrane proteins.
NA-Gly is produced in a wide variety of organs (Huang et al. 2001), which is suggestive that it may play a number of roles and interact with multiple target sites. NA-Gly is metabolised by fatty acid amide hydrolase (FAAH; Huang et al. 2001) and also cyclooxygenase 2 (COX-2) (Prusakiewicz et al. 2002) and one potential role of NA-Gly may be to regulate the activity of these enzymes and thereby influence the production or metabolism of other compounds such as anandamide or various prostaglandins. The production of NA-Gly in regions of the CNS that use glycine as an inhibitory neurotransmitter provides an intriguing possibility that NA-Gly may mediate some of its physiological actions through modulation of the various receptors and transporters involved in glycinergic neurotransmission. In this report, we have focused on the effects of NA-Gly on glycine transporters. Glycine transporters belong to the Na+/Cl–-dependent family of neurotransmitter transporters and there are two subtypes, GLYT1 and GLYT2 (Roux and Supplisson 2000). There are also a number of splice variants of the two subtypes, which differ in their C- and N-termini (Kim et al. 1994; Morrow et al. 1998; Ponce et al. 1998; Hanley et al. 2000). The GLYT2 subtypes are expressed in presynaptic glycinergic neurones of the spinal cord and are thought to contribute to clearance of glycine from the synapse and also providing glycine for packaging into glycinergic synaptic vesicles. Glycine transport by GLYT2 is coupled to the co-transport of 3 Na+ and 1 Cl– (Roux and Supplisson 2000), which ensures that extracellular glycine concentrations at inhibitory synapses are reduced to the low nanomolar level. The GLYT1 subtypes are widely expressed in the central nervous system in glial cells surrounding glycinergic and also glutamatergic synapses (Smith et al. 1992; Zafra et al. 1995). Glycine transport by GLYT1 is coupled to the co-transport of 2 Na+ and 1 Cl– and will reduce the extracellular glycine concentration to the high nanomolar range (Roux and Supplisson 2000). At excitatory glutamatergic synapses, GLYT1 plays a homeostatic role to maintain glycine concentrations for activation of N-methyl-d-aspartate receptors (Bergeron et al. 1998). At inhibitory synapses, GLYT1 plays a role in ensuring rapid clearance of glycine to prevent excessive glycine receptor activity (Gomeza et al. 2003a).
We recently demonstrated that both arachidonic acid and anandamide modulate the glycine transporter, GLYT1a (Pearlman et al. 2003). In this study, we demonstrate that NA-Gly, and also NA-L-Ala, are non-competitive inhibitors of the glycine transporter, GLYT2a, but have no effect on glycine transport by GLYT1b. NA-Gly also has a small inhibitory effect on the related GABA transporter, GAT1.
Materials and methods
NA-Gly, NA-GABA, NA-L-Ala, arachidonic acid, anandamide and R-1 methanandamide were obtained from Cayman Chemical Company (Ann Arbor, MI, USA). The compounds were stored in ethanol at −20°C and stocks were diluted in buffer on the day of use and kept on ice. Restriction enzymes were purchased from Progen Industries Ltd. (Toowong, Qld, Australia). All other chemicals were obtained from Sigma (Sydney, NSW, Australia) unless otherwise stated.
The pcDNA3-GLYT2a was a gift from John Morrow (Organon Laboratories Limited, Lanarkshire, UK) and pcDNA3-GLYT1b was a gift from Dr Marc Caron (Howard Hughes Medical Institute Research Laboratories, Department of Cell Biology and Medicine, Duke University Medical Center, Durham, NC, USA). Professor B. Kanner (Department of Biochemistry, Hadassah Medical School, Hebrew University, Jerusalem, Israel) kindly supplied the plasmid containing pOG51-GAT1.
Complementary DNAs encoding human GLYT1b and human GLYT2a were subcloned into oocyte transcription vector (pOTV). The pOTV-GLYT1b and pOTV-GLYT2a plasmids were linearized with the restriction enzyme SpeI and cRNA was then transcribed from the cDNA constructs with T7 RNA polymerase and capped with 5′7-methyl guanosine using a mMessage mMachine kit (Ambion, Austin, TX, USA). The pOG-GAT1 plasmid was linearized with the restriction enzyme SacII and GAT1 RNA was prepared as above.
Oocyte preparation and electrophysiology
Mature female Xenopus laevis (Xenopus Express, Haute-Loire, France) were anaesthetised with 0.17% 3-aminobenzoic acid ethyl ester and the ovarian lobes were surgically removed. All procedures were in accordance with the Australian National Health and Medical Research Council guidelines for the prevention of cruelty to animals. The lobes were immediately placed in OR2 medium (82.5 mm NaCl, 2 mm KCl, 1 mm MgCl2, 5 mm HEPES, pH 7.5 adjusted with NaOH) on ice and treated with collagenase A (Boehringer, Mannheim, Germany) in OR2 at 20°C for ∼2 h. The lobes were then washed thoroughly with OR2, followed by a wash with ND96 (96 mm NaCl, 2 mm KCl, 1 mm MgCl2, 1.8 mm CaCl2, 5 mm HEPES (Hemi-Na) pH 7.55) and stage V oocytes were isolated. cRNA (50 nL) was microinjected into the defoliculated stage V oocytes which were then incubated at 17°C for 3–4 days in ND96 supplemented with 2.5 mm sodium pyruvate, 0.5 mm theophylline, 50 μg/mL gentamicin.
All experiments were carried out at room temperature (20–22°C). Individual oocytes expressing cRNA were placed in a recording chamber and superfused with ND96 at a rate of ∼7 mL/min. The oocytes were voltage clamped at −60 mV. Whole-cell currents were measured using standard two-electrode voltage clamp techniques using a Geneclamp 500 (Axon Instruments, Foster City, CA, USA) with microelectrodes (0.5–1.0 MΩ) filled with 3 m KCl. Currents were recorded using a MacLab2e or PowerLab2/20 chart recorder (ADInstruments, Sydney, NSW, Australia). Compounds were added to the recording chamber for up to 90 s and current responses were measured. Compounds were washed out of the bath with ND96 for 3–5 min. The maximal ethanol concentration applied to cells was 0.05% (v/v), which did not generate any detectable effects on baseline currents or current measured in the presence of glycine. Application of NA-Gly, NA-L-Ala and NA-GABA at concentrations greater than 100 μm caused fluctuations in the baseline current, which may be as a result of disruption of membrane fluidity and so, in all assays, 100 μm of the lipid was the maximum used. In experiments with staurosporine, oocytes were treated with 1 μm staurosporine for 60 min and then placed in the oocyte recording chamber and transport currents measured as described above.
Current (I) as a function of glycine concentration [(Gly)] was fitted by least-squares analysis to a derivation of the Michaelis–Menten equation, I = [(Gly) × Imax]/[EC50 + (Gly)], using Kaleidagraph V3.0 (Synergy Software, Reading, PA, USA). Imax is the maximal current and EC50 is the concentration of glycine that generates half maximal current. Glycine current as a function of NA-Gly concentration [(NA-Gly)] was fitted by least-squares analysis to the equation I/Imax = 1 − [(NA-Gly)/[(NA-Gly) + IC50] + C. IC50 is the (NA-Gly) at which half maximal reduction in transport current occurs and C is the residual current at maximal inhibition of transport. Values are presented as mean ± SEM and paired t-tests were used to test for significant differences between measurements.
NA-Gly inhibits glycine transport mediated by GLYT2a
Application of glycine to Xenopus laevis oocytes expressing GLYT2a, voltage clamped at −60 mV, generates concentration-dependent inward currents, with an EC50 of 25.2 ± 4.2 μm (n = 4). Co-application of 30 μm glycine with 10 μm NA-Gly reduces the amplitude of the glycine transport current (Fig. 2a), with maximum reduction occurring 60 s after application of NA-Gly. After washout of NA-Gly from the oocyte chamber, subsequent applications of glycine generate currents of similar amplitude to pre-exposure currents, which demonstrates that the inhibitory effects of NA-Gly on glycine transport currents are fully reversible. Activation of protein kinase C has been reported to cause down-regulation of GLYT2a (Fornes et al. 2004), and therefore it is possible that NA-Gly may activate an endogenous G-protein-coupled receptor present in the oocyte membrane that activates protein kinase C and in turn cause the reduction in transport currents. We investigated this possibility by recording glycine transport currents before and after treating the oocytes with the protein kinase C inhibitor, staurosporine (1 μm) for 1 h. In this experiment, 10 μm NA-Gly caused a 61 ± 4% (n = 4) reduction in transport currents before staurosporine treatment and 63 ± 1% (n = 4) after treatment (Fig. 2b). Thus, we can rule out the possibility that NA-Gly leads to activation of an endogenous protein kinase C.
The concentration dependence of NA-Gly inhibition of glycine transport currents was measured and the IC50 for NA-Gly inhibition of GLYT2a is 5.1 ± 3.1 μm, with a maximal inhibition of 87.9 ± 12.5% (n = 7; Fig. 2c). Radiolabelled glycine ([3H]glycine) uptake assays were also conducted. NA-Gly was applied to oocytes expressing GLYT2a, for 1 min to allow equilibration of NA-Gly at the GLYT2a site, and then [3H]glycine uptake was measured. NA-Gly reduces the amount of [3H]glycine uptake by GLYT2a, with an IC50 of 7.8 ± 2.4 μm (n = 5) and caused 72.0 ± 5.3% maximal inhibition (n = 5; Fig. 2c). This observation confirms that the reduction in glycine-induced currents is as a result of the inhibition of transport by GLYT2a.
Glycine concentration–responses were measured in the presence and absence of 10 μm NA-Gly (Fig. 2d). The EC50 for glycine is 25.2 ± 4.2 μm (n = 4) and, in the presence of 10 μm NA-Gly, the EC50 for glycine is 28.0 ± 7.1 μm and the Imax is reduced to 63.0 ± 4.1% (n = 4). These results demonstrate that NA-Gly is a non-competitive inhibitor of GLYT2a.
GLYT2a is also inhibited by other N-arachidonyl- amino acids
We have previously demonstrated that arachidonic acid and anandamide modulate the activity of the glycine transporter GLYT1a (IC50 for arachidonic acid inhibition of GLYT1a: 2.1 ± 0.7 μm; Pearlman et al. 2003). Given that NA-Gly is structurally related to both arachidonic acid and anandamide, we first investigated whether these compounds modulate the activity of GLYT2a. In contrast to NA-Gly, anandamide and arachidonic acid at concentrations up to 30 μm had no measurable effects on glycine transport currents mediated by GLYT2a (Fig. 3). Thus, it appears that the chemical nature of the lipid head group is a critical determinant of the effects on GLYT2a. We further investigated the nature of the lipid head group required for activity at GLYT2a by testing the effects of other members of the N-arachidonyl-amino acid family (Fig. 3). NA-L-Ala also reduced the amplitude of glycine transport currents with similar potency as for NA-Gly (IC50 for NA-L-Ala, 8.0 ± 3.9 μm, n = 7), with complete block of transport currents at concentrations of NA-L-Ala at 100 μm. NA-GABA also inhibited glycine transport currents, but with slightly reduced potency (11.9 ± 6.0 μm, n = 5) and only 62.6 ± 8.0% (n = 5) inhibition at maximal concentrations of NA–GABA.
Interaction between the arachidonic acid derivatives and other neurotransmitter transporters
Arachidonic acid and anandamide show differential effects on GLYT1b compared with GLYT2a (Pearlman et al. 2003; Fig. 3). We investigated whether compounds such as NA-Gly, NA-GABA and NA-L-Ala also show differential effects on two other closely related transporters, the glycine transporter GLYT1b and the GABA transporter GAT-1. Application of 100 μm NA-Gly, NA-GABA and NA-L-Ala to oocytes expressing GLYT1b had no effect on their own, but when co-applied with 30 μm glycine, the amplitude of the glycine transport current was partially reduced (Fig. 4a). It was not possible to investigate these effects with higher concentration of lipids (> 100 μm) because the holding currents at −60 mV did not remain stable.
NA-Gly, NA-GABA and NA-L-Ala were also tested on the GABA transporter GAT1 (Fig. 4b). Application of 30 μm GABA to oocytes expressing GAT1 generates inward currents of similar magnitude to 30 μm glycine-induced currents at GLYT2a. All three compounds had no effect on the current when applied alone. Co-application of 30 μm GABA with NA-Gly or NA-GABA caused a small reduction in GABA transport currents (Fig. 4b). NA-L-Ala caused a greater level of inhibition 62.5 ± 7.6% with an IC50 of 19.9 ± 11.3 μm (n = 4). The rank order of potency at GAT1 is NA-L-Ala >> NA-GABA ≥ NA-Gly.
The results presented in this study demonstrate that NA-Gly is a reversible inhibitor of glycine transport mediated by GLYT2a but not GLYT1b. The effects of NA-Gly on GLYT2a are unlikely to be mediated by an indirect effect as a consequence of protein kinase C-mediated down-regulation of transporter activity and therefore we conclude that the actions of NA-Gly are as a result of a direct interaction with the transporter. NA-Gly is a non-competitive inhibitor of glycine transport by GLYT2a (Fig. 2), which suggests that the NA-Gly binding site on GLYT2a is likely to be distinct from the glycine recognition site.
The NAAAs are lipid soluble compounds with polar head groups and therefore may interact with GLYT2a via the lipid–protein interface or directly via an aqueous accessible site. If NA-Gly was to bind to the lipid–protein interface of the transporter, the effect of glycine transport inhibition is likely to persist for a considerable time. We observed that the effects of NA-Gly, at concentrations up to 100 μm, were fully reversible following a 3-min washout period, which suggests that NA-Gly is most likely to bind to a site that is exposed to the aqueous phase.
The variation in modulation of transport by the different N-arachidonyl derivatives is likely to be as a result of the chemical constituents of the lipid head groups. NA-Gly, NA-L-Ala and NA-GABA contain a nitrogen atom and a carboxyl group in their head group, whilst the inactive compounds, arachidonic acid and anandamide, lack either the carboxyl group (anandamide) or the nitrogen (arachidonic acid). This suggests that both the nitrogen atom and the carboxyl group are required for producing an inhibitory effect. The potency of modulation of the GLYT2a can yield information about the chemical nature of the interaction with the transporter. The rank order of potency for inhibition of glycine transport at GLYT2a is NA-Gly ≥ NA-L-Ala > NA-GABA. NA-L-Ala and NA-Gly have one carbon between the carboxyl group and nitrogen atom of their head groups, while NA-GABA has a longer carbon chain between the carboxyl group and nitrogen atom. The spacing between the carboxyl group and the nitrogen atom appears to be important in determining the potency of modulation of GLYT2a. The effects of the arachidonic acid derivatives were also tested on the closely related transporter GAT1. NA-L-Ala is equally as potent as NA-Gly at GLYT2a, but also inhibits GAT1, whereas NA-Gly has minimal effects at both GAT1 and GLYT1. NA-GABA inhibits all three transporters, but not as potently as NA-Gly and NA-L-Ala at GLYT2a.
An issue that arises from this study is whether the concentrations of NA-Gly used in this study are likely to occur in vivo and whether NA-Gly may act as an endogenous regulator of GLYT2a activity. NA-Gly is produced in highest concentrations in the spinal cord (Huang et al. 2001), which is also a region where GLYT2a is abundantly expressed, but at this stage accurate measurements of free or extracellular concentrations of these lipophilic compounds have not been made. The concentrations of NA-Gly that inhibit GLYT2a (IC50 5.1 ± 3.1 μm) are comparable with the concentrations of NA-Gly required to inhibit FAAH (IC50 of 4.1 ± 7.0 μm) (Huang et al. 2001). The Km value for arachidonic acid at human COX-2 is 5.8 ± 1.3 μm, while for anandamide the Km value is 23.7 ± 5.7 μm (Yu et al. 1997). Although the Km for NA-Gly metabolism by COX-2 has not been reported, it has been demonstrated that NA-Gly is a more potent substrate for COX-2 than anandamide (Prusakiewicz et al. 2002). Therefore, we may expect that the Km value for NA-Gly at COX-2 will be between 5 and 24 μm. Thus, the concentrations of NA-Gly that inhibit GLYT2a are at least comparable with concentration required for activity at other potential sites of action. However, it should be noted that FAAH and COX-2 are intracellular enzymes and may be exposed to very different NA-Gly concentration dynamics than the extracellular site of a glycine transporter.
Plantar administration of NA-Gly reduces the second phase of formalin-induced pain behaviour in rats (Huang et al. 2001) and systemic NA-Gly produces analgesia in the hotplate test in mice (Burstein 1999). A question that arises from the results presented in the current study is: could the analgesic effects of NA-Gly arise via an action at GLYT2? The potency of NA-Gly at GLYT2a is comparable with its potency at other reported targets and therefore it is reasonable to suggest that NA-Gly does have effects on GLYT2a function. Intrathecal application of the glycine receptor antagonist strychnine increases the responsiveness of dorsal horn neurons to noxious/innocuous stimuli (Peng et al. 1996) and produces allodynia (Yaksh 1989). Therefore, it is reasonable to suggest that enhancement of glycinergic neurotransmission may provide analgesia. At present, it is difficult to predict the biological consequences of pharmacological inhibition of GLYT2. GLYT1 is expressed in glial cells in close proximity to the synapse (Zafra et al. 1995) and inhibition of GLYT1 would be expected to enhance glycinergic synaptic neurotransmission. However, GLYT2 is expressed in extrasynaptic regions of neurones (Spike et al. 1997). Knockout of the GLYT2 gene results in a hyperexcitable state, which is thought to arise from a lack of glycine uptake into glycinergic neurones and the subsequent lack of loading of synaptic vesicles (Gomeza et al. 2003b). Short-term partial inhibition of GLYT2 by compounds such as NA-Gly are likely to have very different effects than complete removal of the protein from birth. Partial non-competitive inhibition of extrasynaptic transporters would not prevent loading of glycine in the cytoplasm for storage in vesicles, but would slow the process and may cause transient elevations in extrasynaptic glycine levels. This extrasynaptic elevation of glycine would not affect synaptic transmission, but may generate a tonic stimulation of extrasynaptic glycine receptors (Mori et al. 2002). This idea is speculative, but with the recent development of GLYT1- and GLYT2-selective inhibitors many of these questions can now be addressed. Further ongoing studies in our laboratory will investigate the actions of NA-Gly on phasic and tonic glycinergic neurotransmission.
We thank Suzanne Habjan, Hue Tran and Kong Li for isolation and preparation of Xenopus laevis oocytes and Drs Chris Vaughan and Mark Connor for helpful comments and suggestions. This work was supported by the National Health and Medical Research Council of Australia.
- 1989) Arachidonic acid induces a prolonged inhibition of glutamate uptake into glial cells. Nature 342, 918–920. , , and (
- 1998) Modulation of N-methyl-d-aspartate receptor function by glycine transport. Proc. Natl Acad. Sci. USA 95, 15 730–15 734. , , and (
- 1999) The cannabinoid acids: non-psychoactive derivatives with therapeutic potential. Pharmacol. Ther. 82, 87–96. (
- 2001) Direct inhibition of T-type calcium channels by the endogenous cannabinoid anandamide. EMBO J. 20, 7033–7040. , , , and (
- 2004) The second intracellular of the glycine transporter 2 contains crucial residues for glycine transport and phorbol ester-induced regulation. J. Biol. Chem. 279, 22 934–22 943. , , and (
- 2003a) Inactivation of the glycine transporter 1 gene discloses vital role of glial glycine uptake in glycinergic inhibition. Neuron 40, 785–796. , , , , , and (
- 2003b) Deletion of the mouse glycine transporter 2 results in a hyperekplexia phenotype and postnatal lethality. Neuron 40, 797–806. , , , , , , and (
- 2004) A structure–activity relationship study on N-arachidonoyl-amino acids as possible endogenous inhibitors of fatty acid amide hydrolase. Biochem. Biophys. Res. Commun. 314, 192–196. , , , , and (
- 2000) GABA receptor rho1 subunit interacts with a novel splice variant of the glycine transporter, GLYT-1. J. Biol. Chem. 275, 840–846. , and (
- 2001) Identification of a new class of molecules, the arachidonyl amino acids, and characterization of one member that inhibits pain. J. Biol. Chem. 276, 42 639–42 644. , , et al. (
- 1994) Cloning of the human glycine transporter type 1: molecular and pharmacological characterization of novel isoform variants and chromosomal localization of the gene in the human and mouse genomes. Mol. Pharmacol. 45, 608–617. , , , , , , and (
- 2002) β-Alanine and taurine as endogenous agonists at glycine receptors in rat hippocampus in vitro. J. Physiol. 539, 191–200. , and (
- 1998) Molecular cloning and functional expression of the human glycine transporter GlyT2 and chromosomal localisation of the gene in the human genome. FEBS Lett. 439, 334–340. , , , , and (
- 2003) Arachidonic acid and anandamide have opposite modulatory actions at the glycine transporter, GLYT1a. J. Neurochem. 84, 592–601. , and (
- 1996) Effects of GABA and glycine receptor antagonists on the activity and PAG-induced inhibition of rat dorsal horn neurons. Brain Res. 736, 189–201. , and (
- 1998) Characterization of the 5′ region of the rat brain glycine transporter GLYT2 gene: identification of a novel isoform. Neuroscience Lett. 242, 25–28. , , , and (
- 2002) Selective oxygenation of N-arachidonylglycine by cyclooxygenase-2. Biochem. Biophys. Res. Commun. 296, 612–617. , , and (
- 2000) Neuronal and glial glycine transporters have different stoichiometries. Neuron 25, 373–383. and (
- 1997) Structural requirements for binding of anandamide-type compounds to the brain cannabinoid receptor. J. Med. Chem. 40, 659–667. , , , and (
- 1992) Cloning and expression of a glycine transporter reveal colocalization with NMDA receptors. Neuron 8, 927–935. , , , and (
- 1997) An ultrastructural study of the glycine transporter GLYT2 and its association with glycine in the superficial laminae of the rat spinal dorsal horn. Neuroscience 77, 543–551. , , and (
- 1989) Behavioral and autonomic correlates of the tactile evoked allodynia produced by spinal glycine inhibition: effects of modulatory receptor systems and excitatory amino acid antagonists. Pain 37, 111–123. (
- 1997) Synthesis of prostaglandin E2 ethanolamide from anandamide by cyclooxygenase-2. J. Biol. Chem. 272, 21 181–21 186. , and (
- 1995) Glycine transporters are differentially expressed among CNS cells. J. Neurosci. 15, 3952–3969. , , , , and (