The interrelated lives of NMDA receptors and glycine transporters
Year: 2009
Type: article
Abstract: Most glutamate N-methyl-d-aspartic acid receptors (NMDARs) in the adult CNS probably assemble as dimers of dimers, each one composed of an NR1 and an NR2 subunit (the developing CNS and adult motoneurons also express NR3 subunits, but the stoichiometry with which these assemble to form functional NMDARs remains unclear). The extracellular domain of these subunits contains a region, the S1S2 core, where the interaction between ligand and receptor actually occurs (Furukawa et al. 2005). NMDAR opening, however, is not fully accomplished until other events have also taken place: not only each NR2 subunit needs to bind a glutamate molecule, but also both NR1 subunits have to bind a molecule of a co-agonist (under physiological conditions (Johnson & Ascher, 1987) or d-serine (Schell et al. 1995)). In addition, membrane depolarization is required to relieve the Mg2+ block of the receptor. The particular features of all these steps vary within the NMDAR population. In fact, NMDARs come in different flavours that differ in the type of NR2 subunit they express. Interestingly, it is the identity of the glutamate-binding NR2 subunit that sets the NMDAR affinity not only for glutamate (NR2B > NR2A ≈ NR2D > NR2C; range 53–172 nm), but also for glycine (NR2C ≫ NR2B ≈ NR2D > NR2A; range 0.36–3.7 μm) (Laurie & Seeburg, 1994). This form of negative co-operativity between glutamate- and glycine-binding (and vice versa) ultimately results in a glycine-dependent apparent desensitization of NMDARs, and suggests the existence of an allosteric interaction between NR1 and NR2. Other forms of desensitization occur via glycine-independent mechanisms, and may or may not be triggered by Ca2+ influx through the receptor itself (McBain & Mayer, 1994). The puzzle here is how glutamate and the co-agonist, for example glycine, manage to act in a concerted way to set off NMDARs gating. Despite the initial assumption of glycine being continuously present in the extracellular space at saturating concentrations for NMDARs (1–10 μm), theoretical estimates taking into account the thermodynamics of glycine transport suggest that glycine transporters (GlyT1, GlyT2) can, at equilibrium, lower the glycine concentration down to ∼149 nm (Attwell et al. 1993), well below the transporter affinity for glycine (∼20 μm). This could be particularly relevant if glycine transporters were expressed at high density (150–3000 μm−2) in small dead-end spaces close to NMDAR clusters, because they could potentially desaturate the NMDAR glycine-binding site without necessarily altering the overall ambient glycine concentration (Supplisson & Bergman, 1997). The likelihood of this scenario, however, awaits testing by finely resolved spatial mapping of GlyT1–2 expression. If the NMDAR glycine-binding site is not constantly occupied, how is the co-agonist concentration enhanced during glutamatergic transmission? At present, the two major mechanisms suggested are co-agonist release from astrocytes via non-NMDA and metabotropic glutamate receptor pathways (Oliet & Mothet, 2006) and reverse uptake (Attwell et al. 1993), due to the effect of membrane depolarization on voltage-dependent transport. However, mechanisms that involve signalling via G-protein-coupled receptors or substrate translocation via transporters seem slow in comparison to the time course of synaptic transmission, and probably faster yet-unknown release routes also exist. The potentially crucial role of glycine transporters in regulating NMDAR function does not require the presence of glycinergic synapses, as indicated by the presence of glycine transporters in neurons and glia in regions where glycine receptors are abundant (retina, brainstem, spinal cord) or present at lower levels (hippocampus), but that are enriched with glutamatergic synapses (Zafra et al. 1995). Accordingly, increased NMDAR activation has been observed in hippocampal CA1 pyramidal neurons of GlyT1+/− mice (Gabernet et al. 2005). Clearly, a thorough understanding of synaptic NMDAR function cannot be achieved without a view of glycinergic signalling, particularly of glycine transporters. Although drugs that target the NMDAR glycine-binding site do not appear to interfere with GlyT1–2 (with the exception of glycine), it is not known whether glycine transporter blockers also act on NMDARs. In this issue of The Journal of Physiology, Zhang et al. (2009) suggest that the GlyT1 inhibitor sarcosine is also a weak agonist of the NMDAR glycine-binding site. This is somewhat intriguing because sarcosine, glycine and d-serine share similar chemical structures. The main finding is a dose-dependent potentiation of NMDAR currents, partially inhibited by the NMDARs glycine-binding site antagonist 5,7-dichlorokynurenic acid, and not due to glycine contamination or sarcosine block of GlyT1–2 (the effect persists with the GlyT1 inhibitor NFPS). Sarcosine and glycine, however, may have different mechanisms of action. For example, currents elicited by prolonged NMDA applications desensitize less with subsaturating concentrations of sarcosine than with glycine. An extensive characterization of the mechanistic reasons for this phenomenon is missing, but since replacing extracellular Ca2+ with Ba2+ does not abolish the differences between sarcosine- and glycine-induced desensitization, the effects are not probably attributable to differences in NMDAR Ca2+-dependent inactivation. One possibility is that agonist-binding to NR2 is less efficient in triggering sarcosine- than glycine-dissociation from NR1. This effect would disappear when raising the co-agonist concentration, because now the occupancy of the glycine-binding site would remain high despite the drop in affinity induced by NMDA binding to the receptor. The experimental findings reported by Zhang et al. are indeed in line with this hypothesis. A more in-depth analysis of the biophysical properties of NMDARs in the presence of different co-agonists is probably appropriate, together with a pharmacological characterization of the effects of sarcosine on NMDARs with different subunit composition. However, despite these caveats, the paper by Zhang et al. gives us a useful warning on the potential non-specificity of some of the few available glycine transporter blockers, which might come particularly handy when trying to gain a thorough view of the patho-physiological interplay between glutamatergic synapses and glycine uptake systems.
Source: The Journal of Physiology
Author: Annalisa Scimemi
Language: English
Subfield: Cellular and Molecular Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W2053357561