Up NMDA AMPA Metabotropic References


NMDA Receptors

NMDA receptors are highly permeant for Ca2+, show slower gating kinetics and the channel is blocked in a voltage-and use-dependent manner by physiological concentrations of Mg2+ ions (Mcbain and Mayer, 1994). These properties make them ideally suited for their role as a coincidence detector underlying Hebbian processes in synaptic plasticity such as learning (see later), chronic pain, drug tolerance and dependence (Collingridge and Singer, 1990; Bear and Malenka, 1994; Trujillo and Akil, 1995; Danysz and Parsons, 1995; Collingridge and Bliss, 1995; Dickenson, 1997).

Glycine as a co-agonist

Glycine is a co-agonist at NMDA receptors at a strychnine-insensitive recognition site (glycineB) and it’s presence at moderate nM concentrations is a prerequisite for channel activation by glutamate or NMDA (Danysz and Parsons, 1998). Physiological concentrations reduce one form of relatively rapid NMDA receptor desensitization. Recently it has been suggested that D-Serine may be more important than glycine as an endogenous co-agonist at NMDA receptors in the telencephalon and developing cerebellum. There is still some debate as to whether the glycineB site is saturated in vivo (Danysz and Parsons, 1998) but it seems likely that the degree of NMDA receptor activation varies depending on regional differences in receptor subtype expression and local glycine or D-serine concentrations. Moreover, glycine concentrations at synaptic NMDA receptors could be finely modulated by local expression of specific glycine transporters such as GLYT1 (Supplisson and Bergman, 1997).


The polyamines spermine and spermidine have multiple effects on the activity of NMDA receptors (Johnson, 1996; Williams, 1997). These include an increase in the magnitude of NMDA-induced whole-cell currents seen in the presence of saturating concentrations of glycine, an increase in glycine affinity, a decrease in glutamate affinity, and voltage-dependent inhibition at higher concentrations. Endogenous polyamines could act as a bi-directional gain control of NMDA receptors, by dampening toxic chronic activation by low concentrations of glutamate-through changes in glutamate affinity and voltage-dependent blockade-but enhancing transient synaptic responses to mM concentrations of glutamate (Williams, 1997; Zhang and Shi, 2001).

Molecular Biology

Two major subunit families designated NR1, NR2 as well as a modulatory subunit designated NR3 have been cloned. Most functional receptors in the mammalian CNS are formed by combination of NR1 and NR2 subunits which express the glycine and glutamate recognition sites respectively (Hirai et al., 1996; Laube et al., 1997).

NR1 Subunits

Alternative splicing generates eight isoforms for the NR1 subfamily (Zukin and Bennett, 1995). The variants arise from splicing at three exons one encodes a 21-amino acid insert in the N-terminal domain (N1, exon 5), and two encode adjacent sequences of 37 and 38 amino acids in the C-terminal domain (C1, exon 21 and C2, exon 22). NR1 variants are sometimes denoted by the presence or absence of these three alternatively spliced exons (from N to C1 to C2). NR1111 has all three exons, NR1000 has none, and NR1100 has only the N-terminal exon. The variants from NR1000 to NR1111 are alternatively denoted as NMDAR1e, c, d, a, g, f, “h” and b respectively or NMDAR1-4a,-2a,-3a,-1a,-4b,-2b,-3b and-1b respectively, but the more frequent terminology using non-capitalized suffices for the most common splice variants is NR1a (NR1011 or NMDAR1A) and NR1b (NR1100 or NMDARIG). MRNA for double splice variants in the C1/C2 regions such as NR1011 (NR1a) show an almost complementary pattern to those lacking both of these inserts such as NR1100 (NR1b); the former are more concentrated in rostral structures such as cortex, caudate, and hippocampus, while the latter are principally found in more caudal regions such as thalamus, colliculi, locus coeruleus and cerebellum (Laurie et al., 1995).

NR2 Subunits

The NR2 subfamily consists of four individual subunits, NR2A to NR2D. Various heteromeric NMDA receptor channels formed by combinations of NR1 and NR2 subunits are known to differ in gating properties, Mg2+ sensitivity and pharmacological profile (Sucher et al., 1996). The heteromeric assembly of NR1 and NR2C subunits for instance, has a lower sensitivity to Mg2+ but increased sensitivity to glycine and a very restricted distribution in the brain. In situ hybridization has revealed overlapping but different expression for NR2 mRNA e.g. NR2A mRNA is distributed ubiquitously like NR1 with highest densities occurring in hippocampal regions and NR2B is expressed predominantly in forebrain but not in cerebellum where NR2C predominates. The spinal cord expresses high levels of NR2C and NR2D (Tolle et al., 1993) and these may form heteroligomeric receptors with NR1 plus NR2A which would provide a basis for the development of drugs selectively aimed at spinal cord disorders(Sundstrom et al., 1997). NMDA receptors cloned from murine CNS have a different terminology to those in the rat: z1 remains the terminology for the mouse equivalent of NR1 and e1 to e4 represent NR2A to 2D subunits respectively.

NR3 Subunits

NR3 (NRL or Chi-1) is expressed predominantly in the developing CNS and does not seem to form functional homomeric glutamate-activated channels but co-expression of NR3 with NR1 plus NR2 subunits decreases response magnitude (Sucher et al., 1995; Kinsley et al., 1999; Matsuda et al., 2002). However, NR3A or NR3B do co-assemble with NR1 alone in Xenopus oocytes to form excitatory glycine receptors that are unaffected by glutamate or NMDA, Ca2+-impermeable, resistant to blockade by Mg2+ uncompetitive and competitive antagonists and actually inhibited by the glycine co-agonist D-serine. (Chatterton et al., 2002)

Uncompetitive NMDA receptor antagonists

Antagonists which completely block NMDA receptors cause numerous side effects such as memory impairment, psychotomimetic effects, ataxia and motor dis-coordination as they also impair normal synaptic transmission - a two edged sword. The challenge has therefore been to develop NMDA receptor antagonists that prevent the pathological activation of NMDA receptors but allow their physiological activation. It has been suggested that uncompetitive NMDA receptor antagonists with rapid unblocking kinetics but somewhat less pronounced voltage-dependency than Mg2+ should be able to antagonise the pathological effects of the sustained, but relatively small increases in extracellular glutamate concentration but, like Mg2+, leave the channel as a result of strong depolarization following physiological activation by transient release of mM concentrations of synaptic glutamate (Parsons et al., 1999; Jones et al., 2001). As such, uncompetitive NMDA receptor antagonists with moderate, rather than high affinity may be desirable. Memantine, ketamine, dextromethorphan and possibly felbamate and budipine are clinically-used agents which belong to this category – NB: for the last two it is unsure if uncompetitive NMDA receptor antagonism really contributes to their therapeutic efficacy. Others such as neramexane, remacemide, NPS-1506 and possibly the cannabinoid dexanabinol are at different stages of clinical development. Several promising agents have unfortunately been abandoned at late stages of development, possibly due to the choice of the wrong, too ambious, clinical indications such as stroke and trauma.

Glycine site antagonists

Most full glycineB antagonists (i.e. those without intrinsic partial agonist activity) show very poor penetration to the CNS although some agents with improved, but by no means optimal pharmacokinetic properties have now been developed. GlycineB antagonists have been reported to lack many of the side effects classically associated with NMDA receptor blockade such as no neurodegenerative changes in the cingulate / retrosplenial cortex even after high doses (Hargreaves et al., 1993) and no psychotomimetic-like or learning impairing effects at anticonvulsive doses (Murata and Kawasaki, 1993; Kretschmer et al., 1997; Baron et al., 1997; Danysz and Parsons, 1998). The MSD compound L-701,324 has even been proposed to have atypical antipsychotic effects (Bristow et al., 1996). The improved neuroprotective therapeutic profile of glycineB full antagonists could be due to their ability to reveal glycine-sensitive desensitization (Parsons et al., 1993).

Kynurenic acid is an endogenous glycineB antagonist but it seems unlikely that concentrations are sufficient to interact with NMDA receptors under normal conditions (Danysz and Parsons, 1998; Stone, 2001). However, concentrations are raised under certain pathological conditions (Danysz and Parsons, 1998; Stone, 2001) and interactions with other receptors such as a7 neuronal nicotinic have been reported at lower concentrations (Hilmas et al., 2001). Strategies aimed at increasing kynurenic acid concentrations by for example by giving its precursor 4-Cl-kynurenine, inhibiting brain efflux with probenecid or inhibiting its metabolism have been proposed to be of therapeutic potential (Danysz and Parsons, 1998; Stone, 2001).

D-cycloserine and (+R)-HA-966 are partial agonists at the glycineB site with different levels of intrinsic activity: 57% and 14% respectively in cultured hippocampal neurones (Karcz-Kubicha et al., 1997). Although these systemically-active partial agonists do not induce receptor desensitization (Henderson et al., 1990; Kemp and Priestley, 1991; Karcz-Kubicha et al., 1997) they have favourable therapeutic profiles in some in vivo models (Lanthorn, 1994; Witkin et al., 1997). This may, in part, be due to their own intrinsic activity as agonists at the glycineB site which would serve to preserve a certain level of NMDA receptor function even at very high concentrations (Priestley and Kemp, 1994; Fossom et al., 1995; Krueger et al., 1997).

D-cycloserine shows agonist like features at low doses, while with increasing dosing antagonistic effects predominate (Lanthorn, 1994). Such findings are often falsely interpreted to be “typical” for partial agonists i.e. agonism at low and antagonism at high doses. However, partial agonism actually means that an agent reaches a ceiling, non-maximal effect at higher doses (intrinsic activity) i.e. will antagonise receptor activation by high concentrations of a full agonist but facilitate at low concentrations of a full agonist (Henderson et al., 1990; Karcz-Kubicha et al., 1997). Recent data indicate that the consistent biphasic effects of D-cycloserine seen in vivo may rather be related to different affinities and intrinsic activities at NMDA receptor subtypes. D-cycloserine is a partial agonist for the murine equivalents of NR1/2A and NR1/2B heteromers (38% and 56% intrinsic activity compared to glycine 10 µM) but is more effective than glycine at NR1/2C (130%) (O'Connor et al., 1996). This effect is accompanied by higher affinity at NR1/2C receptors - NR1/2C > NR1/2D >> NR1/2B > NR1/2A (O'Connor et al., 1996). Very similar data were published recently by a different group, except that the intrinsic activity at NR1/2C was even higher (192%) (Sheinin et al., 2001). As such, it is likely that the biphasic effects seen in vivo are due to agonistic actions at NR1/2C receptors at lower doses and inhibition of NR1/2A and NR1/2B containing receptors at higher doses. This receptor subtype selectivity and differential intrinsic activity could well underlie its promising preclinical profile in some animals models.

Although ACPC has been reported to be a partial agonist with very high intrinsic activity, it is probably really a full agonist at the glycineB site and actually behaves as an antagonist in some in vivo models (neuroprotection, anticonvulsive effects) which are likely to be mediated via competitive antagonistic properties at higher concentrations {NahumLevy et al., 1999 #18977} (Skolnick et al., 1989). The consistent observation that chronic treatment with ACPC is neuroprotective could be because it desensitizes or uncouples NMDA receptors (Skolnick et al., 1992; Papp and Moryl, 1996) or may be related to an increase in the relative levels of NR2C expression (Fossom et al., 1995).

NR2B selective antagonists

Ifenprodil and its analogue eliprodil block NMDA receptors in a spermine-sensitive manner and were originally proposed to be polyamine antagonists. It is now clear that both agents are selective for NR2B subunits (Legendre and Westbrook, 1991) and bind to a site that is distinct from the polyamine recognition site, but interact allosterically with this site and the glycineB site. NR2B selective agents may also offer a promising approach to minimize side effects as agents would not produce maximal inhibition of responses of neurons expressing heterogeneous receptors. Thus, cortical and hippocampal neurons express both NR2A and NR2B receptors in approximately similar proportions, but very little NR2C or NR2D. NR2B selective agents therefore block NMDA receptor mediated responses of such neurons to a maximal level of around 30-50% of control. Several studies have shown that ifenprodil and eliprodil reduce seizures and are effective neuroprotectants against focal and global ischaemia and trauma at doses that do not cause ataxia or impair learning (Parsons et al., 1998). These compounds are not devoid of side effects and some companies attempted to improve the selectivity NR2B antagonists by reducing affects at other receptors such as a1 and a2 adrenergic receptors - traxoprodil (CP-101,606) and CP-283,097 showed improved selectivity and in vivo potency (Butler et al., 1997; Menniti et al., 1997; Chenard and Menniti, 1999). However, an unfortunate new side effect has recently been reported, i.e. that some of these agents may produce a prolongation of the QT interval in the cardiac action potential due to blockade of human ether-a-go-go-related gene (hERG) potassium channels (Gill et al., 1999). This would be less of a problem in acute excitotoxicity and traxoprodil is still under development for stroke / TBI.