Postepy Hig Med Dosw. (online), 2011; 65: 338-346
Full Text PDF  

Glutamate NMDA receptors in pathophysiology and pharmacotherapy of selected nervous system diseases
Rola receptorów NMDA w patofizjologii i farmakoterapii wybranych chorób układu nerwowego
Łukasz Dobrek, Piotr Thor
Department of Pathophysiology, Jagiellonian University Medical College Cracow, Poland
Corresponding author
Łukasz Dobrek, M.D., Department of Pathophysiology, Jagiellonian University Medical College, Czysta Street 18, 31-121 Cracow, Poland; e-mail:

Received:  2011.02.07
Accepted:  2011.05.11
Published:  2011.06.07

Glutaminian jest podstawowym neuroprzekaźnikiem pobudzającym, który działa na receptory NMDA. Związek ten jest współodpowiedzialny za regulowanie wielu ważnych fizjologicznych funkcji, wliczając w to uczenie się, pamięć i zachowanie. Nadmiar glutaminianu i nadaktywność receptorów NMDARs wywołuje patologiczne zmiany. Zjawisko neurotoksyczności zależnej od glutaminianu bierze udział w patogenezie wielu zaburzeń neurologicznych. Artykuł pokrótce opisuje rolę glutaminianu w patofizjologii udaru niedokrwiennego mózgu, wybranych chorób neurodegeneracyjnych i schizofrenii oraz omawia obecne i potencjalne znaczenie leków działa­jących na receptory glutaminergiczne w neuropsychofarmakologii.
Słowa kluczowe: glutaminian • receptory NMDA • neurotoksyczność glutaminianu • antagoniści receptora NMDA

Glutamate is the basic excitatory neurotransmitter acting via N-methyl-D-aspartate receptors (NMDARs). It co-regulates many important physiological functions, including learning, memo­ry, and behaviour. An excess of glutamate, as well as NMDAR over-activity, produce pathologi­cal effects. Glutamate-related neurotoxicity is involved in the pathogenesis of many neurological conditions. This article briefly describes the role of the glutamate system in the pathophysiology of brain ischemia, selected neurodegenerative disorders, and schizophrenia. It also reviews the current and potential future status of agents targeting NMDARs in neuropsychopharmacology.
Key words: glutamate • NMDA receptors (NMDARs) • excitotoxicity • NMDAR antagonists

AD - Alzheimer's disease; ALS - Amyotrophic lateral sclerosis; AMPAR - α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor; CJD - Creutzfeld-Jacob disease; EAATs - Excitatory aminoacids transporters; GABA - γ-aminobutyric acid; glyT1 - glycine transporter 1; HD - Huntington disease; KAR - Kainate receptor; LTD - Long term depression; LTP - Long term potentiation; MAPK - Mitogen activated protein kinase; NMDAR - N-methyl-D-aspartate receptor; PD - Parkinson disease; PSD - Postsynaptic density; TBI - Traumatic brain injury.
1. Glutamate receptors
In mammalian central nervous system synapses, glutama­te is the major neurotransmitter mediating excitatory neu­rotransmission. It is released from presynaptic vesicles, diffuses across the synaptic cleft, and acts on both me­tabotropic and ionotropic glutamate receptors located in the presynaptic terminals and postsynaptic membranes of the brain and spinal neurons. Three ionotropic glutamate receptor subtypes can be distinguished, and they are na­med according to their agonists: NMDAR (N-methyl-D-aspartate receptor), AMPAR (α-amino-3-hydroxy-5-me­thylisoxazole-4-propionic acid receptor) and KAR (kainate receptor). NMDARs are composed of protein complexes which form an intrinsic ion channel, permeable to mono­valent cations (including Na+ and K+), and bivalent ones (mostly Ca2+). We also know of a metabotropic glutamate receptor coupled to a G intrinsic membrane protein [12,13].
NMDARs contain four subunits which are a combination of: NR1, NR2, and NR3, (encoded by genes GRIN1, GRIN2A-D, GRIN3A-B, respectively). There is consensus that NMDARs are tetramers composed of two NR1 subunits and two NR2 subunits, less commonly including NR3 subunits. NR1 sub­units exhibit basic NMDAR features and disruption of NR1 genes abolishes NMDAR responses. This demonstrates that NR1 subunits are rather essential. There are eight different NR1 elements, generated by alternative splicing of a single gene. Four (A-D) NR2 compounds can be distinguished, located in various brain regions and playing a modulatory role in regards to NMDARs. It has been shown that the com­bination of NR1 with different NR2 subunits results in di­verse electrophysiological and pharmacological responses. NR1 and NR2A are ubiquitous, NR2B occurs in the fore­brain, NR2C in the cerebellum, with NR2D being the rarest. There is a binding place in the channel pore for Mg2+, and at resting membrane potential, Mg2+ is attached to this bin­ding site, blocking ion flow through the channel [3,12,22].
Generally, NMDARs occur in many but not all cerebral cortex neurons and some cortical astrocytes. They are mo­stly located on dendrites. Immunocytochemical studies have revealed that NMDARs are less frequently found in the 4th layer than in layers 2 to 3 and 5 to 6, where they are preferentially expressed by pyramidal neurons. This is in agreement with the notion that afferent glutaminergic in­put reaches the cerebral cortex through the thalamocortical pathway, formed by axons of the 4th layer. NMDA recep­tors are localised in the postsynaptic membrane, organised by a multi-protein structure called the postsynaptic densi­ty (PSD). These receptors, however, are mobile and move between the synaptic and extra-synaptic pools. The PSD is defined as a type of postsynaptic membrane that conta­ins high concentrations of glutamate receptors, ion chan­nels, kinases, phosphatases and associated proteins [2,13].
Considering the NMDAR molecular structure, one may surmise that they have a common membrane topology with a large extracellular N-terminus, a membrane region composed of three transmembrane segments, and differ in the cytoplasmic C-terminus, depending on the vario­us subunits, which interacts with numerous intracellular proteins [3,22].
Opening of the channel pore by NMDAR requires the si­multaneous binding of the major agonist - glutamate (which play a neurotransmitter role) and the co-agonist - glyci­ne (or D-serine, which act as modulators). The glutama­te binding site is located on the NR2 subunit, whereas the glycine binding site is located on the NR1 subunit. After presynaptic glutamate release, and with sufficient glycine concentration in the synaptic cleft, NMDAR activation takes place. When the membrane is depolarised, the voltage-de­pendent Mg2+ block is removed from the channel interior, allowing for ion to enter. By contrast, AMPARs, composed of various combinations of four subunits (GluR1-GluR4) are permeable to Ca2+ only in the absence of the GluR2 element [3,12,13,22].
The general scheme of NMDA receptor structure together with the potential sites for pharmacological action descri­bed below presents figure 1.
Figure 1. NMDA receptor scheme and selected NMDAR-based pharmacological strategies for treatment [12,22]

NMDARs can be divided into two classes according to the­ir conductance properties: high conductance channels (bu­ilt from NR2A or NR2B) and low conductance ones (for­med by NR2C and NR2D) with reduced sensitivity to Mg2+ block. An influx of extracellular calcium initiates complex signalling pathways compromised of the mitogen-activa­ted protein kinase (MAPK) superfamily that transduces excitatory signals across the neuron. Taking into conside­ration in vitro substances, MAPKs have been also called microtubule associated protein-2 kinase (MAP-2 kinase), myelin basic protein kinase (MBP kinase), ribosomal S6 protein kinase (RSK-kinase) and epidermal growth factor (EGF) receptor threonine kinase (ERT kinase). MAPK ac­tivation was observed in response to NMDARs stimulation, starting with tyrosine phosphorylation of MAPKs. An ad­ditional class of kinases, named MAP kinase kinases and MAPK kinase kinases (MAPKK kinases), are then excited, through an intermediate step involving MAPK stimulation. Usually, three MAPKs are distinguished: extracellular si­gnal-regulated kinase-1 (ERK1 and 2), the Jun N-terminal kinases (JNKs) activating Jun transcription factor, and the p38 MAPKs. The two last ones are called stress-activated kinases (SAPKs), because they are stimulated by stressful conditions. Consequently, the transcription factor cAMP­-Ca2+ response element-binding protein (CREB) causes the expression of genes that encode brain-derived neuro­trophic factor (BDNF) and other factors promoting neuro­nal survival and activity [5,9,24,27]. Detailed information concerning NMDAR and MAPK signalling is given in an excellent review prepared by Haddad [9].
AMPARs, similar to NMDARs, couple to the MAPK pa­thways through similar sets of signalling systems and with Ca2+ entrance resulting in the synthesis of CREB and other transcription factors. However, AMPARs also can activa­te MAPK through Ca-independent mechanisms through the SRC-family tyrosine kinase. In summary, at the mo­lecular level NMDARs interact with MAPKs, creating a neurochemical axis that regulates neuronal functions [6].
As a side note, it should be mentioned that the glutama­te system changes together with development and growth. GABA (γ-aminobutyric acid), the main inhibitory neuro­transmitter in the adult central nervous system, acts as an excitatory neurotransmitter in the early postnatal state, in­fluencing its GABA-A receptors. It has been revealed that GABA-A receptor activation depolarizes neuroblasts and immature neurons in all brain regions. It is unknown why GABA operates as an excitatory agent in neonatal neuro­nes while being an inhibitory one in later stages. Factors responsible for the shift from this excitatory to inhibito­ry action have not been determined so far. Moreover, glu­taminergic transmission is initially purely mediated by NMDARs, without any contribution of AMPA receptors. Thus, these three receptors exhibit a sequential participa­tion in neuronal excitation [1].
2. Glutamate and synaptic plasticity
The glutamate system is thought to be involved in learning and memory processes. It is associated with the phenomenon of synaptic plasticity - the variable efficacy of neurotransmis­sion which enables the brain to store memories and experien­ces. This phenomenon requires gene transcription and pro­tein synthesis to stabilize synaptic changes over time. Both NMDA and AMPA receptors are suspected in regulating sy­naptic plasticity, however the consensus exists that AMPARs are responsible for short-term changes in synaptic strength, while NMDARs affect genes that are necessary for long-term maintenance of these changes. In short, AMPAR depolarisa­tion of the postsynaptic membrane facilitates NMDAR acti­vation, which in turn modulates surface AMPAR presence. Thus, postsynaptic changes develop changes in the synap­tic strength which is characterised by long-term potentiation (LTP) or long-term depression (LTD) [16,25].
LTP is defined as a strengthening of synaptic transmission that is long lasting (at least more than an hour), commonly induced by brief, high-frequency stimulation. LTD is regar­ded as being long-lasting suppression of synaptic strength that is elicited by low frequency stimulation, typically re­sulting from NMDAR activation. LTP is believed to be a key molecular element involved in learning and memo­ry, while LTD is said to be the means by which we obta­in information storage and consolidation in the brain [16].
Both LTP and LTD are forms of associative plasticity based on positive feedback, counteracting both the maximum and minimum synaptic strength changes in an effort to norma­lise neuronal excitability. Decreased neuronal activity le­ads to a homeostatic increase in the strength of excitatory synapses, while increased neuronal activity has the oppo­site effect. This preserves the balance between both exci­tatory and inhibitory synapses, termed synaptic scaling. Under basal conditions, the ratio of the relative activities of NMDARs and AMPARs is about 1:1. During "pre-LTP", the ratio between AMPARs and NMDARs is disturbed be­cause of AMPAR postsynaptic expression. This decreases the NMDAR: AMPAR ratio (1:3) and triggers postsynap­tic removal, a decrease in the number of AMPA receptors, and proportional NMDAR potentiation. The result of these processes is a LTP state. While the ratio between NMDARs and AMPARs is restored (1:1), the synapses become twi­ce as strong, with twice as many AMPARs and NMDARs as in the "pre-LTP"state [25].
3. Glutamate excitotoxicity in pathophysiology of selected neurological disorders. The potential role of NMDAR antagonists in pharmacotherapy
As summarised above, glutamate and glutamate receptors are engaged in cognition and behaviour control. However, an excessive amount of glutamate and over-activation of glutamate receptors leads to neuronal cell injury. Thus, the same processes which are essential and critical for normal neuronal functioning, in excess lead to excitotoxic cell de­ath. This phenomenon is termed glutamate-related excito­toxicity and was first used by John Olney in 1970 [16,21].
The mechanisms of excitotoxicity arise from many factors. In pathological conditions, an excessive Ca2+ influx into the neuron promotes various processes resulting in dendri­tic and/or synaptic damage and cell necrosis or apoptosis. This is the consequence of Ca2+ mitochondrial overload, causing oxygen free radical formation, caspase activation, and intracellular protein degradation. This calcium overlo­ad also causes Ca-dependent activation of neuronal NOS which in turn causes overproduction of toxic peroxynitri­te ion (ONOO-). An important element of excitotoxicity is the stimulation of mitogen-activated protein kinase p38 (MAPK p38) which activates transcription factors affec­ting neuronal apoptosis [13,17].
The phenomenon of excitotoxicity is a target of both neu­roprotective efforts and modern pharmacotherapy since it is implicated in the pathophysiology of many acute and chronic neurological diseases mentioned below.
3.1. Traumatic brain injury
Many factors may lead to an excessive glutamate presen­ce. One of them is mechanical insult - head or spinal cord injuries lead to a sudden, large glutamate release from in­jured neurons. This "fast" excitotoxicity in traumatic bra­in injury (TBI) has been confirmed in both animal mo­dels and human studies. It points to the neuro-protective potential of NMDAR antagonists in TBI treatment. Some experimental evidence supports this hypothesis, although several clinical programmes were terminated premature­ly because of failure to find any benefits in using the same agents as in clinical trials dealing with stroke (see below). Selfotel and dexanabinol, some of the agents studied in TBI, were found to have no impact on mortality or clinical outcomes; however the number of participants was inade­quate to determine unambiguous benefits and risks. Data from other studied agents (aptiganel and eliprodil) rema­in inadequately reported [13,19].
3.2. Ischemic stroke
The next important set of entities manifesting in a sudden and high glutamate delivery are neuronal ischemic events. During an ischemic stroke, energy deprivation of membra­ne protein pumps causes many neurons to lose their abi­lity to maintain ionic homeostasis. This causes their de­polarization, lysis, and/or autodestruction, similar to that observed in traumatic injury. Energy failure causes abnor­mal glutamate accumulation and prolonged synapse acti­vation, mostly because of impaired glutamate re-uptake into astrocytes [11,17,19].
Among all studied entities, the preclinical rationale for acu­te brain ischemia treatment with non-selective NMDAR antagonists was possibly the strongest. Nevertheless, cli­nical studies with these agents have failed so far. The re­asons for this failure are multifactorial. First, there have been many difficulties regarding the methodology of con­ducted clinical trials: patient selection, problems with blin­ding, random allocation of treatment dose, and interactions of drugs with anaesthetic agents. Moreover, experimental studies suggest that NMDAR antagonists are most effecti­ve when given in the pre-ischemia phase or up to two hours after induction of ischemia - patients were often diagno­sed and treated at a later time, beyond the short therapeu­tic time window. Moreover, the levels of the studied agents cannot be easily measured in the brain, and plasma con­centrations determined in previous studies were consisten­tly below values needed for maximal protection as repor­ted in animal models. On the other hand, higher doses of NMDAR antagonists were reported to induce serious side effects: psychomimetic effects (hallucinations, agitations, peripheral sensory disturbances), a centrally mediated in­crease in blood pressure, nausea, vomiting and catatonia. These disadvantages could have been circumscribed if NR2B selective NMDAR antagonists - ifenprodil or tra­xoprodil were administrated since they produce minimal side effects. However, the enthusiasm for these appro­aches has diminished because of their variable neuropro­tective efficacy. This may be associated with reduced af­finity and unbinding from inactivated NMDARs, leading to transient activation of unaffected receptors. Thus, fur­ther studies are required, particularly in order to determi­ne the receptor mechanism. There are also efforts to intro­duce another NMDAR antagonist - gavestinel - in stroke pharmacotherapy [13,19].
At present, the basic treatment for stroke is still thrombo­lytic therapy.
3.3. Neurodegenerative disorders
A high glutamate level is also involved in the pathophy­siology of many slowly progressive neuronal disorders. The concept of "slow excitotoxicity" was developed as a pathogenetic factor explaining the gradual neuronal loss in neurodegenerative diseases such as Alzheimer's dise­ase (AD), Huntington's disease (HD), Parkinson's disease (PD), multiple sclerosis, HIV-associated dementia, and amyotrophic lateral sclerosis (ALS). In these diseases, it is suspected that long-term exposure to moderate levels of glutamate causes NMDARs hyperactivity, resulting in the apoptotic-like cell death of neurons [17,19].
For many years, NMDAR antagonists have been known to be effective in animal models of Parkinson's disease (PD) which could be explained by the pathophysiological premises of PD. The depletion of nigrostratial dopamine produces over-activation of glutaminergic pathways to the stratium, which contributes to dyskinesia. Amantadine, a low-affinity NMDAR blocker, is used as an adjuvant the­rapy in PD (especially because of its additional ability to enhance dopamine release and impair dopamine re-upta­ke). It improves levodopa-induced dyskinesias (see fur­ther). Some preclinical studies have shown that non-se­lective NMDAR antagonists act in a synergistic way with levodopa and dopamine agonists. Remacemide is a sodium channel blocker and its principal metabolite - remacemi­de desglycine - is a low affinity uncompetitive NMDAR antagonist. The reports of some clinical trials suggest im­proved overall motor function, however these results were not statistically significant [10,13,19]. This drug was also tested for treatment of refractory focal seizures, showing its efficacy after 14 to 15 weeks of administration in three clinical trials including over 500 patients. Dizziness was a notable side effect. These agents also have pharmacoki­netic interactions with several conventional anti-epileptic drugs. The exact mechanism of remacemide action in epi­lepsy is unknown, but a general agreement exists amongst researchers that NMDAR participation in the pathogene­sis of epilepsy is without any doubts. A strong argument supporting this hypothesis is the efficacy of felbamate the­rapy [19,26].
A low-affinity uncompetitive antagonist - memantine - has also been studied as a treatment for dementias including Alzheimer's disease (AD, see the last chapter). Considering the role of NMDARs in cognition, it seems irrational that NMDAR antagonists may improve AD symptomatology [13,19]. On the other hand, there are links between exci­totoxicity and AD development. Misfolded mutant prote­ins are thought to be implicated in the pathogenesis of this disease, especially soluble oligomers of β-amyloid (Aβ) peptide and hyperphosphorylated tau proteins. Oxidative stress and intracellular Ca2+ influx both enhance synthesis of these proteins. Moreover, Aβ increases the NMDAR re­sponse and inhibits glutamate re-uptake. Thus, NMDAR disturbances are secondarily co-responsible for AD deve­lopment, also diminishing learning and memory functio­ning [17]. Preclinical studies have reported that a small an­tagonism of NMDARs improves learning of certain tasks, although the precise mechanism underlying this observa­tion remains unknown.
There are also reports indicating that NMDAR distur­bances participate in neurodegeneration associated with Creutzfeld-Jacob disease (CJD). The endogenous cellular prion protein PrPC protects from excitotoxicity by down-re­gulating the NMDAR subpopulation. CJD is characterised by a progressive misfolding of the PrPC form into a patho­logical PrPSc form, that may contribute to a neuroprotec­tive function loss, leading subsequently to excitotoxicity and neurodegeneration [12].
3.4. Schizophrenia
Contrary to previous findings, there is evidence indicating that potentiating NMDARs seems be beneficial for schizo­phrenia treatment. Schizophrenia is a chronic psychiatric condition characterized by positive, negative and cogniti­ve symptoms. Positive symptoms include hallucinations, delusions, disorganised speech and behaviour. Negative symptoms include flattened or restricted affect and lack of motivation. Cognitive symptoms include progressive me­mory and learning disturbances and symptoms associated with cortical processing [14].
The pathophysiology of schizophrenia is associated with several neurotransmitter systems. A common theory exi­sts implicating altered dopaminergic transmission in schi­zophrenia development. However, a different hypothesis implicates glutamate in the pathogenesis of schizophre­nia. This second hypothesis arises from studies revealing that administration of non-competitive NMDAR antagoni­sts (such as phencyclidine or ketamine) disrupts cognitive and behaviour functions, producing a schizophrenia-like syndrome, recapitulating both positive and negative symp­toms. When administrated to schizophrenic patients, these antagonists can worsen symptoms. These findings indica­te that schizophrenia is also characterized by diminished glutamate system and NMDAR activity. Thus, achieving an increase in the activity of the glutamate system through the administration of dopaminolytic agents seem to be a logical therapeutic option [7,14,15,20].
3.5. NMDARs in pain
There are findings supporting a hypothesis that NMDARs are co-responsible for nociception. Animal studies reve­aled that NMDARs are located in unmyelinated and my­elinated axons of peripheral tissues. It has been demon­strated that peripheral nociceptive fibres express NR2B and NR2D subunits of NMDARs, while NR2A subunits appear to be absent from afferent terminals. Consistent with this, local glutamate injection results in nocicepti­ve behaviour. Use of NR2B-selective antagonists poten­tiates NMDAR inhibition and should alleviate pain esthe­sia. NMDAR numbers increase in inflammatory changes, contributing to allodynia and hyperalgesia that both have a peripheral and a central component. Allodynia results from low-intensity stimuli acting via low-threshold affe­rents, generating pain. The phenomenon of hyperalgesia develops from noxious input that generates a pain respon­se but with augmented amplitude and duration. The central sensitisation of the spinal cord is mediated by presynap­tic NMDAR activation. Many primary afferents termina­ting in the dorsal horn express NMDARs and activation of presynaptic NMDARs results in the release of substance P (SP), calcitonin gene related peptide (CGRP), and glu­tamate from primary afferents. This facilitates and pro­longs nociceptive transmission to the central nervous sys­tem, resulting in an elevation of dorsal horn excitability. The NMDAR-induced increased nociception seems also to be related to a disinhibition phenomenon - a blockade of inhibitory mechanisms which should suppress hypere­sthesia. Disinhibition may be a consequence of a reduc­tion in inhibitory neurotransmitters such as GABA (g-ami­nobutyric acid), disturbed or diminished GABA receptors or loss of inhibitory neurons. In physiological conditions, the excessive participation of NMDARs in synaptic trans­mission is prevented by GABA-A receptor mediated hy­perpolarizing currents. Lack of this inhibition after GABA system deprivation and low-intensity stimulation begins to induce central sensitization which never takes place under physiological states [4,19,23].
At present, glutamate antagonists are proposed to be ef­fective in postoperative pain. Postoperative hyperalgesia is a complex modulated by peripheral, spinal and supra-spi­nal level perception, caused by several physiopathologic mechanisms. Postoperative pain involves many neurotran­smitter systems that either facilitate or inhibit nociception of somatic, neuropathic, inflammatory or visceral origins. Ketamine is a well-known NMDAR antagonist agent with a confirmed efficacy in reducing surgery-induced hyperal­gesia. This drug was developed as an anaesthetic agent and has been demonstrated to cause "dissociative anaesthetics." Many small trials confirm the adjuvant role of ketamine in subanaesthetic doses during analgesia. On the other hand, ketamine must be administrated via a systemic route to pro­duce its analgesic effect, excluding the suspected direct neuronal NMDAR blocking. The mechanism of action of systemic ketamine may be related to different elements - this drug produces multiple pharmacologic effects and in­teracts with many systems that mediate analgesic effects unrelated to NMDAR antagonism. Ketamine is an agonist of opioid receptors, exhibits inhibition of neuronal nico­tinic receptors and activates the mono-aminergic descen­ding inhibitory system that modulates nociception in the dorsal horn. Ketamine also suppresses the production of pro-inflammatory cytokines (TNF-α, Il-6) and exerts a di­rect anti-inflammatory effect on macrophages. Thus, these other ketamine actions, rather than pure NMDAR antago­nism, may account for the efficacy of this drug in posto­perative pain management [4,23,28].
Apart from ketamine, several other drugs posses antago­nistic NMDAR properties. Dextrometorphan, a D-isomer of the codeine analogue levorphanol, was shown to inhi­bit central sensitization in experimental studies. However, clinical use has been disappointing. Patients receiving de­xtrometorphan parenterally reported less pain in the ear­ly postoperative period, although the results were inconc­lusive. As a side note, dextrometorphan is being tested in clinical trials involving children suffering from Rett's syn­drome - a neurodevelopmental disease affecting mostly females, characterized by the development of autistic fe­atures, stereotypic hand movements and epileptic attacks [4,23,28]. Other opioid analgesics, such as methadone or buprenorphine, also have anti-NMDAR properties at lo­wer doses than those needed to induce complete analgesia. Anticonvulsant drugs such as gabapentin also display an­ti-hyperalgesic properties and have the ability to modula­te glutaminergic neurotransmission. Nefopam, a centrally acting analgesic, also diminishes glutamate receptor acti­vity and prevents postoperative opioid overconsumption. Selected non-steroidal anti-inflammatory drugs such as ketorolac, demonstrate a central analgesic effect involving modulation of NMDAR activity [4,23,28].
In summary the inhibition of the glutaminergic system is suspected to be involved in the pain-relief mechanism of many drugs. Ketamine may also have a role in pre-empti­ve analgesia for surgical procedures, however, this comes at the expense of a small increase in the risk of psychomi­metic effects. At present, there is no evidence supporting an unambiguous rationale for other NMDAR antagoni­sts in both acute and chronic pain. This area needs further study in the future.
4. NMDAR-targeting agents - current therapies and future perspectives
Early preclinical and clinical trials of glutamate targeting drugs concentrated on NMDARs and three main types of antagonists have been studied: competitive NMDAR an­tagonists (glutamate - e.g. selfotel or glycine binding site - e.g. gavestinel), non-competitive allosteric drugs, and/or NMDAR channel blockers. As was mentioned before, these agents were not effective for the following indica­tions: ischemic stroke, and TBI, because of the restrictions discussed above. Further development of these NMDAR antagonists is unlikely and the initial enthusiasm for this kind of therapy has been abolished [12,13,19].
Lack of success caused a diminished interest in the glu­tamate system as a potential target of modern pharmaco­therapy. It was surprising to find that well-tolerated drugs with multiple mechanism of action, demonstrating benefits for some neurological conditions, also exhibit anti-NMDA properties. These were mentioned above: felbamate, rilu­zole, amantadine and memantine.
Felbamate was initially screened for anticonvulsant acti­vity in experimental models and showed a broad antico­nvulsant profile similar to valproic acid but with less neu­rotoxicity. The complex pharmacodynamic mechanism of felbamate involves inhibition of voltage-dependent Na+ and Ca2+ channels as well as NMDAR antagonism. These drugs seem to be a non-competitive allosteric inhi­bitors with some selectivity for NR2B-containing recep­tors. During clinical trials, felbamate produced no adverse side effects specific to other anti-epileptic drugs or anta­gonists tested in TBI and ischemic stroke. The most com­mon reported side effects of felbamate were nausea, ano­rexia and insomnia. However, pharmacovigilance studies published after the introduction of this agent to common clinical practice revealed two rare but of special importan­ce reactions related to felbamate: aplastic anaemia (with incidence of about 1:8000) and hepatotoxicity (1:26000). These unexpected adverse events have limited felbamate clinical use - nowadays this drug is recommended only in intractable partial seizures and in Lennox-Gastaut syndro­me that is refractory to primary therapy [12].
Riluzol was also developed as an anticonvulsant agent, but findings demonstrating its modulatory effects in NMDAR-mediated neuronal death caused that further development of riluzol was geared towards its neuroprotective func­tions. It was discovered that riluzol is a Na+ channel bloc­ker in NMDAR containing neurons and that it prevents the entrance of neuronal Ca2+, stabilising NMDARs and pro­tecting them from depolarization. This drug also enhanced glutamate clearance from the synaptic cleft by increasing glutamate re-uptake. The most promising results of rilu­zol were in clinical trials obtained for amyotrophic lateral sclerosis (ALS). The drug showed a statistically significant survival advantage and a 9% absolute risk reduction com­pared to placebo. Patients receiving riluzol reported small beneficial effects on limb function. The drug was well tole­rated with asthenia, nausea and elevation of liver enzymes being the most frequent side-effects. Although the efficacy of riluzol in treating ALS was not acceptable, the drug was still approved in many countries for ALS treatment [12].
Amantadine was introduced into clinical practice for the prophylaxis of respiratory infections due to influenza. It was discovered that this drug had anti-Parkinsonian properties through direct dopaminergic activity. Further studies re­vealed that amantadine is also an NMDAR antagonist ac­ting as an open-channel blocker. Further clinical trials are still needed to establish this drug's value in PD [10,12,19].
Memantine was synthesised as a potential hypoglycaemic agent, however it was ineffective in lowering glucose blo­od level. On the other hand it was observed that cognitive functions of diabetic patients treated with memantine were improved, thus this drug was used in the treatment of de­mentia. Memantine, similar to amantadine, was discovered to be an open-channel NMDARs blocker but it was shown to have a relatively low affinity for NMDARs, allowing fast binding but also rapid dissociation from the receptors. Moreover, the action of memantine is voltage-dependent and therefore, this agent did not block NMDAR channels upon strong postsynaptic depolarization (which is speci­fic during physiological conditions). The NMDARs were blocked during moderate, long-lasting depolarization, ob­served in pathological, excitotoxic conditions. Memantine was also shown to inhibit nicotinic cholinergic receptors that, when activated, also contributes to amyloid-β - in­duced tau protein phosphorylation and AD development. This drug was well-tolerated with nausea, diarrhoea, he­adache, insomnia, and dizziness being the most serious adverse effects. It was approved for moderate and severe AD treatment in most of the European countries, the USA and Canada. Clinical trials were carried out on memantine in the treatment of frontotemporal lobar degeneration, PD dementia, cognitive symptoms related to Huntington's di­sease, and cognitive dysfunction in the course of TBI; this drug may also be neuroprotective in these entities [12,19].
It should be mentioned that novel agents focused on AD tre­atment are being developed. There are two other NMDAR targeting agents currently being studied: neramexane and dimebon. Neramexane, similarly to memantine, is an open­-channel NMDAR blocker and displays a similar pharma­cokinetic and comparable clinical tolerability. Dimebon, which was initially classified as an antihistaminic drug, is being studied as a potential agent for AD and HD becau­se it also blocks NMDARs [12].
Prospective NMDAR antagonists and other agents affecting the glutamate system are based on two major strategies, which do not target the extracellular part of NMDARs, as other medications have so far.
The first novel strategy is based on the transporter systems regulating the glutamate amount in the synaptic cleft and the second one targets intracellular proteins which are in­volved in NMDAR signalling pathways.
Glutamate released into the synaptic cleft is then remo­ved using a family of excitatory amino acids transporters (EAATs). The idea of developing glutamate re-uptake ac­tivators arose thanks in part to the theory behind selecti­ve serotonin re-uptake inhibitors. There are five types of EAATs (EAAT1 - EAAT5) with EAAT2 responsible for most of the glutamate turnover in the brain. Agents incre­asing EAAT2 activity or its membrane expression on both neurons and glia might provide new a therapeutic appro­ach in reducing glutamate-mediated excitotoxicity becau­se of glutamate clearance enhancement. It was found that ceftriaxone, a beta-lactam antibiotic, increases EAAT2 ac­tivity in mouse models of ALS. It is currently going thro­ugh clinical trials in regards to this. There are also reports that riluzol, with its complex mechanism of action, also increases the activity of EAAT2 [12].
Among signalling proteins, there are many potential en­zymes (such as protein kinases and phosphatases) which may act as modulatory sites by interfering in signalling do­wnstream to the nucleus. The idea behind such agents al­ready exists in modern oncology. Several molecules dri­ving neoplastic transformation have been identified. This has allowed us to develop designed cancer therapeutics which can inhibit intracellular neoplastic signal transduc­tion. An example of this are monoclonal antibodies (ima­tinib or trastuzumab). One molecular target that regulates NMDAR activity is protein kinase C. In animal studies, it was demonstrated that inhibition of this enzyme impairs spatial memory. Consequently, artificial elevation of pro­tein kinase C activity might be especially effective in neu­rological disturbances with memory impairment (AD and other dementias). Studies concerning protein kinase C ac­tivators are ongoing, with the most advanced studies invo­lving bryostatin-1 and nefiracetam [12].
As mentioned above, there are attempts to increase glu­tamate system function and NMDAR activity in schizo­phrenia. Already some opportunities exist when it comes to enhancing NMDAR function. Some drugs have alre­ady been studied in pre-clinical and clinical trials. Direct glutamate agonists may be used, although their applica­tion is limited because of possible overdosing and excito­toxicity. Other therapeutic opportunities involve the use of NMDAR modulatory site agonists, especially agonists ac­ting on the glycine/D-serine site. When administrated to­gether with classic neuroleptic agents, they significantly improved symptoms. The antibiotic D-cycloserine is a par­tial glycine site agonist that improves glutaminergic activi­ty in schizophrenia, however, further studies are required to establish its usefulness in the treatment of this disease. Pilot studies suggesting that the core symptoms of social impairment in patients with autism may be improved when the patients are treated with D-cycloserine are under way.
An interesting approach in the development of antipsy­chotic agents involves the glutamate system. This appro­ach increases extracellular glycine levels through re-upta­ke inhibition (by blocking the glyT1-transporter), similar to serotonin re-uptake inhibitors in depression pharma­cotherapy. One example of these blockers is sarcosine. Sarcosine was more effective in ameliorating both posi­tive and negative schizophrenic symptoms when compa­red to D-serine. This finding suggests that the central gly­cine level in schizophrenia may be insufficient to saturate NMDARs [8,12,14,18].
The summary of NMDAR - targeting agents discussed is presented in table 1.
Table 1. NMDAR-targeting agents [12,19]

Despite several failures concerning the development of NMDAR targeting drugs, it seems that when based on ge­neral pathophysiological premises, the search for agents preventing excitotoxicity phenomenon is both important and relevant. One should expect increasing knowledge con­cerning the glutamate system, NMDARs and methods of affecting glutaminergic activity, along with further rese­arch into glutamate-targeting agents, as modern pharma­cotherapy for certain neurological disturbances progresses.
[1] Ben-Ari Y., Khazipov R., Leinekugel X., Caillard O., Gaiarsa J.L.: GABA-A, NMDA and AMPA receptors: a developmentally regulated "menage a trois". Trends Neurosci., 1997; 20: 523-529
[2] Conti F.: Localization of NMDA receptors in the cerebral cortex: a schematic overview. Braz. J. Med. Biol. Res., 1997; 30: 555-560
[3] Cull-Candy S., Brickley S., Farrant M.: NMDA receptor subunits: diversity, development and disease. Curr. Opin. Neurobiol., 2001; 11: 327-335
[4] De Kock M.F., Lavand'homme P.M.: The clinical role of NMDA receptor antagonists for the treatment of postoperative pain. Best Pract. Res. Clin. Anaesthesiol., 2007; 21: 85-98
[5] Groc L, Bard L., Choquet D.: Surface trafficking of N-methyl-D-aspartate receptors: physiological and pathological perspectives. Neuroscience, 2009; 158: 4-18
[6] Groc L., Choquet D.: AMPA and NMDA glutamate receptor trafficking: multiple roads for reaching and leaving the synapse. Cell Tissue Res., 2006; 326: 423-438
[7] Grosjean B., Tsai G.E.: NMDA neurotransmission as a critical mediator of borderline personality disorder. J. Psychiatry Neurosci., 2007; 32: 103-115
[PubMed]  [Full Text HTML]  [Full Text PDF]  
[8] Gunduz-Bruce H.: The acute effects of NMDA antagonism: from the rodent to the human brain. Brain Res. Rev., 2009, 60: 279-286
[9] Haddad J.J.: N-methyl-D-aspartate (NMDA) and the regulation of mitogen-activated protein kinase (MAPK) signaling pathways: a revolving neurochemical axis for therapeutic intervention? Prog. Neurobiol., 2005; 77: 252-282
[10] Hallett P.J., Standaert D.G.: Rationale for and use of NMDA receptor antagonists in Parkinson's disease. Pharmacol. Ther., 2004; 102: 155-174
[11] Hoyte L., Barber P.A., Buchan A.M., Hill M.D.: The rise and fall of NMDA antagonists for ischemic stroke. Curr. Mol. Med., 2004; 4: 131-136
[12] Kalia L.V., Kalia S.K., Salter M.W.: NMDA receptors in clinical neurology: excitatory times ahead. Lancet Neurol., 2008; 7: 742-755
[13] Kemp J.A., Mc Kernan R.M.: NMDA receptor pathways as drug targets. Nature Neurosci., 2002; 5 Suppl.: 1039-1042
[14] Kristiansen L.V., Huerta I., Beneyto M., Meador-Woodruff J.H.: NMDA receptors and schizophrenia. Curr. Opin. Pharmacol., 2007; 7: 48-55
[15] Krystal J.H., D'Souza D.C., Petrakis I.L., Belger A., Berman R.M., Charney D.S., Abi-Saab W., Madonick S.: NMDA agonists and antagonists as probes of glutamatergic dysfunction and pharmacotherapies in neuropsychiatric disorders. Har. Rev. Psychiatry, 1999; 7: 125-143
[16] Lau C.G., Zukin R.S.: NMDA receptor trafficking in synaptic plasticity and neuropsychiatric disorders. Nat. Rev. Neurosci., 2007, 8: 413-426
[17] Lipton S.A.: Pathologically-activated therapeutics for neuroprotection: mechanism of NMDA receptor block by memantine and S-nitrosylation. Curr. Drug Targets, 2007, 8: 621-632
[18] Millan M.J.: N-methyl-D-aspartate receptors as a target for improved antipsychotic agents: novel insights and clinical perspectives. Psychopharmacology, 2005; 179: 30-53
[19] Muir K.W.: Glutamate-based therapeutic approaches: clinical trials with NMDA antagonists. Curr. Opin. Pharmacol., 2006; 6: 53-60
[20] Newcomer J.W., Krystal J.H.: NMDA receptor regulation of memory and behaviour in humans. Hippocampus, 2001; 11: 529-542
[21] Olney J.W., Ho O.L.: Brain damage in infant mice following oral intake of glutamate, aspartate or cysteine. Nature, 1970; 227: 609-611
[22] Paoletti P., Neyton J.: NMDA receptor subunits: function and pharmacology. Curr. Opin. Pharmacol., 2007; 7: 39-47
[23] Petrenko A.B., Yamakura T., Baba H., Shimoji K.: The role of N-methyl-D-aspartate (NMDA) receptors in pain: a review. Anesth. Analg., 2003; 97: 1108-1116
[PubMed]  [Full Text HTML]  [Full Text PDF]  
[24] Popescu G.: Mechanism-based targeting of NMDA receptor functions. Cell Mol. Life Sci., 2005; 62: 2100-2111
[25] Rao V.R., Finkbeiner S.: NMDA and AMPA receptors: old channels, new tricks. Trends Neurosci., 2007; 30: 284-291
[26] Sagratella S.: NMDA antagonists: antiepileptic neuroprotective drugs with diversified neuropharmacological profiles. Pharmacol. Res., 1995; 32: 1-13
[27] Stephenson F.A., Cousins S.L., Kenny A.V.: Assembly and forward trafficking of NMDA receptors. Mol. Membr. Biol., 2008; 25: 311-320
[28] Suzuki M.: Role of N-methyl-D-aspartate receptor antagonists in postoperative pain management. Curr. Opin. Anaesthesiol., 2009; 22: 618-622
The authors have no potential conflicts of interest to declare.