Glutamate continues to surprise us. Today's graduate students will find it hard to believe that a mere 25 to 30 years ago proponents of a neurotransmitter role of glutamate were decidedly avant-garde and that the suggestion that glutamate might play a specific role in the pathophysiology of neurological and psychiatric diseases was considered heretic at best. What a difference a few years can make: by the mid-1980s, glutamate was widely accepted to be the major excitatory neurotransmitter in the mammalian central nervous system, excitotoxicity had become a neuroscientific household term, and the pioneers of the day began to advocate the development of “glutamatergic” drugs to combat human brain diseases. These appealing concepts persuaded some of the best minds in the neurosciences to study glutamate neurotransmission in depth. In rapid succession, we learned about the existence of distinct ionotropic and metabotropic receptors and their unique and discretely localized regulatory sites and processes. Molecular biological techniques uncovered a surprisingly diverse family of glutamate transporters, identified receptor subunits, and provided mice with targeted deletion or overexpression of certain glutamatergic functions. The use of molecular tools further solidified the central importance of glutamate neurotransmission in the normal development and function of the mammalian brain and its critical role in diseases of the central nervous system.
The involvement of glutamatergic neurotransmission in brain physiology and pathology, and the ever increasing number of new molecular targets for glutamatergic drug development, has provided a bonanza for pharmacological research. The recent, third ASPET-Ray Fuller Symposium was therefore a timely attempt to review current knowledge in the field and to evaluate emerging concepts with an eye toward pharmacological exploitation.1 The one-and-a-half-day meeting, named in memory of one of the major figures of contemporary psychopharmacology, brought together experts from the preclinical and clinical realm who presented complementary, and often synergistic, data to an audience from academia, government, and industry.
The regulation of the expression and function ofN-methyl-d-aspartate (NMDA) receptors was a major conference topic. Using quantitative immunogold electron microscopy, John Morrison (Mount Sinai University, New York, NY) showed that the number of NR1 subunits of the receptor per synapse increases in the outer molecular layer of the hippocampal dentate gyrus within days after a perforant path transection. This and the demonstration that a similar increase occurs in area CA1 of the hippocampus in old ovariectomized rats provides anatomical evidence for a role of NMDA receptor proliferation in synaptic plasticity. Results from the laboratory of Suzanne Zukin and Michael Bennett (Yeshiva University, New York, NY) demonstrate that protein kinase C (PKC) modulates NMDA receptor trafficking and channel gating. PKC delivers new NMDA receptors to the surface by SNARE-dependent exocytosis. The degree of potentiation of NR1/NR2A receptors is greatest for NR1 splice variants which have the shortest carboxy terminal; these splice variants also exhibit the highest efficiency of insertion at the cell membrane by constitutive exocytosis. Notably, these mechanisms might be exploitable for the development of drugs that affect NMDA receptor function in subtle ways—and only when PKC is up-regulated under certain physiological and pathological conditions. A fundamentally different, novel pharmacological approach to influence NMDA receptors takes advantage of the fact that several endogenous ligands normally regulate receptor function in the brain (Robert Schwarcz, University of Maryland, Baltimore, MD). The actions of glycine, the prototypical coagonist of the NMDA receptor, have been documented most extensively, but several other small molecules compete for the same site. Of these, the agonist d-serine and the antagonists kynurenate and N-acetyl-aspartylglutamate (NAAG) are formed and degraded primarily in astrocytes. The processes of these glial cells closely appose, and often surround, glutamatergic synapses, and each of the three compounds has been demonstrated to influence NMDA receptor function. As pointed out by Schwarcz, targeting astrocytes to influence the metabolic machinery of these factors is beginning to yield effective glutamatergic compounds, such as inhibitors of kynurenine 3-hydroxylase (to enhance kynurenate formation) or NAALADase (to reduce NAAG degradation; Joseph Coyle, Harvard University, Cambridge, MA).
The need to find new and increasingly subtle ways to influence NMDA receptor function was further emphasized by Carol Colton (Georgetown University, Washington, DC), who presented evidence that glutamate, acting through the NMDA receptor, is a crucial mediator of neuronal migration at certain stages of embryonic development. In the trisomy 16 mouse, a model for Down syndrome, the decrease in the number of migrating neurons is linked to a functional defect in NMDA receptors. In agreement with a number of previous studies demonstrating a pivotal role of NMDA receptor activity in synaptic development, these findings highlighted the risks of using NMDA receptor antagonists indiscriminately during pregnancy and early life. If direct NMDA receptor blockade is required for therapeutic intervention, it is therefore clearly preferable to reduce the hazard by administering subunit-specific receptor antagonists. One possibility is to target the NR2B subunit, which is highly expressed in cortex, hippocampus, and striatum, i.e., brain areas that are exquisitely vulnerable to excitotoxic insults. This subunit is also concentrated in the dorsal horn of the spinal cord, an area involved in the processing of painful stimuli. NR2B-specific antagonists have been developed and show promise in animal models of chronic pain and Parkinson's disease (PD) (Frank Menniti, Pfizer R & D, Groton, CT). As explained by Raymond Dingledine (Emory University, Atlanta, GA), these agents are particularly attractive on theoretical grounds. For unknown reasons, compounds such as ifenprodil and its analogs have the ability to potentiate the “H+-effect”, the well documented, voltage-independent reduction in NMDA receptor function with decreasing (more acidic) pH. These drugs are therefore relatively innocuous under physiological conditions but become increasingly effective as neuroprotectants in an acidic environment, such as that encountered in cerebral ischemia and epilepsy.
Mechanisms of excitotoxicity still fascinate the leading researchers in the field of glutamate pharmacology. While the central dogma of excitotoxicity, i.e., the critical role of elevated intracellular free Ca2+ concentrations following excessive stimulation of glutamate receptors, has remained unchallenged, several additional endogenous factors are clearly of essence as well. Mark Mattson (National Institute on Drug Abuse, Baltimore, MD) showed that mild cellular stress, effected by reduced caloric intake, mobilizes endogenous protective principles, such as brain-derived neurotrophic factor (BDNF), which in turn attenuate kainate neurotoxicity. In contrast, 4-hydroxynonenal, a nonenzymatic product of membrane lipid peroxidation, enhances neurotoxicity by phosphorylation of Na+-dependent Ca2+channels. The presentation of Ted Dawson (Johns Hopkins University, Baltimore, MD) drew attention to the importance of the gaseous messenger nitric oxide (NO) in excitotoxicity. NO, coupled with mitochondrial dysfunction and superoxide anion generation, forms peroxynitrate, which in turn results in cell death through the activation of poly(ADP-ribose) polymerase (PARP). PARP inhibitors are neuroprotective, PARP knockout mice are resistant to excitotoxic insults, and vulnerability is restored when PARP is reintroduced into these mice. The role of free radicals in excitotoxicity, Dawson explained, is further illustrated by the fact that colocalized manganese superoxide dismutase conveys protection against NO in neurons containing NO synthase. Interestingly, the effects of 4-hydroxynonenal, NO, and PARP are selective for NMDA receptor-mediated excitotoxicity and do not generalize to α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors.
AMPA receptors mediate rapid excitatory synaptic transmission. Their distinct regulation includes phosphorylation of individual subunits by multiple kinases and binding to a variety of specifically interacting proteins. Richard Huganir (Johns Hopkins University) described the identification of specific binding proteins using a yeast two-hybrid system and elaborated the cascade of events that jointly control AMPA receptor trafficking. GRIPs (glutamate receptor-interacting proteins) and other proteins, as well as additional regulatory peptides, determine the proper synaptic targeting of AMPA receptors. This rapidly occurring (5–15 min) receptor mobilization, which is responsible for the activation of so-called “silent” synapses during long-term potentiation, was studied further by Roger Nicoll (University of California, San Francisco). He reported work with stargazer, an ataxic and epileptic mutant mouse, that lacks AMPA receptors at cerebellar mossy fiber-granule cell synapses. Stargazin, the mutated protein, normally evokes synaptic AMPA responses by two sequential and complementary mechanisms. First, the protein guides the receptor to the cell surface independent of the receptor's PDZ domain. Second, after binding to the PDZ domain on the C terminus of the receptor, stargazin carries the receptor protein to the synapse. The recent cloning of five members of the stargazin family, which are all selectively expressed in the brain, raises the possibility that stargazin-like processes might play an important role in the intracellular cycling of AMPA receptors.
The regulatory proteins and peptides described by Huganir and Nicoll do not affect NMDA receptor function and in fact appear to be quite specific for certain AMPA receptor subtypes. This, together with the discrete localization of these subtypes (for example, of GluR3 receptors on GABAergic neurons in area CA1 of the hippocampus; Morrison), further increases the complexity of the AMPA receptor family. In physiological terms, these intricate anatomical features and elaborate regulatory mechanisms probably define the role of AMPA receptors in learning and memory processes. Because of its ability to render heteromeric AMPA receptor assemblies Ca2+-impermeable, the GluR2 subunit of the AMPA receptor, in particular, is also involved in brain pathology. Thus, as reported by Zukin, GluR2 expression is down-regulated in vulnerable neurons before seizure- or ischemia-induced cell death. Conversely, experimental knockdown of GluR2 function by antisense oligonucleotides targeted to the GluR2 gene induces neuronal death and enhances neuronal vulnerability to an ischemic insult. These studies, taken together, emphasize the importance of AMPA receptor subunits as targets for the development of promising neuroactive drugs, ranging from cognition enhancers to neuroprotectants. There can be little doubt about the tremendous potential of this approach, and there was a consensus expectation among speakers and other symposium participants that major advances in the field will be made in the near future.
Excitatory amino acid (EAA) pharmacology is not limited to ionotropic glutamate receptors. G-protein-coupled (metabotropic; mGlu) receptors modulate glutamatergic neurotransmission by presynaptic, postsynaptic, and glial mechanisms. Receptor categorization according to second messenger linkage (groups I–III) and subgrouping according to pharmacological properties, cellular localization, and in several cases cloning (mGlu1–8), appears to be nearing completion. The major challenge at this stage is the development of subtype-selective receptor agonists and antagonists to define the biological role of each receptor in vitro and in vivo. As explained by Roberto Pellicciari (University of Perugia, Italy), the design of selective mGlu receptor ligands is greatly aided by the use of computer modeling. By identifying subtle steric differences in evolutionary divergent residues outside the well conserved glutamate binding pocket, it is possible to build subtype-selective pharmacophore models, which can in turn be applied to the synthesis of novel compounds. One successful example of this approach, a novel mGlu1 antagonist capable of reducing neuronal damage in global and focal ischemia models, was introduced at the meeting. Darryle Schoepp (Lilly Research Laboratories, Indianapolis, IN) focused a significant portion of his presentation on potent, systemically active mGlu2/3 ligands (the agonists LY354740 and LY379268 and the antagonist LY341495). Studies in relevant animal models support the idea that agonists at these receptor subtypes hold potential for the reversal of glutamate hyperfunction in pathological situations. Schoepp also indicated that powerful high-throughput screens have recently begun to yield useful research tools, such as 2-methyl-6-(phenylethynyl)pyridine, a highly potent and selective mGlu5 receptor antagonist. The expected advent of an entire battery of novel, mGlu receptor-specific compounds guarantees exciting—and maybe initially somewhat confusing—times for mGlu receptor aficionados.
Glutamate transporters, the prime determinants of ambient synaptic glutamate concentrations, too, are a heterogeneous family of proteins. They derive from different genes, show differential cellular localization, and have different pharmacological and kinetic properties. In addition, specifically colocalized glutamate transporter-interacting proteins (GTRAPs) stimulate glutamate uptake, probably by membrane stabilization (Jeffrey Rothstein, Johns Hopkins). EAAT2 (or GLT-1, using an alternate nomenclature), a ubiquitous astrocytic transporter, is believed to play a particularly dominant role in regulating extracellular glutamate levels, although the perisynaptic neuronal localization of EAAT1 and EAAT4 suggests important functions of these proteins as well (Morrison; Rothstein). The importance of EAAT2 can be demonstrated, for example, in a mouse model of amyotrophic lateral sclerosis (ALS), where overexpression of the transporter substantially prolongs the animals' life spans. Interestingly, fashionable neuroprotective drugs (such as immunophilins) or endogenous factors [such as glia-derived neurotrophic factor (GDNF)] increase EAAT2 protein and activity, indicating that a reduction of synaptic glutamate may contribute to the remarkable therapeutic potential of these compounds.
Kevin Behar (Yale University, New Haven, CT), the final speaker in the basic science session, reviewed recent methodological advances in assessing brain glutamate metabolism in vivo. As recognized many years ago, neurotransmitter glutamate is subject to a dynamic cycle involving neuronal and astrocytic processes (the “glutamate-glutamine shuttle”). The significance of glutamate and glutamine (and GABA) fluxes can now be studied quantitatively in animals and humans by NMR after i.v. [13C]glucose and [15N]NH3infusions. It turns out that the flux rate is linearly related to glucose oxidation, i.e., that glutamate cycling is activity-dependent. This work not only raises questions concerning the role of the various glutamate receptors and transporters in cerebral glutamate metabolism but highlights the active interplay between neurons and astrocytes. In practical terms, the new methodology provides an attractive, noninvasive tool to assess the effects of pharmacological agents on brain glutamate function.
The clinical significance of abnormal glutamate function, and the therapeutic potential of a variety of selective corrective measures, was emphasized by most symposium speakers. In addition, several presentations concentrated on specific brain diseases and other medically relevant problems. James McNamara (Duke University, Durham, NC) described the pathophysiology of Rasmussen's encephalitis, a childhood neurodegenerative disease that may result in part from an autoimmune attack against the GluR3 receptor. In vitro studies using mixed neuron-glia cultures and anti-GluR3 IgG revealed cytotoxicity, which required formation of a “membrane attack complex”. A series of follow-up experiments showed that neurons succumb later than astrocytes because they contain protective, complement-activated regulatory proteins (such as CD59). This work, which may have implications for a host of inflammatory, seizure, and neurodegenerative disorders, adds another level of complexity—immunological compromise—to glutamatergic pathophysiology. Further support of this link came from Ronald Dubner (University of Maryland), who studies inflammation-induced pain mechanisms. In animal models of hyperalgesia caused by inflammation, NMDA and AMPA/kainate receptor antagonists reduce central sensitization, i.e., the long-term changes that contribute to the amplification and persistence of pain. Region-specific changes in the composition of certain NMDA and AMPA receptor subunits are likely to be involved in both development and maintenance of hyperalgesia. Mechanistically, Dubner conceptualized persistent pain as the result of neuronal plasticity, akin to long-term potentiation at forebrain sites. He also proposed a circuit model in which discrete receptor changes in spinal dorsal horn neurons and subsequent receptor up-regulation in the brainstem eventually account for central sensitization at the level of the spinal cord.
Two devastating neurological disorders, PD and ALS, have long been associated with cerebral glutamate dysfunction. In PD, several lines of evidence suggest an involvement of both NMDA and AMPA receptors in nerve cell death and motor symptoms (Timothy Greenamyre, Emory University). For example, animal models of PD present with region-selective and specific alterations in NR1 and GluR1 receptors, and, perhaps more importantly, chronic NMDA receptor blockade attenuates 6-hydroxydopamine-induced neuronal loss in the substantia nigra. Moreover, systemic administration of AMPA or NMDA receptor antagonists has pronounced antidyskinetic effects in Parkinsonian monkeys and potentiates the effects of dopaminergic drugs. Based on these and similar animal experiments, clinical trials were initiated with the low-affinity NMDA channel blockers remacemide and amantadine. In a substantial subpopulation of PD patients, these drugs have a striking, beneficial effect on motor function and reduce the side effects of dopaminergic medication. In monotherapy or as adjuvants, ionotropic glutamate receptor antagonists therefore hold great promise for the treatment of PD. So far, no similar clinical studies have been performed in ALS. However, as reported by Rothstein, approximately 65% of ALS patients express an abnormal, truncated (possibly alternatively spliced) form of EAAT2, and these transporters are selectively and profoundly decimated in the patients' motor cortices. These data, taken together with the above-mentioned prolonged life span of ALS mice overexpressing EAAT2, seem to justify an antiglutamatergic strategy for the treatment of ALS.
The link between glutamatergic abnormalities and central nervous system dysfunction is not restricted to the realm of neurology. Drug abuse, for example, traditionally discussed almost exclusively in the context of adaptations of monoaminergic systems, is now understood to involve several important glutamatergic components as well (Marina Wolf, Chicago Medical School, Chicago, IL). Behavioral and cellular adaptations induced by chronic administration of drugs of abuse (i.e., addiction) can be explained by persistent AMPA receptor changes in a circuitry involving the prefrontal cortex, the nucleus accumbens, and the ventral tegmental area. Addiction may involve maladaptive synaptic remodeling and strengthening, using mechanisms not unlike those that effect long-term potentiation. Wolf's work implies that subtle, possibly subunit-specific interference with AMPA receptor function constitutes a promising approach in the fight against drug abuse. Schizophrenia, in contrast, has been hypothetically linked to a hypofunction of NMDA receptors. This concept is based primarily on the notion that the dissociative anesthetics phencyclidine or ketamine, which are noncompetitive NMDA receptor antagonists, cause the full range of characteristic schizophrenia symptoms when administered to normal subjects. NMDA receptor abnormalities, and changes in the brain content and disposition of endogenous glycine/NMDA receptor modulators such as kynurenate and NAAG, exist in schizophrenia and may play critical roles in the pathophysiology of the disease. As reviewed by Coyle, modulators of the glycine coagonist site of the NMDA receptor, such as glycine, d-cycloserine, and d-serine, reduce the negative symptoms and improve cognitive function in schizophrenic patients receiving typical antipsychotic medication. These encouraging results provide critical information, which can be used to further explore the pathophysiology of schizophrenia, and will certainly stimulate additional clinical trials based on specific glutamatergic interventions.
In spite of the breadth of new information presented, the relatively brief symposium could not do justice to the wealth of exciting developments in the area of glutamate neuropharmacology. Discoveries in the preclinical arena are made at a breathtaking pace; novel, specific, and highly efficacious compounds for hypothesis testing become available in short intervals; and interesting ideas for new clinical applications arrive often and unexpectedly. Some of the concepts emphasized by several speakers, e.g., the importance of neuron-glia interactions and of receptor- and transporter-associated binding proteins, are certainly here to stay, while others might not. The outstanding success of the symposium notwithstanding, these considerations suggest that it will be necessary to reassess the status of this dynamic research area in the not too distant future. There is much to look forward to.
- Received November 27, 2000.
- Accepted November 28, 2000.
↵1 ASPET-Ray Fuller Symposium: “The Neuropharmacology of Glutamate”, October 14–15, 2000, Baltimore, Maryland. Organizing committee: Joseph T. Coyle (Harvard University), Ted M. Dawson (Johns Hopkins University), Jeffrey F. McKelvy (Merck Research Laboratories, La Jolla, CA), and Robert Schwarcz (University of Maryland).
- protein kinase C
- Parkinson's disease
- nitric oxide
- poly(ADP-ribose) polymerase
- α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
- excitatory amino acid
- amyotrophic lateral sclerosis
- metabotropic glutamate
- N-acetylated α-linked acidic dipeptidase
- γ-aminobutyric acid
- The American Society for Pharmacology and Experimental Therapeutics