Introduction

Top Abstract Introduction Heterogeneity of NMDA Receptors Diversity of Excitotoxic... Cloning of NMDA Receptors:... Application to Excitotoxicity:... Understanding of Excitotoxicity... Conclusions References

Glutamate acts postsynaptically at several receptor types that are named for their prototypic pharmacological agonists (Dingledine et al., 1999). One major subtype of glutamate receptor involved in fast synaptic excitation is the AMPA (-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) receptor, a ligand-gated cation channel with crucial roles in synaptic physiology and some disease processes. Another glutamate receptor, the N-methyl-D-aspartate (NMDA) receptor, is also important to neurologists and neuropathologists because of its unique physiological and pathophysiological roles. This receptor is crucial to many forms of a process known as excitotoxicity, during which the inability to respond properly to elevations in synaptic concentrations of glutamate overexcites neurons, leading to neuronal death (Lipton and Rosenberg, 1994). Since many neurological disorders have been proposed to have excitotoxic components, development of NMDA receptor antagonists is a major area of pharmaceutical investigation (Table 1) (Lipton and Rosenberg, 1994). Although approaches to treatment of these disorders using NMDA receptor antagonists have been effective in animal models, successful therapy in humans has been limited by the severe side effects of complete NMDA receptor blockade including psychosis, nausea, vomiting, memory impairment, autonomic instability, and neuronal cell death (Lipton and Rosenberg, 1994; Ikonomidou et al., 1999). Improved understanding of the basic processes involved in excitotoxicity could lead to more selective therapies.

Recent evidence demonstrates that excitotoxicity is a diverse process, which may reflect the heterogeneity of NMDA receptors themselves or related components of neuronal cell biology. NMDA receptors vary both regionally and developmentally in the brain (reviewed by Lynch et al., 1997). Understanding the basis of such heterogeneity could potentially provide selective pharmacological targets, the inhibition of which might prevent excitotoxicity without the side effects of complete NMDA receptor blockade. The cloning of multiple NMDA receptor subtypes almost 10 years ago explains much of the biochemical basis of receptor heterogeneity and offers the possibility of using information about different receptor subtypes for better understanding of the process of excitotoxicity. In the present review, we will discuss the new understanding of NMDA receptor subtypes from molecular biological knowledge of the NMDA receptor, and ways in which heterologous expression of NMDA receptor subunits may clarify the complex process of excitotoxicity.

	Heterogeneity of NMDA Receptors

Top Abstract Introduction Heterogeneity of NMDA Receptors Diversity of Excitotoxic... Cloning of NMDA Receptors:... Application to Excitotoxicity:... Understanding of Excitotoxicity... Conclusions References

The NMDA receptor is a ligand-gated ion channel requiring simultaneous activation by two agonists (glutamate and glycine) for channel opening (Dingledine et al., 1999). Ion passage also requires depolarization, because magnesium directly blocks the ion channel in a voltage-dependent manner. Although the physiological significance is unclear, the receptor is also modulated by polyamines such as spermine and spermidine in a biphasic manner. At low micromolar concentrations, polyamines promote channel opening by increasing the affinity of the receptor for glycine as well as by removing tonic proton inhibition (Lynch et al., 1997; Dingledine et al., 1999). Polyamines block the channel in a voltage-dependent manner at higher concentrations that are probably not normally achieved extracellularly in vivo. Three other types of endogenous compounds (zinc, redox modulators, and nitric oxide) inhibit the NMDA receptor allosterically through different sites (reviewed by Lynch and Guttmann, 2001).

With the emphasis on reducing injury in excitotoxicity, a large number of NMDA receptor antagonists have been produced, including agents that block the glutamate site (CGP 39653), the glycine site (7-chlorokynurenic acid), or the ion channel itself (dizocilpine, phencyclidine) (Lynch et al., 1997; Dingledine et al., 1999). Other compounds originally characterized for their ability to inhibit other receptors (ifenprodil, haloperidol, amitriptyline, and amantidine) also inhibit the NMDA receptor. Ifenprodil and haloperidol act as allosteric modulators of proton inhibition of the receptor, whereas amitriptyline and amantidine act as low affinity channel-blocking agents (Lynch et al., 1997; Dingledine et al., 1999).

These diverse pharmacological antagonists produce different effects and side effects when given to animals, suggesting heterogeneity among NMDA receptors in brain. Such heterogeneity in native receptors has been confirmed by anatomical and developmental evidence. For example, in ligand binding studies, cerebellar NMDA receptors are less sensitive to agents that produce open channel block as well as agents that competitively block the glutamate binding site (Lynch et al., 1997). Similarly, neonatal NMDA receptors are more sensitive to ifenprodil and may have a weaker magnesium block than NMDA receptors from adults (Williams et al., 1993; Lynch et al., 1997; Dingledine et al., 1999; Kirson et al., 1999). These data demonstrate that multiple NMDA receptor subtypes exist in the brain, providing pharmacological targets for therapy with fewer side effects than are seen with traditional NMDA receptor antagonists.

	Diversity of Excitotoxic Processes

Top Abstract Introduction Heterogeneity of NMDA Receptors Diversity of Excitotoxic... Cloning of NMDA Receptors:... Application to Excitotoxicity:... Understanding of Excitotoxicity... Conclusions References

The diverse nature of the numerous disorders in which excitotoxicity has been implicated demonstrates that there are also likely to be variations in the basic process of excitotoxicity (Fig. 1). Excitotoxicity may be involved in the pathophysiology of a large number of disorders (Table 1), including chronic disorders such as epilepsy and Huntington's disease as well as acute disorders such as stroke. The initiating event in the excitotoxic process can be vascular or ischemic (stroke), genetic (as presumed in Huntington's disease), metabolic (hyperglycinemia), or immunologic (as proposed in human immunodeficiency virus encephalopathy). The different inciting events most likely lead to the different temporal and anatomic features seen in these disorders. Although the agonist precipitating excitotoxicity typically is presumed to be glutamate, in specific metabolic disorders or models it may be glycine or quinolinate (Lipton and Rosenberg, 1994; Lynch et al., 1997).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Excitotoxicity, glutamate receptors, and multiple messenger systems. Glutamate is capable of activating multiple receptor systems. A crucial component of excitotoxicity is mediated through the NMDA receptor, which allows high levels of calcium entry. Such calcium selectively activates a variety of downstream events, many of which are toxic to the cell when excessive levels of activation are reached. Modified from Lynch and Dawson (1994) with permission (Lippincott-Williams and Wilkins).

There also appear to be many complex, variable features between defined animal and cell culture models of excitotoxicity. Variability in models of excitotoxicity can arise at the level of the receptor inasmuch as some models of disease depend largely on AMPA receptors (Brorson et al., 1994). In other models, calcium entry selectively through the NMDA receptor is required for excitotoxic cell death (Tymianski et al., 1993), and NMDA receptor-mediated uptake of calcium by mitochondria is required in some models (Stout et al., 1998). In addition, even within NMDA receptor-dependent models of excitotoxicity, the activation of NMDA receptors gives rise to multiple processes that may be toxic to the cell under certain conditions, including reactive oxygen species production, calpain activation, and NO synthesis (Brorson et al., 1995; Ayata et al., 1997). Glutamate can also cause oxidative injury in some situations through nonreceptor-mediated mechanisms (Schubert and Piasecki, 2001). These distinct intracellular processes may play equally damaging roles under different situations. Although NO is a crucial mediator of NMDA receptor-induced excitotoxicity in many paradigms, models exist in which nitric oxide is unnecessary for excitotoxic cell death (Munir et al., 1995). Animals lacking nNOS remain susceptible to NMDA receptor-induced cell death to some degree (Ayata et al., 1997; Dawson and Dawson, 1998). The complexity of excitotoxicity is further illustrated by the observation that apoptosis rather than necrosis, the classical mode of cell death in excitotoxicity, also plays a role in some excitotoxic processes (Zhang et al., 1998). In some paradigms specific events along the excitotoxic cascade direct the cell toward apoptotic or necrotic events (Bonfoco et al., 1995).

Some of this complexity in mechanisms of glutamate-induced cell death in vivo reflects specific differences in glutamate receptor subtype content. For example, in ischemic lesions, forebrain neurons are much more susceptible to ischemia than cerebellar granule neurons, even though the levels of NMDA receptor are similar between these cells. This difference in susceptibility correlates with the variation of NMDA receptor properties between the cerebellum and the forebrain, with cerebellar receptors having smaller single-channel conductance values and being less sensitive to effects of competitive antagonists of the glutamate site (Lynch et al., 1997). Thus, heterogeneity in NMDA receptor combinations and properties may correlate with differences in patterns of excitotoxicity. Similarly, in the cerebellum, the Purkinje cells are usually the most susceptible to ischemia, likely reflecting a specific calcium-permeable AMPA receptor made in these cells (Brorson et al., 1994). Thus, the presence of specific receptor subtypes from both ionotropic glutamate receptor families may alter the susceptibility of cells to excitotoxic processes.

Another neuron specific protection from excitotoxicity is the relative sparing of nNOS-containing neurons in both cell culture models of excitotoxicity as well as in human diseases associated with excitotoxic pathophysiology. In many models of excitotoxicity, NO, synthesized by calcium/calmodulin-regulated nNOS, is the major mediator of neuronal damage (Dawson and Dawson, 1998). While NMDA receptors have been suggested to be the major receptors involved in synthesis of nitric oxide, the nNOS-containing neurons are spared in such models. This seems surprising given that the toxic NO presumably should be in highest concentration in the neurons in which it is synthesized, although the selective expression of manganese superoxide dismutase in nNOS-containing neurons may explain this unexpected protection (Dawson and Dawson, 1998). However, although not yet understood, the specific properties of the NMDA receptors in nNOS containing neurons might also contribute to this observation as well, particularly if features beyond simple NO production are crucial to the process of cell death (Dawson and Dawson, 1998).

A final variable component of excitotoxicity is the role of messenger systems in blocking or potentiating the process of excitotoxicity. Because second messenger systems modulate many receptors including the NMDA receptor, they represent an alternative pharmacological site where NMDA receptor-mediated effects could be controlled without directly blocking the ion channel. Protein kinase C (PKC) physiologically modulates NMDA receptors in either a potentiating or an inhibitory manner (Grant et al., 1998). Mechanistically, this results from a potentiation of peak NMDA receptor currents by PKC as well as an enhancement of calcium-dependent inactivation of the receptor by PKC (Lu et al., 2000). PKC has similar variable roles in excitotoxic paradigms. Because NMDA receptors are potentiated by PKC in most cells, PKC activation could be important for continued activation of NMDA receptors to a level required for excitotoxicity (Calabresi et al., 1999; Bruno et al., 2001). However, in some excitotoxic paradigms, agents acting through PKC attenuate excitotoxicity (Koh et al., 1991; Pizzi et al., 1999; Bruno et al., 2001). One possible explanation is that activation of PKC is required only for a specific branch of cell death such as apoptosis. PKC has been invoked as a mediator of apoptosis in some systems and a potentiator in others, perhaps depending on the isozyme of PKC that is present (Lin et al., 1997). Thus, the role of PKC in excitotoxicity is again complex, likely reflecting heterogeneity in both the type of neurons the features of NMDA receptors that are present. Understanding the characteristics of distinct NMDA receptor subtypes and how they are modulated by second messenger systems could address these possibilities.

	Cloning of NMDA Receptors: Molecular Biological Approaches to Understanding NMDA Receptors and Excitotoxicity

Top Abstract Introduction Heterogeneity of NMDA Receptors Diversity of Excitotoxic... Cloning of NMDA Receptors:... Application to Excitotoxicity:... Understanding of Excitotoxicity... Conclusions References

The cloning of multiple NMDA receptor cDNAs that contribute to functional NMDA receptors has facilitated understanding of the properties of NMDA receptor subtypes. The initial clone, designated NR1, reproduces many (but not all) features of native NMDA receptors when expressed in Xenopus oocytes, but much more limited features when introduced into mammalian cells (Dingledine et al., 1999; reviewed by Lynch and Guttmann, 2001). This most likely results because of the presence of endogenous complementary subunits in oocytes but not in mammalian cells. NR1 exists as eight splice variants that are developmentally and regionally expressed in the brain (Lynch et al., 1997; Dingledine et al., 1999). Other cDNAs coexpressed with NR1 are required for production of fully functional NMDA receptors in mammalian cells (Lynch et al., 1997). The cDNAs isolated in this manner encode a second subunit group, designated NR2, which exists in four forms produced from separate genes (NR2A, NR2B, NR2C, and NR2D). NR2 subunits alone have no known electrophysiologic activity in any system but combine with NR1 to produce fully functional receptors in either Xenopus oocytes or mammalian cells. The NR2 subunits are heterogeneously distributed in the brain, and their distribution correlates with variations in the functional properties that define NMDA receptor heterogeneity. For example, the NR2C subunit is selectively made in the cerebellum, a region in which NMDA receptors are less sensitive to antagonists of the glutamate sites and agents that produce open channel block (Lynch et al., 1997). Neonatal animals typically express only NR2B and NR2D, potentially accounting for the differences in adult and neonatal receptors. Thus, these subunits are differentially expressed in a manner that may explain pharmacological differences in native NMDA receptors.

The proteins encoded by these cDNAs share specific features. Both NR1 and all NR2 proteins are transmembrane proteins with three complete transmembrane regions and an intramembrane loop between the first and second complete transmembrane regions. This structure is predicted based on epitope mapping and homology to the X-ray crystallographic structure of AMPA receptors (Dingledine et al., 1999). The amino terminal region of each protein is extracellular and contains putative ligand binding regions and multiple glycosylation sites. The carboxyl-terminal region is intracellular and may control regulation of the receptor by second messenger systems, perhaps through multiple tyrosine and serine phosphorylation sites. This intracellular domain anchors NMDA receptors to specific cellular locations and attaches distinct messenger systems to the receptor (Niethammer et al., 1996; reviewed by Lynch and Guttmann, 2001). The final C-terminal amino acids connect the receptor to the postsynaptic protein PSD 95 or other postsynaptic proteins (Niethammer et al., 1996). However, the NR1 and NR2 subunits differ structurally in many ways. The NR1 protein is more than 300 amino acids shorter than NR2C and NR2D, and 500 amino acids shorter than the NR2A and NR2B. The extra peptide sequence in NR2 subunits is located intracellularly in the C-terminus, a region of low homology between different NR2 subunits. Since the function of this domain is likely to be differential localization and activation, this suggests that different subunits may be associated with different responses to messenger systems.

	Application to Excitotoxicity: Direction of NMDA Receptor Antagonists to Specific Subunits

Top Abstract Introduction Heterogeneity of NMDA Receptors Diversity of Excitotoxic... Cloning of NMDA Receptors:... Application to Excitotoxicity:... Understanding of Excitotoxicity... Conclusions References

Expression of NMDA receptor subunits in heterologous systems has allowed the understanding of the basis of the pharmacological heterogeneity of the NMDA receptor. The NR1 subunit is the glycine binding subunit, whereas the glutamate binding subunit is the NR2 subunit (Dingledine et al., 1999). Combinations of different NR1 and NR2 subunits give rise to receptors with different properties. For example, the NR2A subunit conveys high affinity for glutamatergic antagonists, whereas receptors containing the NR2C subunit have low affinity for these agents (reviewed by Lynch et al., 1997). NR2C is expressed in high levels in the cerebellum, explaining the relative insensitivity of cerebellar NMDA receptors to glutamatergic antagonists and agents that block the open channel. The atypical antagonist ifenprodil and its derivatives have high affinity for receptors containing the NR2B subunit, consistent with the higher affinity of neonatal receptors for ifenprodil and the high levels of the NR2B subunit made in the neonatal period (Williams et al., 1993). The interaction between the NR1 and NR2 subunits also regulates glycine agonist affinity as receptors containing NR2B have a higher affinity for glycine and glycine antagonists (Benke et al., 1999). In addition, future design of antagonists must include the consideration that some NMDA receptors contain more than one NR2 subunit in a heterotrimeric receptor (Chazot et al., 1994). Understanding these properties allows the design of agents for specific receptor subtypes.

Subtype-specific antagonists might allow blockade of a crucial number or subset of NMDA receptors. This would facilitate cell preservation during excitotoxic events while not causing complete blockade, likely reducing side effects. The NR2B-selective derivatives of ifenprodil appear to be best suited for this purpose (Chenard and Menniti, 1999). Based on a common structure, these agents appear to be effective in cell culture models of excitotoxicity and in anti-ischemic therapy. Interestingly, animals treated with these newer derivatives of ifenprodil do not appear to develop the severe side effects seen with agents such as dizocilpine (MK801) or phencyclidine. This suggests that the receptors blocked by these agents in vivo (presumably those containing the NR2B subunit) represent a subset of receptors that are capable of mediating excitotoxic responses but do not appear sufficient for producing the intolerable effects when blocked. Thus, agents such as these represent a method for successful therapy in humans.

Use of these agents in animals has also revealed some surprising findings. These agents appear to act in regions (such as the cerebral cortex) where the NR2B subunit is not necessarily the main NR2 subunit present and at times in development when NR2B is not the major subunit. Conceivably, they might be even more efficacious (but also more toxic) when given at developmental stages (such as neonates) where the NR2B receptor is the major receptor of the brain. In addition, they have proven very effective in alleviating another "toxic" response mediated by excessive release of glutamate. Severe abnormal movements termed dyskinesias are associated with chronic L-DOPA therapy in Parkinson's disease. Although all NMDA receptor antagonists partially block these devastating side effects of chronic therapy, the NR2B-selective antagonists are particularly efficacious (Steece-Collier et al., 2000). Thus, development of these agents represents a major area of drug advancement resulting from knowledge of the molecular biology of the NMDA receptor.

Similarly, other sites on the NMDA receptor represent potential sites for subtype-selective inhibition. Because the affinity of the NR1/2B subtype for glycine is higher than that of NR1/2A or NR1/2C receptors, use of glycine site antagonists represents another subtype-specific approach to prevention of excitotoxicity (Benke et al., 1999). In addition, the redox site on the NMDA receptor has subtype-specific effects, with NR2B-containing receptors responding distinctly from NR2A containing receptors (reviewed by Lynch and Guttmann, 2001). Finally, the zinc site on NR2A represents an additional site for subtype-specific therapy (Lynch and Guttmann, 2001). No selective clinical drugs are presently available for these sites on the NR2A receptor subtype, but they too may be a focus of drug development in the near future.

	Understanding of Excitotoxicity from Cloned Receptors: Cell Death Mediated by Recombinant Receptors

Top Abstract Introduction Heterogeneity of NMDA Receptors Diversity of Excitotoxic... Cloning of NMDA Receptors:... Application to Excitotoxicity:... Understanding of Excitotoxicity... Conclusions References

The crucial role of NMDA receptors in excitotoxicity has been confirmed by transfection of cloned NMDA receptors into cell lines. While expression of non-NMDA receptors is not usually toxic, transfection of NMDA receptors into cell lines is accompanied by cell death (Chazot et al., 1994; Cik et al., 1994; Anegawa et al., 1995; Boeckman and Aizenman, 1996; Raymond et al., 1996; Anegawa et al., 2000; Rameau et al., 2000). This result is independent of the method of transfection and may be noted in cell counting assays, enzymatic assays for cell death, or fluorescence-based cytotoxicity assays. It is seen in several different cell lines such as HEK293, HEK293t, CHO and COS cells. Presumably, the cell death depends on activation by glutamate and glycine as cell culture media routinely contains sufficient levels of these amino acids to activate the NMDA receptor (Anegawa et al., 1995). This cell death is blocked by antagonists of the glutamate, glycine, and channel sites of the NMDA receptor, matching the properties of excitotoxicity in vivo. The molecular biological features demonstrate that toxicity requires NMDA receptors of appropriate constitution. Nonfunctional NMDA receptors such as homomeric NR1a receptors do not produce this effect, whereas NR1/2A and NR1/2B receptors cause transfected cells to die. In contrast, much lower levels of cell death occur with NR1/2C or NR1/2D receptors, even though the whole cell currents in these cells are generally similar in size to those in cells transfected with NR1/2A and NR1/2B. This corresponds with data from the brain, in which cells of the cerebellum and nNOS-positive neurons of the forebrain, both of which contain higher levels of NR2C or NR2D, are spared in focal ischemia and other models of excitotoxicity (Standaert et al., 1996). Thus, the NR2 subunit primarily produced by cells may be a mediator of excitotoxicity or at least a marker of those cells most susceptible to excitotoxic injury.

Other studies have used the "excitotoxicity" of transfected cells to better define the properties of NMDA receptors that mediate excitotoxicity. Although AMPA receptors are not usually toxic when transfected into cells, they may become toxic when transfected cells are treated with cyclothiazide, an agent that blocks desensitization of AMPA receptors (Raymond et al., 1996). This illustrates a crucial property of NMDA receptors linking them with excitotoxicity. NMDA receptors desensitize to a smaller degree than do AMPA receptors. Thus, glutamatergic stimulation through NMDA receptors may be more toxic because of its relative lack of desensitization. NMDA receptor desensitization is also more complex with components that depend on glycine concentration, zinc concentration, and effects on intracellular processes (Krupp et al., 1998; Villarroel et al., 1998). How NMDA receptor desensitization affects excitotoxicity is not yet well defined.

However, transfected cell models also demonstrate that those features of the receptor outside the ligand binding or ion channel regions may play crucial roles in excitotoxicity. Removal of the final 400 amino acids of the NR2A receptor decreases the level of cell death in transfected cells without altering the change in intracellular calcium level in response to agonists (Anegawa et al., 2000). This links the processes of excitotoxicity to the mediation of intracellular events that occur selectively through the C terminus of the receptor, providing a rationale for the crucial role of NMDA receptor-mediated calcium increases in neuronal excitotoxicity (Tymianski et al., 1993). Similarly, cotransfection with PSD 95 increases the size of agonist-evoked currents in these cells but does not increase levels of cell death (Rutter and Stephenson, 2000). Although the reason for this is unclear, binding of the receptor to PSD 95 could affect the attachment of signaling systems to the receptor and indirectly influence the degree of cell death. In addition, changing the splice variants of the carboxyl-terminal domain of the NR1 subunit modulates cell death, most likely by altering the calcium/calmodulin-dependent inactivation of the receptor (Lu et al., 2000; Rameau et al., 2000).

This transfected cell system also defines the cellular properties that mediate the toxic response induced by NMDA receptor stimulation. As in neurons, toxicity in transfected cells depends on increased intracellular calcium. The receptors that cause toxicity in transfected cells (NR2A and NR2B) also pass significantly higher fractional levels of calcium and produce much higher intracellular calcium responses to agonists than those receptors that are not toxic (Burnashev et al., 1995; Grant et al., 1997). Intracellular calcium responses of the nontoxic receptors NR2C and NR2D are also more affected by depletion of intracellular stores, a calcium source not usually associated with excitotoxicity (Grant et al., 1997, 1998). Mutants of NR2A that pass lower levels of calcium also produce lower levels of cell death although this effect may not reflect calcium permeability alone (Cik et al., 1994; Anegawa et al., 1995; Boeckman and Aizenman, 1996; Rameau et al., 2000). Cotransfection of calcium binding proteins, which blunts the size of the intracellular calcium response to NMDA receptor stimulation, also decreases levels of toxicity (Rintoul et al., 2001). Heterologous expression systems generally do not have high levels of nNOS, showing that elevated calcium may kill cells in a manner that does not involve NO in this system. However, other types of free radical production are a component of NMDA receptor mediated cell death in these cell lines, similar to that noted in neuronal toxicity (Anegawa et al., 2000). Thus, the transfected cell system recapitulates many of the properties of excitotoxicity in vivo, and an examination of transfected cells may be useful for studying the mechanisms involved in excitotoxicity.

Theoretically, the transfected cell system also may provide clues to the relative role of second messenger systems in the excitotoxic process (Fig. 2). Because PKC potentiates activity of many forms of NMDA receptors, inhibitors of PKC might block excitotoxic disease. However, in heterologous systems the effect of PKC on NMDA receptor activity is subtype-dependent, with NR2A- or NR2B-containing receptors being potentiated, whereas NR2C- or NR2D-containing receptors are inhibited or unaffected. This potentially provides a mechanism for complex NMDA receptor subtype-specific effects of PKC in excitotoxicity (Grant et al., 1998). Data examining the blockade of cell toxicity in neurons and in transfected cells support this possibility. In lesions of the striatum, a brain region containing high levels of NR2B, potentiation of NMDA receptors through metabotropic glutamate receptors appears to increase cell death (Calabresi et al., 1999). In transfected cells, NR2A-dependent cell death is blocked by PKC inhibitors (Wagey et al., 2001). However, in the cerebellum, which has high levels of NR2C subunits, activation of PKC blocks cell death through excitotoxicity (Pizzi et al., 1999). Thus, understanding the subunit-specific properties of NMDA receptor subtypes explains the role of PKC in modulating excitotoxicity.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Excitotoxicity and multiple receptor subtypes. The distinct properties of NMDA receptor subtypes appear to differentially contribute to excitotoxicity based on transfected cell models. NR2A- and NR2B-containing receptors give rise to larger calcium rises, are potentiated by PKC, and give rise to greater levels of toxicity. NR2C and NR2D produce lower calcium rises, are inhibited by PKC (with appropriate NR1 forms), and give rise to much lower levels of toxicity. Modified from Lynch and Dawson (1994) with permission (Lippincott-Williams and Wilkins).

A final goal of an excitotoxicity model is not only to model the basic process but also to understand the mechanisms of specific diseases. Recent studies show that expression of mutant huntingtin together with NMDA receptors in HEK-293 cells may model some characteristics of Huntington's disease (HD), frequently viewed as a slow version of excitotoxicity. Transfection of mutant huntingtin with some NMDA receptor subunit combinations potentiates the cell death, as well as the currents mediated by NMDA receptor stimulation. No such effect is seen with transfection of normal huntingtin (Zeron et al., 2001). These effects are NMDA receptor subtype-selective and only observed with NR2B-containing receptors (Chen et al., 1999; Zeron et al., 2001). Interestingly, the NR2B subunit is the major NR2 subunit of the striatum, where the effects of HD are most profound (Standaert et al., 1996). Thus, this simple transfected cell system provides a physiologically relevant model for development of new agents for HD specifically, and defines a mechanism by which NMDA receptor stimulation can interact with other cellular processes to provide a disease model. Further studies may allow the modeling of the role of receptor subtypes in other complex excitotoxic situations such as other disease models.

	Conclusions

Top Abstract Introduction Heterogeneity of NMDA Receptors Diversity of Excitotoxic... Cloning of NMDA Receptors:... Application to Excitotoxicity:... Understanding of Excitotoxicity... Conclusions References

Since the emergence of the molecular biology of the NMDA receptor, the role of different NMDA receptor subtypes in basic excitotoxicity has been revealed, and subtype-specific agents have been developed to some degree. Further improvement directed at specific disease processes may facilitate the development of NMDA receptor antagonists for clinical use. For example, NR2B receptors appear to be selectively up-regulated in giant cells of tuberous sclerosis, whereas dysplastic neurons of tuberous sclerosis selectively synthesize NR2B and NR2D (White et al., 2001). Future therapies for epileptogenicity of tubers might be directed to NR2D, the subtype selectively produced by the epileptogenic cells (the dysplastic neurons). Such studies will need to differentiate between the subtypes made as a component of the disease rather than as a compensatory mechanism or epiphenomenon of the disease. Regardless, such analyses may eventually make possible successful therapy of excitotoxicity directed against NMDA receptor subtypes in humans.