Noble Meets Nouveau: A Shared Anesthetic Binding Site for Xenon and Isoflurane on a Glutamate Receptor

XENON is of noble extraction, as old as the hills, ubiquitous but rare and elusive; it is present as a trace gas in earth’s atmosphere at only 0.05 ppm. Isoflurane, on the other hand, is a compound born in the atomic age, a polyhalogenated methyl ethyl ether crafted by Ross C. Terrell, Ph.D., in 1965 after advances in organic fluorine chemistry facilitated by development of the atomic bomb.1Despite extreme differences in their structures and properties, both xenon and isoflurane have the striking ability to induce general anesthesia, although the mechanisms involved are far from clear. It is therefore remarkable that this noble gas and upstart organic chemical have now been found by Dickinson et al.  2to frequent the same binding site on the N -methyl-d-aspartate (NMDA) subtype of glutamate receptor.

N -Methyl-d-aspartate receptors are major postsynaptic membrane receptors for glutamate, the principal excitatory neurotransmitter in the mammalian central nervous system.3NMDA receptors, defined pharmacologically by their selective activation by the agonist NMDA, are a subtype of glutamate-gated ion channels that open to permit cation entry upon binding of the agonist glutamate (or NMDA) to an NR2 subunit and of the coagonist glycine to the NR1 subunit. NMDA receptors are involved in memory, pain, neurodegeneration, and cerebral ischemia as a result of their capacity to allow Ca²⁺entry upon activation, hence their critical relevance to anesthesiology. This landmark study greatly advances our understanding of the mechanisms of general anesthetic interactions with NMDA receptors by identifying a novel binding cavity in the NR1 subunit for these two markedly different anesthetics.

Accumulating evidence indicates that there is no universal target that explains all actions of every general anesthetic, or even of a single anesthetic agent.4Neurotransmitter-gated ion channels, particularly receptors for γ-aminobutyric acid (GABA) and glutamate, are modulated by most anesthetics, at both synaptic and extrasynaptic sites, while additional receptors and ion channels, such as specific Na⁺and K⁺channels, are also recognized as important targets. General anesthetics act as positive or negative allosteric modulators of these ion channels at clinically effective concentrations. Most inhaled anesthetics, including all of the volatile ether anesthetics (e.g. , isoflurane, sevoflurane, desflurane, enflurane), some of the alkanes (e.g. , halothane), and most intravenous anesthetics (e.g. , propofol, thiopental, etomidate) enhance GABA^Areceptor function by increasing channel opening to enhance inhibition at both synaptic and extrasynaptic receptors. In addition to their prominent effects on GABA^Areceptors, volatile anesthetics selectively depress excitatory synaptic transmission presynaptically, where their principal action seems to be a reduction in glutamate release. In contrast, the nonhalogenated inhaled anesthetics (e.g. , xenon, nitrous oxide, cyclopropane), as well as the intravenous anesthetic ketamine, have little or no effect on the GABA^Areceptors but depress excitatory glutamatergic synaptic transmission postsynaptically via  NMDA glutamate receptor blockade. The study by Dickinson et al.  is significant for identifying a shared mechanism by which both xenon and isoflurane inhibit NMDA receptors through competitive antagonism of the binding of the obligatory coagonist glycine, which interacts with a distinct ligand binding site and is required for full receptor activation by glutamate.

Identification of inhaled anesthetic binding sites on putative target proteins has been fraught with difficulty due to the low affinity of the interactions between anesthetic and receptor and the absence of atomic resolution structures of pharmacologically relevant molecular targets. As a result, most atomic resolution anesthetic binding sites have been identified in well-characterized model proteins for which three-dimensional atomic resolution structures are already available.5Nevertheless, considerable evidence obtained from molecular modeling and site-directed mutagenesis indicates that the molecular volumes of anesthetic target sites influence modulation of GABA^Aand glycine receptors by these drugs.6This evidence supports the notion that anesthetics modify receptor function by occupying a critical volume within a binding cavity. Internal cavities within proteins are important for conformational flexibility and ligand-induced signal transduction such that occupation of these cavities might alter receptor function. Dickinson et al.  2took an elegant approach to identify sites of interaction of xenon and isoflurane with the NMDA receptor by combining molecular modeling of xenon binding to the NR1 glycine binding domain with evidence for inhibition of receptor function obtained by patch clamp electrophysiology of heteromeric NMDA receptors consisting of the prevalent NR1/NR2A and NR1/NR2B subunit combinations.

Starting with the published crystal structure of the NR1 ligand-binding domain,7molecular modeling was used to identify putative xenon binding sites on the NMDA receptor, an innovative approach in the identification of anesthetic receptor interactions. This approach identified a putative xenon binding site that contained up to three xenon atoms and overlapped the known binding site for the coagonist glycine. Moreover, a molecule of isoflurane would fit into this putative xenon binding site. Pharmacologic analysis of recombinant NMDA receptor currents revealed that inhibition by xenon occurred by a mixed competitive/noncompetitive mechanism with glycine, whereas inhibition by isoflurane was competitive with glycine. A point mutation (F639A) in the second transmembrane domain of the NR1 subunit has been shown recently to reduce inhibition by xenon and isoflurane.8In support of a glycine competition mechanism, Dickinson et al.  showed that the reduced inhibition of NR1(F639A)/NR2A receptors by xenon and isoflurane results from the increased glycine affinity of the mutant receptor and therefore from reduced competition by the anesthetics. These important observations suggest that both anesthetics inhibit NMDA receptors by competing with glycine, the first evidence for direct competitive inhibition of ligand binding to a ligand-gated ion channel by an anesthetic.

Nitrous oxide is also known to inhibit NMDA receptors, and its pharmacologic properties may resemble those of xenon.9In view of a recent report that xenon and nitrous oxide share common binding sites within hydrophobic cavities of model proteins considered to be prototypes for the NMDA receptor,10it will be interesting to determine whether nitrous oxide interacts with the glycine binding site in a manner similar to that of xenon. Any differences observed between the two anesthetics could be relevant to the neurotoxicity of nitrous oxide compared with the purely neuroprotective effects of xenon.

The identification of binding sites for xenon and isoflurane on the NMDA receptor is an impressive step, but important questions remain regarding the pharmacologic relevance of these findings, particularly with regard to isoflurane. Although NMDA receptor blockade seems likely to mediate the anesthetic and neuroprotective effects of xenon,11,12postsynaptic effects of isoflurane at glutamatergic synapses are relatively minor compared with the presynaptic reduction in glutamate release observed in hippocampal brain slices13and isolated nerve terminals.14Endogenous neuronal NMDA receptors undergo complex regulation by endogenous agonist and antagonist ligands and ions (e.g. , glutamate, glycine, Zn²⁺, Mg²⁺, polyamines), receptor scaffolding and targeting proteins, and multisite phosphorylation, all of which could affect anesthetic sensitivity in vivo . Therefore, pharmacologic effects observed in vitro  with cloned receptors expressed in nonneuronal cells must be verified in intact neuronal synapses.

The competitive nature of the xenon and isoflurane effects on glycine indicate that they will have reduced effects in the presence of high extracellular glycine concentrations, in the absence of depolarization sufficient to relieve the voltage-dependent Mg²⁺block of NMDA receptors, and/or with specific receptor subtype combinations. In this regard, the noncompetitive component of the blockade by xenon, which was not observed for isoflurane, might enhance the importance of NMDA block by xenon in both normal and pathologic conditions associated with high agonist concentrations. One of the most important questions that can now be addressed concerns the significance of these interactions between xenon and isoflurane and the NMDA receptor in terms of the overall pharmacology of these general anesthetics. The solution to this problem should be facilitated by the development of mutant “knock-in” mice expressing the NR1(F639A) point mutation—these mice would be predicted to be resistant to xenon and/or isoflurane anesthesia if this site is important.

The structural information provided by the molecular modeling will facilitate site-directed mutagenesis of the 12 amino acid residues shown to surround the putative anesthetic binding cavity. This will identify residues critical to anesthetic interactions and provide additional evidence for this cavity as a binding site for isoflurane and xenon, and might suggest interesting residues to mutate in vivo . Such evidence is particularly important for isoflurane in the absence of supporting structural or modeling data. Mutations that increase amino acid side chain volume would be predicted to destabilize and ultimately exclude anesthetic binding. These supporting studies should provide further support for this milestone in molecular anesthesiology. In addition to identifying a receptor locale frequented by both noble and nouveau anesthetic gases, Dickinson et al.  have identified at atomic resolution a common site of interaction for two dissimilar anesthetics that underlies their shared NMDA receptor blocking effects.

Departments of Anesthesiology and Pharmacology, Weill Cornell Medical College, New York, New York. hchemmi@med.cornell.edu

The author thanks Neil L. Harrison, Ph.D. (Professor, Departments of Anesthesiology and Pharmacology, Weill Cornell Medical College, New York, New York), for his helpful comments in preparing this Editorial View.