| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Waggoner Center for Alcohol and Addiction Research, University of Texas at Austin, Austin, Texas
Address correspondence to Cecilia M. Borghese, PhD, WCAAR, University of Texas at Austin, 2500 Speedway, MBB 1.124, Austin, TX 78712-1095. Address e-mail to cborghese{at}mail.utexas.edu Address reprint requests to R. Adron Harris, PhD, at the same address.
Consensus concerning the relevant targets for volatile anesthetics is far from being reached: even cherished proteins of long standing, such as 
-aminobutyric acid type A (GABAA) receptors, have their determined detractors as well as staunch defenders. In addition to accumulating evidence indicating anesthetic effects on the favorite candidate proteins, it is necessary to determine the in vivo relevance of such effects, to try to eliminate neuronal players that do not mediate a specific anesthetic effect. Several compounds that fail to follow the Meyer-Overton rule (1,2) have been proposed as useful tools to this end. These halogenated alkanes are structurally related to well known anesthetics, and possess oil/gas partition coefficients that, according to the rule, should guarantee them anesthetic capabilities. But these compounds do not go by the book: they are unable to induce immobility at predicted minimum alveolar anesthetic concentrations (MACs), they do not decrease desflurane MAC when coapplied (3), and they have convulsant properties (3,4). The most studied nonimmobilizer is 1,2-dichlorohexafluorocyclobutane (F6, also called 2N). It also differs from anesthetics in that it does not affect thermoregulation (5), and does not depress breathing (conversely, it is a respiratory stimulant) (6). There is one important anesthetic quality that F6 retains: it suppresses learning (7,8), in a way not attributed to an alteration in pain perception (8).
So the Meyer-Overton rule stands corrected: lipophilicity is not the only determinant of anesthetic capability; hydrophilicity must also be considered. There is considerable evidence that conventional anesthetics act at an interface between water/membrane/protein, so both hydrophilic and lipophilic characteristics are necessary to exert their effects. Nonimmobilizers do not possess the right combination of hydrophilicity and lipophilicity that allows them to concentrate at the water-membrane interface; instead, they accumulate in the membrane hydrocarbon core (9). Therefore, they cannot access and/or interact with the "immobility" sites. However, they can bind to a hydrophobic site to exert amnesic effects (10). A specific site for mediating even the convulsant effect seems likely in light of the differential effect of cis and trans isomers of F6 (11).
Nonimmobilizers have been tested on several major candidates for anesthetic targets, namely, ligand-gated ion channels (GABAA, glycine, serotonin-3, glutamate, and neuronal nicotinic acetylcholine receptors [nAChRs]) and G protein-coupled receptors (12), as well as on potassium channels (13,14), expressed in Xenopus laevis oocytes. F6 inhibits serotonin-2A currents (15), as well as muscarinic M1 receptor currents (16), so these G protein-coupled receptors are not likely to be involved in the immobilizing effect of volatile anesthetics. However, for all other receptors, nonimmobilizers have been reported not to have any appreciable effect and thus have not eliminated these ligand-gated ion channels as mediators of immobility.
In nonneuronal nAChRs, F6 proved to have similar effects to classical anesthetics, either quantitatively smaller in Torpedo nAChR (17) or equally active in mouse muscle nAChRs (18).
For neuronal nAChRs, the picture appears more complicated: as already mentioned, a study on human neuronal nAChRs reported no effect of F6 in several subtype combinations expressed in X. laevis oocytes (19), whereas another report found inhibition of nAChRs present in PC12 cells and in medial habenula neurons (20).
In this issue of Anesthesia & Analgesia, Dr. Raines’ group (21) from Harvard Medical School reports inhibition of both rat and human neuronal nAChRs by F6. This is not only an important finding by itself, but it also settles the controversy between the earlier reports on nAChRs, and the reason behind the previous discrepancy raises an important issue in anesthetic research.
The relevance of the F6 inhibition of 
4ß2 nAChRs is evident once it is pointed out that all volatile anesthetics (and some IV anesthetics) are potent blockers of the neuronal heteromeric nAChRs (12). If both volatile anesthetics and nonimmobilizers inhibit nAChRs, then it follows that the immobilization produced by inhaled anesthetics cannot be mediated by blockade of cholinergic transmission through nAChRs. It also gives sustenance to the hypothesis that the amnesic effects of volatile anesthetics and nonimmobilizers could be mediated by their actions on nAChRs, which are known to be strongly implicated in learning and memory processes (22). A point that needs further clarification is the relationship between the anesthetic or nonimmobilizer concentration needed to impair learning and memory, and the nAChR inhibition observed at that concentration. Sonner et al. (8) have reported suppression of classical fear conditioning at 0.2x predicted MAC F6 and 0.3x MAC desflurane. At 0.2x MAC, isoflurane inhibits 25% of acetylcholine-induced currents in chicken 4ß2 receptors expressed in X. laevis oocytes; in the same expression system, 0.2x predicted MAC F6 inhibits 10% of rat 4ß2 and 30% of human 4ß2 current (21). The apparent discrepancy between the concentration necessary for total learning suppression and the relatively small percentage of nAChRs blocked at that point deserves closer examination. The convulsant effect of F6 is probably unrelated to its action on neuronal nAChRs, because seizures are triggered by nicotinic agonists, not antagonists (23).
The observations published in this issue by Raines et al. (21) prompted us to take a second look at our earlier studies, which had shown no effect of F6 on neuronal nAChRs. To our surprise, we observed effects of F6 that were quite similar to those documented by Raines et al. Our previous negative results on the neuronal nAChRs could have been attributable to an unrecognized loss of nonimmobilizer. Although our general procedure for preparing and applying F6 and 2,3-dichlorooctafluorobutane (F8, another nonimmobilizer) to the nAChRs was the same as in our studies on the other receptors, nonimmobilizer concentrations in the recording chamber were not determined in all cases, and the experiments on the nAChRs did not include measuring bath concentrations. This brings us to a most emphatic recommendation: when dealing with volatile compounds, checking the concentration that actually reaches the cell is critical, especially when the saline/gas partition coefficient is relatively small, as is the case with nonimmobilizers. Another caveat that applies to some preparations is that O2 levels have to be controlled, because volatile compounds of low potency displace a substantial amount of O2 from solution and effects of hypoxia may be misinterpreted as anesthetic actions (24). This is not an applicable issue to receptors expressed in X. laevis oocytes, which display normal responses to the respective ligands even in the absence of O2 (25).
The first whiff of nonimmobilizers probably set Overton and Meyer spinning in their graves; however, they would likely admit that these compounds are not only useful to identify relevant, plausible targets for anesthetics, but also to dissect the mechanisms underlying the different actions of anesthetics.
Acknowledgments
Supported in part by Grant GM47818 from the National Institutes of Health.
References
4ß2 neuronal nicotinic acetylcholine receptors. Anesth Analg 2002; 95: 573–7.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
