Hypothesis

Recent evidence supplies new insights regarding the two universal effects of inhaled anesthetics: 1) immobility in response to a noxious stimulus and 2) amnesia. We hypothesize that these two effects result from actions at separate molecular and anatomic sites and that they are produced by different mechanisms. We propose that inhaled anesthetics cause immobility in response to noxious stimuli by an action in the spinal cord at an interface between polar and nonpolar regions. Such a site might be an interfacial region adjacent to membranes or proteins. In contrast, we propose that production of amnesia occurs at a supraspinal site and occurs in a nonpolar environment. An example of such a nonpolar site could be the interior of a phospholipid bilayer or a hydrophobic pocket within a protein. A Definition of Anesthesia All inhaled compounds producing the reversible state we call anesthesia share two characteristics and only two: the capacity to provide 1) immobility in response to a noxious stimulus and 2) amnesia 1. Anesthetics have other clinically relevant effects, including analgesia and hypnosis 2, but such effects do not uniquely define the anesthetic state. Historically, anesthetics were assumed to provide a continuum of effects, amnesia occurring at low anesthetic concentrations, and immobility resulting from the same process but at higher anesthetic concentrations. We propose a different hypothesis: disparate mechanisms at disparate molecular and anatomical sites cause immobility and amnesia. We presume that one mechanism applicable to all inhaled anesthetics causes immobility and that a separate mechanism produces amnesia. Evidence for Two Separate Sites and Mechanisms Variations among anesthetics in the ratio of the anesthetic partial pressure producing immobility to that causing amnesia provide strong evidence that different mechanisms underlie these effects 3. The finding that certain lipid-soluble vapors and gases do not suppress movement in response to noxious stimuli, a finding contrary to the Meyer-Overton hypothesis (see below), provides still more compelling evidence. Not only do such inhaled compounds fail to cause immobility, they do not decrease the immobilizing dose-requirement for a conventional anesthetic such as desflurane or isoflurane 4,5. However, in one experimental model of learning, they were found to cause amnesia 6. Thus, inhaled compounds may be divided into two groups. Compounds that produce both immobility and amnesia might be called "full anesthetics." Compounds that fail to produce immobility but do cause amnesia might be called "nonimmobilizers." We have previously called nonimmobilizers "nonanesthetics" 4,5,7. The failure of nonimmobilizers to produce immobility conflicts with the Meyer-Overton hypothesis 8,9. That hypothesis predicts that anesthesia results from the presence of a certain concentration of molecules in a lipid-like (hydrophobic) site; i.e., the anesthetizing partial pressure correlates inversely with the lipophilicity of that compound 10. The lipophilicity of many nonimmobilizers indicates that they should produce immobility at partial pressures below their vapor pressures, but they do not 4. Our recent finding that two nonimmobilizers produce amnesia, as defined by suppression of the ability of rats to learn in classical conditioning experiments 6, documents a basic difference in mechanisms of the two components of anesthesia-amnesia and suppression of movement in response to noxious stimuli. The amnestic effect of nonimmobilizers suggests that a key molecular selectivity factor present at the site mediating immobility is lacking in the site mediating amnesia. That is, anesthetics and nonimmobilizers share some property that produces amnesia, but the difference in their capacity to produce immobility indicates a difference in some other fundamental property (a key molecular selectivity factor). Our hypothesis that full anesthetics act on separate anatomical and molecular sites to produce immobility versus amnesia is consistent with recent reports that full anesthetics act primarily on the spinal cord to produce immobility. Decerebration does not decrease the anesthetic partial pressure required to cause immobility 11,12. However, descending modulatory systems influence anesthetic actions on the cord, decreasing the anesthetic requirement when an anesthetic is delivered primarily to the cord as opposed to both cord and cerebral structures 13,14. In contrast, amnesia probably results from an effect on supraspinal centers. Suppression of Movement The Meyer-Overton hypothesis suggests that conventional anesthetics produce immobility by an action in the hydrophobic core of neuronal membranes. However, this hypothesis is less convincing if a broad range of anesthetics is considered, including transitional compounds, such as cyclohexane 15 or CF2 Cl-CF2 Cl 4, which cause immobility at partial pressures considerably higher than those predicted by the Meyer-Overton hypothesis. It also fails to explain the lack of action of nonimmobilizers. Since anesthetics distribute throughout the membrane and surrounding aqueous phases 16-19, a site different from the membrane interior might be more consistent with current data. However, we cannot exclude the possibility that there are more specific requirements at the hydrophobic binding site that are not met by nonimmobilizers. Full anesthetics have both a substantial lipophilicity and a hydrophilicity that usually considerably exceed those for nonimmobilizers. In contrast, although nonimmobilizers may have a substantial lipophilicity, they exhibit a low hydrophilicity 4,5,7. The properties of full anesthetics are consistent with production of immobility by an action at an interfacial region between water and neuronal membranes or proteins. Such a view is supported by detailed, molecular-level simulations of solubilities across a water-membrane interface. These simulations yield a good correlation between interfacial concentrations and the capacity to produce immobility, even for compounds that deviate from the Meyer-Overton hypothesis 16,18. The observed correlation means that interfacial concentrations of full anesthetics at anesthetizing partial pressures are approximately the same. Most nonimmobilizers differ from full anesthetics in their inability to achieve these concentrations at any partial pressure 18. Significant general anesthetic concentrations at the water-membrane interfaces have been observed experimentally 17,20,21, and attempts have been made to explain how these interfacial molecules could produce anesthesia by nonspecific mechanisms 22,23. Although the Meyer-Overton hypothesis implies a lipid site of action, most investigators now focus on a direct action at sites on membrane proteins 24. Several results suggest that anesthetics suppress movement by binding to proteins in specific receptor(s). Suppression of movement caused by one full anesthetic, isoflurane (CHF2-O-CHClCF3), appears to be stereoselective, the (+) isomer being about 50% more potent than the (-) isomer in the rat 25. The effects of this anesthetic on specific receptors 26 and on potassium conductance 27 are also stereoselective. Full anesthetics and compounds that are solely amnestic differ in their effects on putative protein sites of action. Full anesthetics, but not nonimmobilizers, enhance the effect of gamma-amino butyric acid (GABA) on GABAA receptors expressed in oocytes 28,29, GABAA receptors derived from mouse brain 30, and GABAA-mediated responses in spinal cord 31,32. Equally important, the enhancement occurs at partial pressures that produce immobility in the whole animal. Similarly, full anesthetics alter responses mediated via glutamate receptors in intact spinal cord 31,32 and expressed in oocytes, whereas nonimmobilizers do not 28. Finally, full anesthetics increase desensitization of nicotinic acetylcholine receptors, but nonimmobilizers do not 33. In summary, it appears that anesthetics probably suppress movement in response to noxious stimuli by binding to specific receptor/channel complexes. We propose that this anesthetic site of action has both polar and nonpolar components. Such a site might be located near the water-membrane or water-protein interface or near an amphipathic site within a protein. Production of Amnesia The hippocampus provides the substrate underlying some forms of memory, possibly spatial memory, and perhaps, in humans, memory for recent events. Full anesthetics can suppress impulse transmission through the hippocampus, particularly through CA1 neurons, a finding consistent with the capacity of inhaled anesthetics to produce amnesia. Nonimmobilizers also can suppress such transmission 34 and, as indicated above, can cause amnesia 6. These findings suggest that the molecular site of action for amnesia differs from that which mediates immobility. They also suggest that amnesia results from an action at a nonpolar site, a site different from that which mediates immobility. The low affinity of nonimmobilizers for water indicates that their capacity to produce amnesia does not result from an action on a polar site. Present findings do not exclude the possibility that anesthetics cause amnesia by a nonspecific effect rather than by binding to a specific receptor. In summary, full anesthetics and nonimmobilizers may produce amnesia by acting on a nonpolar site. An example of a nonpolar site would be the interior of a phospholipid bilayer or a hydrophobic pocket within a protein. Cautions and Requirements for Future Studies A putative site of anesthetic action must respond to the anesthetic partial pressures that produce the effect (immobility or amnesia) in vivo. Partial pressures well above or below these are of little relevance. Thus, consistent with the suggestion that GABAA receptors are relevant to the production of immobility, the maximum enhancement of the effect of GABA on GABAA receptors occurs at anesthetic partial pressures required to produce immobility in vivo. This also makes the GABAA receptor a less likely target for the amnesia component of anesthesia because the partial pressures of full anesthetics that cause amnesia in vivo have only a slight effect on GABAA receptors. Satisfying the above requirement leaves questions unanswered. What molecular locations of full anesthetic molecules produce immobility, and what locations of nonimmobilizers and full anesthetics produce amnesia? What mechanical changes produce these specific or nonspecific effects? How are the mechanical changes accomplished? Why do these mechanical changes cause immobility and amnesia? If our hypotheses prove correct, they provide a basis for pursuing these finer details of the mechanisms of anesthesia.