Dr. Sonner is to be congratulated for discussing “the origin and evolution of the response to inhaled anesthetics”.1 His hypotheses concerning this subject have led to the spirited discussion reflected in the editorials in this issue of the journal concerning both the maintenance of an anesthetic response throughout phylogeny and what might be the basis for that maintenance. If one is going to discuss evolution of the anesthetic response, however, one must step back and further consider what we know about the genetic control of anesthetic responses across phylogeny. Genetics and evolution are inseparable, seamlessly intertwined. Genetic studies of anesthetic sensitivity in widely divergent organisms can serve as the window through which evolution can be viewed.
Evolutionary maintenance of the anesthetic response through natural selection could arise by two distinct mechanisms. The first is that the anesthetic response mimics a response found in nature (quiescence) and which is itself under positive selection pressure. A second possibility is that anesthetics could disrupt a conserved physiologic function that was the force behind the selection. In this case, the anesthetic response would be indirectly conserved only because the affected function was conserved. We will consider each of these possibilities separately below.
First, consider the possibility that the anesthetic response is itself selected for in nature. For the anesthetic response to be evolutionarily maintained, that response must be favored for survival. However, anesthetics cause a physiologic state that would seem to not be favored, at least in the extreme. Apparently, all organisms are inhibited or made quiescent by volatile anesthetics. Although it is difficult to imagine that this quiescence is favored in most organisms, there are examples of favored quiescence (such as hibernation or sleep). However, most animals do not hibernate and normal sleep does not really resemble the anesthetized state. The selected quiescence could be more moderate, simply avoiding hyperactivity or limiting deleterious activity. In this case, it is not hard to envision positive selection associated with quiescence. Of course, it is possible for evolutionary forces to select for a quiescent response to a rarely occurring, but detrimental, compound if quiescence would lead to limitation of the detrimental exposure. The idea would be that the organism lies down or otherwise limits activity until the toxic purple cloud passes by. However, if this agent, either endogenous or exogenous in nature, is truly rare, then the dose-response curve to it should not be extremely narrow as described by Sonner (1, see below).
In our opinion, the strongest argument for a favored anesthetic-like response is the universal finding that inhibitory channels tend to be enhanced by volatile anesthetics, whereas excitatory channels tend to be inhibited.2 If anesthetics simply affect another physiologic characteristic (like disrupting the function of a transmembrane region), then why are the channels affected in opposite manners? Here it may be that we are misled about this observation by our selection of which agents to study. We tend to focus on those gases that act as anesthetics; there are many similar agents that are not anesthetics (nonimmobilizers) or are even epileptics.3 Taken as a broad group, these gases affect membrane proteins, but do not all have the characteristic effects in opposite directions for inhibitory and excitatory channels. Our view of anesthetics is skewed because we are interested only in those that cause anesthesia and thus are more likely to affect excitatory and inhibitory channels in opposite directions. We ignore those agents that behave differently and thus are not anesthetics. This indicates that the finding of opposite effects on inhibitory and excitatory channels is preselected and not a general characteristic of small hydrophobic molecules.
The counter argument is that agents which are good anesthetics in one organism are generally good anesthetics in all organisms. Therefore, a group of gases must have the opposite effects on channels in all organisms. This could certainly arise because of a conserved response to naturally occurring molecules which the specific anesthetic gases imitate. In this light, the idea that metabolic diseases and compounds mimic anesthesia is interesting. Again, one should consider genetics with evolution. It is not necessary, or useful, to postulate a conserved response to an abnormal metabolite, which is not significantly present in normal organisms. Metabolic pathways are strongly affected by natural selection. The organism should only be selected for in response to the array of normal metabolites at physiologically occurring concentrations. Again, as noted above, if quiescence would limit the production of toxic metabolites, then such a response could be favorable.
Finally, the hypothesis that neurotransmitters in high (but normally occurring) concentrations may serve to inhibit their effect is both novel and is easy to envision as having a strong selective advantage.4 However, it is difficult to design an experiment to test the role of neurotransmitters or metabolites as the endogenous anesthetics. One possibility is discussed at the end of this editorial.
Next consider the possibility that anesthetics disrupt a conserved physiologic function that was itself the force behind the selection, i.e., that the selection of the anesthetic response is indirect. If an underlying common mechanism gives rise to the similar effects of anesthetics in all organisms, then the genetics of those responses should give a hint as to that mechanism. First, however, one should consider the similarities between “normal” animals of different species. In all organisms tested, the dose-response curves for anesthetic responses (of admittedly widely different end-points) are relatively steep.5 The steepness of the dose-response curves implies that, whatever the target(s) of anesthetics is/are, there is either not much variation in the population or there are many similar targets that contribute to the end-point.6,7 The finding could arise because all targets are exposed to similar milieus/forces or that the one or few targets are very tightly conserved physiologically. Protein cavities8 are examples of the first class; membrane bound proteins could represent either class (since they all must function within a lipid environment resulting in rather conserved transmembrane regions).
There is at least one case in which isogenic organisms are exposed to anesthetics, i.e., the nematode C. elegans. In that organism, isogenic strains with identical environmental exposures have Hill coefficients for two different end-points similar to humans for minimum alveolar anesthetic concentration (where genetic variation is expected to be great).9,10 This strongly supports the model that the variations of some critical components are quite small in the general population. However, in recombinant inbred strains, the variation in response can be quite large,11,12 which implies that coordinate evolution of target components may be important, and not of any particular component. When proteins function in a complex, their changes are limited by the proteins with which they interact. However, large changes in the interactions can occur when the proteins are mixed as in recombinant inbred strains. Knowledge of which proteins interact with which other proteins (called the interactome) is therefore important in models of what might constitute an anesthetic target. This subtle point, that interactions between molecules are crucial, rather than the molecules themselves, may offer an important clue.
Genetic variation in anesthetic sensitivity has been found in multiple organisms, yeast, nematodes, fruit flies, rodents, and humans. Yeast may seem to be an unlikely model system for studying the effects of inhaled anesthetics. However, the simplicity of yeast, and the ease with which it can be genetically manipulated, offers advantages for studying the basis of inhaled anesthetic action. In general, genes that change the ability of inhaled anesthetics to stop growth in yeast are generally in pathways that control nutrient transport/availability in some way and are membrane bound.13 The nematode Caenorhabditis elegans is also an excellent genetic model and has a well-defined nervous system. Studies in the nematode have identified multiple genes that can change the animal’s behavior in inhaled anesthetics. At first glance, these gene products appear widely different from each other affecting ion channels, mitochondrial proteins, and mechanisms controlling fusion of synaptic vesicles.14–16 A unifying view could relate the genes identified to date as operating together at a presynaptic location. The fruit fly, Drosophila melanogaster, offers several advantages as a genetic system, including a high density of genetic markers. The studies to date in Drosophila also indicate that ion channels and presynaptic proteins are important in affecting anesthetic sensitivity.16,17 All potential targets identified above play a role in nervous system function in mammals. However, the importance of these gene products in determining anesthetic sensitivity in mammals is not yet established.
At first, the question “Why study mammals?” seems to have an obvious answer. Mammals such as mice or rats possess one unique advantage: they are more similar to humans than are simpler model organisms. However, it is difficult to do blinded forward screens for potential anesthetic targets in mice or rats. Rather, they serve as systems for testing candidate targets felt to be likely involved in anesthetic responses. Such studies have shown that some ion channels do have effects on whole animal anesthetic sensitivity.18,19
The work described above has been notable in not identifying a unified theory for what the anesthetic targets are; however, they have indicated that membrane-associated proteins do affect sensitivity and that the presynapse is likely an important “supertarget” (a phrase we first heard coined by Rod Eckenhoff). However, we are left with the question of what molecules are conserved across phylogeny that may give rise to the conservation of anesthetic response. If the second possibility (conservation of another characteristic that gives rise to the anesthetic response) is correct, then we need to group those mutants changing sensitivity to understand what molecular functions are important. The conservation of cell membranes, of transmembrane regions of membrane-associated proteins, or of protein cavities would seem to be strong candidates for conserved features affected by anesthetics. If the first possibility is instead correct (the anesthetic response itself is conserved), then small molecules resulting from metabolic pathways are present in all organisms, as discussed by Sonner.1 Even molecules such as H2S, not normally present in high quantities in oxygen using organisms, but known to cause a hibernation-like state,20 must be considered. The fact that a naturally occurring compound such as H2S can have such profound effects on activity and on survival under hypoxic conditions21 may serve as a critical clue as to how the anesthetic response first arose. Neurotransmitters, as suggested by Cantor,4 are also a class of molecules that could affect membrane protein interactions.
The above arguments would seem to not generate a good experiment to test a hypothesis. However, for the first time perhaps, the accumulated data may point to a new, uniquely advantageous, model organism. If the targets or anesthetic molecules are so tightly selected for, it will be difficult to generate mutations with large effects in model organisms. The presumption is that such mutations will likely be lethal or at least affect function enough that testing anesthetic sensitivity would be difficult. However, multiple coordinate changes in an organism, accumulated over time, may give rise to smaller phenotypic changes but which do affect anesthetic sensitivity. What is needed is an organism with a very wide genetic diversity (rather than one or a few normal wild types). The wide genetic diversity, preferably with groups separated genetically for long periods of time, would allow for mapping of multiple loci contributing to small changes in anesthetic sensitivity. Of course, the organism should have a sequenced genome. For the first time, it appears that humans might be the correct model organism, for all the same reasons that the species is usually not a good one, too much genetic diversity. The hypothesis should be made to support one’s favorite theory (metabolites as anesthetics, proteins as targets, etc.). This should be followed by a study of the anesthetic sensitivity of an extremely large human population to be correlated to a detailed polymorphism map to identify those characteristics controlling anesthetic sensitivity. The underlying hypothesis would be that there are outliers in the general population, missed when determining anesthetic minimum alveolar concentrations with the rather large Hill coefficients. Characterizing the genotypes of any such patients will greatly help in the identification of important pathways that determine anesthetic sensitivity in humans.
1. Sonner JM. A hypothesis on the origin and evolution of the response to inhaled anesthetics. Anesth Analg 2008;107:849–54
2. Dilger JP. The effects of general anaesthetics on ligand-gated ion channels. Br J Anaesth 2002;89:41–51
3. Eger EI II, Koblin DD, Sonner J, Gong D, Laster MJ, Ionescu P, Halsey MJ, Hudlicky T. Nonimmobilizers and transitional compounds may produce convulsions by two mechanisms. Anesth Analg 1999;88:884–92
4. Cantor RS. Receptor desensitization by neurotransmitters in membranes: are neurotransmitters the endogenous anesthetics? Biochem 2003;42:11891–7
5. Eckenhoff RG, Johansson JS. On the relevance of “clinically relevant concentrations” of inhaled anesthetics in in vitro experiments. Anesthesiology 1999;91:856–60
6. Sonner JM. Issues in the design and interpretation of minimum alveolar anesthetic concentration (MAC) studies. Anesth Analg 2002;95:609–14
7. Sonner JM, Gong D, Eger EI II. Naturally occurring variability in anesthetic potency among inbred mouse strains. Anesth Analg 2000;91:720–6
8. Eckenhoff RG. Promiscuous ligands and attractive cavities: how do the inhaled anesthetics work? Mol Interv 2001;1:258–68
9. Crowder CM, Shebester LD, Schedl T. Behavioral effects of volatile anesthetics in Caenorhabditis elegans
. Anesthesiology 1996;85:901–12
10. Morgan PG, Sedensky MM, Meneely PM, Cascorbi HF. The effect of two genes on anesthetic response in the nematode Caenorhabditis elegans.
11. van Swinderen B, Galifianakis A, Crowder M. Common genetic determinants of halothane and isoflurane potencies in Caenorhabditis elegans.
12. Deitrich RA, Draski LJ, Baker RC. Effect of pentobarbital and gaseous anesthetics on rats selectively bred for ethanol sensitivity. Pharmacol Biochem Behav 1994;47:721–5
13. Palmer LK, Wolfe D, Keeley JL, Keil RL. Volatile anesthetics affect nutrient availability in yeast. Genetics 2002;161:563–74
14. van Swinderen B, Saifee O, Shebester L, Roberson R, Nonet ML, Crowder CM. A neomorphic syntaxin mutation blocks volatile-anesthetic action in Caenorhabditis elegans.
Proc Natl Acad Sci USA 1999;96:2479–84
15. Kayser EB, Morgan PG, Sedensky MM. GAS-1: a mitochondrial protein controls sensitivity to volatile anesthetics in the nematode Caenorhabditis elegans.
16. Humphrey JA, Hamming KS, Thacker CM, Scott RL, Sedensky MM, Snutch TP, Morgan PG, Nash HA. A putative cation channel and its novel regulator: cross-species conservation of effects on general anesthesia. Curr Biol 2007;17:624–9
17. Walcourt A, Scott RL, Nash HA. Blockage of one class of potassium channel alters the effectiveness of halothane in a brain circuit of Drosophila. Anesth Analg 2001;92:535–41
18. Heurteaux C, Guy N, Laigle C, Blondeau N, Duprat F, Mazzuca M, Lang-Lazdunski L, Widmann C, Zanzouri M, Romey G, Lazdunski M. TREK-1, a K+
channel involved in neuroprotection and general anesthesia. EMBO J 2004;23:2684–95
19. Quinlan JJ, Ferguson C, Jester K, Firestone LL, Homanics GE. Mice with glycine receptor subunit mutations are both sensitive and resistant to volatile anesthetics. Anesth Analg 2002;95:578–82
20. Blackstone E, Morrison M, Roth MB. H2
S induces a suspended animation-like state in mice. Science 2005;308:518
21. Blackstone E, Roth MB. Suspended animation-like state protects mice from lethal hypoxia. Shock 2007;27:370–2