Why Can All of Biology Be Anesthetized?

Eckenhoff, Roderic G. MD

Anesthesia & Analgesia:
doi: 10.1213/ane.0b013e31817ee7ee
Anesthetic Pharmacology: Letters & Announcements
Chinese Language Editions
Author Information

From the Department of Anesthesiology and Critical Care, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania.

Accepted for publication April 23, 2008.

Address correspondence and reprint requests to Roderic G. Eckenhoff, MD, Department of Anesthesiology and Critical Care, University of Pennsylvania School of Medicine, Philadelphia, PA 19104. Address e-mail to roderic.eckenhoff@uphs.upenn.edu.

Article Outline

When everything that lives can be immobilized, inhibited, or anesthetized with the same simple compounds, all within roughly a 10-fold concentration range, biology is trying to tell us something important. There are several possibilities for this remarkable degree of conservation, some of which are enumerated in this set of papers, but most boil down to either (a) evolutionary pressure for a specific anesthetic target or targets, or (b) evolutionary pressure for a biological feature that is hijacked by anesthetics.

It would seem that no one favors the first notion very strongly, although Professor Sonner suggests an indirect approach to it by proposing that the targets are ion channels accessed via interactions with the lipid membrane.1 He should be applauded for the courage to resurrect this notion, now with a hypothetical linkage that accommodates conservation. In this case, the selection pressure is the need for a means to minimize over-stimulation (excitoxicity, perhaps) using small molecules, essentially an extension of Professor Cantor’s ideas.2 As I understand the model, the transmitter first stimulates the ion channel and then reaches concentrations in the synapse that influence properties of the lipid membrane to exert a form of feedback inhibition on the protein. But, given the extreme degree of degeneracy of molecules that are proposed by Professor Sonner to produce this membrane effect, it seems likely that the organism(s) would have encountered many during evolution. It then becomes difficult to imagine that this degenerate feature, which renders an organism incapable of growing, reproducing or defending itself, could select for anything but elimination. There are, after all, other strategies for feedback inhibition or effect reversal which biology has already used (transporters, catalytic enzymes, feedback circuitry, etc). Professor Morgan’s idea of a feedback control for fermenting organisms is certainly a reasonable possibility.3 But then it is not clear why the selection has remained so strong in organisms that minimally use this sort of biochemistry (mammals), and that have, in fact, become more sensitive instead of less (compared to yeast, for example). Further, despite the similarity when drawn as stick figures, the simple alcohols have very different physicochemical texture as compared to the inhaled anesthetics. It has not yet been shown that they bind the same protein sites as do the inhaled anesthetics, and they probably distribute in lipid membranes differently. So what molecules could have produced the evolutionary pressure, and why has evolution made us more, rather than less, sensitive?

The second possibility, raised by both Professor Crowder4 and myself, is the possibility that there are features on or in anesthetic targets (be they protein or lipid - see later) that have other important roles which have been independently selected for, and which the anesthetics hijack to alter function or activity of the target. It seems to me that this is where the money is. And the prime candidate for the selected-for feature is the protein cavity. Protein cavities are essentially packing defects in a protein after it has folded into a three-dimensional structure.5 A slice through almost any protein will reveal a Swiss Cheese-like character, a variety of sizes and shapes of such holes. Occasionally, they contain small molecules, which may or may not be native ligands, but they are often empty, or contain a few water molecules. So why are they there? Surely biology could have figured out how to pack proteins better if cavities were not important. The most likely explanation is that they are there (perhaps, have evolved) to allow protein motion6 or small ligand migration.7 Proteins must be able to move to alter their conformations between active and inactive forms. If they were perfectly packed, this would be very difficult, as the energy required to break the packing, and the time required to create new and perfect packing, would be prohibitive. Cavities allow for these transitions with lower energy requirements. In biophysics-speak, they destabilize the protein structure because of lower contact surface area and allow for easier shifts of side chains and other structural elements because of less steric constraint (i.e., more “elbow room”). That they have been selected for from the very beginning, or more likely, are simply an intrinsic feature of amino acid polymers, is apparent from the fact that all proteins, even of ancient origin, have them.8 Further, a simple analysis of these packing defects has revealed that a very large number of them can readily accommodate molecules the size and shape of anesthetics.9 The diversity in protein cavities matches the enormous range of shapes and sizes of compounds with anesthetic activity, and thus might explain how so many different small molecules can cause anesthesia. Most are hydrophobic, and yet contain the occasional polar residue, and have highly conserved stereochemistry by virtue of all l-amino acids and right-handed helices. Finally, the number of cavities that could accommodate very large molecules like the neurosteroids is small, since such large empty spaces in proteins are energetically costly. This may explain why a response to neurosteroids is somewhat less conserved than the smaller anesthetics. It could also be that these cholesterol-like molecules indeed do have a bilayer mechanism of action fundamentally different from the inhaled ones.

So how does anesthetic occupancy of pre-existing protein cavities alter protein function? Because if cavities exist to allow motions, motions must alter cavity size and shape, providing a basis for small molecule selectivity. Filling the cavities with a small molecule, even transiently, must therefore impede conformational transitions and cause protein dysfunction. But this does not mean that all proteins are inhibited by small hydrophobic molecules, merely that the conformer with the most attractive cavities is favored, be it an active or inactive conformer.

And cavities are not just found inside proteins, but also at interfaces of proteins, like oligomers and other protein-protein complexes. Analysis of such structures has revealed that a large number of such protein-protein interfaces are uncomplemented; in other words, they do not fit perfectly together and they contain interfacial cavities.10 Why? Probably to weaken the strength of the complex so that it can be reversed. If the affinity were too strong, the complex could not be broken on a biologically useful time scale, rendering the signaling ineffective. Occupancy of such interfacial cavities by an anesthetic might therefore alter signaling by enhancing the strength of the interaction, or reducing its strength via competitive interactions. The interactome, in fact, has probably the richest potential for transducing biologically important anesthetic effects.

Thus it seems likely that protein cavities have been strongly selected for, ever since proteins themselves evolved. They allow for normal protein function, and more importantly for the topic at hand, they have evolved to be empty, not to be a binding site for some undiscovered molecule. Nature considered this a safe strategy to allow protein function precisely because anesthetics were uncommon. That is, up until about 160 yr ago, when Morton and Long came along. If cavities need to be empty for normal activity, then filling them with an anesthetic is likely to cause protein dysfunction and alter the kinetics of protein-protein interactions.11

Are there selected-for features in lipid membranes that can explain conservation of anesthetic sensitivity as well? Absolutely. Lipid bilayers are highly dynamic structures that can readily accommodate most or all anesthetics. They are remarkably conserved across biology, and the mechanisms of their communication with proteins embedded in them are also likely to be highly conserved. Unfortunately, most such mechanisms are hypothetical because of the extreme difficulty of testing them. One attractive candidate, the lateral pressure hypothesis,12 rests on two very fundamental certainties. First, that lateral pressure is nonuniform across the membrane and, second, that small molecule distribution is also nonuniform. The question, then, becomes one of magnitude, and the sensitivity of the ion channel (or receptor) to these pressure profiles. This is difficult to study experimentally, but some of these ideas have been examined by molecular dynamic simulations, and the jury is still out. For example, lateral pressure modulation accounts well for the wide range of structures that cause anesthesia and for the observation that nonimmobilizers, which distribute differently in the bilayer,13 do not. But there are flies in the ointment. Hexane, which causes anesthesia, distributes much like the nonimmobilizers.14 It is likely that cyclopropane does as well. How about more polar small molecules like the neurotransmitters as suggested by Professor Sonner?1 The distribution of a charged molecule like γ-aminobutyric acid in the bilayer is likely to be very different from either halocarbon anesthetics or nonimmobilizers. And if very polar molecules can be anesthetic as suggested, why not the sucrose control, or the innumerable other small molecules that equilibrate across interstitial spaces? Nevertheless, from an evolutionary perspective, features of the lipid membrane and its interaction with ion channels and receptors are likely to have been conserved in all organisms, and are therefore still a viable candidate as an explanation for conserved responses to anesthetics.

Finally, we must acknowledge that a distributed protein theory can explain all the unique features of anesthetics and anesthesia that Professor Sonner uses as justification to return to lipid. Moreover, protein theories have several distinct, evidence-based advantages. First, anesthetics do bind to many proteins in specific, well-defined, and largely preexisting sites.15–17 I am not aware of a single example of a membrane protein whose function is altered by anesthetics and where an absence of binding has been rigorously confirmed (granted, this is often difficult). Further, protein sites can both distinguish stereoisomers of anesthetics (this has not been shown for lipid membranes yet), and nonimmobilizers from anesthetics.17 The small variance in effect within a population (large Hill slopes) can easily be explained by more than a few protein targets contributing to the effect11,18 consistent with either the cavity or lipid hypothesis. Therefore, when taken together, I believe the evidence weighs in favor of protein cavities as the fundamental feature underlying conservation of the anesthetic response.

Back to Top | Article Outline


1. Sonner JM. A Hypothesis on the Origin and Evolution of the Response to Inhaled Anesthetics. Anesth Analg 2008;107:849–54
2. Cantor RS. Receptor desensitization by neurotransmitters in membranes: are neurotransmitters the endogenous anesthetics? Biochemistry 2003;42:11891–7
3. Sedensky MM, Morgan PG. Genetics and the evolution of the anesthetic response. Anesth Analg 2008;107:855–8
4. Crowder CM. Does Natural selection explain the universal response of metazoans to volatile anesthetics? Anesth Analg 2008;107:862–3
5. Hubbard SJ, Gross KH, Argos P. Intramolecular cavities in globular proteins. Protein Eng 1994;7:613–26
6. Hubbard SJ, Argos P. A functional role for protein cavities in domain: domain motions. J Mol Biol 1996;261:289–300
7. Lamb DC, Arcovito A, Nienhaus K, Minkow O, Draghi F, Brunori M, Nienhaus GU. Structural dynamics of myoglobin: an infrared kinetic study of ligand migration in mutants YQR and YQRF. Biophys Chem 2004;109:41–58
8. Liang J, Edelsbrunner H, Fu P, Sudhakar PV, Subramaniam S. Analytical shape computation of macromolecules: II. Inaccessible cavities in proteins. Proteins 1998;33:18–29
9. Byrem WC, Armstead SC, Kobayashi S, Eckenhoff RG, Eckmann DM. A guest molecule - host cavity fitting algorithm to mine PDB for small molecule targets. Biochim Biophys Acta Proteins 2006;1764:1320–4
10. Hubbard SJ, Argos P. Cavities and packing at protein interfaces. Protein Sci 1994;3:2194–206
11. Eckenhoff RG. Promiscuous ligands and attractive cavities; How do the inhaled anesthetics work? Mol Interv 2001;1:258–68
12. Cantor R. The lateral pressure profile in membranes: a physical mechanism of general anesthesia. Biophys J 1997;36:2339–44
13. Koubi L, Tarek M, Bandyopadhyay S, Klein ML, Scharf D. Effects of the nonimmobilizer hexafluroethane on the model membrane dimyristoylphosphatidylcholine. Anesthesiology 2002;97:848–55
14. King GI, Jacobs RE, White SH. Hexane dissolved in dioleoyllecithin bilayers has a partial molar volume of approximately zero. Biochemistry 1985;24:4637–45
15. Franks NP, Jenkins A, Conti E, Lieb WR, Brick P. Structural basis for the inhibition of firefly luciferase by a general anesthetic. Biophysical J 1998;75:2205–11
16. Bhattacharya AA, Curry S, Franks NP. Binding of the general anesthetics propofol and halothane to human serum albumin. J Biol Chem 2000;275:38731–8
17. Liu R, Loll PJ, Eckenhoff RG. Structural basis for high affinity volatile anesthetic binding in a natural 4-helix bundle protein. FASEB J 2005;19:567–76
18. Eckenhoff RG, Johansson JS. On the relevance of “clinically relevant anesthetic concentrations” in in vitro studies. Anesthesiology 1999;91:856–60
© 2008 International Anesthesia Research Society