Some scientific reflections on possible mechanisms of general anaesthesia.
The answer is contingent on both the essence of science and the research problem involved. This paper will first make some general points about science and its procedures and, in an attempt to align our understanding of the actions of anaesthetic drugs with the pharmacodynamics of virtually all other drugs, then seek to dispel the common notion that the mode of administration of any drug, but especially of anaesthetic agents as gas, liquid or solid, is in any way relevant to its mode of action. Accordingly, this paper will make its points through citation of published scientific data which the author considers throw reliable light on a number of important general mechanisms and sites of action of anaesthetic drugs, which might be relevant to their clinical action. It does not seek to deal with minutiae which are of more pressing importance to research scientists than to informed clinicians. Nor does it claim to be an exhaustive survey of the literature.
GENERAL RELEVANT SCIENTIFIC PRINCIPLES
Science is our attempt to understand the physical and biological world which we inhabit by two processes: 1) the most careful and accurate observation of which we are capable; and 2) subsequent rigorous analysis of those observations. However, this is not simply strolling into the world and looking around carefully. It is most profitably done when we have a specific question in our minds. And there's the rub, because, to achieve success, we need to be agreed on what the question is and what it is about. The first question, therefore, has to be: What is general anaesthesia? The easy answer might appear to be that it is the loss of consciousness. But what is 'consciousness'? This question has troubled philosophers and scientists for millennia: it is the Golden Fleece of modern neuroscience. And until we answer it, we cannot really formulate a meaningful question about general anaesthesia. In short, we do not remotely know what it is we are trying to understand, hence we cannot approach the fundamental question scientifically.
Furthermore, since we cannot say what consciousness is, we must guard against the mostly poorly-resisted temptation to say that there is a unitary means by which that nebulous concept is 'lost'. The neuroscientist, Wolf Singer (1), considers consciousness to be a "binding response" to the fact that, as he says, "even simple stimuli evoke responses in numerous neurons with differing but overlapping feature preferences" and hence, "responses evoked by a particular stimulus need to be identified and bound together for further joint processing and must not be confounded with responses to other, nearby stimuli"; in this way consciousness (or "phenomenal awareness, the ability to be aware of one's sensations and feelings") is "characterized by the dynamic binding of feature-specific cells into functionally coherent cell assemblies, which as a whole represent the constellation of features defining a particular perceptual object" (2). In a paper which is, inevitably, as speculative as it is scientific, Tononi and Koch (3) talk rather of "consciousness as Integrated Information", which seems essentially the same concept.
It would follow, therefore, that interruptions and perturbations at a wide variety of sites and, possibly, through strikingly disparate means, would terminate that 'emergent' result which we term consciousness. In short, it may be futile to look for 'the' mechanism of general anaesthesia.
There is a second major issue with the understanding of the actions of anaesthetic drugs. It relates to the nature of science as an activity, in particular the analysis of and the search for an explanation of our observations. When, as with anaesthesia, the problem seems intractable, false explanations can come to be held tenaciously and the Meyer-Overton Hypothesis, which has had such a powerful influence on thinking in this field to the extent that it is often fallaciously termed the Meyer-Overton Rule, is, I believe, a classical example of misinterpretation of scientific results. (See Figure 7 for an example of the fallacy.)
THE "LIPID-SOLUBILITY" (MEYER-OVERTON) HYPOTHESIS
This concept was formulated at a time when almost nothing was known about cell membranes, except that they are rich in lipids. Furthermore, this was when lipid chemistry had developed quite well but protein chemistry was virtually non-existent. We now have a clear idea of the composition and structure of cell membranes (of all kinds): they are bi-laminar leaflets of lipid with proteins dotted about, a little like studs in a leather belt, except that while some of the proteins are fixed in position, others have some mobility (Figure 1; Singer and Nicholson (4)). Even this representation is a simplification since, for clarity, it omits the diverse and functionally important polysaccharides which are bonded to the proteins. Concepts of membrane functions, therefore, must take serious account of protein functions (pumps, receptors, channels etc.) as well as of the contributions of the lipids.
Though I argue that, like the vast majority of other drugs, anaesthetics are most likely to act at specific protein sites, the participation of membrane lipids cannot be discarded as irrelevant; nor, as is often the case, can they be regarded as homogeneous in composition and significance. Two properties, one chemical and the other physical, should be considered here. The first is the lipid inhomogeneity of those membranes; the other is the importance of membrane fluidity, which is likely to influence the mobility of integral proteins as they alter their configuration and move from functionally inactive to active states (so-called "allosteric" change).
The two lipid laminae are decidedly heterogeneous in their chemistry (Figure 2; Mathews and van Holde (5)) and this ordered state (a condition of low entropy) is, according to the important Second Law of Thermodynamics, an energy-requiring and therefore not an accidental condition: it is the result of animophospholipid translocase activity in the cell membranes (Figure 3; DeVaux et al (6)). The chemical and physical properties of these membrane leaflets are, accordingly, different. Prokaryote cells (and perhaps eukaryotes as well) also control the fluidity of their membranes. Figure 4A summarises an experiment by Vigh et al (7) in which bacteria were cultured at different temperatures: as ambient temperature fell, the stiffness of the membrane increased and, accordingly, the organism began to transcribe the gene for a desaturase enzyme, i.e. fatty acids of the cell membrane became increasingly unsaturated (with a lower melting point). When a different experiment was done, in which the fatty acids were first hydrogenated (i.e. their melting point was increased), then likewise more desaturase was produced and membrane fluidity increased (Figure 4B). The fact that this involves a genetic mechanism underlies the importance of the physical properties of the membrane, properties which, according to the Meyer-Overton concept, can be changed by "classical" anaesthetic agents.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
So a lipid-directed action of anaesthetics can be set aside only with some discretion. Nonetheless, apart from what has already been said, its other serious disadvantage is that such a gestalt action on all cells has to be deemed unlikely, in view of the selectivity of anaesthetic action and the subtle differences in pharmacodynamic profiles from anaesthetic agent to agent. In summary, therefore, the thesis of this review is that the Meyer-Overton concept is fallacious: the experimental data are correct but the interpretation of the results is misguided. The false focus is on lipid solubility. The attention should, rather, be on its reciprocal: hydrophobicity (or water insolubility).
[FIGURE 4 OMITTED]
THEORIES BASED ON ALTERED BEHAVIOUR OF WATER MOLECULES
Relationships with water molecules are also a crucial aspect of the "molecular theory", which the Nobel Prize-winning chemist, Linus Pauling (8) and his American colleague Miller (9) proposed virtually simultaneously over 40 years ago. Their concept involved the promotion by the anaesthetics of quasi-spherical assemblies or "clathrates"--akin to Buckminster-Fullerine ("Bucky-Balls")--which are comprised of water molecules held together by hydrogen bonds. These structures were putatively stabilised by the anaesthetic molecule located at their centre. The means by which these constructs achieved anaesthesia were not made clear by these eminent scientists and it must be said that their notions of neural function were, even by the standards of the time, rather fanciful; Pauling even suggesting that microcrystals of water--essentially ice--were formed which compromised the "encephalonic activity" of the nervous system. Nonetheless, he recognised that any properties of the lipids would be changed only slightly by anaesthetics, i.e. the lipids were probably unimportant in the process, but postulated that interactions of these "microcrystals" with charged side-chains of membrane proteins were likely to be of real significance. He produced a plot of anaesthetic potency, correlating with the equilibrium partial pressure of the hydrate crystals incorporating the anaesthetics, which was no less impressive than the Myer-Overton plots but no more explicative, thereby (Figure 5).
These ideas fell quickly from favour, but may well still have a point to make especially since, in the same year, Featherstone et all (10) demonstrated interactions of "inert" anaesthetic gases with proteins, and later physical studies by Enders (11) drew a few related ideas together attractively. As a physicist, Enders studied the effects on the properties of synthetic phosphatidylcholine bilayers (a protein-free simulacrum of the cell membrane) of a series of volatile anaesthetic agents at "medically relevant" concentrations (trichloroethylene, chloroform, methoxyflurane, diethyl ether, n-hexane) and found that all of them produced a prompt and reversible reduction in the dielectric function of these membranes. He ascribed this action to "a decrease in the Debye relaxation frequency" (inter alia, a measure of water diffusion which behaves inversely proportionally to viscosity, being a time-measure of the redistribution of water molecules in the water lattice) "and, correspondingly, an increase in viscosity of those water molecules which are localised at the membranes' surfaces" but not (as implied in the Pauling and Miller hypotheses) solely to an effect on the water "surrounding" the cell membranes.
[FIGURE 5 OMITTED]
We now know that the three-dimensional conformation of proteins is crucial to their function and that activation (as mentioned previously) commonly involves a conformational ("allosteric") change, whether for example in myofibrils during muscle contraction, activation of membrane-bound adenylate cyclase or of retinal rhodopsin. Any impediment to this conformational change will interfere with protein function probably because active sites in the protein are not brought into a more favourable reactive location. We also know that the surrounding water molecules play a role in such changes. "It is", according to Lehninger (12), "the tendency of the surrounding water molecules to relax into their maximum-entropy state that brings about the transition of the polypeptide chain from a random unfolded state to a highly ordered tertiary conformation". The "microcrystals" which the anaesthetics promote represent a state of low entropy (i.e. high order) which must, therefore, have an effect on the conformation of functional proteins.
So those who are attracted to the "lipid solubility" notion are not to be idly disregarded and certainly theories involving membranes should not be set aside too readily (see also the data in Table 1), while total rejection of the Pauling hypothesis might also be inappropriate.
CAN OLDER THEORIES BE RECONCILED AND SUSTAINED?
This nettlesome and dichotomous problem of anaesthesia has encouraged some thinkers in the field to hold the view that there are at least two distinct modes of action, with the gaseous agents in a special category. One of those grounds is the argument (most recently put to me in a personal communication  by Professor Steven Shafer), that those agents must have a distinct mechanism of action on account of the high concentrations at which they act. But is that really so? Data in Table 1 might challenge that notion.
Certainly it shows that nitrous oxide ([N.sub.2]O) is the least potent of those eight listed agents, especially when considered in terms of the alveolar gas tensions which correlate with a common level of anaesthesia (here E[D.sub.50]). However, this measure is (despite anaesthetists' enduring devotion to the MAC concept) probably not an entirely valid criterion. Whatever the mechanism (or mechanisms) of actions of these agents, it must be considered a chemical reaction and such reactions are driven by the relevant energy of the reactants (most commonly thought of in terms of the Law of Mass Action, where the molarity of the reactants, as well as the temperature, is a crucial factor). Certainly, too, when the molarity of the anaesthetic agents in the aqueous phases (extracellular and intracellular fluids) is considered, [N.sub.2]O is again the least potent of these drugs (Table 1). However, consideration of the molarity in the various lipid phases (e.g. the plasma membranes of neurons, their myelin sheaths and intracellular membranes) presents an entirely different picture with [N.sub.2]O, for instance, more potent than chloroform, halothane and enflurane.
In the lipid phase the potency ratio (from the most to the least potent drug) is 3.3, whereas in the aqueous phase it is 80, with different drugs at the extremes in each group. That 80:1 ratio may seem so large as to indicate disparate actions, yet when the barbiturates are considered (Table 2)--with which there is, surely, some agreement on the likely site of action--that potency ratio is 487, a far more spectacular figure (especially since these drugs are a single chemical family).
The far lower potency ratio when the lipid phase is considered (3.3 cf 80) might also be an encouragement to those who prefer the notion of a lipid site of action rather than one at a locus in a protein. The data, though, do not clinch an argument that incompatible fundamental actions are involved.
THEORIES BASED ON REACTIONS WITH SPECIFIC PROTEIN SITES ('RECEPTORS')
Despite the ideas which I have outlined so far, a more important principle, and one not idly to be set aside, is the fact that increasingly (but especially since Ahlquist's definition of alpha and beta adrenergic effects in 1948--indeed elsewhere in pharmacological thought, it is universally) we believe that drug actions are exerted at quite specific 'receptor' sites and that those receptors are loci in protein molecules.
So some relevant questions become: a) can anaesthetic agents react with proteins; and b) if so, what might the mechanism (or mechanisms) of that reaction be? Then c) what might be the functional consequences of those reactions?
In trying to answer that last question, we need to keep in mind the panoply of functions which proteins serve: anabolic and catabolic enzymes, channels and transporters, specialised receptors (for chemicals or various forms of energy in the internal or external worlds), contractile proteins, hormones, antibodies amongst them. Accordingly, the anaesthetic 'receptors' in question may be integral proteins of membranes (plasma or intracellular membranes of diverse sorts, including those in the cell nuclei) or they may be in solution in the extracellular or intracellular fluid, including hormones and various plasma proteins.
Some of the most important modern work bearing on the first of those questions (i.e. can anaesthetics react significantly with proteins?) has been done in the London laboratory of Franks and Lieb (summarised in 1984 (14) and 1994 (15)). I include here one clear and cogent experiment. They used a fire-fly protein luciferase, an enzyme which, with its substrate luciferin, leads to light emission--so the reaction can be followed well (it has a history as an assay for adenosine triphosphate). They followed the reaction in pure protein solutions (i.e. in the absence of any lipid) with different concentrations of halothane and made Lineweaver-Burk plots (Figure 6 (14)). The result was, as expected of a typical Michaelis-Menten enzymatic reaction, invariably a straight line, each with a different slope that was contingent on the halothane concentration, i.e. the quiddity of the reaction was unaltered, merely the velocity. Even more important is the fact that the lines converge to a common point on the y-axis, a result which is classically indicative of competitive inhibition of the enzyme. This outcome means that halothane acts at the very same site on the enzyme protein as the substrate (luciferin) does. Therefore that site is the formal equivalent of a drug receptor.
The point of this experiment is not that the luciferase is a target for anaesthetic action; it is, rather that the action of halothane occurs at a specific locus in a protein. Nevertheless, there is a striking parallelism in these actions--anaesthesia and luciferase inhibition (Figure 7B)--and it is certainly no less convincing than the 'classical' Meyer-Overton relationship (Figure 7A). This concept is further reinforced by the differential pharmacological actions of stereoisomers of anaesthetic drugs which would be predicted to react differently with functional protein sites. Franks and Liebs (15) reported appropriate differences in the potencies of isoflurane isomers while Carmody et al found that barbiturate isomers exerted opposite effects on nociception in conscious animals (16).
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
A different kind of study by McCrae et al provides complementary support for the argument in favour of a protein targets (17). Citing strains of mice that had been bred to achieve different sensitivities to nitrous oxide (one strain 50% more sensitive), they examined sensitivity to halothane and enflurane in mice which had been selectively bred for differential sensitivity to diazepam (in different tests there was a 9- to 14-fold difference in the dose of the benzodiazepine which was needed to elicit similar effects). In these experiments, the mice that were more sensitive to diazepam were also more sensitive to the gaseous anaesthetics: with halothane they were 20.9% more sensitive than the 'diazepamin-sensitive' mice, and with enflurane they were 22.7% more sensitive than the less sensitive strain. Such differences in sensitivity are most unlikely to be due to lipid differences (and, in any case, in the diazepam studies the different groups of mice had essentially identical brain concentrations of the drug while manifesting different sensitivities to it).
EVIDENCE FOR ACTIONS AT PARTICULAR FUNCTIONAL LOCATIONS IN THE NERVOUS SYSTEM
So what might the effects be when these various agents act at functional protein loci which, plausibly, could be regarded as anaesthetic sites?
[FIGURE 8 OMITTED]
An effect on impulse propagation down axons would be one such and, indeed, several gaseous/ volatile anaesthetics substantially reduce the opening of the voltage-gated sodium-ion channels which are the basis for action potentials (Figure 8, Rehberg et all'). The figure shows the effects of desflurane (on channels originating from the central nervous system), but these authors reported essentially the same effect with halothane, sevoflurane, isoflurane, enflurane and diethyl ether at clinically relevant concentrations. In a subsequent paper they showed an identical action by propofols (19). All of these agents have a further important action on sodium ion channels. It should be remembered in this context that those ion channels have three states: open (also known as "activated"), closed (but able to be opened by depolarisation) and "inactivated" (which could also be described as 'locked'), with the channels in that last category (which is, itself, a consequence of depolarisation) being unable to contribute to any axonal activity. Rehberg and Duch found (18) that, whereas the inactivation (or 'locking' process) normally began in their cells at about -75 mV (and increased with further depolarisation: Figure 9), in the presence of the anaesthetic (diethyl ether in the figure, but essentially the same for the other volatile anaesthetics) this inactivation is intensified and essentially begins around -100 mV These authors found the same phenomenon with propofol.
[FIGURE 9 OMITTED]
Taken together, both of those consequences for the sodium ion channels would markedly reduce the amplitudes of any action potentials (thereby probably slowing conduction velocity and the attainable impulse frequency). That action would also reduce the secretion of synaptic transmitter because that process is determined by influx of [Ca.sup.2+] ions into the terminal and the opening of those [Ca.sup.2+] channels is a function of the amplitude of the action potential, i.e. on the extent of the opening of the [Na.sup.2] channels.
In any case, such disparate effects--channel opening and inactivation--at two presumably distinct functional sites on the channel protein strongly argue for targeted actions on proteins, rather than some generalised lipid effect.
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
The second functional effect is reduced transmitter release, for which there might be an additional direct anaesthetic action on [Ca.sup.2] channels. Figure 10, which summarises results obtained with isolated synaptosomes (20), shows that barbiturate directly inhibits the entry of calcium ions (triggered here by three distinct methods with essentially identical results), an effect which would (see above) reduce transmitter release from synaptic terminals of the kind from which these synaptosomes were obtained. Since the brain concentration of pentobarbitone during anaesthesia is about 0.16 mM (21), there would be, as a direct effect of barbiturate at this concentration, about a 25% reduction of [Ca.sup.2+] influx (using these data of Blaustein and Ector), again indicating a significant reduction in transmitter secretion and, accordingly, in synaptic efficacy. Such a presynaptic effect on [Ca.sup.2+] channels would plausibly explain the diminished synaptic transmission which Weakley found when modest doses of thiopentone or pentobarbitone were administered to spinalised and paralysed (but otherwise unanaesthetised) cats (22). In the myotatic reflex, the transmitter is glutamate at the excitatory synapse between the la fibres and the motor neurons in the ventral horn of the spinal cord; these drugs reduced its quantal release by a presynaptic mechanism, leading to 'inhibition' through reduced efficacy of an excitatory synapse. Incidentally, this finding is contrary to the popular notion that barbiturates act exclusively at GABA receptors (and, therefore, postsynaptically: the data of Blaustein and Ector (20) also invalidate that view).
Such an action is also seen with volatile anaesthetic agents. Figure 11 shows the action of methoxyflurane, which has little or no effect on 'basal' (or unstimulated) [Ca.sup.2] influx into these isolated adrenal medullary cells, but has a profound effect on the action of carbachol (which mimics the action of the normal transmitter at this splanchnic synapse: acetylcholine) and on the influx which is induced by the depolarising action of 77 mM [K.sup.+] ions (23). As would be expected, the secretion of catecholamine by these cells was also reduced (the carbachol-induced secretion being reduced by 50% at the highest dose of methoxyflurane). Clinically effective blood levels of methoxyflurane are upwards of 640 [mu]molar". The actions of halothane, enflurane and isoflurane were essentially the same in these experiments.
The functional consequences of such reductions in neurotransmitter secretion would, obviously, depend on whether an excitatory or an inhibitory synapse were involved.
Those previous effects are all examples of presynaptic mechanisms. Post-synaptic effects have also been found experimentally. Typically, the released transmitter binds to a protein receptor on the postsynaptic membrane and, because that binding perturbs the charge relationships in the protein molecule (relationships which determine its three-dimensional configuration), commonly protein conductivity channels open in that membrane for particular ions. Depending on the ions involved and the electro-chemical forces acting on them, currents will flow into or out of the cell (while the channel remains open), leading to a voltage change and, therefore, a change in the excitability of the postsynaptic cell, i.e. its propensity to generate an action potential. Modern techniques have allowed those currents to be measured, notably with a technique known as "patch-clamping", and the opening times and the frequency of the opening to be determined.
This result (Figure 12; Wachtel (24)) shows a cell in which such currents have been measured as they flow through individual channels which are linked with acetylcholine receptors. Note, in the left panel, that the process is stochastic, i.e. though the receptor-channel complex is constantly exposed to the transmitter, it is not constantly open: there is a randomness in the temporal spacing of open periods and these periods are, themselves, of inconstant duration. The magnitude of the current pulse is however constant, i.e. the conductivity of the individual channel does not change. The depolarisation of the cell is contingent on the charge transfer, which is determined by the duration of the current pulses and their frequency. The records in the right panel indicate that the duration of the opening of these post-synaptic channels by acetylcholine was reduced by about 90% by ketamine; accordingly, synaptic efficacy would be seriously compromised. Torda and Gage likewise reported that ketamine, as well as barbiturates, reduces synaptic transmission at the cholinergic neuromuscular junction through a postsynaptic effect, providing yet another example of a barbiturate action beyond GABA receptors (25).
[FIGURE 12 OMITTED]
At other synapses with different transmitters, by contrast, synaptic function can be greatly enhanced by anaesthetic drugs. When, as with the results shown in Figure 13, such a synapse is inhibitory (as here, with GABA as the transmitter) we have a plausible and readily understood link with an anaesthetic action. This illustration shows the actions of three anaesthetic drugs in cells which have been transfected with the genes for the human GABA receptor. Those drugs have quite disparate chemical structures, but all of them substantially increase the synaptic efficacy of GABA (the anaesthetic drug etomidate has a similar potentiating effect [data are not shown]). It should be recalled that binding studies (e.g. with radioactive GABA) indicate quite specific sites of binding of this inhibitory agonist (i.e. to receptors). Furthermore, Belelli and his colleagues found decidedly lesser sensitivities for those potentiating effects when they transfected the cells with genes for GABA receptors from the fruit-fly, Drosophila melanogaster (indicating that the membrane protein subserves the synaptic effects, given that the cells' lipids were not affected in the transfection process). In recent years, a great deal of experimental attention has been given to this receptor. Like the nicotinic cholinoceptor, it is a pentameric structure and modern techniques of genetic engineering (in particular the production of "knock-out" and "knock-in" mice) have been employed to determine the influences of those five protein strands on the properties of the receptors. Mostly such detail is beyond the scope and intent of the present paper, but has been usefully summarised by Rudolph and Antkowiak in particular".
[FIGURE 13 OMITTED]
Actions on "second messenger" systems
There is a second class of transmitter action, one which does not involve the previously discussed direct opening of conductance channels, but achieves transmission through what are called "second messengers". In such systems, when the transmitter (or receptor agonist) binds to the post-synaptic receptor it produces a configurational change in an inwardly-facing protein of the plasma membrane of the target cell. That protein is termed a "G-protein" because of the fact that its enzymatic activity is determined by its binding to a guanyl nucleotide: it is active when bound to GTP (guanosine triphosphate) from the cytoplasm but inactive when bound to GDP. It exchanges between the GTP and GDP according to whether the agonist in the extracellular fluid is bound to its receptor. When activated, the G-protein dissociates a sub-unit which binds to, and thereby, depending upon the type of cell involved, activates either a) adenyl cyclase which then catalyses the hydrolysis of cytosolic adenosine triphosphate to the second messenger cyclic AMP; or b) phospholipase C which then promotes the hydrolysis of the membrane constituent, phosphatidylinositol bisphosphate, to two second messengers: inositol trisphosphate and diacyl glycerol.
[FIGURE 14 OMITTED]
In certain muscarinic synapses in the central nervous system, binding by intracellular GTP reduces the affinity of the receptor for its agonist(28) (Figure 14B). This change in binding affinity is abolished by halothane (Figure 14bB). Obviously, that would influence the production of second messenger and hence the efficacy of the synapse.
One of the roles of diacyl glycerol is the activation of an enzyme called protein kinase C (PKC) which, inter alia, phosphorylates ion channels in cell membranes (and by so doing changes their configuration and electrical properties, e.g. as conductive channels). Figure 15 shows that anaesthetic alcohols significantly inhibit PKC (with a potency [[EC.sub.50]] that is a function of the chain length) and that enflurane and propofol do so as well (29); an earlier paper from this group had reported PKC inhibition by halothane also (30). Their measurements of the energetics for the transfer of an alcohol molecule to the PKC binding site strongly suggested that this occurred from an aqueous medium to a hydrophobic site.
[FIGURE 15 OMITTED]
Such inhibition of PKC would be expected to affect neuronal excitability in a functionally relevant way.
Enzymatic actions generally
Another important enzyme--which achieves hydroxylation in hepatic microsomes and mitochondria in various tissues and is exceedingly important in numerous detoxification reactions--is cytochrome [P.sub.450]. Unlike most enzymes (PKC is also an exception to this generalisation), this complex is inhibited by various anaesthetic agents. Figure 16 shows that in a series of compounds of the alkan-1-ol class (a saturated aliphatic chain with the alcohol on the terminal carbon) all members (at least up to C=14) inhibit [P.sub.450] and that (up to C=11), there is increasing potency as chain length increases(31). Furthermore, there is a strong parallel between the inhibitory concentrations (open circles) and anaesthetic potency (filled circles). This property holds (according to the data of LaBella et al, 1997) in the five different alcohol series which they studied (those with the -OH group in the terminal carbon, on the penultimate carbon, on both terminal carbons and in two series of aromatic alcohols), though the anaesthetic potencies showed some variation in the different chemical classes.
[FIGURE 16 OMITTED]
Even if it is not obvious how this inhibition might contribute to the action of anaesthetics, the effect may be important in their toxicity. The essential point, though, is that there are several anatomical and functional sites where anaesthetic actions might plausibly be exerted.
Until we have a better understanding of the basis of consciousness, we are most unlikely to be able to achieve a gestalt theory of general anaesthesia (or theories, because there may need to be several). Nevertheless, we can reasonably postulate that the brain is central to consciousness; hence a focus on generalised neural effects of anaesthetics--as I (and others) have sought to do--is most likely to be fruitful. This paper considered several such operational sites where particular functional proteins might be affected by a diversity of anaesthetic drugs and, in doing so, emphasised that the mode of administration is not a valid criterion for making distinctions about those actions. Those sites included ion channel proteins of axons (both for action potential propagation and the [Ca.sub.2+] influx which drives transmitter secretion), post-synaptic receptor sites and some of the proteins involved with the production of second messengers as well as on a few other enzymes.
Finally, it argued that it is also more than time that reference to the Meyer-Overton theory (or 'law', as some authors overstate it) should be expunged from anaesthetic textbooks and teaching and that it be replaced by the concept of an action at hydrophobic sites in important target proteins. This is not only consistent with contemporary pharmacological principles but is also far more logical, given the spectrum of the wider physiological actions of anaesthetics, a diversity which seems entirely incompatible with a homogeneous action in lipids.
In summary, this review considers some reasons, notably the enduring problem of defining consciousness, for the lack of any agreed gestalt theory of general anaesthesia. It discusses published reports which give support to a lipid-based mechanism, but rejects them to prefer a protein-based mechanism, which is consistent with the receptor theory that is the foundation of contemporary pharmacology. Anaesthetic actions on axonal and transmitter release sites are considered as well as possible postsynaptic actions at transmitter receptors and channels, as well as loci relating to second messengers and particular enzymes. The paper does not support the concept of distinct mechanisms of action based on the mode of administration of anaesthetic drugs, i.e. whether they are gaseous or non-gaseous.
This is a substantially expanded version of a presentation to the Annual Scientific Meeting of the Australian and New Zealand College of Anaesthetists, Sydney (May 2008).
Address for reprints: Dr J. J. Carmody, 22 William St, Roseville, NSW 2069.
Accepted for publication on September 25, 2008.
(1.) Singer W Consciousness and the binding problem. Ann NY Acad Sci 2001; 929:123-146.
(2.) Kreiter AK, Singer W Stimulus-dependent synchronization of neuronal responses in the visual cortex of the awake macaque monkey. J Neurosci 1996; 16:2381-2396.
(3.) Tononi G, Koch C. The neural correlates of consciousness: an update. Ann NY Acad Sci 2008; 1124:239-261.
(4.) Singer SJ, Nicolson GL. The fluid mosaic model of the structure of cell membranes. Science 1972; 175:720-731.
(5.) Mathews CK, van Holde KE. Biochemistry, 2nd ed. San Francisco: Benjamin Cummings, 1996.
(6.) Devaux PF, Zachowski A, Morrot G, Cribier S, Fellmann P, Geldwerth D et al. Control of the transmembrane phospholipid distribution in eukaryotic cells by aminophospholipid translocase. Biotechnol Appl Biochem 1990; 12:517-522.
(7.) Vigh L, Los DA, Horvath I, Murata N. The primary signal in the biological perception of temperature: Pd-catalyzed hydrogenation of membrane lipids stimulated the expression of the desA gene in Synechocystis PCC6803. Proc Natl Acad Sci USA 1993; 90:9090-9094.
(8.) Pauling L. A molecular theory of general anesthesia. Science 1961; 134:15-21.
(9.) Miller SL. A theory of gaseous anesthetics. Proc Natl Acad Sci USA 1961; 47:1515-1524.
(10.) Featherstone RM, Muehlbaecher CA, DeBon FL, Forsaith JA. Interactions of inert anesthetic gases with proteins. Anesthesiology 1961; 22:977-981.
(11.) Enders A. The influence of general, volatile anesthetics on the dynamic properties of model membranes. Biochim Biophys Acta 1990; 1029:43-50.
(12.) Lehninger AL. Biochemistry, 2nd ed. Worth Publishers 1975.
(13.) Firestone LL, Miller JC, Miller KW Tables of physical and pharmacological properties of anesthetics. In: Roth SH, Miller KW eds. Molecular and cellular mechanisms of anesthetics. Plenum Medical Book Company 1986.
(14.) Franks NP, Lieb WR. Do general anaesthetics act by competitive binding to specific receptors? Nature 1984; 310:599-601.
(15.) Franks NP, Lieb WR. Molecular and cellular mechanisms of general anaesthesia. Nature 1994; 367:607-614.
(16.) Carmody J, Knodler L, Murray S. Paradoxical modulation of nociception in mice by barbiturate agonism and antagonism: is a GABA site involved in nociception? Fur J Neurosci 1991; 3:833-838.
(17.) McCrae AF, Gallaher EJ, Winter PM, Firestone LL. Volatile anesthetic requirements differ in mice selectively bred for sensitivity or resistance to diazepam: implications for the site of anesthesia. Anesth Analg 1993; 76:1313-1317.
(18.) Rehberg B, Xiao YH, Duch DS. Central nervous system sodium channels are significantly suppressed at clinical concentrations of volatile anesthetics. Anesthesiology 1996; 84:1223-1233.
(19.) Rehberg B, Duch DS. Suppression of central nervous system sodium channels by propofol. Anesthesiology 1999; 91:512-520.
(20.) Blaustein MP, Ector AC. Barbiturate inhibition of calcium uptake by depolarized nerve terminals in vitro. Mol Pharmacol 1975; 11:369-378.
(21.) Carmody JJ, Graham GG, Ruigrok MA. Stress in mice increases intrinsic pentobarbitone sensitivity by a predominantly pharmacodynamic mechanism. Clin Exp Pharmacol Physiol 1991;18:703-710.
(22.) Weakly JN. Effect of barbiturates on 'quantal' synaptic transmission in spinal motoneurones. J Physiol 1970; 204:63-77.
(23.) Pocock G, Richards CD. The action of volatile anaesthetics on stimulus-secretion coupling in bovine adrenal chromaffin cells. Br J Pharmacol 1988; 95:209-217.
(24.) Wachtel RE. Ketamine decreases the open time of single-channel currents activated by acetylcholine. Anesthesiology 1988;68:563-570.
(25.) Torda TA, Gage PW Postsynaptic effect of i.v. anaesthetic agents at the neuromuscular junction. Br J Anaesth 1977; 49:771-776.
(26.) Belelli D, Callachan H, Hill-Venning C, Peters JA, Lambert JJ. Interaction of positive allosteric modulators with human and Drosophila recombinant GABA receptors expressed in Xenopus laevis oocytes. Br J Pharmacol 1996; 118:563-576.
(27.) Rudolph U, Antkowiak B. Molecular and neuronal substrates for general anaesthetics. Nat Rev Neurosci 2004; 5:709-720.
(28.) Aronstam RS, Anthony BL, Dennison RL Jr. Halothane effects on muscarinic acetylcholine receptor complexes in rat brain. Biochem Pharmacol 1986; 35:667-672.
(29.) Slater SJ, Kelly MB, Larkin JD, Ho C, Mazurek A, Taddeo FJ et al. Interaction of alcohols and anesthetics with protein kinase C-alpha. J Biol Chem 1997; 272:6167-6173.
(30.) Slater SJ, Cox KJ, Lombardi JV, Ho C, Kelly MB, Rubin E et al. Inhibition of protein kinase C by alcohols and anaesthetics. Nature 1993; 364:82-84.
(31.) LaBella FS, Chen QM, Stein D, Queen G. The site of general anaesthesia and cytochrome P450 oxygenases: similarities defined by straight chain and cyclic alcohols. Br J Pharmacol 1997; 120:1158-1164.
J. J. CARMODY *
School of Medical Sciences (Discipline of Physiology) and Centre for Values, Ethics and Law in Medicine, The University of Sydney; formerly School of Physiology and Pharmacology, University of New South Wales, New South Wales, Australia
* M.B., B.S., M.D.
TABLE 1 Calculated pharmaco-chemical values for selected gaseous and volatile anaesthetic agents (calculations by the present author, based on data from Firestone et al (13); 1986) Agent Alveolar [ED.sub.50]: aqueous [ED.sub.50]([section]) phase drug molarity (mM) Nitrous oxide 768 17.6 Ethyl chloride 304 4.46 Cyclopropane 76 0.89 Diethyl ether 15.2 9.82 Enflurane 12.8 0.59 Chloroform 6.1 1.36 Halothane 5.9 0.22 Trichloroethylene 4.3 0.39 Agent [ED.sub.50]: lipid phase molarity (mM) Nitrous oxide 58.6 Ethyl chloride 25.0 Cyclopropane 43.3 Diethyl ether 50.9 Enflurane 72.6 Chloroform 83.2 Halothane 68.3 Trichloroethylene 50.5 The values for these calculations were based on measurements from various sources including man, dog, rat and tadpole. (Firestone et al, 1986; with permission.) ([section]) [ED.sub.50] expressed as mmHg, alveolar gas. TABLE 2 [ED.sub.50] values for several barbiturate anaesthetics (tadpole and mouse data) Drug [ED.sub.50] molarity (mM) Thiopentone 0.03 Secobarbitone 0.08 Pentobarbitone 0.16 Hexobarbitone 0.16 Amylobarbitone 0.44 Butobarbitone 0.56 Phenobarbitone 3.30 Barbitone 14.6 Pentobarbitone ([section]) 0.14 (plasma) 0.16 (brain) Data from Firestone et al (13) (1986), with permission. ([section]) [ED.sub.100] mouse data from Carmody et al, 1991; all other values from tadpoles ([ED.sub.50]). The dose-response curve in mice is extremely steep: the dosage required to achieve anaesthesia in 100% of animals was 35 mg/kg i.p.
|Printer friendly Cite/link Email Feedback|
|Publication:||Anaesthesia and Intensive Care|
|Article Type:||Clinical report|
|Date:||Mar 1, 2009|
|Previous Article:||Fibreoptic intubation under general anaesthesia--a simple method using an endotracheal tube as a conduit.|
|Next Article:||A randomised, double-blind comparison of three different volumes of hypobaric intrathecal bupivacaine for orthopaedic surgery.|