Current assessment of targets and theories of anaesthesia

B. W. Urban

Klinik für Anästhesiologie und spezielle Intensivmedizin, Universitätsklinikum Bonn,Sigmund-Freud-Straße 25, D-53127 Bonn, Germany

Abstract

Br J Anaesth 2002; 89: 167–83

Keywords: anaesthesia, depth; anaesthesia, general; anaesthetics volatile; anaesthetics gases; anaesthetics i.v.; analgesia; brain, synapses; complications, amnesia; theories of anaesthetic action

In this final article I shall try to summarize new scientific insights resulting from and in connection with the Sixth International Conference on Molecular and Basic Mechanisms of Anaesthesia (MAC2001) regarding the targets at which anaesthetics act, the mechanisms underlying their actions at these targets, and the implications for theories of anaesthesia.

The first section of this review will consider primarily the targets of anaesthetic action that were presented at MAC2001, and the second section will summarize mechanisms of anaesthetic actions at these sites that were presented at the meeting and in earlier reviews. Discussion of whether these targets or mechanisms are relevant to general anaesthesia will be reserved for the final section, where theories of anaesthesia are considered.

Targets of anaesthetic action

As has been shown by the MAC2001 articles in this issue, in the conference proceedings110 and articles from previous conferences,31 32 93 95 97 there are numerous targets of anaesthetic actions at all levels of integration within the central nervous system (CNS). Even when targets are not part of anaesthesia-relevant circuits, such as the firefly luciferase,34 the nicotinic acetylcholine receptor (nAChR) of the electric stingray33 74 and the sodium channels of squid giant axons29 and the electric eel119, they may still serve as a useful model to elucidate mechanisms that are also at work in targets that are part of such anaesthesia-relevant circuits. Therefore, the question of whether or not these are relevant to theories of anaesthesia will be discussed in the section of this review entitled Theories of general anaesthesia. Here I shall briefly summarize what is already known and mention aspects that were emphasized during the conference.

Molecular targets
Meyer and Overton had independently discovered 100 yr ago that anaesthetic action correlates with the lipophilicity of anaesthetic drugs.71 82 This discovery has been interpreted as pointing to membranes as important sites of anaesthetic actions. Consequently, experimental in vitro models have concentrated mainly on membranes and membrane proteins and receptors. A goal of this kind of approach is the identification of specific receptors and molecular sites for anaesthetic action within the CNS. This approach considers that anaesthetic drugs are molecules that must act on other molecules and molecular structures of the CNS, such as lipids and membrane proteins, before they can give rise to functional changes within the CNS. This approach, it was hoped, would not only identify the essential molecular mechanisms of anaesthesia but also help to pinpoint the neuronal components within the CNS that are essential for general anaesthesia.

Figure 1 lists identified sites of anaesthetic interaction with membrane proteins embedded in lipid bilayers and with the bilayer that were discussed during MAC2001.22 74 86 100 107 Within the bilayer, anaesthetics may prefer the interface between the lipid and the aqueous phase, between the lipid and the membrane protein, or the hydrophobic core of the lipid bilayer itself, depending on the physicochemical nature of the anaesthetic.107 112 Anaesthetics may bind to protein-binding sites exposed to the aqueous phase, be they at the water–protein interface, within water-filled crevices inside membrane proteins or in the lumen of ion channels,22 100 or even within aqueous domains or water channels.107 Anaesthetics may bind within the core of the membrane protein itself, between hydrophobic {alpha}-helices37 and hydrophobic or lipophilic pockets,107 on which a great deal of emphasis was placed at the MAC2001 conference.74 107 Anaesthetics may interfere in the interactions between subunits of a protein or between different proteins.86 107 Sandorfy98 has pointed out that protein sites involve not only amino acids but also the carbohydrates that are covalently attached to membrane proteins.



View larger version (59K):
[in this window]
[in a new window]
 
Fig 1 Direct and indirect molecular anaesthetic target sites on ion channels.

 
Although these sites are purely hydrophilic, amphipathic or even purely hydrophobic, much interest focused on anaesthetic interaction sites that are amphipathic or lipophilic, i.e. sites that have some polar components besides a hydrophobic component.107 This is in line with earlier studies, for example the work by Taheri and colleagues103 in which they characterized the physicochemical nature of the site of anaesthetic action by determining what solvent best represents the site of action of inhaled anaesthetics in humans, rats and dogs; they concluded that lecithin was probably more representative of the site of action of these anaesthetics than the other solvents.

All members of the main families of ion channels and their many subtypes are affected by anaesthetics, including sodium channels, potassium channels, calcium channels (both voltage- and ligand-sensitive), glutamate receptors [N-methyl-D-aspartate (NMDA), {alpha}-amino-3-hydroxy-5-methylisoxazole propionic acid (AMPA) and kainate], the novel P2X receptors, nAChR, 5-HT3 receptor channels, GABAA receptor channels and glycine receptor channels.22 107 Although much attention was focused on ion channels in the past, there is good reason to investigate, for example, metabotropic receptors, which modulate synaptic transmission and partly bind the same ligands as ligand-gated ion channel receptors.86 Other protein targets now under investigation include protein kinases and phosphatases, G-proteins86 and protein pumps.83 86 Many subtypes exist on which anaesthetics may even have opposite effects.

Subcellular targets
Anaesthetics act on axons and dendrites and presynaptic and postsynaptic membranes as well as on the somatic membranes of neurones and glia.91 They act on many intracellular structures, such as the neurotransmitter release system,91 the calcium homeostasis and buffering system,86 126 second-messenger cascades86 and mitochondria.79

Cellular targets
The myth that anaesthetics act primarily on cells of the nervous system, on neurones, had already been contradicted at the beginning of the 20th century by Overton,82 who showed that anaesthetics acted on muscle cells as well as on plant cells. At the MAC2001 conference, additional evidence was provided that glial cells,123 skeletal126 and cardiac14 myocytes, endocrine cells80 and cells of the immune system68 121 are also targets.

Local microcircuits
The MAC2001 conference provided many examples of anaesthetic actions on microcircuits within slices from dorsal root ganglion,48 100 spinal cord,45 61 thalamus,101 hippocampus,4 7 11 12 56 81 96 99 cortex4 and cerebellum4 as well as in neuronal networks grown in culture.21

Systems
The rapidly progressing techniques of functional imaging of brain activity5 53 64 76 115 in conjunction with in vivo electrophysiological techniques5 114 have confirmed the findings of earlier studies92 that inhalation anaesthetics, to a greater extent than i.v. anaesthetics, affect all areas of the CNS. Besides a concentration-dependent general decrease in whole-brain activity during anaesthesia, imaging studies indicate that a number of discrete brain structures are related to the effects of anaesthetics, including the spinal cord, brainstem, cerebellum, midbrain and thalamus, midbrain reticular formation, basal ganglia, superior frontal gyrus, anterior cingulate gyrus, posterior cingulate, basal forebrain, insular cortex, prefrontal cortex, parietal and temporal association areas, occipitoparietal association cortices and occipital cortex.53 Functional imaging techniques are beginning to help identify key brain structures that appear to play important roles in the different clinical endpoints produced by anaesthetics. The peripheral nervous system100 also provides targets for anaesthetic actions, as does the endocrine system80 and the immune system.68 121 The so-called selectivity of anaesthetic action for the CNS or even for the brain appears to be based on historical prejudice rather than on scientific data, which is sparse on targets outside the CNS.

Mechanisms of anaesthetic action

Having briefly reviewed target sites of anaesthetic action, we shall now address the question of what mechanisms underlie the interaction between anaesthetics and these targets. Again, discussion of whether or not these anaesthetic mechanisms are relevant to general anaesthesia will be deferred until the final section, where theories of anaesthesia are considered. Suffice it to say at this point that, if the emphasis is on elucidating a mechanism of anaesthetic action, it is not necessarily important whether or not this mechanism occurs in an anaesthesia-relevant setting, whether it occurs at clinically relevant concentrations, and whether or not it is very potent in the experimental system in which it is studied.

Molecular: actions on proteins
Specific receptors
Originally, the term ‘receptor’ was a purely functional, theoretical construct within a theory of drug action that had been developed in order to describe the difference between specific and non-specific drug actions.24 49 A drug interacts specifically if it undergoes a specific reaction with a specific reagent. A drug interacts non-specifically if it shows physical interactions or acts as a solvent or foreign body.24 According to receptor theory, drugs bind to a macromolecule and as a result of this specific binding they trigger a specific biological effect.49 No longer are these macromolecules hypothetical postulates—in many cases we know their molecular structure. Today, receptors are defined as binding sites with a functional correlate.120 Macromolecules are considered to be receptors only if occupation of the binding site by an agonist alters the function of the macromolecule.

Pharmacologically, receptors are characterized by their specific ligands, agonists and antagonists.120 In order to demonstrate that a biological action of a drug results not from non-specific actions but from actions on a receptor, the effects of slight changes in the steric properties of the drug molecule can be investigated. Slight changes in steric properties will hardly change the physical or chemical properties of the drug at all. But because the drug has to match the receptor-binding site in the same way that a key fits a lock, even small steric changes may prevent the key from fitting the lock. Non-specific effects, on the other hand, depending on particle number or solvent behaviour, are not affected if the physical and chemical properties of the drug are changed only slightly. Stereoisomers are molecules possessing the same molecular formula but differing in the spatial arrangement of their atoms (though not in which atoms are joined to which other atoms). Stereoisomers that are mirror images of each other are called enantiomers. Stereoisomers have identical chemical and physical properties, but the arrangement of the atoms is such that no rotation in (three-dimensional) space can turn one stereoisomer into the other. This subtle difference may mean that two stereoisomers cannot bind to the same site and consequently that they cannot produce the same functional changes. If the stereoisomers of a drug produce different functional changes in a macromolecule, this is taken as important evidence that the drug acts on the macromolecule via a receptor.

While there may be substantial potency differences between the stereoisomers of i.v. anaesthetics both in vivo and in vitro,35 stereoisomers of the volatile anaesthetic isoflurane have only a small differential potency in the whole animal,35 in the activation of TOK1 potassium channels123 and in interactions with other ion channels.35 It has been argued previously that halogenated ethers may show weak specific receptor interactions because of their ability to form hydrogen bonds52 108 (Fig. 4A in Urban and Bleckwenn’s review).111 The fact that a direct measure of the binding of volatile anaesthetic to protein is complicated by weak affinity and therefore rapid kinetics17 also appears to confirm that specific receptor interactions for inhalation anaesthetics are weak.




View larger version (52K):
[in this window]
[in a new window]
 
Fig 4 Anaesthetic actions on a hypothetical neurone. Anaesthetic responses depend on networks. This is illustrated here for the spatial and temporal integration of excitation and inhibition within a model neurone before and during anaesthesia. Resting fibres are shown grey, black indicates propagating excitation. The number of arrows indicates the frequency of incoming signals, and non-aligned arrows on incoming fibres indicate a temporal shift in incoming signals. Apart from receiving excitatory and inhibitory input, the excitability of the model neurone is assumed to be controlled by tonic modulatory input (no input=low excitability, high frequency input=high excitability). Anaesthetics may act on the neurone directly by modifying incoming signals (presynaptically or postsynaptically), indirectly by influencing some upstream component so that the incoming modulatory signal is modified, or by a combination of both types of action. Whether the model neurone fires action potentials depends on how the excitatory, inhibitory and modulatory inputs are connected; only an arbitrary sample of possible outcomes is shown.

 
Cavities: hydrophobic pockets
Much attention has focused on preformed protein cavities as binding sites for inhaled anaesthetics. There is considerable experimental support for the hypothesis that small molecules can bind in cavities formed between {alpha}-helices in proteins.107 As the binding energies of anaesthetics to these sites of action appear to be small, these molecules bind, presumably adventitiously, to pre-existing cavities or sites. Therefore, the binding event is not expected to cause an ‘induced fit’ in a protein site or even to provide substantial reorganization of an internal cavity. The effect on protein stability of the binding of an anaesthetic to these cavities depends on their native sizes: proteins with intermediate pre-existing cavities are destabilized, which is interpreted as preferential binding of the anaesthetic to less stable intermediates with enlarged cavities; proteins containing larger cavities are stabilized by the anaesthetic, indicating binding to the native state.27 74

The volume of the cavity or binding pocket may change with the conformation of the protein, as was shown for glycine receptor channels, where the volume of binding pockets differs between the resting (smaller) and activated (larger) states.50 In GABAA receptor channels, the volume of the cavity and the proposed anaesthetic binding site is estimated as between 250 and 370 Å3, suggesting a common site of action for the anaesthetics isoflurane, halothane and chloroform.

Cavities: hydrophilic crevices
However, not only amphipathic cavities but also water-filled crevices in proteins seem to be important. Akabas and colleagues1 infer that, as the membrane-spanning segments undergo GABA-induced conformational changes, water-filled crevices extend from the extracellular surface into the interior of the protein. In order to avoid a vacuum that is energetically unfavourable, water moves into the crevices and cavities that form in the membrane-spanning domain during GABAA receptor gating, thus facilitating conformational changes. By preferentially filling these crevices and cavities, anaesthetics could thus stabilize receptor conformations other than the resting state, increasing the probability of channel opening.1 Anaesthetics may thus enter proteins not only by diffusion through the water-filled lumen of the ion channel or dissolution in the phospholipid bilayer followed by transfer through the lipid–protein interface of the ion channel, but also by transfer to an annular ring formed by the four-component interface of the ligand-binding and transmembrane domains of the protein, the phospholipid bilayer and the interfacing water layer.107

Interfaces: protein–lipid
Lipid bilayer membranes consisting of a bimolecular leaflet of lipid molecules constitute the backbone of a biological membrane. Lipid bilayers by themselves are not electrically excitable as they are perfect insulators, permitting no ion flow across the membrane. Therefore, electrically excitable membranes also contain many different proteins. Integral membrane proteins are essential for mediating a great number of physiological functions. In order to carry these out successfully, membrane proteins must perform properly within—and necessarily interact with—the lipid membrane, in which they change conformation while carrying out their complex functions. Indeed, there is much evidence for a strong effect of the properties of lipid bilayers on the function of membrane proteins.86 107

Pure lipid bilayers are changed by anaesthetics in many ways. Depending on the physicochemical nature of the anaesthetic, membrane physical structure and properties, such as thickness, surface tension, surface electric potential, fluidity and membrane disorder, may be changed and phase transitions may occur.112 Nash79 comments on the fact that lipid targets of anaesthetic action have fallen from favour, but that he knows of no decisive experiment that eliminates them from contention, especially if one acknowledges the possibility that subtle alterations of bilayers by volatile anaesthetics might influence the function of proteins embedded in them. Recent studies using site-directed mutations in ligand-gated ion channels suggest that a primary point of action of anaesthetics is in the transmembrane domain of these channels.107 Another example is provided by certain protein kinases; anaesthetics might operate at the protein–lipid interface by changing the lateral pressure profile.86

Interfaces: protein–protein
It has also been suggested that anaesthetics might be able to act at the interface between protein subunits or between different proteins so as to disrupt, for example, allosteric transitions at domain–domain interfaces of protein kinases, or to prevent agonist-induced dissociation of receptors from heterotrimeric G-proteins.86

Conformational states
As the anaesthetic binds to its receptor it may either occlude an ion channel,100 compete for a ligand or interact with the receptors in an allosteric fashion.22 Anaesthetics binding or interacting with proteins may either induce new protein conformations or stabilize existing conformations. In the former case, new states would have to be incorporated into the state diagram of the protein, whereas in the latter the rate constants of transitions between the various states would be altered.22 The state diagrams become even more complex if, as is the case for local anaesthetic interactions with sodium channels, the different protein states differ in their binding affinities and rate constants. Hille55 called this the modulated receptor hypothesis, which is formally the same as the model of conformation-dependent binding affinities of allosteric enzymes proposed by Monod and colleagues.77 In general, there are many more states and transition rate constants than can be determined from the available data, and so the state diagrams are usually simplified.22 Thus, a state diagram may be consistent with the data it describes; whether it actually represents physical reality will depend on independent measurements of each individual rate constant separately, which are usually not available currently.

However, the main concept behind such transition schemes is the idea that the preferential stabilization of protein conformations may represent a fundamental mechanism of inhaled anaesthetic action.74 107 In the case of ion channels, for example, the binding of small molecules could affect receptor function by changing the equilibrium between the conformations of resting channels and those in the open/desensitized state. If the ligand were to bind to most or all conformations of the receptor (the shapes or volumes of the binding sites differing between conformations), a given ligand could fit into a site in one conformation better than another and provide more stability for that conformation.107

Ion channels
Figure 2 shows correlations of IC50 (concentration of half-minimal inhibition) values for anaesthetic effects on ion channels with the octanol–water partition coefficients of the anaesthetics. In this figure, different anaesthetic functional endpoints and different ion channel subtypes are grouped together in order to show the overall trend. The interactions of anaesthetics with voltage-gated sodium, potassium and calcium channels and with ligand-gated excitatory nACh receptor and 5-HT3 receptor and inhibitory GABAA receptor channels follow Meyer–Overton correlations with relatively similar slopes that are all significantly different from zero. Quite clearly, all three types of ligand-gated ion channels respond more potently to anaesthetics than do the voltage-gated channels. Nevertheless, even the voltage-gated ion channels do respond at clinical concentrations of anaesthetics.



View larger version (28K):
[in this window]
[in a new window]
 
Fig 2 Meyer–Overton correlations for different ion channels, including sodium channels,10 13 15 16 36 38 41 51 70 87–90 104 113 potassium channels,13 15 25 39 40 42 43 46 52 70 calcium channels,18 19 54 105 106 nACh receptor channels,20 23 26 40 44 116–118 5-HT3 receptor channels10 58 69 124 and GABAA receptor channels.9 57 62 85 125 The number in parentheses after each channel name indicates the number of different ion channel subtypes contributing. Only anaesthetics in clinical use are included.

 
Comparing the recent edition of the standard reference book on ion channels55 with previous editions and looking at publications from previous International Conferences on Molecular Mechanisms of Anesthesia31 32 93 95 97 shows the great increase in the number of newly discovered ion channel types and subtypes, and new information on how their functions are regulated, modulated or affected by anaesthetics. Who is to say that we have reached the end of the discovery of new subtypes of ion channels? The number of existing and expressed ion channel subtypes must be in the hundreds, and the number of possible combinations that could be formed from the subunits of the various types of ion channels that are currently known runs into many thousands. Clearly, it is not possible to screen every existing subtype of ion channel for its response not to just one but a range of anaesthetics.

Before it was known how many different ion channels and membrane proteins existed, there was the hope that the great variety of anaesthetic responses could result from the combination of a few types of anaesthetic interactions with a few different molecular targets—membrane proteins and ion channels among them. It was known that a vast range of physiological changes and responses could be observed in an organism under anaesthesia and that various stages of anaesthesia were distinguishable clinically. Is there an analogy with chemistry? There are almost countless chemical reactions and chemical compounds, yet the periodic table of the chemical elements reveals an underlying structure that is quite simple: a limited number of elements in the periodic table can interact with each other through a limited number of bonds (covalent bonds, ionic bond, hydrogen bond, etc.), the reactions taking place in different solvents. Yet, the result is an enormous variety of chemical reactions. By analogy with the chemical elements, in anaesthesia proteins and anaesthetics could constitute the basic building blocks, membranes correspond to solvents, and anaesthetic interactions take the place of chemical reactions. If one were able to identify the participating proteins and to understand the nature of the key anaesthetic interactions, it might be possible, as in chemistry, to explain the bewildering variety of anaesthetic responses with relatively few rules.

In the previous section we saw that some of the key anaesthetic interactions are being revealed (though so far in small numbers, an example being hydrophobic pockets), but there is an indication that there are many more molecular players than there are atoms in the periodic table of the elements. However, we must be aware that each box in the periodic table of anaesthesia (Fig. 3), representing a particular type of protein, actually contains many different subtypes of the protein and that the subtypes can be very different in their anaesthetic responses. This is in contrast to atomic isotopes, which are quite similar in their physical and chemical properties. In contrast to chemistry, where there appears to be only one type of bond between two atoms at any one time, there are different anaesthetic actions on the same molecule simultaneously. This has been shown for every ion channel that has been studied in detail. It would appear that in anaesthesia the underlying molecular structure is not as simple as that of the periodic table of the elements in chemistry.



View larger version (31K):
[in this window]
[in a new window]
 
Fig 3 Periodic table of anaesthesia (table of molecular elements in anaesthesia). The table contains the molecular components involved in anaesthetic action, including the inhalation anaesthetics (Gas), the intravenous anaesthetics (I.V.), endogenous compounds (ENDO), noble gases (Inert), receptors as defined by the International Union of Pharmacology (IUPHAR) and ion channels as originally defined by the Ion Channel Network, including extracellular ligand-gated (ELG), intracellular ligand-gated (ILG), inward rectifying potassium (INR), junctional (JUN), voltage-gated (VLG), and miscellaneous (MIS) ion channels. Gas: (from left to right) nitrous oxide, diethyl ether, chloroform, halothane, enflurane, isoflurane, desflurane, sevoflurane, cyclopropane, divinyl ether, methoxyflurane, fluroxene, ethyl-chloride, trichloroethylene, alcohol. I.V.: (from left to right) thiopental, amobarbital, methohexital, propofol, etomidate, ketamine, midazolam, flunitrazepam, droperidol, morphine, fentanyl, remifentanil, cocaine, lidocaine, bupivacaine. ENDO: (Top to bottom) nitric oxide, carbon monoxide, carbon dioxide, endorphin, enkephalin. Ion channels (from the top of each column): ELG: 5-HT3, ATP-gated (P2X), AMPA, and kainate, NMDA glutamate receptor, nicotinic ACh receptor, GABAA receptor, glycine receptor. ILG: ryanodine, InsP3-sensitive Ca2+-release receptor, cAMP-activated cation channel, cGMP-activated cation channel, CFTR channel, Ca2+-activated K+ channel. INR: ATP-inhibited K+ channel, G/ACh muscarinic-activated K+channel Kirinwardly rectifying K+ channel, Ifhq native hyperpolarization-activated cation channel. JUN: connexins. VLG: Ca2+ channel, C1-- channel, Ke (Keag, Kelk, Kerg) ether a-go-go K+ channel, Kv delayed rectifier K+ channel, Na+ channel. MIS: mechanosensitive channel; mitochondrial membrane channel, nuclear membrane channel; aquaporins; synaptophysin channel. IUPHAR receptors: (left to right, first row) muscarinic ACh receptor, adenosine receptor, adrenoceptors, angiotensin receptor, bradykinin receptor, cannabionoid receptor, chemokine receptor, cholecystokinin receptor, corticotropin-releasing factor receptor, dopamine receptor; (left to right, second row) endothelin receptor, excitatory amino acid receptor, histamine receptor, serotonin receptor, melanocortin receptor, melatonin receptor, neuropeptide Y receptor, nucleotide receptor (P2X receptor, P2Y receptor), opioid receptor, prostanoid receptor; (left to right, third row) protease-activated receptor, somatostatin, vasoactive intestinal peptide and pituitary adenylate cyclase-activating polypeptide receptor, vasopressin and oxytocin receptor.

 
Signal integration
It has been shown previously how different anaesthetic effects on sodium or potassium channels are integrated to account for the overall suppression of sodium currents and potassium currents by anaesthetics.108 This integration of anaesthetic action also occurs at the subcellular, cellular and network levels. Rebecchi and Pentyala86 give the example of G-protein heterotrimers, which are made up of {alpha}, ß and {gamma} subunits, each of which occurs as several subtypes. As almost every {alpha} subtype can combine with almost any ß or {gamma} subtype, with the currently known number of subtypes there are more than 1600 possible combinations of heterotrimers. When it is considered that there are more than 2000 different receptor genes encoded in mammals and that some receptors can interact with more than one G-protein subtype, the number of potential combinations becomes truly enormous. Further layers of combinatorial control and complexity are added as these receptors form homo- and heterodimers, suggesting mechanisms for modulating signal output, coincidence detection and signal integration. As {alpha} subunits are subject to phosphorylation, they can give rise to functionally distinct phosphorylated states of each G-protein. Which combinations should be studied? Which combinations are relevant to mechanisms of anaesthesia? The higher we go up the ladder of integration within the CNS the more combinations become possible. We cannot study them all. We may wish to argue that perhaps there are only a very few key molecular targets that matter. Choosing to study only a few molecular systems in detail does not imply that the others do not matter. Perhaps the time has come not only to ascend this ladder but also to descend—to start from the clinical functional endpoints of anaesthesia, identify the neuronal networks involved and then identify the combination of molecular players that are relevant to these circuits.

Subcellular, cellular and higher levels
One reason why there has been much more research at the molecular level (e.g. on ion channels) is the observation that at the next higher level of complexity the results are often contradictory. Figure 4 illustrates this point: whether a depressant molecular anaesthetic action leads to inhibition or excitation at the higher levels of integration within the CNS or whether it has no consequence at all depends entirely on how these inputs are connected. Figure 4 makes the further point that a mechanism may be relevant to anaesthesia, although it is not affected by anaesthetics directly (in the words of Nash,79 the mechanism is ‘mediated’) but indirectly through modulatory signals originating from systems that are affected by anaesthetics. The neuronal networks do matter; sensitivity to anaesthetic action depends on what functional endpoints are measured and what feedback there is in the system.86 For voltage-clamped ion channels, for example, there cannot be any interaction with other ion channels through membrane depolarization. At this level of investigation, at which the direct effects on the ion channel are measured, concepts such as receptor reserve and feedback have little meaning. Rebecchi and Pentyala 86 illustrate the complexity of signal integration with an example of a second-messenger cascade involving G-protein-coupled signals that rely on many components operating in series. He points out that in tracing the cascade of events towards the response elements, the further away one gets from agonist binding the more likely some step has already been saturated with respect to the preceding signal. Significant receptor reserve must be depleted before an anaesthetic will have a measurable effect. Changes in signal output caused by anaesthetics are likely to be suppressed in systems where feedback is strong.

It is unlikely that all the targets that modulate intrinsic excitability have been identified,61 and so the task of moving up from the lower levels to higher levels of integration within the CNS appears daunting. The possible combinations are almost endless. We may reduce their number by additionally looking at the problem from the other end—by identifying neuronal networks and circuits that play roles in clinical endpoints of anaesthesia, i.e. cognitive functions and reactions to noxious stimuli.

Theories of general anaesthesia

Categories of hypotheses
Millar72 had already pointed out in 1975 that the question ‘How does anaesthesia occur?’ is unanswerable in the absence of a definition of consciousness. Perhaps we can paraphrase this by saying that the question ‘How does general anaesthesia occur?’ is unanswerable in the absence of a definition of general anaesthesia or its components (see review by Urban and Bleckwenn in this issue of BJA).111

Although there are many ways of defining general anaesthesia,111 for any one definition there are four possible models explaining general anaesthesia or one of its components:92

There is only one neuronal lesion which is caused by the same molecular mechanism;

There are many neuronal lesions caused by the same molecular mechanism;

The same neuronal lesion is caused by many molecular mechanisms;

There are many neuronal lesions caused by many molecular mechanisms.

The first two models belong to the class of what are called unitary hypotheses; the second two belong to the class of multisite hypotheses.

Unitary hypotheses
Overton assumed that the mechanisms of clinical anaesthesia and the mechanisms of anaesthetic actions on cells were basically the same for non-specific narcotics (for definition see the review by Urban and Bleckwenn in this issue of BJA).111 In this sense, Overton was a proponent of the unitary hypothesis when he stated (page 69):66 ‘Furthermore, it is highly probable that the mechanism of ether or chloroform narcosis, for example, remains substantially the same in the ganglia cells, the ciliary cells, and in the plant cells as well’. Overton did not assume that anaesthetics cause only one lesion. With the additional observation that no two anaesthetics acted alike, he stated (page 178):66 ‘We know that two non-specific narcotics hardly ever have exactly the same effect, since different nerve constituents undergo narcosis in a somewhat varying order’. Thus, whereas the mechanism should be the same, the effects (or lesions) may be different. Similarly, when Halsey discussed the unitary hypothesis for inhalation anaesthetics (equivalent to Overton’s non-specific narcotics) three-quarters of a century later,28 he stated that a ‘unitary hypothesis’ of anaesthesia did not require a single gross site of action, as there was evidence for several gross sites. For him, the unitary hypothesis required identity of action at the molecular level. Thus, the first possible model was never seriously advocated.

Another example of a unitary hypothesis is the GABA hypothesis, which will be discussed later. This states that enhancement of activity at GABAA receptors is an important component, perhaps the only component, of relevant mechanisms in anaesthesia. Results obtained from genetic studies are now being used as evidence that the unitary hypothesis can be dismissed in the nematode.79 Even for this second kind of unitary hypothesis it should be true that all anaesthetics producing a particular kind of lesion should show the same correlation when the effect is correlated with the underlying single mechanism;111 there should be no exceptions.

Multisite hypotheses
Multisite hypotheses, which allow that many molecular mechanisms may cause one or many neuronal lesions, do not hold that all anaesthetics should show the same correlation between the anaesthetic endpoint and the mechanism-related endpoint. Thus, if certain inhalation anaesthetics do not have much effect on GABAA receptors, this does not imply that GABAA receptors are not important molecular players in the clinical components of anaesthesia. Another example is provided in Figure 5, which shows a good correlation of IC50 values for Kv3.1 potassium channel block with clinical concentrations of anaesthetics. The fact that the correlation does not hold for the opioids does not negate the possibility that these potassium channels themselves play a role or are a model for a relevant target site in anaesthesia, but it does indicate that opioids work through a different mechanism (opioid receptors). Studies of proteins,27 74 108 second-messenger signalling,86 the spinal cord,61 brain slices,4 genetics79 and functional imaging53 all come up with data consistent with multisite theories of anaesthetic action.



View larger version (12K):
[in this window]
[in a new window]
 
Fig 5 Correlation of IC50 for Kv3.1 potassium channels40 43 with clinical concentration.

 
Molecular and cellular theories
Molecular or cellular theories of general anaesthesia will be defined as those theories that are based on anaesthetic mechanisms at the molecular or cellular level that has been proposed to be responsible for generating the state(s) of general anaesthesia. At the first International Conference on Molecular Mechanisms of Anaesthesia, Woodbury and colleagues proposed a molecular mechanism of general anaesthesia, with the justification that ‘even in such a complicated system as the human body, the course of a single reaction can be observed if it is the rate-determining step of a measurable phenomenon’.122

One of the early cellular theories of anaesthesia was that of Meyer and Overton, who made the assumption ‘that narcosis resulted from the modification (caused precisely by the absorption of foreign compounds) of the normal physical state of the lecithin and cholesterol-related compounds in the cell’ (page 70).66 They considered it possible that the cellular effect depended either on the number of molecules absorbed or on their volume. Subsequently, Ferguson30 thought it was the thermodynamic activity that mattered, while Mullins78 later made the case that it should be molecular volume after all. However, none of them stated how a change in the cellular properties of neurones would translate into clinically observable effects (Fig. 6). There have been many other molecular and cellular theories of anaesthesia since then, some of which gave rise to controversies such as whether anaesthetics interfere with membrane protein function by binding directly to proteins or whether the main mode of action is indirect, by changing the physicochemical properties of the lipid membrane, in which anaesthetics readily dissolve.67 73 92 102



View larger version (15K):
[in this window]
[in a new window]
 
Fig 6 How do molecular anaesthetic actions modify brain functions? The networks that are responsible for translating molecular effects into clinically observable effects are still unknown.

 
Discussions of molecular mechanisms of anaesthesia have been dominated by the view that enhancement of activity at GABAA receptors is an important component, perhaps the only component, of the mechanisms that are relevant in anaesthesia. There has been much research on GABAergic neurotransmission, as becomes obvious when reading the reviews in this issue of BJA. However, it is equally clear that some clinical anaesthetics, such as xenon, nitrous oxide and cyclopropane, have little effect on GABAA receptors.22 Genetic studies79 and evidence from brain imaging53 do not support an exclusive role for GABAA receptors in anaesthesia, although they support the hypothesis that the in vivo effects of anaesthetics are mediated at least in part through GABAergic mechanisms. Other receptors and ion channels are also involved, such as glutamate receptor channels and cholinergic neurotransmitter systems.5 Genetic studies on Drosophila even point to voltage-gated sodium channels79 as factors that may affect anaesthetic sensitivity.

No effective antagonist for inhalation anaesthesia has been reported so far. This does not mean that anaesthesia cannot be brought about by receptor action. Opiates can be antagonized by naloxone, benzodiazepines by flumazenil and non-depolarizing muscle relaxants by neostigmine or edrophonium. Therefore, in theory it should be possible to antagonize i.v. anaesthesia induced by a combination of opiate, benzodiazepine and muscle relaxant by giving a combination of antagonists for each substance. This form of anaesthesia would be brought about by receptor action only, in this case the simultaneous action of several receptors. However, experience has shown that routine administration of naloxone after a predominantly narcotic-based anaesthetic procedure often results in hypertension, tachycardia and acute and severe pain.8 This failure may result from the narcotic acting on several opiate receptors with differing pharmocodynamics, while the antagonists for the different receptors possess pharmacokinetic and pharmacodynamic properties different from those of the agonist. Consequently, although an antagonist for inhalation anaesthetics has not been found yet, this does not imply that these anaesthetics do not act at specific receptors. Rather, inhalation anaes thetics may simply act at several receptors that differ in their pharmacodynamic properties and require several different antagonists that have still not been found.

In summary, molecular and cellular theories of anaesthesia will remain incomplete and controversial unless these molecular and cellular effects in vitro are associated with areas and neuronal networks within the CNS. It must be spelt out not only how they produce in vivo effects at the cognitive/behavioural level, such as hypnosis, amnesia, analgesia/antinociception and suppression of movement response to noxious stimulation, but also why other anaesthetic actions have no clinical effect.

Component theories
The statement by Millar72 that general anaesthesia cannot be explained until there is a definition, or perhaps an understanding, of consciousness applies equally to other components of anaesthesia, such as immobility, amnesia, analgesia and the suppression of stress responses to noxious stimuli. Here, as in the fields of consciousness, awareness and sleep, neuroscience lacks basic knowledge and is not yet capable of providing a satisfactory description of molecular and higher mechanisms that could explain, for example, the various stages of sleep, the different forms of memory, or of pain sensations and reactions. Consciousness, sleep, pain and memory represent brain states and cognitive functions that are relevant to understanding anaesthesia, yet their detailed molecular and network description is still an enormous challenge to neuroscience. On the other hand, the appropriate response to this situation need not be to wait until such explanations are provided—the field of anaesthesia could be instrumental in helping neuroscientists to answer these questions.

In order to simplify the problem, a sensible approach would be to look at the individual components of general anaesthesia. This approach has indeed been taken for the issues of immobility, antinociception, amnesia and awareness.5 53 61 114 However, there is a caveat: these components of anaesthesia are not independent of each other, but interact. Pain will increase vigilance and awareness. Noxious stimuli during otherwise adequate anaesthesia will result in a shift to increased arousal, as indicated by increased neuronal activity in the reticular formation and thalamus, and EEG desynchronization.5 There is interaction between the spinal cord (thought to be mainly responsible for immobility in anaesthesia) and the brain cortex (thought to be mainly responsible for awareness and hypnosis).5 53 61 114

Immobility
The spinal cord plays a leading role in the clinical anaesthetic endpoints of immobility in response to a noxious stimulus and has been revealed as the dominant CNS locus for determining the MAC (minimal alveolar concentration) for volatile anaesthetic agents, although supraspinal modulatory influences also play a role.5 61 114

There is good evidence for multiple targets in the spinal cord. Volatile anaesthetics, barbiturates, nitrous oxide and propofol depress reflex activity (nocifensive movements) by both suppression of the excitability of spinal motor neurones and by suppression of responses of spinal nociceptive neurones.114 Direct actions of anaesthetic agents, in particular volatile agents, on motor neurones are thought to contribute greatly to immobility, with multiple targets for anaesthetic actions on motor neurones themselves, on elements presynaptic to them and on a number of different ion channels, which may vary from agent to agent.61 While the actions of anaesthetic agents on GABAA and glycine receptors are important, they are not, particularly for volatile agents, the sole or even the major contributors to anaesthetic depression of motor neurone excitability. Actions on these receptors may be more important for some i.v. agents than for volatile agents, for which actions on glutamate receptors on motor neurones are probably important.60 Much is still unknown about sites for volatile agents on glutamate and other receptors61 and about the way that anaesthetics affect the neuronal networks that generate complex moving patterns (escape instead of simple withdrawal responses).5

Analgesia and antinociception
Although analgesia and pain sensations have an emotional component relating to the evaluation of nociceptive stimuli, the effects of anaesthetics on this aspect have not received much attention; antinociception has been studied instead. Like immobility, antinociception by anaesthetics is processed to a large extent at the spinal level, but it is also subject to supraspinal modulation. The thalamus and anterior cingulate cortex appear to be targets for opioid and nitrous oxide analgesia, and it appears that analgesics such as opioids and nitrous oxide modulate pain perception by affecting different neural circuits.53

At the point where anaesthetics may abolish immobility, there may still remain haemodynamic responses to noxious stimuli in the presence of certain anaesthetics, such as isoflurane and barbiturates. The answer to the question of why some anaesthetics suppress haemodynamic responses to noxious stimulation while others do not may be related to the observation that certain anaesthetics (isoflurane, halothane, nitrous oxide and barbiturates) at low concentration appear to have hyperalgesic effects.5 Clearly, a hypothesis of action for this component is still a long way off.

Amnesia and memory
Whereas the suppression of motor and autonomic responses to noxious stimuli and the block of sensory information transfer concerning stimuli affecting the body surface (such as sound, touch and nociceptive agents) may well be largely controlled at subcortical sites, the loss of consciousness (hypnosis) and the block of memory formation (amnesia) requires interactions of anaesthetics with cortical activity. Memory formation involves a variety of sites in the brain, among them the hippocampus, amygdala, prefrontal cortex and other cortical sensory and motor areas. Of the two broad categories of memory, explicit memory appears to be more sensitive to anaesthetics than implicit memory, but the specific sites where this action occurs have not been distinguished yet.5 Anaesthetic effects on memory are even more complex, as learning and memory formation during anaesthesia depend on the circumstances under which stimuli are presented. Fear conditioning to a tone was less sensitive to isoflurane than fear conditioning to context (e.g. the surrounding environment).5 Here also, as for the other functional components discussed above, there are differences among anaesthetics: at equipotent MAC concentrations, isoflurane appears to be more potent than nitrous oxide as regards the production of unconsciousness and the reduction of explicit memory formation.5 No comprehensive theory of amnesia or memory is yet in sight.

Hypnosis and consciousness
The physiological processes and anatomical sites that give rise to consciousness, awareness, arousal and attention are still poorly understood. Several sites have been proposed to participate in consciousness and awareness, including the cerebral cortex, thalamus and reticular formation.6 53 All of these structures are affected by anaesthetics directly, but there is also evidence to suggest that ascending traffic from the spinal cord to the brain may modulate the effects of anaesthetic agents on arousal.61 Functional imaging studies confirm earlier studies showing that higher anaesthetic concentrations are required to suppress subcortical rather than cortical structures and that the thalamus and midbrain reticular formation may be key targets for drug-induced loss of consciousness.53 They also support a role for anaesthetics in interacting antagonistically with muscarinic signalling in the CNS, which is known to be involved in the modulation of consciousness and tends to enhance wakefulness. There is evidence indicating that inhalation anaesthetics may interact with specific functional systems within the cortex, such as systems that synchronize brain activities4 and the specific neuronal networks responsible for task-induced brain activation, rather than causing a global decrease in functional activation.53

Relevant targets and mechanisms
As Nash79 points out, there might be substantial differences between organisms in the degree to which target function must be altered so as to produce the assayed endpoint, i.e. different safety factors may be involved. In addition, there might be substantial differences in the potency with which anaesthetics disrupt proteins or other target functions in different organisms. Thus, knowledge of the neuronal networks involved in any critical functional endpoint or component of general anaesthesia is essential before it can be decided whether or not in vitro targets or mechanisms are relevant and contribute significantly to general anaesthesia. These neuronal networks must no longer be treated as black boxes. The black boxes must be opened and the components of the networks identified and characterized. Furthermore, many researchers now accept that it is reasonable to consider anaesthesia as the sum of several contributions. As the neuronal networks underlying many of the components of general anaesthesia discussed in the previous sections have now begun to be characterized,5 53 61 79 114 it does not seem very helpful either to search for one ion channel, membrane site or enzyme that is most sensitive107 to anaesthetics or to label anaesthetic targets and mechanisms as relevant or irrelevant to general anaesthesia.

Integrated theories
Integrated theories of general anaesthesia are theories that do not treat the several components of general anaesthesia independently but provide an integrating basis for their explanation.

The sleep hypothesis
The sleep hypothesis that was considered by Overton and his predecessors (page 57)66 is an example of an integrated theory. In sleep, many physiological functions are affected simultaneously and differ from their waking behaviour. There are phases of sleep in which arousal is more difficult than during other phases, when motor programmes appear to be executed in the brain cortex while the limbs do not move, and when there can be insensibility to a range of external sensory inputs.59

The sleep hypothesis is revived periodically. That certain brain states of anaesthesia and natural sleep are comparable has been suggested again recently, subsequent to findings that general anaesthetics induce delta activity in the EEG.4 During non-REM sleep, synchronized cortical delta activity is present that appears to originate in the thalamus where, during delta sleep, cortical neurones seem to be driven and synchronized by thalamic relay neurones. On the basis of the EEG classification of the stages of sleep, an EEG classification of anaesthesia has been proposed84 and has even been developed into a means of monitoring anaesthesia.63

The preprogrammed response hypothesis
Pursuing the idea of a sleep hypothesis, Overton observed that ‘The phenomena of artificial narcosis have such similarities to those of natural sleep that one is forced quite involuntarily to ask the question whether or not natural sleep is caused by a narcotic-acting substance produced by the organism itself. In weighing such a hypothesis, one’s thoughts turn all the more to carbon dioxide, since it would be absolutely necessary for the hypothetical compound to be able to pass out of the ganglia cells’ (page 148).66 Besides carbon dioxide, there are other endogenous substances with anaesthetic properties, including ammonia, carbon monoxide, nitric oxide, opioid peptides and certain steroid hormones.

It is fascinating that a single substance, such as diethyl ether, can switch off bodily functions in such a coordinated way that physiological functions essential to life are lost last (at high ether concentrations). A theory explaining this phenomenon, i.e. a theory of ‘etherization’, may be a good starting point for developing a theory of general anaesthesia.

Depending on the drug concentration, diethyl ether anaesthesia is characterized by the loss of pain sensation, memory, consciousness, sensory stimuli, eyelid and corneal reflexes, muscle tone, somatosensory and autonomic reflexes, the swallowing reflex, respiration and cardiac function. It seems truly amazing that, despite such a wide range of actions, the inhalation of diethyl ether does not cause the physiological functions of the patient to be thrown into chaos. Malignant hyperthermia,47 which results from acute uncontrolled increases in skeletal muscle metabolism, is an example of chaos created by a general anaesthetic. Instead, diethyl ether seems to cause a very orderly switching-off procedure, non-essential physiological functions being abolished long before vital functions are lost. This may suggest that anaesthetics trigger already existing mechanisms that have been preprogrammed by evolution to respond to anaesthetic-like substances.

The LIPOID hypothesis
The observation that general anaesthetic potency correlates with the lipophilicity of general anaesthetic drugs (Figs 2 and 7) may even suggest that they trigger various, still unexplored mechanisms that exert control over the concentration of lipophilic substances.109 Lipophilic substances can readily cross membranes and interrupt not only any membrane-based function (including that of membrane proteins) but also other cellular proteins. There are many endogenous lipophilic substances, for example ammonia, carbon monoxide, carbon dioxide, nitric oxide, hydrogen sulphide, alkanes (methane etc.), acetone, urea, steroids, enkephalins and endorphins, cannabinoids (anandamide etc.) and other arachidonic acid derivatives. Fatty acid amides are putative endogenous ligands for anaesthetic recognition sites in the mammalian CNS.65 Inhalation anaesthetics are a good example of how dangerous such compounds can become at high concentrations, because they have a very narrow therapeutic index (typically about 3) and become lethal very quickly. Thus, control of the concentration of lipophilic substances may be important to any organism and integrative mechanisms may have developed at all levels of integration within the CNS. Examples of systemic responses may be elimination responses, such as increases in respiration and perfusion, vasodilation, coughing, sweating, tearing and emesis. Molecular and cellular responses may consist of the increased metabolism of drugs (enzyme induction) or reduced sensitivity of critical proteins to lipophilic agents. One might even speculate that particularly critical and ubiquitous proteins, such as axonal sodium channels, may have evolved to become quite insensitive to the actions of lipophilic substances.



View larger version (28K):
[in this window]
[in a new window]
 
Fig 7 Meyer–Overton correlations of anaesthetic actions with octanol–water partition coefficients are observed at all six levels of CNS integration. (A) Molecular: inhibition of peak currents from sodium channels.87 88 90 (B) Subcellular: depression of compound action potentials from frog sciatic nerve.102 (C) Cellular: depression of firing rates of sensory neurones (muscle receptor organ) from crayfish.94 (D) Depression of spontaneous firing in rat neocortical brain slices.3 (E) Block of somatosensory potentials in rats.2 (F) Clinical concentrations in general anaesthetic procedures.75

 
The LIPOID (lipophilic intrusion protection organization—integrative design) hypothesis of anaesthesia109 postulates that general anaesthesia is brought about by lipophilic anaesthetics triggering some or many, depending on the complexity of the anaesthetic molecule, of the different mechanisms controlling the concentration of lipophilic substances within an organism. The main purpose of presenting this highly speculative hypothesis is to illustrate that classes of theories of general anaesthesia are imaginable that are compatible with there being many simultaneous actions of anaesthetics rather than a select few.

Conclusion

So far, the search for unitary or simple mechanisms of anaesthesia has failed. Considering the many different anaesthetic effects that have been discovered in vitro and in vivo, there are two ways of responding. One is that the search for simple mechanisms should be continued in order to obtain proof that only a few anaesthetic sites and actions are really relevant and that the others do not matter. Alternatively, integrated explanations should be sought that reconcile many simultaneous anaesthetic targets and actions with a still functional organism.

The fact that we still do not have generally accepted hypotheses for the mechanisms of general anaesthesia is not a consequence of a lack of attempts to tackle this problem, but seems rather a reflection of its complexity. A change of paradigm may be called for. More attempts have to be made to open the black boxes, and more research effort must be directed to the identification and investigation of the neuronal networks relevant to particular components of anaesthesia. Until such networks can be identified and in vitro mechanisms tested in these networks in vivo, it seems futile to speculate on the relevance of in vitro mechanisms for general anaesthesia.

In the preceding reviews in this issue of BJA, several modes of anaesthetic action have been described that differ not only between i.v. and inhalation anaesthetics, but also between different modern inhalation anaesthetics. Therefore, depending on the techniques and anaesthetics used, anaesthetic procedures may also differ significantly in how (well) they achieve different clinical endpoints. There is a need for well-controlled studies of clinical outcome, using different general anaesthesia procedures, in which theories of general anaesthesia can be tested.

Acknowledgements

The author wishes to thank Zita Dorner and Markus Bleckwenn for help in preparing the figures.

References

1 Akabas MH, Horenstein J, Williams DB, Bali M, Bera AK. GABA- and drug-induced conformational changes detected in the GABAA receptor channel-lining segments. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

2 Angel A, Gratton DA. The effect of anaesthetic agents on cerebral cortical responses in the rat. Br J Pharmacol 1982; 76: 541–9[Abstract]

3 Antkowiak B. Different actions of general anesthetics on the firing patterns of neocortical neurons mediated by the GABAA receptor. Anesthesiology 1999; 91: 500–11[ISI][Medline]

4 Antkowiak B. In vitro networks: cortical mechanisms. Br J Anaesth 2002; 89: 102–11[Abstract/Free Full Text]

5 Antognini JF, Carstens E. In vivo characterization of clinical anesthesia and its components. Br J Anaesth 2002; 89: 156–66[Abstract/Free Full Text]

6 Antognini JF, Schwartz K. Exaggerated anesthetic requirements in the preferentially anesthetized brain. Anesthesiology 1993; 79: 1244–9[ISI][Medline]

7 Asahi T, Hirota K, Sasaki R, Yamazaki M. Synaptic and non-synaptic pathways in rat hippocampal slices in vitro. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

8 Bailey PL, Egan TD, Stanley TH. Intravenous opioid anesthetics. In: Miller RD, Cuchiara RF, Miller ED, Reves JG, Roizen MF, Savarese JJ, eds. Anesthesia. Philadelphia: Churchill Livingstone, 2000; 273–376

9 Banks MI, Pearce RA. Dual actions of volatile anesthetics on GABAA IPSCs: dissociation of blocking and prolonging effects. Anesthesiology 1999; 90: 120–34[ISI][Medline]

10 Barann M, Göthert M, Fink K, Bönisch H. Inhibition by anaesthetics of 14C-guanidinium flux through the voltage-gated sodium channel and the cation channel of the 5-HT3 receptor of N1E-115 neuroblastoma cells. Naunyn Schmiedebergs Arch Pharmacol 1993; 347: 125–32[ISI][Medline]

11 Baudoux S, Empson RM, Richards CD. General anesthetics do not prevent action potential invasion of axon branches or synaptic boutons. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

12 Benkwitz C, Roewer N, Antkowiak B. Anesthetic action on neuronal network activity in hippocampus: electrophysiology and Ca-imaging. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

13 Benoit E. Effects of intravenous anaesthetics on nerve axons. Eur J Anaesthesiol 1995; 12: 59–70[ISI][Medline]

14 Bosnjak ZY. Anesthetic-induced cardiac preconditioning. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

15 Bräu ME, Sander F, Vogel W, Hempelmann G. Blocking mechanisms of ketamine and its enantiomers in enzymatically demyelinated peripheral nerve as revealed by single-channel experiments. Anesthesiology 1997; 86: 394–404[ISI][Medline]

16 Bräu ME, Vogel W, Hempelmann G. Fundamental properties of local anesthetics: half-maximal blocking concentrations for tonic block of Na+ and K+ channels in peripheral nerve. Anesth Analg 1998; 87: 885–9[Abstract]

17 Chan K, Meng QC, Johansson JS, Eckenhoff RG. Low-affinity analytical chromatography for measuring inhaled anesthetic binding to isolated proteins. Anal Biochem 2002; 301: 308–13[ISI][Medline]

18 Charlesworth P, Pocock G, Richards CD. Calcium channel currents in bovine adrenal chromaffin cells and their modulation by anaesthetic agents. J Physiol (Lond) 1994; 481: 543–53[Abstract]

19 Croxton TL, Hirshman CA. Inhibitory effects of thiopental, ketamine, and propofol on voltage-dependent Ca2+ channels in porcine tracheal smooth muscle cells. Anesthesiology 1995; 83: 1274–82[ISI][Medline]

20 de Armendi AJ, Tonner PH, Bugge B, Miller KW. Barbiturate action is dependent on the conformational state of the acetylcholine receptor. Anesthesiology 1993; 79: 1033–41[ISI][Medline]

21 Dickinson R, de Sousa SLM, Lieb WR, Franks NP. Anaesthetic actions on hippocampal autapses. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

22 Dilger JP. The effects of general anesthetics on ligand-gated ion channels. Br J Anaesth 2002; 89: 41–52[Abstract/Free Full Text]

23 Dilger JP, Vidal AM, Mody HI, et al. Effects of general anesthetics on acetylcholine receptor channels. Prog Anesth Mech 1995; 3: 50–5

24 Dipalma FR. Drill’s Pharmacology in Medicine. New York: McGraw-Hill, 1954

25 Dreixler JC, Jenkins A, Cao YJ, Roizen JD, Houamed KM. Patch-clamp analysis of anesthetic interactions with recombinant SK2 subtype neuronal calcium-activated potassium channels. Anesth Analg 2000; 90: 727–32[Abstract/Free Full Text]

26 Dybek A. Vergleich der Anästhetikawirkungen an einem Acetylcholinrezeptor und einem Kaliumkanal der humanen Neuroblastomzelle SH-SY5Y mittels Patch-Clamp- und schneller Superfusionstechnik. PhD dissertation. Bonn, Germany: University of Bonn, 1999

27 Eckenhoff RG. Promiscuous ligands and attractive cavities. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

28 Eger EI. Anesthetic Uptake and Action. Baltimore: Williams and Wilkins, 1974

29 Elliott JR, Haydon DA. The actions of neutral anaesthetics on ion conductances of nerve membranes. Biochim Biophys Acta 1989; 988: 257–86[ISI][Medline]

30 Ferguson J. The use of chemical potentials as indices of toxicity. Proc R Soc Lond B Biol Sci 1939; 127: 387–404

31 Fink BR. Molecular Mechanisms of Anesthesia. New York: Raven Press, 1975

32 Fink BR. Molecular Mechanisms of Anesthesia 2. New York: Raven Press, 1980

33 Forman SA, Miller KW. Molecular sites of anesthetic action in postsynaptic nicotinic membranes. Trends Pharmacol Sci 1989; 10: 447–52[ISI][Medline]

34 Franks NP, Lieb WR. Do general anaesthetics act by competitive binding to specific receptors? Nature 1984; 310: 599–601[ISI][Medline]

35 Franks NP, Lieb WR. Which molecular targets are most relevant to general anaesthesia? Toxicol Lett 1998; 100–101: 1–8

36 Frenkel C, Duch DS, Recio-Pinto E, Urban BW. Pentobarbital suppresses human brain sodium channels. Brain Res Mol Brain Res 1989; 6: 211–6[ISI][Medline]

37 Frenkel C, Duch DS, Urban BW. Molecular actions of pentobarbital isomers on sodium channels from human brain cortex. Anesthesiology 1990; 72: 640–9[ISI][Medline]

38 Frenkel C, Urban BW. [Interactions of intravenous anesthetics with human CNS ion channels. Electrophysiologic studies with a new type of voltage clamp technique.] Anaesthesist 1994; 43: 229–34[ISI][Medline]

39 Friederich P. Untersuchungen zum Einfluss von Allgemeinanästhetika auf einen menschlichen neuronalen Kaliumstrom mit der Ganzzell Patch-Clamp Methode. MD dissertation. Bonn, Germany: University of Bonn, 1997

40 Friederich P, Benzenberg D, Trellakis S, Urban BW. Interaction of volatile anesthetics with human Kv channels in relation to clinical concentrations. Anesthesiology 2001; 95: 954–8[ISI][Medline]

41 Friederich P, Reckziegel G, Urban BW. Ist Ketamin ein Lokalanästhetikum? Wirkung auf humane Neuronale Na-Ströme. Anasthesiol Intensivmed 1999; 40: S111:272.1

42 Friederich P, Trellakis S, Urban BW. [Epileptic seizures from etomidate? The human Kv1.1 potassium channel in humans.] Anasthesiol Intensivmed Notfallmed Schmerzther 2001; 36: 100–4[ISI][Medline]

43 Friederich P, Urban BW. Interaction of intravenous anesthetics with human neuronal potassium currents in relation to clinical concentrations. Anesthesiology 1999; 91: 1853–60[ISI][Medline]

44 Furuya R, Oka K, Watanabe I, Kamiya Y, Itoh H, Andoh T. The effects of ketamine and propofol on neuronal nicotinic acetylcholine receptors and P2x purinoceptors in PC12 cells. Anesth Analg 1999; 88: 174–80[Abstract/Free Full Text]

45 Grasshoff C, Antkowiak B. Effects of volatile anesthetics on action potential firing of spinal neurons in organotypic slice cultures. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

46 Gray AT, Winegar BD, Leonoudakis DJ, Forsayeth JR, Yost CS. TOK1 is a volatile anesthetic stimulated K+ channel. Anesthesiology 1998; 88: 1076–84[ISI][Medline]

47 Gronert GA, Antognini JF, Pessah IN. Malignant hyperthermia. In: Miller RD, Cuchiara RF, Miller ED, Reves JG, Roizen MF, Savarese JJ, eds. Anesthesia. Philadelphia: Churchill Livingstone, 2000; 1033–52

48 Gruss M, Henrich M, König P, Hempelmann G, Vogel W, Scholz A. Ethanol modifies action potentials in a subtype of small sensory neurones by activation of BKCa-channels. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

49 Hardman JG, Gilman AG, Limbird LE. The Pharmacological Basis of Therapeutics. New York: McGraw-Hill, 1996

50 Harris RA, Mascia MP, Lobo IA. Sites of anesthetic action on a ligand-gated ion channel. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

51 Haydon DA, Elliott JR, Hendry BM, Urban BW. The action of nonionic anesthetics substances on voltage-gated ion conductances in squid giant axons. In: Roth SH, Miller KW, eds. Molecular and Cellular Mechanisms of Anesthetics. New York: Plenum Medical Book Company, 1986; 267–77

52 Haydon DA, Urban BW. The actions of some general anaesthetics on the potassium current of the squid giant axon. J Physiol (Lond) 1986; 373: 311–27[Abstract]

53 Heinke W, Schwarzbauer C. In vivo imaging of anaesthetic action in humans: Approaches with PET and fMRI. Br J Anaesth 2002; 89: 112–22[Abstract/Free Full Text]

54 Herrington J, Lingle CJ. Kinetic and pharmacological properties of low voltage-activated Ca2+ current in rat clonal (GH3) pituitary cells. J Neurophysiol 1992; 68: 213–32[Abstract/Free Full Text]

55 Hille B. Ion Channels of Excitable Membranes. Sunderland, MA: Sinauer Associates, 2001

56 Hirota K, Asahi T, Sasaki R, Yamazaki M, Roth SH. The agent- and pathway-specific actions of general anesthetics in hippocampus in vitro. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

57 Huidobro-Toro JP, Bleck V, Allan AM, Harris RA. Neurochemical actions of anesthetic drugs on the gamma-aminobutyric acid receptor–chloride channel complex. J Pharmacol Exp Ther 1987; 242: 963–9[Abstract]

58 Jenkins A, Franks NP, Lieb WR. Actions of general anaesthetics on 5-HT3 receptors in N1E-115 neuroblastoma cells. Br J Pharmacol 1996; 117: 1507–15[Abstract]

59 Kandel ER, Schwartz JH, Jessell TM. Principles of Neural Science. New York: Elsevier, 1991

60 Kendig JJ. Barbiturates: active form and site of action at node of Ranvier sodium channels. J Pharmacol Exp Ther 1981; 218: 175–81[Abstract]

61 Kendig JJ. In vitro networks: subcortical mechanisms. Br J Anaesth 2002; 89: 91–101[Abstract/Free Full Text]

62 Krasowski MD, Harrison NL. The actions of ether, alcohol and alkane general anaesthetics on GABAA and glycine receptors and the effects of TM2 and TM3 mutations. Br J Pharmacol 2000; 129: 731–43[Abstract/Free Full Text]

63 Kreuer S, Biedler A, Larsen R, Schoth S, Altmann S, Wilhelm W. The Narcotrend—a new EEG monitor designed to measure the depth of anaesthesia. A comparison with bispectral index monitoring during propofol–remifentanil anaesthesia. Anaesthesist 2001; 50: 921–5[ISI][Medline]

64 Kurata J, Thulborn KR, Gyulai FE, Firestone LL. Functional magnetic resonance imaging reveals temporal dynamics of cerebral pain processing. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

65 Laws D, Verdon B, Coyne L, Lees G. Fatty acid amides are putative endogenous ligands for anaesthetic recognition sites in mammalian CNS. Br J Anaesth 2001; 87: 380–4[Free Full Text]

66 Lipnick RL. Studies of narcosis and a contribution to general pharmacology. Charles Ernest Overton. New York: Chapman and Hall and the Wood Library-Museum of Anesthesiology, 1991

67 Little HJ. How has molecular pharmacology contributed to our understanding of the mechanism(s) of general anesthesia? Pharmacol Ther 1996; 69: 37–58[ISI][Medline]

68 Loop T, Humar M, Hoetzel A, et al. Thiopental inhibits the activation of nuclear factor-B. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

69 Machu TK, Harris RA. Alcohols and anesthetics enhance the function of 5-hydroxytryptamine3 receptors expressed in Xenopus laevis oocytes. J Pharmacol Exp Ther 1994; 271: 898–905[Abstract]

70 McGivern J, Scholfield CN. General anaesthetics and field currents in unclamped, unmyelinated axons of rat olfactory cortex. Br J Pharmacol 1990; 101: 217–23[Abstract]

71 Meyer H. Zur Theorie der Alkoholnarkose. Arch Exp Pathol Pharmakol 1899; 42: 109–37

72 Millar RA. Anaesthetic actions. Br J Anaesth 1975; 47: 335, 349[ISI][Medline]

73 Miller KW. The nature of the site of general anesthesia. Int Rev Neurobiol 1985; 27: 1–61[ISI][Medline]

74 Miller KW. The nature of sites of general anaesthetic action. Br J Anaesth 2002; 89: 17–31[Abstract/Free Full Text]

75 Miller RD. Anesthesia. Philadelphia: Churchill Livingstone, 2000

76 Mintun MA. In vivo molecular brain imaging: approaches with positron emission tomography. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

77 Monod J, Wyman J, Changeux JP. On the nature of allosteric transitions: a plausible model. J Mol Biol 1965; 12: 88–118[ISI][Medline]

78 Mullins LJ. Some physical mechanisms in narcosis. Chem Rev 1954; 54: 289–323[ISI]

79 Nash HA. In vivo genetics of anaesthetic action. Br J Anaesth 2002; 89: 143–56[Abstract/Free Full Text]

80 Nishimura S, Yokono S, Taenaka N, Mashimo T. General anesthesia suppresses nocturnal melatonin secretion. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

81 Orser BA, Joo DT, Zhu G, et al. Increased desensitization of AMPA receptors may cause hypersensitivity to anesthetics in GluR2 null mutant mice. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

82 Overton E. Studien über die Narkose. Jena: Fischer, 1901

83 Pentyala S, Halpern-Lewis J, et al. Intravenous anesthetics reduce airway resistance by their action on plasma membrane calcium pump. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

84 Pichlmayr I, Lips U, Kuenkel H. The Electroencephalogram in Anesthesia. Fundamentals, Practical Applications. Examples. New York: Springer Verlag, 1984

85 Pistis M, Belelli D, Peters JA, Lambert JJ. The interaction of general anaesthetics with recombinant GABAA and glycine receptors expressed in Xenopus laevis oocytes: a comparative study. Br J Pharmacol 1997; 122: 1707–19[Abstract]

86 Rebecchi MJ, Pentyala SN. Anesthetic actions on other targets: protein kinase C and guanine nucleotide-binding proteins. Br J Anaesth 2002; 89: 62–78[Abstract/Free Full Text]

87 Rehberg B, Bennett E, Xiao YH, Levinson SR, Duch DS. Voltage- and frequency-dependent pentobarbital suppression of brain and muscle sodium channels expressed in a mammalian cell line. Mol Pharmacol 1995; 48: 89–97[Abstract]

88 Rehberg B, Duch DS. Suppression of central nervous system sodium channels by propofol. Anesthesiology 1999; 91: 512–20[ISI][Medline]

89 Rehberg B, Wittmann M, Urban BW. Sodium channels (from rat, mouse, and man) in neuroblastoma cells and different expression systems have similar sensitivities to pentobarbital. Neurosci Lett 1999; 264: 81–4[ISI][Medline]

90 Rehberg B, Xiao YH, Duch DS. Central nervous system sodium channels are significantly suppressed at clinical concentrations of volatile anesthetics. Anesthesiology 1996; 84: 1223–33[ISI][Medline]

91 Richards CD. Anaesthetic modulation of synaptic transmission in the mammalian CNS. Br J Anaesth 2002; 89: 79–90

92 Richards CD. The mechanisms of general anaesthesia. In: Norman J, Whitwam JG, eds. Topical Reviews in Anaesthesia. Bristol: Wright, 1980; 1–84

93 Richards CD, Winlow W. Molecular and Cellular Mechanisms of General Anesthesia. New York: Elsevier, 1998

94 Roth SH. Differential effects of anesthetics on neuronal activity. In: Fink BR, ed. Molecular Mechanisms of Anesthesia 2. New York: Raven Press, 1980; 119–28

95 Roth SH, Miller KW. Molecular and Cellular Mechanisms of Anesthetics. New York: Plenum Medical Books Company, 1986

96 Roth SH, Samanani N, Archer DP. Receptor-mediated excitatory effects of general anesthetics. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

97 Rubin E, Miller KW, Roth SH. Molecular and cellular mechanisms of alcohol and anesthetics. Ann N Y Acad Sci 1991; 625: 1–848[ISI]

98 Sandorfy C. The possible role of oligosaccharides in the mechanisms of anaesthesia. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

99 Sasaki R, Hirota K, Yamazaki M. Barbiturates accelerate the recovery from brief hypoxia in area CA1 of rat hippocampal slices. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

100 Scholz A. Mechanisms of (local) anaesthetics on voltage-gated sodium and other ion channels. Br J Anaesth 2002; 89: 52–61[Abstract/Free Full Text]

101 Schwarz SKW, Puil E. CNS Effects of local anesthetics: novel lidocaine actions in the thalamus. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

102 Seeman P. The membrane actions of anesthetics and tranquilizers. Pharmacol Rev 1972; 24: 583–655[ISI][Medline]

103 Taheri S, Halsey MJ, Liu J, Eger EI, Koblin DD, Laster MJ. What solvent best represents the site of action of inhaled anesthetics in humans, rats, and dogs? Anesth Analg 1991; 72: 627–34[Abstract]

104 Thill M, Rehberg B, Urban BW. Differences in isoflurane sensitivity between sodium channels of different origin. Eur J Physiol 1999; 437: R77

105 Todorovic SM, Lingle CJ. Pharmacological properties of T-type Ca2+ current in adult rat sensory neurons: effects of anticonvulsant and anesthetic agents. J Neurophysiol 1998; 79: 240–52[Abstract/Free Full Text]

106 Todorovic SM, Perez-Reyes E, Lingle CJ. Anticonvulsants but not general anesthetics have differential blocking effects on different T-type current variants. Mol Pharmacol 2000; 58: 98–108[Abstract/Free Full Text]

107 Trudell JR, Bertaccini E. Molecular modelling of specific and non-specific anaesthetic interactions. Br J Anaesth 2002; 89: 32–40[Abstract/Free Full Text]

108 Urban BW. Differential effects of gaseous and volatile anaesthetics on sodium and potassium channels. Br J Anaesth 1993; 71: 25–38[ISI][Medline]

109 Urban BW. Order despite a multitude of molecular anaesthetic actions. Eur J Anaesthesiol 1995; 12: 1–4[Medline]

110 Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

111 Urban BW, Bleckwenn M. Concepts and correlations relevant to general anaesthesia. Br J Anaesth 2002; 89: 3–16[Abstract/Free Full Text]

112 Urban BW, Frenkel C, Duch DS, Kauff AB. Molecular models of anesthetic action on sodium channels, including those from human brain. Ann N Y Acad Sci 1991; 625: 327–43[ISI][Medline]

113 Urban BW, Haydon DA. The actions of halogenated ethers on the ionic currents of the squid giant axon. Proc R Soc Lond B Biol Sci 1987; 231: 13–26[ISI][Medline]

114 Vahle-Hinz C, Detsch O. What can in vivo electrophysiology in animal models tell us about mechanisms of anaesthesia? Br J Anaesth 2002; 89: 123–43[Abstract/Free Full Text]

115 Veselis RA, Reinsel RA, Feshchenko VA. Novel insights into drug–memory interactions using functional neuroimaging (FNI) techniques. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

116 Violet JM, Downie DL, Nakisa RC, Lieb WR, Franks NP. Differential sensitivities of mammalian neuronal and muscle nicotinic acetylcholine receptors to general anesthetics. Anesthesiology 1997; 86: 866–74[ISI][Medline]

117 Wachtel RE. Relative potencies of volatile anesthetics in altering the kinetics of ion channels in BC3H1 cells. J Pharmacol Exp Ther 1995; 274: 1355–61[Abstract]

118 Wachtel RE, Wegrzynowicz ES. Kinetics of nicotinic acetylcholine ion channels in the presence of intravenous anaesthetics and induction agents. Br J Pharmacol 1992; 106: 623–7[Abstract]

119 Wartenberg HC, Wang J, Rehberg B, Urban BW, Duch DS. Molecular actions of pentobarbitone on sodium channels in lipid bilayers: role of channel structure. Br J Anaesth 1994; 72: 668–73[Abstract]

120 Watson SP, Girdleston D. TIPS on nomenclature. Trends Pharmacol Sci 1994; 16: 15–6[ISI]

121 Wittmann S, Daniels S, Fröhlich D. Thiopentone and propofol inhibit the expression of FMLP receptors in human neutrophils in vitro. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

122 Woodbury JW, D’Arrigo JS, Eyring H. Molecular mechanism of general anesthesia: Lipoprotein conformation change theory. In: Fink BR, ed. Molecular Mechanisms of Anesthesia. New York: Raven Press, 1975; 253–75

123 Yost CS, Kindler CH. Tandem pore (2P) domain K+ channels as anesthetic targets. In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press

124 Zhou Q, Lovinger DM. Pharmacologic characteristics of potentiation of 5-HT3 receptors by alcohols and diethyl ether in NCB-20 neuroblastoma cells. J Pharmacol Exp Ther 1996; 278: 732–40[Abstract]

125 Zimmerman SA, Jones MV, Harrison NL. Potentiation of gamma-aminobutyric acidA receptor Cl current correlates with in vivo anesthetic potency. J Pharmacol Exp Ther 1994; 270: 987–91[Abstract]

126 Zink W, Graf BM, Martin E, Fink RHA, Kunst G. Modulation of intracellular calcium homeostasis by tetracaine in skinned skeletal muscle fibres—reasons for its moderate myotoxicity? In: Urban BW, Barann M, eds. Molecular and Basic Mechanisms of Anesthesia. Lengerich: Pabst Science Publishers, 2002 in press