anaesthetic mechanisms General anaesthetics are among the most useful, the most dangerous, the least specific, and the least understood of the major drugs. They are useful because they allow surgery without sensation or any memory of the trauma or pain. They are dangerous because the dose that can kill is not much larger than the dose needed for anaesthesia. They are non-specific in that substances ranging from the noble gas xenon to barbiturates and steroids can produce general anaesthesia, and furthermore in that general anaesthetics affect many functions. Finally, the mechanisms of anaesthetic action are still not known and we also do not know which of the actions are critical to allow surgery.

Anaesthesia can be defined as ‘the reversible, non-specific suppression of organized function in a living system’ ( Meyer 1937). Clinical or general anaesthesia is only one example — but it is the example of overwhelming practical importance. Surgery requires: loss of consciousness, prevention of any memory of the trauma (amnesia), pain relief (analgesia), and muscle relaxation. All of this must be achieved without stopping the heart and with ventilation of the lungs maintained — either by leaving normal breathing unaffected, or artificially. Most general anaesthetics are capable of producing the desired combination of effects, but only at precisely controlled concentrations. In early days, the dose of chloroform or of ether had to be sufficient to achieve all the required effects, but in modern practice the anaesthetic drug is used only to produce the loss of consciousness and amnesia. Pain relief and relaxation of muscles, which would require higher and hence more dangerous concentrations of the anaesthetic, are usually secured by using more specific analgesics and muscle relaxants. The relaxants suppress normal breathing, so patients must be artificially ventilated.

The physiologist and the clinician have tended to approach the study of general anaesthesia by asking two questions: where do general anaesthetics act, and which function do they suppress? Suggestions for regions of the brain that could be the site of action have included the brain stem reticular formation, the cerebral cortex, and the thalamus. When sensory receptors are stimulated there are electrical responses that can be observed in the cortex. Anaesthetics applied at clinical concentrations to certain layers of the cortex, or to relay cells in the thalamus, delay the appearance of the responses and reduce their amplitudes. In one view anaesthetics ‘turn a valve’ shutting off the sensory input to the cortex.

action potential propagation in nerve fibres, release of neurotransmitters, and the responses of nerve cells to them, have all been suggested as the critical function that anaesthetics must suppress. Long distance communication in the nervous system occurs via ‘all-or-none’ action potentials which are designed to be reliable. High concentrations of anaesthetics can be shown experimentally to block these action potentials — but the levels required would be lethal in a patient. Subtler effects can be observed at the synapses where nerve cells come into contact with each other. Here communication occurs by release of chemicals — the neurotransmitters — which then act on the receptor sites of other nerve cells to stimulate their activity. At many synapses the transmitters are now known. Two of these have been proposed as playing critical roles in general anaesthesia: glutamate produces excitation of the cells it acts upon, and some anaesthetics decrease this excitatory effect; by contrast, GABA (gamma-aminobutyric acid) tends to inhibit electrical activity in the cells it acts on, and many anaesthetics potentiate this effect. It is proposed that by increasing an inhibitory effect, the level of activity in many portions of the nervous system can be reduced, which might very plausibly be a mechanism for general anaesthesia. Interaction with GABA receptors is virtually certain to be an important effect in the general anaesthesia produced by the barbiturates, certain steroids, and the benzodiazepines (drugs related to valium).

Physical chemists and pharmacologists, on the other hand, fascinated by the wide range of effects produced by anaesthetics, have been more concerned with different but closely-related questions that could be crucial for the development of more effective and safer anaesthetics: what are the molecular properties of the anaesthetic receptors or targets on cell membranes? And what are the properties that make a molecule a potent anaesthetic?

How do we characterize anaesthetics and their receptors when all that is known initially is that an effect is produced? There are two main approaches: determination of the relation between the chemical structure of the anaesthetics and their activity, and the measurement of binding of the anaesthetics to preparations which might contain the targets on which they act. About the only firm rules for the structure of anaesthetics are that they must be primarily hydrophobic (not soluble in water but soluble in lipids) and that the molecules must not be too large. The largest anaesthetic drugs are similar in size to cofactors for proteins and to the steroid lipids that form part of the structure of membranes. Because so many substances can satisfy these requirements, structure–activity relations have not been very helpful, and attention has been concentrated on the alternative approach.

A common method is to compare the potency of a drug at producing an effect with its binding to tissues, to homogenates of tissues, or to models that might contain the receptors (or at least sites that look like the receptors). Where possible, similar studies are carried out for every type of receptor. To apply this method, two key types of experiment must be performed. First, the dose or concentration needed to produce the effect must be determined. Given the great variety of effects produced by anaesthetics, many different effects or end points can and have been used — including the suppression of light flashes by fireflies (see later), the increased survival of red blood cells made to swell by exposure to dilute solutions, and the prevention of convulsions in deep sea divers. However, the two most commonly employed measures have been the loss of righting reflex in animals and loss of flinching response to surgical incision in people. If goldfish are released in tanks containing a range of concentrations of the alcohols, from ethanol through octanol, the concentration of the alcohol at which 50% of the goldfish lose the ability to keep upright and to swim can be determined. Now remember one important property of anaesthesia is that it is reversible (the patient must survive!) — so after the test, if the goldfish are put back into clean water they quickly recover. A common technique for measuring the righting reflex in mice is a rolling cage very much like the exercise wheels often provided for pet gerbils. The difference is that here the wheel is made to turn and the mouse must respond to stay upright. In this instance, as in most clinical use of anaesthetics, the anaesthetic is added as a gas or vapour.

Regardless of the end point chosen to indicate that the anaesthetic has taken effect, the experimental data for the first type of experiment are then a list of the concentrations, either in the water or in the air inhaled, at which a variety of anaesthetics produce equal effect — the C_an. An anaesthetic which produces an effect at a lower concentration is said to be more potent. Thus for a series of anaesthetics applied in the same way, potency is defined as the reciprocal of the concentration, 1/C_an.

The second key experiment is measurement of the amount bound. The principles are the same as for any type of drug, but there is an important difference. Most useful drugs, other than anaesthetics, bind with high affinity to specific cell membrane receptors. With a little luck in the choice of preparation, if measurements are repeated first for very low concentrations of drug and then for increasing concentrations, at first the amount bound increases, but it then reaches an apparent limit as the specific sites become saturated — that is, when there are no more free sites to allow more drug to bind. When this pattern is seen, the interaction between the drug and the site can be described by the concentration required for half-maximal binding. But for anaesthetics the binding is weak and no saturation is seen. Binding increases linearly with the applied concentration of the drug, and the amount that is bound depends only on the partition (or distribution) coefficient — the ratio of concentration of the drug in the chosen preparation to its concentration in the water or air used to deliver it. The higher this coefficient is, the lower the concentration that is required to reach a chosen level of binding. In other words, if different anaesthetics are applied by the same route until each produces the same amount bound, the concentration of each that is required is inversely proportional to its own partition coefficient. Many different preparations have been used in the binding experiments, ranging from simple solvents, like octanol and olive oil, to lipid vesicles, the membranes of red blood cells, and specific proteins.

One of the most striking and famous correlations in biology was first reported in 1901 by Meyer and by Overton, who found that anaesthetic potency was directly proportional to the partition into olive oil. Very similar results can be obtained in the goldfish experiment described above. In 1937 K. H. Meyer restated this result in a way which more clearly draws out its significance:
[Anaesthesia] commences when any chemically indifferent substance has achieved a certain molar concentration in the lipoids of the cell. This concentration depends on the nature of the animal or cell, but is independent of [the anaesthetic].

Lipoids are the hydrophobic parts of a cell (the cell membranes including the lipid bilayers and the proteins embedded in them). Meyer's rule, that anaesthetic effect depends solely on how many anaesthetic molecules are bound to the lipoids, does not explain anaesthesia, but it does tell us that the site of action is strongly hydrophobic; this almost certainly implies that anaesthetics act within membranes.

Theories for the mechanism of action have tended to divide into two groups: anaesthetics alter some physical property of the lipid membranes, which in turn alters the function of the proteins which form the ion channels in the membranes; or anaesthetics bind directly to membrane proteins and alter their function. This question remains unresolved, very possibly because the correct answer is both — but to varying degrees for different anaesthetics.

There are many aspects of cell membrane structure and function that anaesthetics could change and that could in turn alter nerve cell excitability. Proposals have included membrane volume expansion, changes in thickness of the membrane, changes in its tension, or changes in its fluidity. Perhaps the most important reason why many people favoured theories involving indirect effects, such as changes in fluidity, was that they ‘explained’ the need for anaesthetics to be hydrophobic — but work on fireflies showed that interactions directly with proteins could also ‘explain’ this need. Fireflies are found in the eastern US. On warm early summer evenings, they fly about emitting flashes of light from their tails. Presumably the purpose is to attract a mate, though it often also attracts small kids with jars. The flashes of light are produced by a change in the form of the molecules of the protein luciferase. Light emission from luciferase is inhibited by anaesthetics at concentrations that are remarkably close to those needed to produce general anaesthesia. The anaesthetics bind in a hydrophobic pocket which is the normal binding site for a necessary cofactor. These data lead to two important conclusions. Firstly, we must not take the models too seriously. None would claim that there is firefly luciferase in our brains any more than anyone would claim that we have pools of olive oil in our brains. Thus none are claiming that binding to a pocket in this particular protein or partitioning into a pool of olive oil is responsible for all anaesthetic effects. But — and this is the second point — since one such pocket has been found, there may well be others, and binding in these pockets may be the key factor in clinical general anaesthesia. It will be interesting indeed if membrane proteins turn out to have such hydrophobic pockets. Perhaps the binding sites for barbiturates and steroids on the GABA receptors are examples waiting for a more complete molecular description.

S. B. Hladky

See also membrane receptors.