Dr. Raines correctly points out that the effects of either 1 MAC isoflurane or 1 MAC halothane on tryptophan side-chain mobility in bovine serum albumin is comparable to what follows from a 1°C reduction in temperature. Changes in lipid fluidity in the presence of anesthetic molecules can be mimicked by small variations in temperature, and this has been argued to indicate that lipids are an implausible site of anesthetic action. 1By analogy then, as noted by Dr. Raines, the same line of argument would suggest that a protein target would be an equally unlikely in vivo site of anesthetic action.
If anesthetics interact directly with protein targets and alter their function, then binding must influence either the structure of the protein or its dynamics. Alternatively, anesthetics may compete with native ligands for their binding sites. 2The latter mode of action does not appear to apply in the case of ligand-gated ion channels, since anesthetics increase the affinity of the native neurotransmitter. 3The limited information on protein structural changes induced by bound anesthetic molecules indicates that secondary structure is not altered. 4–7It is therefore likely that a bound anesthetic molecule instead perturbs the tertiary structure of the protein, or perhaps the quaternary structure. Examples of the latter structural change are the effects of halothane and diethyl ether on the aggregation state of the membrane-bound Ca-ATPase. 8,9The 2.2 Å X-ray crystal structure of firefly luciferase with two bound bromoform molecules revealed that anesthetic binding caused minimal overall protein structural changes. 7Another possibility is that anesthetics may not alter protein structure but instead modify amino acid side-chain dynamics, 10which are intimately related to protein function. In line with changes in dynamics is the finding that bromoform binding caused a neighboring histidine residue (H310) in the firefly luciferase crystal structure to become less mobile. 7High resolution X-ray crystallography will be required to detect the small structural changes that are likely in the case of weakly interacting ligands such as the volatile general anesthetics. A promising alternative approach for examining the effect of a bound anesthetic molecule on protein structure and side-chain mobility is to perform molecular dynamics simulations. 11
The effect of anesthetic agents on bovine serum albumin dynamics suggested a potential mechanism for how protein function is altered. It was proposed that anesthetic binding traps the protein in a substate with a lower free energy minimum than the native substate, and therefore effectively prevents the conformational changes required for normal protein activity at a physiologic temperature. 10Studies with additional proteins will reveal whether this model has wide applicability. It certainly remains plausible that physical properties of lipids and proteins not amenable to experimental analysis at present are responsible for the clinical effects of inhaled anesthetics. One example of this is the lipid-based mechanism of general anesthetic action proposed by Cantor. 12However, until such theories can be experimentally tested, we remain optimistic that currently available biophysical tools will provide useful information, and generate testable hypotheses, regarding how these important clinical agents may exert their effects on the central nervous system.