Robert S. Cantor
1. Work in progress:
Receptor desensitization by neurotransmitters in membranes:
are neurotransmitters the “endogenous” anesthetics?
We propose a novel explanation of general anesthesia, related to the desensitization of ligand-gated ion channels, that would explain two of the most troubling peculiarities of general anesthesia: the remarkably small variability of sensitivity both within the human population and across a broad range of animal phyla, and the striking ability of anesthetics to selectively inhibit excitatory channels, while activating inhibitory channels. It is hypothesized that in addition to rapid, saturable binding of a neurotransmitter to its receptor, resulting in activation, the neurotransmitter also acts indirectly by diffusing into the membrane and changing one or more membrane properties, which induces a change in the conformational equilibria of the receptor (desensitization). Unlike binding, this slower indirect mechanism is both nonsaturable (there is no limit to the resulting shift in protein activity with increasing membrane concentration of neurotransmitter) and nonspecific (each neurotransmitter will, in principle, affect all receptors in the membrane.) For proteins modeled as having only resting and active conformational states, time-dependent ion currents are predicted that exhibit many of the characteristics of desensitization both for inhibitory and excitatory channels. If, by design, receptors regulate the time course of ion currents by this mechanism, then (a) mutations that significantly alter receptor sensitivity to this effect would be lethal and (b) to act constructively, excitatory receptors would have been engineered to be inhibited, and inhibitory channels activated. All molecules that affect the membrane similarly but do not bind to receptors would thus inhibit excitatory channels and activate inhibitory channels; i.e., cause anesthesia. If so, then the endogenous anesthetics are the neurotransmitters, the survival advantage conferred by their proper membrane-mediated action on receptors explaining the selection pressure for anesthesic sensitivity.
The simplest in vitro functional test of this hypothesis would be electrophysiological: use a procedure designed to test the ability of an anesthetic to alter the activity of a particular receptor, but replace the anesthetic with a neurotransmitter that does not bind to that receptor. For example, using a membrane patch containing a single excitatory receptor (e.g., nAChR), preincubate the patch with a different neurotransmitter (e.g., GABA) and determine the dose-response curve (or the time-dependence of the ion current at various agonist concentrations) for the normal agonist (acetylcholine); if the proposed mechanism is valid, the presence of the wrong neurotransmitter (regardless of whether its receptor is inhibitory or excitatory) should inhibit the ion current significantly. Conversely, preincubation of an inhibitory receptor with a different neurotransmitter would be predicted to potentiate the response of the receptor to its normal agonist, and at high enough concentrations, presumably to activate the channel in the absence of its agonist.
Other membrane proteins undergo conformational transitions in which specific binding of a triggering molecule rapidly activates the protein, with inactivation occuring on longer time scales. As Jim Sonner has recently pointed out [Anesth. Analg. (2002) 95, 609-614], the broadly conserved sensitivity to anesthetics across different animal phyla indicates that the selection pressure is likely to be of very distant common ancestry. Thus, it is tempting to speculate that the double mechanism of action hypothesized above - fast specific binding followed by nonspecific membrane-mediated modulation - may have been developed early in evolution as a useful mechanism for tuning the time dependence of the activity of such proteins, which has the advantage of only requiring a single molecule. If so, then this generic mechanism may have been adapted in many different contexts, including ligand-gated ion channels in excitable cells, and could thus explain the distant common ancestry of anesthetic sensitivity.
The plausibility of this mechanism depends on the details of the kinetics and thermodynamics of membrane partitioning of neurotransmitter, and that its presence in the membrane alters bilayer properties similarly to anesthetics, sufficiently so to modulate receptor conformational equilibria. Partitioning equilibria and possibly the accompanying changes in membrane properties could presumably be probed using calorimetric, micromechanical, or 31P NMR methods. However, at present no data exist on membrane/water partition constants for acetylcholine or the zwitterionic amino acid neurotransmitters. However, it has recently been shown that charged organic molecules partition to a far greater extent into membranes than expected; largely residing in the head group region and the hydrophobic domain adjacent thereto. Neurotransmitters might be expected to partition similarly, affecting head group packing, presumably increasing the out-of-plane component of head-group dipole moments, which would result in large increases in repulsive interactions, and which would thus be expected to cause a net redistribution of the lateral stresses in the membrane from the interior toward the aqueous interfaces, even at low membrane concentrations of neurotransmitters. (The lateral pressure profile, its dependence on bilayer composition, and its mechanism of influence on protein conformational equilibria, are discussed in detail below.) If it is such changes in the lateral pressure profile that cause the shift in protein equilibria, then the overall effect of neurotransmitters would be expected to be similar to that of anesthetics: shifting lateral pressures out of the membrane interior.
2. Recent research:
By what mechanism do changes in bilayer composition modulate conformational equilibria of intrinsic membrane proteins?We have suggested an indirect mechanism, involving the lateral pressure profile, i.e., the depth-dependent distribution of lateral stresses in the membrane.
The idea is simple:1. As a membrane protein undergoes a conformational transition (e.g., from an "inactive" to an "active" state) its overall shape will change. In particular, its cross-sectional area in the transmembrane region may change nonuniformly, i.e., it expands or contracts laterally, to a degree that varies with depth in the bilayer.
2. The bilayer applies lateral pressure against the protein, the magnitude of which also depends on depth within the bilayer. The distribution of these lateral stresses depends sensitively on bilayer composition.
3. To accomplish its conformational change, the protein must thus perform mechanical work. The total work is the sum of the work done at different depths - a function both of the lateral pressure and the change in cross-sectional area of the protein at that depth. A change in bilayer composition will redistribute the pressures, which changes the mechanical work of the conformational change, resulting in a shift in protein conformational equilibrium (modulated protein activity.) For example, an increase in lateral pressure at a depth where the change in protein cross-sectional area is largest causes an increase in mechanical work, and thus shifts the protein equilibrium back towards the inactive state, i.e., protein inhibition.
For this hypothetical mechanism to be tested and to have predictive value, it is necessary to determine both
While it is possible to predict the shifts in lateral pressure
different lipid composition, the second part is more elusive: there
almost no direct experimental information on the change in the
area (as a function of depth in the bilayer) that accompanies protein
I. Calculations of the lateral pressure profile for varying bilayer composition
Initially, simple lattice statistical mechanical calculations were used to predict the effects of small solutes of varying interfacial activity on the pressure distribution, resulting in a novel nonspecific (i.e., indirect) molecular mechanism of anesthesia. This approach predicts some of the well-known failures of the Meyer-Overton correlation - the anomolously low anesthetic potency of molecules which are highly soluble in membranes - e.g., overly hydrophobic molecules such as perfluorinated alkanes, and also long-chain alkanols.
This putative mechanism by which bilayer composition influences protein activity has been explored in much greater detail through extensions and improvements in the lattice statistical thermodynamic methodology. Resulting calculations predict large redistributions of lateral pressure to accompany variation in chain length, degree and position of chain unsaturation, head group repulsion, and incorporation of cholesterol and interfacially active solutes. Combinations of compositional changes are found that compensate with respect to bilayer thickness, thus eliminating effects of hydrophobic mismatch, while still effecting significant shifts of the pressure profile. It is also predicted that the effect on the pressure profile of addition of short alkanols can be reproduced with certain unnatural lipids. These results suggest possible roles of cholesterol, highly unsaturated fatty acids and small solutes in modulating membrane protein function and suggest unambiguous experimental tests of the pressure profile hypothesis. In part as a test of the methodology, equilibrium molecular areas and area elastic moduli have also been calculated for bilayers of varied composition.
Unfortunately, there is little direct information on the change in shape of the transmembrane region of any protein. So, we considered various geometric models of such conformational transitions. For both a generic model, and for specific models that describe likely cooperative rearrangements of alpha-helices in bundles, it is found that the conformational equilibrium depends on the first and second integral moments of the lateral pressure distribution. In addition to revealing the possible physical underpinnings of the well-known correlation between protein activity and the "non-lamellar" tendency of bilayer lipids, this dependence on moments of the pressure profile allows for prediction of the relative effects of different lipid compositional changes even in the absence of information on specific protein shape changes. Effects of variation in acyl chain length, degree and position of unsaturation, and addition of cholesterol and small interfacially-active solutes (n-alkanols) are compared to effects on activity of proteins such as rhodopsin.
When the common practice of fitting dose-response data to the Hill equation is applied to the predictions of the pressure-profile mechanism, the fits are found to be reasonably good, with large Hill coefficients. Since this would commonly be interpreted as evidence of the existence of multiple sites with strong positive cooperativity, it is argued that caution must therefore be exercised in the interpretation of titration data in the absence of direct evidence of the existence of binding sites. The form of the titration curve predicted from this lateral pressure mechanism is shown to be quite general for indirect mechanisms. It is also shown that this form is the same as would obtain from classical models of binding cooperativity such as that of Monod, Wyman and Changeux in the limit of an infinite number of sites with vanishingly small site affinity.
IV. Breaking the Meyer-Overton rule: predicted effects of varying stiffness and interfacial activity on the intrinsic potency of general anesthetics.
Exceptions to the Meyer-Overton rule are commonly cited as evidence against indirect, membrane-mediated mechanisms of general anesthesia. However, another interpretation is possible within the context of an indirect mechanism in which solubilization of an anesthetic in the membrane causes a redistribution of lateral pressures in the membrane, which in turn shifts the conformational equilibrium of membrane proteins such as ligand-gated ion channels. It is suggested that compounds of different stiffness and interfacial activity have different intrinsic potencies, i.e., they cause widely different redistributions of the pressure profile (and thus different effects on protein conformational equilibria) per unit concentration of the compound in the membrane. Calculations incorporating the greater stiffness of perfluoromethylenic chains and the large interfacial attraction of hydroxyl groups predict the higher intrinsic potency of short alkanols than alkanes, the cutoffs in potency of alkanes and alkanols and the much shorter cutoffs for their perfluorinated analogs. Both effects, increased stiffness and interfacial activity, are present in unsaturated hydrocarbon solutes, and the intrinsic potencies are predicted to depend on the magnitude of both effects and on the number and locations of multiple bonds within the molecule. Most importantly, the intrinsic potencies of polymeric alkanols with regularly spaced hydroxyl groups are predicted to rise with increasing chain length, without cutoff; such molecules should serve to distinguish unambiguously between indirect mechanisms and direct binding mechanisms of anesthesia.
Partition coefficients of short n-alkanols between bilayers of different lipid composition (equivalently, the variation in bilayer/water partition coefficients) are calculated as a function of lipid acyl chain length and unsaturation, the strength of lipid head group repulsions, and the addition of cholesterol. Predictions are obtained from a statistical thermodynamic approach using a mean field lattice model identical to that used recently to calculate the lateral pressure profile in fluid bilayers. Increasing length, and particularly increasing cis-unsaturation of the acyl chains are predicted to increase the bilayer/water partition coefficients of short-chain alkanols, whereas addition of cholesterol is predicted to have the opposite effect. The magnitude of the shifts are predicted to be significantly larger for lipids with head groups with little or no repulsions, such as PE, than for more strongly repulsive head groups such as PC.
Redistributions of lateral pressures resulting from changes in bilayer composition can affect other membrane equilibria. Similarly to the protein conformational changes discussed above, the aggregation of membrane peptides is characterized by a nonuniform change in lateral excluded area. In particular, some membrane peptides, such as Alamethicin, form barrel-stave aggregates with a broad probability distribution of size (number of peptides in the aggregate) that has been shown to depend on the characteristics and composition of the lipid bilayer. Using thermodynamic analysis analogous to that developed for protein conformational equilibria, along with a simple geometric model of aggregates of kinked cylindrical peptides and with results of previously calculated lateral pressure distributions, it is possible to predict the effects of changes in bilayer characteristics on aggregate size distributions, in good qualitative agreement with experimental results.