THE MEMBRANE AS THE SITE OF ANESTHETIC ACTION

The electrical activity (i.e., transfer of ions) immediately underlying the transmission of nervous impulses occurs principally at the plasma membranes of nerves. Because inhaled agents disrupt this transmission, synaptic or axonal membranes, or both, are usually assumed to be the primary sites of anesthetic action. A membrane site of action is also consistent with the hydrophobic or amphipathic theories of anesthesia because plasma membranes consist largely of hydrophobic or amphipathic components.

Electrophysiologic studies reveal the effects of anesthetics on the flow of ions through excitable membranes. In isolated axons, the conduction of the nervous impulse requires the sequential flow of sodium and potassium ions through selective transmembrane channels. Excitation produces a rapid increase in sodium conductance to a peak (i.e., activation process) followed by a slower decline in sodium conductance to zero (i.e., inactivation process). Although channels in peripheral neurons tend to be relatively insensitive to anesthetics,^[2] clinical concentrations of inhaled anesthetics may decrease the magnitude of sodium ^[110] and potassium^[111] currents in channels isolated from brain.

Electrophysiologic recordings of single membrane channels activated by a neurotransmitter can be obtained by forming a seal between the tip of a glass micropipette and a membrane patch of a few square micrometers. Using this patch-clamp technique, unitary acetylcholine receptor channel currents can be measured in the absence and in the presence of inhaled anesthetics ( Fig. 4-11 ).^{[112] [113]} Isoflurane decreases the average open duration of this membrane channel, and such an accelerated decay of the membrane current may decrease the net charge transferred across the membrane and impair transmission (see Fig. 4-11 ).^{[112] [113]}

In contrast to the inhibitory effects of anesthetics on ion flow through membranes described earlier, other experiments suggest that anesthetics may act by enhancing the conductance of certain ions through membranes. Volatile anesthetics enhance the inhibitory current responses produced by GABA (see Fig. 4-7 ),^{[49] [114]} presumably by increasing the flow of chloride ions through GABA receptor-channel complexes in neuronal membranes.^[114] Inhaled agents may also enhance potassium conductance, resulting in membrane hyperpolarization

118

Figure 4-11 Examples of unitary acetylcholine receptor channel currents before, during, and after local microperfusion with 5 mM isoflurane. Channel openings are represented by downward directions of the current trace. When isoflurane is removed (bottom panel), the channels are indistinguishable from those observed under control conditions. Notice that pre-equilibrated and 10-fold lower concentrations of isoflurane also decrease the average open duration of acetylcholine receptor channel currents.^[113] (From Brett RS, Dilger JP, Yland KF: Isoflurane causes "flickering" of the acetylcholine receptor channel: Observations using the patch clamp. Anesthesiology 69:161, 1988.)

The requirement of an intact plasma membrane for the transmission of nervous impulses and the abilities of anesthetics to disrupt the flow of ions through plasma membranes point to a membrane site of anesthetic action. Nevertheless, anesthetics could act indirectly at a cytoplasmic site. For example, anesthetics may alter the calcium-accumulating activity of cellular organelles (e.g., mitochondria ^[76] ) and thereby alter the levels of intracellular free calcium. Such alterations in intracellular free calcium could influence the conductance properties of excitable membranes and alter the presynaptic release of neurotransmitters.^[76]

This discussion strongly suggests but does not prove that anesthesia results from an association of inhaled agents with the plasma membranes of nerves. If this is so, which components of the plasma membrane are altered by the anesthetics? Biologic membranes consist of a cholesterol-phospholipid bilayer matrix that is approximately 4 nm thick. Peripheral proteins are weakly bound to the exterior hydrophilic membrane, and integral proteins (e.g., ion channels) are deeply embedded in or pass through the lipid bilayer ( Fig. 4-12 ). Synaptic plasma membranes are approximately one-half protein and one-half lipid by weight. If, as implied by the Meyer-Overton rule, inhaled agents bind to hydrophobic sites, anesthetics could act on the nonpolar interior of the lipid bilayer, at hydrophobic pockets in proteins extending outside or embedded into the lipid bilayer, or at the hydrophobic interface between intrinsic membrane proteins and the lipid matrix (see Fig. 4-12 ).

Attempts to better understand the penetration of inhaled agents into, and their interaction with, membrane sites have led to an examination of isolated membrane components. These experiments were greatly aided by the discovery that phospholipids dispersed in an aqueous medium spontaneously form bilayers comprising the surfaces of spherical structures (i.e., liposomes). These phospholipid bilayers act as a permeability barrier to ions and are similar to those found in biomembranes. In contrast, membrane proteins are often difficult to isolate and purify, and membrane proteins typically exhibit an impaired function unless surrounded by a boundary layer of lipid. Nevertheless, advances have resulted in the biochemical and electrophysiologic characterization of several protein receptor or ionophores thought to permit the passage (i.e., tunneling) of ions through membranes during excitation. However, only limited information is available concerning direct binding of inhaled agents with these membrane protein ionophores, and many experiments that examine anesthetic-protein interactions employ soluble proteins as model systems. Such proteins are easy to prepare in reasonable quantities but may not mimic precisely the natural proteins responsible for ion translocation.

119

Figure 4-12 Four possible target sites for inhaled anesthetic molecules (solid circles) in a neuronal membrane include the lipid bilayer as a whole (a), lipids at a protein-lipid interface (b), a protein site bounded by lipid (c), and a protein site exposed to an aqueous environment (d). (Adapted from Franks NP, Lieb WR: What is the molecular nature of general anesthetic target sites? Trends Pharmacol Sci 8:169, 1987.)