Axons are cylinders of axoplasm encased in the axonal membrane, which is similar to other plasma membranes (Fig. 2-2). Axons are always enveloped by a Schwann cell. Many unmyelinated axons lie within invaginations of a single Schwann cell (Fig. 2-2C), in contrast to the thick myelin sheath of a myelinated axon, which is formed by a single Schwann cell wrapped many times around the axon (Fig. 2-2B). This myelin sheath is interrupted periodically at the nodes of Ranvier, where the extracellular medium has access to the axolemma (Fig. 2-2A).

Nerve impulses travel along nonmyelinated axons as a uniform wave, similar to the way a flame progressively ignites the fuse of a firecracker. The nerve impulse, or action potential, is a change in the electrical voltage across the membrane that is due to changes in the permeability of ionic channels in the axon membrane (2,3). In nonmyelinated axons, these permeability changes occur relatively uniformly along the axon, supporting a wave of inward ion current that underlies the depolarization of the nerve impulse. In myelinated axons, however, changes in the axon’s membrane permeability and the associated inward currents occur only at the nodes of Ranvier; the myelinated internode of the axon is depolarized by the passive spread of current from the nodes (4). Thus, impulse conduction in these axons is continuous but not homogeneous.

The modern fluid mosaic model depicts the plasma membrane as a bilayer of phospholipid molecules with their fatty acyl chains facing each other, thus making the inner portion of the membrane hydrophobic (Fig. 2-3). The surfaces of the membrane facing the cytoplasm and extracellular fluid are formed by the charged and polar hydrophilic groups of the phospholipid molecules. Globular proteins are also present, and some of these penetrate through the entire thickness of the membrane. Ionic channels are composed of such transmembrane proteins, as are various transporters and receptors for peptide and nonpeptide transmitters and hormones. Such proteins are heavily glycosylated by carbohydrate groups attached covalently to their extracellular surface and often are covalently bound to lipophilic fatty acids that stabilize their transmembrane domains in the hydrocarbon core of the membrane bilayer (5). An intracellular matrix of cytoskeleton is frequently linked to membrane proteins by other specific proteins. All these attachments secure active proteins within the membrane and also can respond dynamically to cellular activity and metabolism to change the protein populations on the cell membrane. As described later (in Basic Pharmacology), local anesthetic molecules are distributed primarily near the interface of membranes, with aqueous solutions that surround them, and also at protein–membrane interfaces and, sometimes, in the ion-conducting pores of membrane channels. Figure 2-3 shows lidocaine’s membrane and protein distribution at the density of drug that occurs for the critical blocking content, which corresponds to one lidocaine molecule for about 100 phospholipid molecules.

In clinical practice, local anesthetics must diffuse across a number of structures before reaching their site of action in the axonal membrane (6). Peripheral nerves contain both afferent and efferent axons (Figs. 2-1 and 2-4). These individual axons and their Schwann cells are linked loosely together by a delicate layer of fine connective tissue (endoneurium) that allows the easy diffusion of most local anesthetics.

Bundles of axons are enclosed in a squamous epithelial cellular “sheath,” the perineurium, which comprises several layers of cells and acts as a semipermeable barrier to local anesthetics (7,8,9). One or more perineurial bundles are covered by an outermost, easily permeable, connective tissue layer, the epineurium. This layer also carries the nutritional blood vessels of larger nerves. Factors that have an important influence on local anesthetic diffusion to the axons include the thickness of the perineurium, the presence or absence of myelin, the size of the axons, and the anatomic position of the axons, either closer to the outer, more superficial mantle layers of the nerve or deeper within the inner “core” sections of the nerve.

Two factors combine to create electric potentials in cells: (a) concentration gradients of ions across membranes and (b) selective permeation of ions through membranes. The gradients are diffusional forces that tend to move the ions; the selective changes in permeability permit that tendency to be manifested as ionic current. Energy from the cell’s metabolism is used to create and maintain the gradients.

The concentration of potassium ions (K⁺) inside a cell is about 10 times greater than the extracellular K⁺ concentration, and vice versa for sodium ions (Na⁺). A special protein in the membrane (the Na/K pump) actively transports K⁺ into the cell and Na⁺ out of the cell, using adenosine triphosphate (ATP) as the source of energy (10). In the resting cell membrane, a selective permeability to K⁺ ions exists, permitting the net efflux of a small number of K⁺ ions, thus leaving the axoplasm electrically negative (polarized) while making the outside electrically positive. This accounts, for the most part, for the cell’s “resting potential,” which typically equals -60 to -70 mV (Fig. 2-5).

At rest Na⁺ ions tend to flow into the axon, both because the resting axon is electrically negative inside and because the Na⁺ ions are more concentrated outside. During a nerve impulse, a selective permeability to Na⁺ arises when specific Na channels in the axon membrane are opened (2,3). Figure 2-5 shows the contributions of ionic currents during a conducted nerve impulse. The large inward Na current (I_Na⁺) accounts for the depolarizing phase of the impulse. Opening of Na⁺ channels occurs as the membrane is initially depolarized from the resting potential; that is, as the potential becomes less negative. This opening, a result of complex channel activation, is an intrinsic behavior of Na⁺ channels, due to charged amino acids that are part of the channel molecule’s structure (11). As the membrane depolarizes and the potential becomes less negative, more Na⁺ channels open, and they open more rapidly. As more channels open, more Na⁺ ions enter the cell, and the resulting depolarization is accelerated (Fig. 2-6). This positive feedback cycle accounts for the “regenerative” behavior of nerve impulses (2).

Action potentials are brief, transient events. Repolarization of the membrane to the negative resting value occurs because of three factors: (a) The driving force moving Na⁺ into the cell diminishes as the axoplasmic potential becomes less negative and approaches the Nernst potential for Na⁺(∼ +50 mV), (b) the sodium channels eventually close (inactivate) during a depolarization, and (c) voltage-dependent K channels open and permit a large, outward K⁺ current that returns the axoplasmic potential to or beyond, its resting value (Fig. 2-5). In simple terms, impulses have a depolarization phase and a repolarization phase. Inward currents, carried by Na⁺ ions, depolarize the cell, whereas outward currents, carried by K⁺ ions, repolarize the cell. The integrated result of all these factors is diagrammed in Figure 2-6, which summarizes the ionic contribution to the nerve impulse.

With each action potential, a pulse of Na⁺ ions enters the cell and a pulse of K⁺ leaves it. Small changes in the concentrations of these ions can build up during repeated discharge and thus lessen the ion gradients. However, the Na⁺/K⁺-ATPase (Na⁺/K⁺ pump), which is activated by elevation of intracellular Na⁺, will return this Na⁺ to the extracellular space and, reciprocally, lower the extracellular K⁺, removing it to the intracellular compartment, using the energy of ATP. This “electrogenic” pump generates an outward current during this process and thereby hyperpolarizes the cell, contributing to an after-potential (10) and shaping the patterns of impulses that occur in constant or intermittent trains of discharge.

Inward current entering the axon during the depolarizing phase of an impulse flows within the conducting medium of the axoplasm and thereby spreads to adjacent, inactive regions. These adjacent regions are depolarized by this “local circuit” current, usually to levels far in excess of those for threshold conditions,
and the regenerative impulse then extends to this region, generating its own inward current. The propagating wave of potential spreads far along axons during any one instant, over distances that are much larger than the fiber’s diameter. For example, if an impulse lasts for 1 msec, from the beginning of its depolarizing phase to the time of initial repolarization, and is conducted at a velocity of 1 m/sec, as in the small, nonmyelinated C fibers (of diameter less than 1 μm), then the action potential extends over 1 mm along the axon, 1,000 times the axon’s diameter. For large myelinated axons, 20 μm in diameter, impulses travel at 60 to 100 m/sec and thus extend over distances of 60 to 100 mm. Inward currents from all the active nodes add together as they spread toward inactive regions, ensuring that impulse propagation will continue. Therefore, complete block of impulses in about three to five nodes in sequence is necessary for the total prevention of impulse propagation in myelinated fibers (4,12,13). Furthermore, the “action current” flowing into the nerve under the impulse is five to ten times that required to depolarize the next segment to threshold; this ratio is referred to as the margin of safety for impulse conduction. Due to this large margin of safety, about 80% of the Na channels in an axon membrane must be inhibited in order to fully block a propagated impulse.

In the face of reduced membrane excitability due to disease, altered metabolism, or drug action, impulses may still occur but will be conducted more slowly. When the net inward current is decreased, the action potential amplitude will be smaller, the rate of impulse depolarization slower, and the velocity of conduction lower. This condition often worsens as an impulse attempts to propagate through a diseased or drugged axon, a situation termed decremental impulse conduction. If the decrement is sufficiently great, conduction failure occurs; if not, then only conduction slowing along the affected axonal region occurs, and normal impulse conduction resumes if the decremented action potential is able to excite the contiguous normal region of the axon. This situation certainly occurs for nerve conduction during the onset and regression of a local anesthetic block. Even in an effectively, fully blocked axon, conduction fails decrementally as an impulse enters the anesthetic-exposed region of the nerve rather than stopping abruptly in a very short distance.

Under physiologic conditions, most local anesthetics exist in rapid equilibrium between the neutral and protonated, charged species (Fig. 2-9). The ratio of ionized (LH⁺) to nonionized (L) molecules is given by the Henderson-Hasselbalch equation:

where the pK_a is the value of pH at which the concentrations of the unionized and ionized species are equal. The pK_a values of clinically used local anesthetics range from about 7.7 to 9.5 (27), so that in solution all these drugs are charged more often than they are uncharged. Protonation and deprotonation in solution are very rapid processes, occurring about 10³ times per second. When the drug is buried in a membrane or bound to a protein molecule, however, these reactions can be much slower.

Both ionized and nonionized drug species can inhibit Na⁺ channels. Quaternary derivatives of local anesthetics, which are permanently charged, when placed inside a neuron are potent blockers of impulses and of Na⁺ currents (28,29). Interestingly, these drugs have very slowly developing effects when present outside of the axon; their constant charge and low hydrophobicity greatly restrict their passage into and through the membrane.

Perpetually uncharged local anesthetics also effectively block impulses and Na⁺ channels (30,31,32), demonstrating that ionization is not essential for pharmacologic activity. The ability of tertiary amine local anesthetics to block Na⁺ channels in single cells is faster and greater when the drug is applied externally at alkaline external pH (33,34,35), or when it is applied by direct internal perfusion at neutral or slightly acidic internal pH (36). Both of these observations are consistent with a model in which the neutral form of the anesthetic dissolves in and passes through the axon membrane and, having reached a cytoplasmic region, becomes protonated.

Changes in the binding and dissociation of local anesthetics by direct mutation of certain amino acids of the large, α-subunit of the channel demonstrate that the drug binds in the inner “pore” of the Na channel (37,38,39,40) (Fig. 2-10). One intriguing finding is that dissociation of anesthetic molecules from the closed channel is clearly sensitive to the pH outside the nerve, but relatively insensitive to the intracellular pH (34). Thus, when it is bound in the Na⁺ channel, a local anesthetic molecule can interact with extracellular protons. Since the changes in channel structure that underlie its gating also contribute to changes in local anesthetic affinity (see the earlier section, Mechanisms of Anesthetic Action),
other factors that influence gating can modify drug blockade. Among these are the concentration of calcium (Ca²⁺) ions outside the cell (19) and the identity of the smaller β-subunits that are associated with the large α-subunit (41).

Drug hydrophobicity is also a determinant of potency. The more hydrophobic drugs are more potent blockers of action potentials (42,43) and of Na currents (44,45,46). This increased potency results from a faster apparent association rate, as well as from a slower dissociation rate, for the binding of the local anesthetics to their channel sites (21,44). In other words, the more hydrophobic local anesthetics can reach the site more easily and appear to leave the site more slowly than the less hydrophobic drugs. Molecular mass of the anesthetics, however, also affects the dissociation of the charged local anesthetics, but not of the neutral ones (45). These results support the concept of two pathways to and from the anesthetic site (19). One pathway may be through the channel’s pore (the hydrophilic route, more favorable to charged molecules) and the other may be through the membrane (the hydrophobic route, accessible primarily to uncharged drug molecules). Whether these separate routes eventually lead to a single or several sites of action is unclear at this time.

From this discussion, it appears that pH affects anesthetic potency on isolated neurons by two opposing actions. An alkaline extracellular pH favors the neutral, membrane-permeant form of the drug, facilitating the rise of drug concentration in the membrane and thus inside the cell, whereas an acidic extracellular pH favors the more potent protonated blocking species when drug is present at the site of action. Both neutral and charged forms of anesthetics participate in impulse blockade. Of course, during most clinical procedures, the functional potency of local anesthetics is determined largely by the fraction of injected drug that actually passes into the nerve bundle, and conditions that favor the penetrating, neutral species are desirable (6,47). Other local anesthetics probably have about the same unfavorable distribution properties.

Exactly how do local anesthetics interfere with the operation of Na⁺ channels? Such questions are addressed by a combination of physiologic and structural studies. Much of this study is anchored on the observation that anesthetics inhibit open and inactivated channels more potently than resting closed channels, and that intentional mutation of specific amino acid residues believed to line the inner region of the pore (37,39) (along the transmembrane helices labelled S6 in Fig. 2-10), leads to changes in the affinities of local anesthetics for different states of the channel (48,49,50).

Does a local anesthetic molecule inhibit the channel by “plugging” the pore and preventing ions from passing, or by interfering with the conformational changes that underlie channel opening? Both possibilities may occur during the blockade of impulses by tertiary amine drugs, because the channel regions that bind the drug are coupled to those regions that sense the changes in membrane potential and that initially move in order to activate the channel. Other substances that pass through the channel as well as binding within it, such as hydrogen (H⁺) and certain cations, also change the gating, providing precedents for the coupled interaction between the pore-forming and the gating domains. Some local anesthetics interfere with the conformational restriction of gating regions (i.e., the molecular equivalent of inactivation), inducing an “inactivation-like” state by a drug-dependent mechanism rather than stabilizing the normal conformational changes that occur during inactivation.

The diameter and myelination of a nerve fiber are correlated with its impulse physiology as well as its message-carrying function (82,83). Nerve fibers are categorized into three major anatomic classes (Table 2-1): myelinated somatic nerves (A-fibers), myelinated preganglionic autonomic nerves (B-fibers), and nonmyelinated axons (C-fibers). The B- and C-fibers are small, ranging in diameter from about 2 μm to less than 1 μm, respectively, whereas the A-fibers vary in diameter from about 4 to 20 μm.