Advances in Neurobiology of the Neuromuscular Junction: Implications for the Anesthesiologist

Ease of experimental study makes the vertebrate NMJ the synapse whose formation and function is best understood. The first part of this section summarizes molecular mechanisms involved in NMJ formation, with emphasis on the regulation of acetylcholine receptor (AChR) expression in the subsynaptic membrane. For more extensive accounts, see recent reviews by Sanes and Lichtman 1and Duclert and Changeux. 2The issue of how motor axons are guided to innervate the correct muscle fibers is not addressed in this review.

Committed myogenic cells and motor neurites arrive at the region where a muscle is to be formed at approximately the same time during development. Myogenic cells of mesodermal origin, after migration to such sites, divide to fuse into multinucleated myotubes expressing many contractile and synaptic proteins as part of their developmental program. Motor axons extending along peripheral nerves are followed by Schwann cells. 3Synaptic transmission begins within minutes after the growth cone contacts a myotube and is mediated initially by nicotinic AChRs (nAChRs) expressed constitutively along the entire myotube surface. Such nAChRs, termed “fetal” because of their expression early in development, are assembled from five subunits termed α, β, γ, and δ, each encoded by a different gene. 4In response to bound acetylcholine, nAChRs flicker rapidly between open and closed states for approximately 5 ms (apparent open times) and allow Na⁺, K⁺, and Ca²⁺ions to flow across the muscle membrane down their electrochemical gradients. Dissociation of acetylcholine closes the channel. 5The long burst duration of the fetal nAChRs’ channel combined with the high electrical input resistance of the myotubes allows single acetylcholine quanta to elicit action potentials in the myotubes (fig. 1). 6At early stages of neuromuscular development, muscle fibers receive input from several motor axons at a single synaptic site. All neural inputs except a single motor nerve withdraw as the NMJ matures. 7 

Synapse maturation involves the formation of a motor nerve terminal with densely packed synaptic vesicles containing the transmitter acetylcholine. Postsynaptic differentiation is characterized by the formation of a postsynaptic apparatus anchoring nAChRs at a density of 10,000/μm²in the subsynaptic muscle membrane. Unlike nAChRs in the nonsynaptic membrane, synaptic nAChRs become metabolically stabilized, their half-lives in the membrane increasing from approximately 1 to 10 days. 8The basal lamina (BL) enveloping the muscle fiber contains molecular components important to synapse formation, maintenance, and function. The postsynaptic region is further characterized by the presence of cytoskeletal and membrane proteins thought to be involved in its structural maintenance, the anchoring of AChRs and of voltage-activated sodium channels, as well as by the accumulation of several myonuclei. Subsynaptic myonuclei selectively begin to express a new nAChR subunit, ϵ, at the synapse, 9,10giving rise to a new, functionally distinct nAChR subtype (termed “adult”) with the subunit composition α²βϵδ in the synaptic muscle membrane. 4This mature nAChR has shorter burst duration and a higher conductance to Na⁺, K⁺, and Ca²⁺than the fetal nAChR. 11As discussed below, fetal nAChRs gradually disappear both from synaptic and nonsynaptic muscle membranes. Schwann cells cap the entire synaptic structure.

Signals from the nerve are twofold: (1) the nerve-induced propagated action potentials affect muscle fibers along their entire length, and (2) released or membrane-bound molecules act locally in the region of the NMJ.

Electrical activity down-regulates the synthesis of nAChRs in all but the subsynaptic myonuclei. 12Electrical activity also induces Ca²⁺influx through L-type Ca²⁺channels, which mediates metabolic stabilization of the synaptic nAChRs 13,via  unknown mechanisms. Impulse activity also affects synapse elimination. Specifically, blockade of the electrical activity in the motor nerve delays the withdrawal and thus the reduction of synaptic inputs converging on a single fiber. One factor involved appears to be the relative synaptic strength of the competing axons. Pharmacologic blockade of nAChRs selectively in the subsynaptic domain occupied by one contending terminal will cause that input to withdraw, 14and during normal synapse elimination, the axon withdrawal is preceded by loss of its subsynaptic domain. 15This is consistent with the notion that activation of nAChRs and the associated Ca²⁺influx may result in a competitive advantage. Both a reduction in acetylcholine release 16and the activity-dependent rearrangement of subsynaptic cartels that can occur independently of the presence of a nerve 17may shift the balance between inputs. Accordingly, synchronous stimulation of all inputs converging on a myofiber suppresses elimination of polyinnervation. 18 

Signaling molecules, believed to be of neural origin, regulate the differentiation of a presynaptic nerve terminal and a subsynaptic apparatus. In particular, agrin and neuregulin bind to the synaptic portion of the muscle fiber BL. 19,20 

Nitkin et al.  21originally purified agrin from BL of the synapse-rich electric organ of Torpedo californica  based on its ability to induce aggregates or clusters of nAChRs expressed constitutively in the membrane of cultured myotubes. Colocalized with nAChR clusters were several components of the postsynaptic apparatus as well as acetylcholinesterase, suggesting a role for agrin in the regulation of postsynaptic differentiation. 22However, soluble agrin did not affect nAChR gene transcription.

Molecular cloning showed that agrin is a 200-kd protein that, in its native form, is expressed as a 600-kd heparansulfate proteoglycan. 23–25Splice variants of agrin have different abilities to cluster nAChRs in myotubes. 26,27Specifically, neurally derived agrin cluster nAChRs, whereas isoforms expressed in skeletal muscle, kidney, and blood vessels do not induce myotubes to form nAChR clusters. Functional mapping shows that an 8, 11, or 19 amino acid splice insert within the C-terminal-most 20 kd of agrin is essential to nAChR clustering activity. 28Alternative splicing at the N-terminus results either in a secreted isoform exhibiting strong binding to laminin, i.e. , to the BL, 29or in an isoform that is inserted into the cell membrane and whose function is not known. 30 

A receptor-coupled tyrosine kinase, termed MuSK (for muscle-specific kinase), appears to mediate agrin-induced clustering of nAChRs. 31Mice lacking MuSK and mice lacking agrin have very similar phenotypes. 32,33They lack NMJs, and their motor axons, rather than forming short branches from a central nerve trunk, wander along the entire length of muscle fibers without making synapses. The pathway downstream of MuSK mediating agrin-induced nAChR clustering is not known. Activation of MuSK by agrin phosphorylates nAChR β subunits, but this is not sufficient for clustering. 34 

An important role is played by rapsyn, a 43-kd peripheral cytoplasmic membrane protein that is associated in a 1:1 ratio with the β subunit of synaptic nAChRs. 35When coexpressed with nAChRs, rapsyn causes their clustering. Furthermore, mice lacking rapsyn die at birth because their NMJs do not cluster nAChRs and lack several other components of the subsynaptic membrane and cytoskeleton. 36Synapse-specific aggregation of MuSK, and of synaptic BL components as well as synaptic nAChR gene expression, appear normal, consistent with the idea that MuSK forms a primary scaffold to which other components are attached by rapsyn. Rapsyn may also serve to link components of the signaling pathway activated by agrin. 37 

Studies of mouse mutants suggest roles for several other synapse-specific membrane and cytoskeletal proteins in NMJ development and maintenance. For example, the membrane-spanning dystrophin–glycoprotein complex (DGC), comprising α and β dystroglycans as well as several other components, bind extracellularly to laminin and intracellularly to dystrophin to link the extracellular matrix to the cytoskeleton. This provides mechanical stability to the muscle along its entire length, with mutations causing different forms of muscular dystrophy. 38Agrin and several synapse-specific isoforms of laminin 39as well as rapsyn bind to the DGC. Deletion of distinct components of the synaptic DGC demonstrate the implication of the DGC in the maintenance of the synapse. 40 

Although the molecular signals mediating the selective stabilization of one and the elimination of other nerve inputs to the developing NMJ are not well understood, thrombin derived from muscle prothrombin, the endogenous thrombin inhibitor nexin-1, as well as thrombin receptors may shape these neural inputs. 41,42Evidence supporting this hypothesis is derived from in vivo  43and in vitro  studies. 44The latter suggest that muscle-derived thrombin activates a protease activated receptor (PAR-1). 45This G-protein–coupled receptor may then activate protein kinase C, which leads to reduced insertion and stability of nAChRs at the endplate surface. 46According to the hypothesis of Balice-Gordon and Lichtman, 14endplate areas undergoing loss of nAChRs would also lose neuronal inputs. In addition to thrombin, Ca²⁺-sensitive proteases may shape the NMJ. 47This suggests a role for Ca²⁺influx through the ϵ nAChR in stabilization of the NMJ.

As indicated above, muscle electrical activity down-regulates expression of AChR subunit genes. Therefore, the maintenance of a high concentration of nAChRs in the subsynaptic muscle membrane requires that the nAChR subunit genes α, β, δ, and ϵ are transcribed selectively in subsynaptic myonuclei of innervated, electrically active muscle fibers. A signal from the nerve and bound to the synaptic portion of the BL 48,49appears to regulate this transcription. Neuregulin 1 (NRG-1), 50a product of the nrg-1  gene, is believed to be the nerve signal. By alternative mRNA splicing, this gene codes for a number of growth and differentiation factors with many different functions in development. They all share one epidermal growth factor–like domain that mediates their biologic activity by activating receptor kinases termed ErbBs. ErbB receptors are concentrated in the subsynaptic muscle membrane. 51NRG-1 isoforms are expressed by motor neurons 52as well as muscle fibers. 53NRG-1 isoforms accumulate in the synaptic BL, probably by binding to agrin and other heparansulphate proteoglycans 54that are induced by agrin. 53NRG-1–activated nAChR subunit gene transcription is mediated via  the activation of phosphatidylinositol 3-kinase and mitogen-activated protein kinases 55,56and regulatory elements, termed N-box, in their promoters that are similar to those conferring nerve-induced, synapse-specific expression to reporter genes in vivo . 57–59The DNA binding factors involved are growth-associated binding proteins (GABPα/β), members of the Ets family of transcription factors. 60,61NRG-1 also induces the expression of voltage-gated sodium channels. 62 

Surprisingly, neural agrin alone, when attached to culture substrate or to BL, but not when applied in soluble form, can induce nAChR gene transcription in cultured myotubes or in nonsynaptic muscle region in vivo , respectively. Importantly, this occurs in the absence of a nerve terminal and thus of NRG-1 from neurons. 63,64Neural agrin further induces the formation of a postsynaptic-like membrane exhibiting all the hallmarks of a normal postsynaptic apparatus, including the formation of folds, the accumulation of myonuclei and membrane and cytoskeletal proteins, as well as MuSK, NRG-1, and ErbB receptors 65–67(fig. 2). The inhibition of agrin-induced transcription of nAChR ϵ subunit gene by forced overexpression of inactive mutants of ErbB2 in cultured myotubes is consistent with the idea that agrin organizes a NRG–ErbB receptor pathway that, in turn, activates nAChR gene transcription, with NRG-1 originating from muscle. 53The multiple effects of agrin are all mediated by activation of MuSK. 68Recent experiments suggest that activation of MuSK induces not only the clustering of MuSK and ErbBs, but also of the transcription of their genes (Moore C [Diploma Biology, Basel, Switzerland], Brenner HR [Department of Physiology, University of Basel, Switzerland], unpublished observations, October 2000).

In summary, neural agrin alone acting via  MuSK can organize the induction of a postsynaptic apparatus, including the synthesis of proteins that control the synthesis of other synaptic components. Agrin is therefore the only master organizer of synaptic development to be identified. However, it is not clear whether NMJ development depends on the supply of NRG-1 from motor neurons or whether NRG-1 is supplied by the muscle fiber. Furthermore, other neural factors are likely involved in subsynaptic differentiation, consistent with the observation that nAChR density in ectopic, nerve-free postsynaptic membranes induced by agrin appears lower than at normal synapses (Brenner HR [Department of Physiology, University of Basel, Switzerland], unpublished observations, July 1996).

Little is known regarding the identity of factors affecting presynaptic differentiation, but three candidates with activities on cultured neurons consistent with such roles are present in synaptic BL. Fibroblast growth factor 2, when coated to beads and muscle agrin, stimulates the accumulation of vesicles in cultured neurites. 69–71However, mice lacking muscle agrin have normal NMJs. 72A laminin β chain, β2, in the context of synapse-specific laminin-11, 73stops motor neurite outgrowth 39and, in vivo , prevents glial entry into the synaptic cleft. 73Synapse-specific accumulation of laminin β2 is regulated by neural agrin, 65again mediated via  MuSK activation. 68Neurotrophins secreted by muscle fibers, activating trk  B receptors localized in the synaptic muscle membrane, are required for the maintenance of the postsynaptic nAChR-rich region. 74 

Unlike motor neurons, Schwann cells do express ErbB receptors, and they depend on neuronal NRG-1 for their survival required for the maintenance of motor neurons. 3Consequently, NRG-1 expressed by subsynaptic muscle regions may influence synapse formation indirectly. Consistent with this notion, injection of NRG-1 into neonatal muscle causes a redistribution of Schwann cells, a loss of synaptic sites, and growth of motor neurons throughout the muscle. 75The role of Schwann cells is more obvious during reinnervation after nerve cuts. Terminal Schwann cells sprout processes on denervation. These are used by regenerating motor axons as guides to leave the endplate domains and, driven by factors from denervated fibers, 76reach other synapses. In this way they cause transient polyneuronal innervation of individual endplates. 77 

The function of the endplate nAChR depends on five subunit proteins that combine to form the pentameric unit (fig. 3). The α subunit was the first to be purified. Subsequent analyses of amino acid sequence, as well as accessibility of synaptic nAChRs to ligands, revealed that both the N and C termini of the α-subunit protein protrude beyond postsynaptic membranes into the extracellular space. Repeated clusters of hydrophobic amino acid residues suggested that between its N and C termini, the α subunit formed four membrane-spanning helices, M1 through M4. 78Extensive sequence homology with α facilitated characterization of four additional subunit proteins contributing to nAChR structure. 79,80We now know that the nAChR of adult mammalian skeletal muscle is a pentameric complex of two α subunits in association with a single β, δ, and ϵ subunit. These subunits interact to form a transmembrane pore as well as the extracellular binding pockets for acetylcholine and other agonists or antagonists. The M2 transmembrane-spanning segment of each subunit lines the cation selective pore. 81The extracellular binding sites for acetylcholine and antagonists such as curare form at the interface of the N-terminal domain of the αδ and the αϵ subunits. 82,83In the absence of acetylcholine or other agonists, the stable closed state of this pore normally precludes flow of cations down their electrochemical gradient. A major function of the ϵ and γ subunits is to stabilize this closed state. 84Simultaneous binding of two acetylcholine molecules to a nAChR 85initiates conformational changes that open the pore. 86,87The duration of this open state depends on the duration of dual occupation by acetylcholine.

An interesting, although poorly understood observation is that developing mammalian muscle contains a γ rather than an ϵ subunit (fig. 1). 4Although the role, if any, of the developmentally programmed ϵ for γ subunit shift is not understood, these subunits determine pharmacologic 88,89and physiologic 11properties of the muscle nAChR. Mutations of the human ϵ subunit gene give rise to congenital forms of myasthenia gravis (MG). 90In addition, endplates of mice lacking the ϵ subunit gene begin to degenerate within 2 weeks after birth. This myasthenic condition leads to death within 3 months. 91–93These findings suggest that understanding the role of the ϵ subunit will improve therapeutic management of the healthy and diseased NMJ. Furthermore, such understanding may also clarify the significance of subunit changes to the health of central nervous system synapses, the function of which is mediated by ligand-operated ion channels structurally related to the muscle nAChR. 94,95 

Genes coding for the γ and ϵ subunit map to human chromosomes 2 and 17, respectively. 96The regulation and timing of human γ and ϵ gene expression, as well as the subunit shift of the nAChR, is unexplored. Although the γ to ϵ subunit shift occurs for all mammalian species studied, most information has been acquired from rodents.

Muscle precursor cells of 12-day rat embryos 97express mRNAs encoding α, β, δ, and γ subunits. Formation of the NMJ at embryonic days 15 to 17 initiates accumulation and decline of these mRNAs, respectively, below the junctional and extrajunctional sarcolemma. Two to three days later, mRNA encoding the ϵ subunit is first detectable in subendplate nuclei. During the first 2 weeks after birth, the levels of mRNA encoding the γ and ϵ subunits change in a reciprocal fashion. 98As previously discussed, neural factors control subunit gene expression. In particular, NRG-1 binds to ErbB receptors located on the endplate membrane. 51,99,100This activates tyrosine kinase to phosphorylate GABP α/β, which binds to the promoter sequence of the ϵ subunit gene. A single nucleotide mutation in this promoter sequence reduces its affinity for the GABP. The reduced synaptic specific expression of the ϵ subunit leads to a myasthenic condition in humans 90and the mouse. 58 

Although functioning, mature, ϵ-subunit–containing nAChRs are observed at endplates of 1-day-old rodents 101(McArdle JJ [Professor of Pharmacology and Physiology, Newark, NJ], unpublished observations, November 2000; data provided in the form of an abstract presented to the Society for Neuroscience), substitution for the γ nAChR is gradually completed within 3 weeks after birth. 11,102,103Thus, the ϵ for γ subunit shift occurs during the dynamic phase of synaptogenesis. 1Copopulation of developing or reinnervating endplates with ϵ and γ nAChRs causes endplate currents that have a fast and slow component of decay. Because of its briefer apparent open time, the ϵ nAChR is responsible for the fast component of endplate current decay. 104At the same time, activation of the ϵ nAChR will increase Ca²⁺concentration within the subsynaptic cytoplasm. 105,106Because Ca²⁺is an essential second messenger, the ϵ form of nAChRs may have evolved to allow highly localized Ca²⁺influx to regulate nearby mechanisms that determine the architecture and function of the NMJ. On the other hand, excessive activation of the ϵ nAChR, as during cholinesterase inhibition, 107may overload the endplate with Ca²⁺, which initiates degenerative processes. Similar to Ca²⁺-mediated glutamate neurotoxicity, 108prolonged activation of the ϵ AChR may increase the concentration of Ca²⁺in the cytoplasm below the endplate membrane to activate degenerative processes. 109For example, Ca²⁺-activated calpain, DNase, or phospholipase may degrade molecules essential to synaptic stability. 110As in the case of N -methyl-d-aspartate–induced neurotoxicity, subendplate mitochondria may be stimulated to produce reactive oxygen species that initiate degenerative processes. 111Strong support for a necrotic effect of increased influx through the mature nAChR comes from studies of mutations discovered in patients with slow-channel congenital myasthenic syndromes (SCCMS).

The skeletal muscle weakness and fatigue of SCCMS is associated with degeneration of the motor endplate. 112Diverse mutations of different nAChR subunits contribute to the SCCMS. Initial studies attributed the SCCMS to mutation within the ϵ and β subunits, which slow channel closure in the presence and allow spontaneous openings in the absence of acetylcholine. 113Mutations of the α subunit, which increase the affinity of the nAChR for agonist, decrease the agonist dissociation rate, allowing repeated channel openings. 114The net effect of these gain-of-function mutations is to prolong the open state of the nAChR. This allows what normally is physiologic activation of the NMJ to overload the postsynaptic region with Ca²⁺and initiate necrosis. In addition to the resultant loss of junctional clusters of nAChR, depolarization–desensitization block of the endplate occurs because the prolonged synaptic potentials summate temporally. An open channel blocker of the nAChR, quinidine sulfate, is therapeutically efficacious in SCCMS because it normalizes the open duration of slow channel mutants. 115 

In addition to gain-of-function mutations that contribute to SCCMS, the α and ϵ subunit demonstrate loss-of-function mutations that contribute to another congenital myasthenic syndrome. 113,116These mutations decrease the rate of channel opening and increase the closure rate. This loss of nAChR function reduces the safety factor for synaptic transmission. Just as for ϵ subunit knockout mice, the endplate region is simplified in patients with the loss-of-function mutations. However, in contrast to ϵ knockout mice, expression of the γ subunit is up-regulated in the human condition. This up-regulation preserves the human phenotype. It is interesting to note that up-regulation of the γ subunit does not occur in autoimmune MG. 117 

In addition to the physiologic consequences summarized above, subunit mutations also modify the pharmacologic sensitivity of the nAChR. A striking example is the choline sensitivity of nAChRs in SCCMS. 118Normal nAChRs do not respond to plasma concentrations of this ordinary metabolite, but mutated nAChRs in SCCMS are activated. Such activation worsens the cationic overload of the motor endplate, which is responsible for endplate degeneration in the SCCMS. Recent evidence suggests that nitric oxide synthase inhibitors may have the potential to provide therapeutic benefit in SCCMS. 119 

In view of the Overton-Meyer lipid theory of general anesthetic action, it is useful to consider biochemical studies suggesting an influence of membrane lipids and cholesterol on the function of the nAChR. Early biochemical studies suggested an influence of cholesterol on the function of the reconstituted nAChR. 120,121The isolated nAChR has a particularly high affinity for cholesterol. 122Furthermore, functional insertion of isolated nAChRs into artificial membranes requires cholesterol. 123The postsynaptic membrane is rich in cholesterol. 124These observations suggest novel posttranslational processing of newly synthesized nAChRs. Only after nAChRs are inserted into the postsynaptic membrane and charged with cholesterol do they become fully active. 125Cells deficient in sphingolipid biosynthesis are unable to insert normal concentrations of nAChR into their membrane. 126Reduction of membrane cholesterol dramatically increases the input resistance of muscle fibers, allowing for greater endplate depolarization in response to acetylcholine. 127,128 

The effect of cholesterol on nAChR function is not attributable to an action on the bulk lipid of the membrane. 129Rather, cholesterol may interact with either nonannular sites within subunits of the nAChR that are not part of the lipid–protein interface, 122or with the immobilized lipid-belt region surrounding the nAChR. 130The α M1 and M4 transmembrane domains and the γ M4 domain appear to form the cholesterol “binding” domain. 131The lipid-soluble steroid promegestone 132and organochlorine insecticides 133may noncompetitively block the nAChR by acting at these protein–lipid interfaces. Furthermore, amino acid substitutions in the vicinity of the protein–lipid alters channel gating kinetics. 134The specificity of such putative sites for cholesterol is apparently not high since other neutral lipids maintain nAChR function. 135 

It is generally accepted that the synthesis and release of acetylcholine involves a cycle of events (fig. 4). Acetylcholine is first formed in the cytoplasm of the nerve terminal from acetyl coenzyme A and choline in a reaction catalyzed by the soluble enzyme choline acetyltransferase. An energy-dependent “transporter” then accumulates acetylcholine within vesicles. Acetylcholine is packed at superosmotic concentrations (approximately 300 mm) within the lumen of the vesicle, together with adenosine triphosphate (ATP), proteoglycans, H⁺, Mg²⁺, and Ca²⁺ions. 136The acetylcholine:ATP molar ratio in synaptic vesicles has been estimated to range from 10:1 to 1:1. 137,138Each vesicle appears to contain 5,000–10,000 molecules of acetylcholine. The acetylcholine contained in a single vesicle is often referred to as a “quantum” of transmitter. Release of acetylcholine is a Ca²⁺-dependent process and is triggered by an increase in the concentration of free Ca²⁺within the nerve terminal. This results from the opening of voltage-gated Ca²⁺channels by the depolarization of the nerve impulse. In addition to Ca²⁺channels, several forms of potassium channel are present in the nerve terminal, including voltage-gated and Ca²⁺-activated potassium channels. The potassium channels are likely to limit the duration of nerve terminal depolarization and hence the extent of Ca²⁺entry and transmitter release. In addition to acetylcholine, ATP is also released and subsequently hydrolyzed within minutes to adenosine in the junctional cleft. 137Adenosine in the cleft binds to prejunctional P¹purinoceptors, 139which depress neuromuscular transmission via  a G-protein–mediated Ca²⁺channel inhibition. 140P²purinoceptors, sensitive to ATP but not to adenosine, have been identified in the muscle.

Synaptic vesicle (SV) exocytosis occurs in successive steps: docking vesicles attach to the presynaptic active zone. SVs then undergo a priming reaction to become capable of responding to a Ca²⁺signal. The action potential causes membrane depolarization and a sharp increase in internal Ca²⁺concentration through voltage-gated Ca²⁺channels and direct release from intracellular Ca²⁺stores. This Ca²⁺signal triggers the fusion of SVs with the presynaptic membrane and subsequently exocytosis. The sequence of exocytosis is very rapid (< 0.3 ms). Fusion results in the release of a “quantum” of several thousand acetylcholine molecules into the synaptic cleft. The synaptic cleft is very narrow (around 50 nm), and acetylcholine can diffuse this distance in a few microseconds to reach the postsynaptic membrane. Some of acetylcholine molecules bind to the nAChRs on the postsynaptic membrane, while the rest are rapidly hydrolyzed by the acetylcholinesterase present in the synaptic cleft to choline and acetate. Choline is recycled into the terminal by a high-affinity uptake system, making it available for the resynthesis of acetylcholine. The hemicholinium-3 inhibits the later mechanism. After exocytosis, the membrane components of the SVs are recovered by endocytosis and recycled for future use.

The released acetylcholine binds to α subunits of the AChRs. These ligand-gated cation channels allow sodium to enter and depolarize the muscle cell membrane at synaptic sites. This local depolarization leads to the activation of nearby voltage-gated sodium channels, which amplify and propagate action potentials across the surface of the muscle fiber and into the transverse tubules where Ca²⁺channels are present at high density. 141The dihydropyridine receptors (DHPRs) in the transverse system membrane act as voltage sensors, detecting the depolarization and opening adjacent type-1 ryanodine receptor (RyR1)–Ca²⁺-gated, Ca²⁺-release channels in the apposing sarcoplasmic reticulum membrane by some protein–protein interaction. 142,143DHPR-RyR1 coupling is not yet fully elucidated. 144Several endogenous effectors, such as Ca²⁺, Mg²⁺, adenine nucleotides, calmodulin, and nitric oxide, are known to regulate RyR1 function. 143,145After DHPR-RyR1 coupling, the RyR1 release large amounts of Ca²⁺from the sarcoplasmic reticulum, resulting in muscle contraction. The translation of electrical signaling at the surface membrane into intracellular Ca²⁺release from the sarcoplasmic reticulum is known as excitation–contraction coupling. 146Binding of Ca²⁺to the troponin complex alters the interactions between tropomyosin and the contractile machinery, allowing the proper interaction between actin molecules and myosin heads. Thus, muscle contraction occurs via  myofilament sliding. As the sodium channel openings subside, chloride enters the cell through more slowly opening voltage-sensitive chloride channels, to return the muscle membrane potential to its resting level (approximately −70 to −90 mV). 141Molecular leakage (nonquantal leakage) and quantal leakage of acetylcholine from the nerve are events that are unrelated to nerve impulse. 147 

At the NMJ, SVs are specialized secretory organelles used for fast signaling between nerve and muscle. There are two pools of vesicles, a readily releasable store (active pool) and a reserve store. Electron microscopic studies demonstrate that the majority of SVs are sequestered in the reserve pool in a filamentous network believed to be composed mainly of actin, synapsin (an actin-binding protein), and spectrin (fig. 4). 148,149Synapsin I binds vesicles to the presynaptic cytoskeleton (actin filaments and microtubules). 148Mice lacking synapsins are viable and fertile with no gross anatomic abnormalities, but they are prone to seizures and are unable to properly regulate synaptic transmission. For example, repetitive stimulation of synapses at physiologic frequencies result in massive synaptic depression. 148,150This suggests that the SV cycle is unable to mobilize appropriately during repetitive stimulation. 148,150 

The synaptic vesicles possess a diverse set of specialized proteins that can be divided into two functional classes: proteins involved in the uptake of neurotransmitters (transport proteins) and proteins that mediate SV membrane traffic such as docking, fusion, and budding. 151It is believed that intrinsic and peripheral membrane proteins of SVs are imported from the cell body via  axonal transport. 152Sudhof 151developed a structural model of the vesicle membrane (fig. 4). Although many proteins have been implicated in the process of exocytosis, the overall mechanism is still not completely understood.

After an action potential and Ca²⁺influx, phosphorylation of synapsin I is activated by cyclic adenosine monophosphate–dependent protein kinase and by calcium-calmodulin activated protein kinases I and II (fig. 4). 153This weakens binding between SVs and the cytomatrix, allowing mobilization of SVs from the reserve pool into the active pool lying close to the plasma membrane. SVs then attach to the presynaptic plasma membrane in a process known as docking. Synaptotagmins, synaptophysins, and the SV associated membrane protein (VAMP, or synaptobrevin) are integral vesicular membrane proteins involved in the docking process of SVs within a specialized region termed the active zone. The active zone is characterized by the presence of electron-dense regions on both the presynaptic and postsynaptic plasma membrane that contain clusters of Ca²⁺channels. 154 

Synaptotagmin I is believed to be the main Ca²⁺-binding protein, and it has the ability to bind multiple Ca²⁺ions. 155Synaptotagmin I is involved in localizing SVs to synaptic zones rich in voltage-gated Ca²⁺channels 156or stabilizing vesicles in the docked state at the presynaptic membrane. 157Synaptotagmin I is therefore essential for the fast component of neurotransmitter release. 155Mice deficient in synaptotagmin I lack fast, but not slow neurotransmission. 155 

The formation of a core complex of three synaptic proteins (the SNARE complex) is an essential step for the docking process. Two of these proteins are from the plasma membrane: SNAP25 (synaptosome-associated membrane protein of 25 kd) and syntaxin 1 (or HPC1). The third protein is from SVs (synaptobrevin) (fig. 4). 151,158The core complex forms the anchor for a cascade of protein–protein interactions required for exocytosis to occur. However, controversy exists as to which proteins function in docking, fusion, or both. 159Recent evidence suggests that the SNARE complex is perhaps only one of several protein complexes involved in vesicle targeting and fusion. Synaptotagmin I also interacts with the plasma membrane proteins syntaxin 160and neurexins. 161 

The SV proteins are common targets for environmental toxins. The neurexins include one of the receptors for α-latrotoxin (black widow spider venom), a toxin that causes massive neurotransmitter release. Cleavage of SNAP25, syntaxin 1, or synaptobrevin by clostridial neurotoxins (which include tetanus and botulinum toxins) results in inhibition of exocytosis. 159Botulinum toxins are zinc endoproteases that are used clinically for treatment of muscle dystonia and for spastic disorders. Since approval of type-A botulinum toxin by the US Food and Drug Administration in December 1989 for three disorders (strabismus, blepharospasm, and hemifacial spasm), the number of indications for its use has increased greatly and now includes numerous focal dystonias, spasticity, tremors, cosmetic applications, and migraine and tension headaches. 162Treatments can be repeated several times without major side effects, such as the development of an immune response. Synaptotagmin is not a known substrate for any neurotoxin, but it may be targeted by antibodies found in Lambert-Eaton myasthenic syndrome (LEMS). 163 

Further “mutation” or “priming” events are required to convert a docked vesicle into a fusion-competent, readily releasable vesicle. At the priming stage, the system becomes competent to undergo fusion on an increase in Ca²⁺concentration. A family of low-molecular-weight guanosine triphosphate–binding proteins, termed rabs, are involved in vesicle attachment to acceptor membranes. 164Rab3A is required to maintain a normal reserve of SVs, to facilitate accelerated exocytosis during repetitive stimulation when SV recycling becomes rate limiting. Triggering SV exocytosis leads to dissociation of rab3A from SVs. This dissociation is inhibited by botulinum and tetanus toxins. In mice lacking rab3A, synaptic transmission persists but is more susceptible to fatigue and is less plastic, a phenotype consistent with altered vesicle availability at active zones. 164 

A fundamental step in synaptic transmission is the fusion of SVs with the plasma membrane and the release of their content. Fusion occurs within a few hundred microseconds of Ca²⁺entering the nerve terminal via  presynaptic voltage-gated Ca²⁺channels. 165Ca²⁺triggers exocytosis by participating in one or more reactions that catalyze vesicle fusion. Recent evidence suggests that vesicle fusion is mediated by two proteins with opposite actions: synaptotagmin, which probably serves as the Ca²⁺sensor, 151and rab3, which limits the number of vesicles that can be fused as a function of Ca²⁺to allow a temporally limited, repeatable signal. However, it is not yet clear how vesicle fusion is triggered by Ca²⁺-bound synaptotagmin, and it is possible that one of the several proposed interactions with SNARE proteins could be important. 166At the NMJ, the release of acetylcholine contained inside one vesicle causes a miniature endplate potential. These miniature endplate potentials have small amplitudes (0.5–1 mV) that are normally insufficient to trigger action potentials. A nerve impulse causes the release of approximately 20–200 quanta, depending on the species, within a fraction of a millisecond. The endplate potential is generated by electrical summation of many miniature endplate potentials synchronously discharged from the active zones. The peak amplitude of the endplate potential is 15–20 mV.

After fusion, the SV membrane is recovered via  endocytosis. However, because exocytotic vesicle membranes contain unique proteins, endocytosis must retrieve them selectively. Three mechanisms have been proposed. 167The first mechanism suggests that endocytosis in nerve terminals is based on a membrane-budding process that requires the formation of coated pits and coated vesicles. This seems to involve some sort of coating protein that is widely assumed to be clathrin or “accessory” proteins (dynamin, endophilin, and synaptojanin). Many of these proteins have now been characterized in considerable detail. 168Synaptotagmin also appears to be involved in this proposed mechanism. After pinching off the membrane, the clathrin-coated vesicles uncoat and another cycle starts again. The second mechanism also proposes that clathrin-coated vesicles transit through endosomes and other intermediates, from which functional SVs are then formed. The third proposed mechanism is the “kiss-and-run” hypothesis. It attempts to explain the rapid retrieval of SVs after exocytosis. According to this hypothesis, the SVs empty within fractions of a millisecond as their low-molecular-weight contents escape through the fusion pore. The fusion pore then closes, the vesicle reaccumulates transmitter from the cytoplasm, and is once again ready to participate in synaptic transmission. 169This model implies that vesicles do not lose their identity during exocytosis and that a new vesicle is formed by the rapid reclosure of a transient fusion pore. 170SVs then accumulate acetylcholine by active transport. Recent data suggest that kiss-and-run operates in parallel with the classic coated-vesicle recycling. 171Recycling SVs appear to be incorporated into the releasable pool from which they have roughly the same probability of release as the preexisting vesicles. 172The entire SV cycle takes approximately 1 min. 173 

At the NMJ, acetylcholinesterase (enzyme classification 3.1.1.7) is a type-B carboxylesterase enzyme responsible for rapid hydrolysis of released acetylcholine, thereby controlling the duration of receptor activation. 174Approximately 50% of the released acetylcholine is hydrolyzed during the time of diffusion across the synaptic cleft before reaching nAChRs. The efficiency of acetylcholinesterase depends on its fast catalytic activity. Acetylcholinesterase ranks as one of the highest catalytic efficiencies known. It can catalyze acetylcholine hydrolysis (4,000 molecules of acetylcholine hydrolyzed per active site per second) at near diffusion-limited rates. 174The active site lies near the bottom of a deep and narrow cleft that reaches halfway into the protein. Acetylcholine must enter this cleft in the enzyme that is blocked by a mobile ring of molecules more than 97% of the time. Molecular dynamics simulations showed that the entrance to the cleft opens and shuts so frequently that any acetylcholine molecules lingering nearby have ample chances to diffuse in. 175The molecular dynamics simulations also showed that the motions of the channel extend from the region outside the acetylcholinesterase enzyme to the active site. These fluctuations in the width of the channel are required to allow acetylcholine to move from the outside into the active site. They also contribute to the selectivity of the enzyme, by slowing the entrance of substrates that are larger than acetylcholine. 175 

Acetylcholinesterase is highly concentrated at the NMJ but present in a lower concentration throughout the length of muscle fibers. 176In mammals, acetylcholinesterase is encoded by a single gene. It has been localized to chromosome 7q22 in humans. 177Much of the acetylcholinesterase at the NMJ occurs in the asymmetric or A12 form consisting of three tetramers of catalytic subunits covalently linked to a collagen-like tail. Asymmetric acetylcholinesterase is bound to the junctional BL. 178The distribution of acetylcholinesterase molecules on the synaptic BL closely matches the distribution of nAChRs. 179 

Acetylcholinesterase is regulated, in part, by muscle activity and by the spontaneous or nerve-evoked depolarization of the plasma membrane. 180Fast muscles express severalfold higher levels of acetylcholinesterase activity than slow muscles, and this is correlated with the relative abundance of acetylcholinesterase mRNAs in these muscles. Drugs that block membrane depolarization, such as the sodium channel antagonist tetrodotoxin, decrease accumulation of acetylcholinesterase. 181In contrast, sodium channel agonists such as veratridine dramatically increase acetylcholinesterase assembly. 182After denervation, there is a large decrease in the density of acetylcholinesterase molecules at the NMJ that can be restored by electrical stimulation of the denervated muscles or by their reinnervation either at the original 183or at ectopic sites. 184 

In addition to hydrolysis of acetylcholine, acetylcholinesterase has other functions such as nerve growth-promoting activities 185and modulation of nAChRs. 186 

The importance of the enzyme is illustrated by the following conditions. Congenital acetylcholinesterase deficiency results in a disabling congenital myasthenic syndrome. 187This subset of congenital myasthenic syndrome is caused by genetic defect in the collagenic tail of acetylcholinesterase that attaches the enzyme to the BL of the endplate. 187On the other hand, inhibition of the enzyme, e.g. , by nerve gas, results in prolonged exposure of nAChR to acetylcholine, causing desensitization of nAChR and a depolarization block at physiologic rates of stimulation. 188Chronic fatigue is a symptom of Gulf War syndrome, a disorder proposed to result from exposure to acetylcholinesterase inhibitors. 189Partial inhibition of acetylcholinesterase, e.g. , by overexposure to insecticides, results in excessive influx of Ca²⁺through the nAChRs ion channel, which leads to local necrotic myopathy and an endplate myopathy. 107Oximes are clinically important reactivators of acetylcholinesterase that can prevent these degenerative effects of insecticide intoxication. 190Nevertheless, acetylcholinesterase inhibitors are therapeutically useful for antagonism of residual neuromuscular block and for symptomatic treatment of patients with MG.

The active surface of the acetylcholinesterase is best viewed as having two subunits, the anionic site and esteratic site. 191The anionic site is concerned with binding and orienting the substrate molecule. 191The esteratic site is responsible for the hydrolytic process. 191A second “anionic” site, which became known as the “peripheral” anionic site, was proposed based on binding of bis-quaternary ammonium compounds. 192Binding of ligands to the peripheral anionic site causes inactivation of the enzyme, although the mechanism of inhibition is not clear. There is also evidence for a role of the peripheral anionic site of acetylcholinesterase in neurite regeneration and outgrowth and in the growth and differentiation of spinal motor neurons. 193 

Neostigmine and edrophonium are the most commonly used anticholinesterases in the operating room. Edrophonium is a prosthetic inhibitor that binds to the anionic site on the acetylcholinesterase by electrostatic attachment and to the esteratic subsite by hydrogen bonding. The dissociation half-life of this reaction is less than 0.5 min. 194The in vivo  activity of edrophonium is predicted to be rapid in onset, and, clinically, edrophonium has a more rapid onset of action than neostigmine. 195Neostigmine and pyridostigmine are oxydiaphoretic (acid transferring) inhibitors of acetylcholinesterase. Neostigmine and pyridostigmine transfer a carbamate group to the acetylcholinesterase, which forms a covalent bond at the esteratic site. The dissociation half-life of the carbamate-enzyme bond of neostigmine is at least 7 min. 194However, it should be noted that the pharmacologic actions of neostigmine and edrophonium are not limited to enzyme inhibition. 196,197Evidence suggests that the direct influences of the acetylcholinesterase inhibitors on neuromuscular transmission independent of enzyme inhibition involve at least three distinct, although possibly interacting mechanisms: (1) a weak agonist action, (2) the formation of desensitized receptor-complex intermediates, and (3) the alteration of the conductance properties of active channels.

Aging is associated with progressive decrease in skeletal muscle mass and strength (sarcopenia) caused by reduction of anabolic hormone concentrations, decline in muscle protein turnover, and other neuromuscular alterations. 198Between 20 and 80 yr of age, the cumulative decline in skeletal muscle mass amounts to 35–40%. The loss of muscle mass is not associated with weight loss because of a corresponding increase in fat. 199Loss of muscle mass, particularly the preferential loss of type II fibers, 199results in diminished strength and power-generating capacity. 200This has been attributed to structural changes in myosin caused by protein oxidation. 201The issue of whether skeletal muscle oxidative capacity declines with age remains controversial. The aging process also includes a slowing of time and rate of relaxation of skeletal muscle probably caused by decreased rates of maximal sarcoplasmic reticulum Ca²⁺uptake and sarcoplasmic reticulum Ca²⁺-ATPase activity. 202Although the loss of muscle mass associated with aging may be of multifactorial etiology, it is modifiable through resistance training.

In the soleus muscle of old mice, SV density declined to 32% of adult values. 203However, no electromyographic decrements were seen at trains of 10 Hz stimulation. 204The decreased SVs density was accompanied by an increase in the quantal content of transmitter release in the soleus (but not diaphragm or sternomastoid muscles) of old mice. 205The rate constant of transmitter turnover in old mice was also found to be more than twice that in adults. 206The increased transmitter turnover seems a compensation for diminished SVs. With increasing age, an increase of the number of RyR1 uncoupled from DHPR has been found in humans. 207Uncoupling of DHPR-RyR1 leads to a significant reduction in the amount of releasable Ca²⁺in skeletal muscles from old humans.

As mentioned previously (see Signals from the Nerve), the exchange of trophic factors by motor neurons and muscle fibers maintains the NMJ. Neurotrophic factors (muscle-derived trophic factors acting on motor neurons) and myotrophic factors (motor–neuron-derived trophic factors acting on muscle fibers) may play a role in the generation of secondary myotubes and the maturation of NMJs during development. 208It has been suggested that the expression of the trophic factors and their receptors (trk  B) might be altered with age, resulting in synaptic dysfunction and cell death. 209TrkB is a family of transmembrane proteins composed of a tyrosine kinase that serve as receptors for brain-derived neurotrophic factor, neurotrophin-3, and neurotrophin-4/5. Knockout mice lacking brain-derived neurotrophic factor or neutrotrophin-3 did not exhibit significant motor neuron loss, but mice lacking trk  B had significant reduced numbers of motor neurons. 210 

Aging is associated with a reduction in total muscle fiber number. A substantial selective atrophy of fast, glycolytic type II fibers was observed with aging. 199It is believed that type II fibers have a reduced reinnervation capacity compared with type I fibers. The specific force developed by both fast- and slow-twitch single intact muscle fibers declines with aging, and more significantly in the former. 211In humans (aged > 60 yr), reduction in number of α-motor neurons and their myelinated axons were observed in lumbar ventral roots. 212Axonal atrophy is probably caused by a reduction in the expression and axonal transport of cytoskeletal proteins in the peripheral nerve. For NMJs of humans, aging is associated with a decrease in motor unit numbers. 213In some junctions, motor neurons regenerate by sprouting and formation of new sites. 209The newly formed synaptic sites appear to be unstable, with many disappearing within several weeks. 209Greater complexity of terminal arborization is observed in the elderly NMJs than those in the adult. 214Areas of axonal contact become progressively more scarce with advancing age, leading to a reduction of the effective area of synaptic contact in the NMJ. This can result in a decline in the trophic interaction of nerve and muscle and in impairment of stimulus transmission. 215The repeated cycles of retraction and compensatory outgrowth probably represent the altered balance between degeneration and regeneration of nerve terminals. 204The greater complexity of terminal arborization may reflect an adaptive and reactive response at the NMJ in an attempt to preserve synaptic area and to compensate for the loss of adhesion of nerve terminals to the synaptic matrix or surrounding Schwann cells. 19The cellular mechanisms underlying these changes are unclear, although a deficiency of actin has been implicated by some investigators. 204 

The adaptive process to aging at the NMJ includes increase of transmitter release despite reduced supply of synaptic vesicles, functional reactive sprouting after partial denervation, and maintenance of nerve terminal integrity in the face of increased outgrowth and retraction. 204Although function may be initially preserved, the increasing extent of adaptation means a progressively more fragile system. Increased fragmentation and loss of active synaptic areas can lead to deterioration of NMJ structure and function. Therefore, the capacity of skeletal muscle to generate force declines with age. 216In the elderly, the diaphragm undergoes significant reduction in specific force. 217This would increase the workload on the diaphragm. 218 

Not only the functional changes at the NMJ but also the multitude of physiologic changes that accompany the aging process (decreases in total body water, glomerular filtration and renal blood flow, liver mass and splanchnic blood flow, and serum albumin concentrations, and increases in fat) affect the action of neuromuscular blockers in the elderly. The onset of nondepolarizing neuromuscular blockers is delayed in the elderly compared with the young. This has been attributed to slower biophase equilibration. 219However, there have been conflicting reports of the pharmacodynamics and pharmacokinetics of neuromuscular blockers in the elderly. Ornstein et al.  220reported minor differences in the pharmacokinetics of cisatracurium in elderly patients that were not associated with alterations in recovery after a single dose of cisatracurium. Other investigators noted that the duration of action of mivacurium was prolonged in the elderly by approximately 30% as compared with young adults. 221A decrease in butyrylcholinesterase activity may be the reason for the longer duration of action of mivacurium in the elderly.

Rupp et al.  222noted that elderly patients had significantly decreased plasma clearances and volumes of distribution of vecuronium, whereas elimination half-life and recovery index were not different when compared with that of their younger counterparts. In contrast, other investigators reported that both spontaneous recovery 223,224and elimination half-life of vecuronium were prolonged and plasma clearance of vecuronium was reduced in older versus  younger patients. 224Similar results were reported with rocuronium; however, there was no difference in the elimination half-lives between the two groups. 225 

It appears that the prolonged duration of action of neuromuscular blockers in the elderly patients is secondary to altered pharmacokinetics. The pharmacokinetics and pharmacodynamics of compounds primarily dependent on spontaneous degradation via  Hofmann elimination (for example, cisatracurium) are not markedly affected by advancing age. In contrast, the action of steroidal neuromuscular blockers, agents dependent on organ elimination, is prolonged in the elderly.

The duration of action of neostigmine and pyridostigmine is reported to be prolonged in the elderly, probably because of reduction in plasma clearance. 226,227However, it has been shown that a greater dose of neostigmine is required in the elderly than in the young to produce adequate antagonism of neuromuscular blockade. 228Aging is associated with a prolongation of the elimination half-life and a reduction of the plasma clearance that resulted in higher plasma concentration of edrophonium. 229However, this was not reflected in an increased duration of antagonism in the elderly as compared with younger patients.

The physiology and pharmacology of the NMJ is pivotal to many aspects of the practice of anesthesiology, including intraoperative care, intensive care unit (ICU) treatment, and pain management. The complexities of normal neuromuscular transmission described above are altered in many pathologic states. The NMJ can be affected by reduced central neuronal activity in spinal cord trauma, stroke, and states of prolonged inactivity. Deficient primary motor neuron activity in Guillain-Barré syndrome (GBS) and amyotrophic lateral sclerosis (ALS) causes changes in the neuromuscular unit. Diseases such as the Lambert-Eaton syndrome, a myasthenic syndrome, as well as exogenously administered magnesium and certain antibiotics, result in reduced presynaptic release of acetylcholine. MG and rare congenital nicotinic channelopathies produce postsynaptic abnormalities of skeletal muscle receptor function. Ion channel dysfunction in skeletal muscle has emerged as pivotal in understanding the etiology of neuromuscular disorders. For instance, sodium and chloride channelopathies are now linked to myotonia and periodic paralysis. Mutations of Ca²⁺channels at the sarcoplasmic reticulum have been identified in some cases of malignant hyperthermia. 230The progress in the fields of molecular genetics and cellular electrophysiology has changed the traditional clinical classification of the large and heterogeneous group of neuromuscular disorders.

This subject was reviewed in 1992 by Martyn et al . 231A classic principle of pharmacology suggests that decreased exposure to an agonist results in postsynaptic receptor up-regulation (increases in number), whereas increased agonist exposure results in receptor down-regulation (decreases in number). 231Therefore, diseases that cause reduced neuronal input will result in an up-regulation of nAChRs in skeletal muscle (table 1). In conditions of primary myopathy, including many of the muscular dystrophies, there is an increase in the number of postsynaptic nAChRs on the basis of chronic muscle regeneration. Nicotinic receptor up-regulation is complicated by the existence of two forms of nAChRs (mature and fetal nAChRs) in muscle tissue (as discussed in Development of the Neuromuscular Junction and The γ to ϵ Subunit Shift). Up-regulation of nAChRs, found in states of functional or surgical denervation, is characterized by the spreading of fetal type (α²βδγ) receptors at extrajunctional sites. 232This is noted within 48 h after partial denervation. 233Furthermore, after denervation, in addition to the usual mature isoform of the Na⁺channel, an immature isoform of the Na⁺channel is expressed on the muscle membrane. 234The fetal-type nAChRs are resistant to nondepolarizing neuromuscular blockers and more sensitive to succinylcholine. 235When depolarized, the immature isoform has a prolonged open channel time that exaggerates the K⁺efflux. 234A positive correlation was found between the number of nAChRs and the intensity of the hyperkalemia after administration of succinylcholine. 236In contrast, reduced expression of the postsynaptic nAChR results in resistance to depolarizing and sensitivity to nondepolarizing neuromuscular blockers. 237 

Nevertheless, there exist multiple reports in the clinical literature of “increased sensitivity” to nondepolarizing neuromuscular blockers in patients with actual or functional denervation. 238,239The resolution to this apparent paradox likely lies in the well-studied margin of safety for neuromuscular transmission. The later is defined as the fraction of AChRs that could be pharmacologically blocked before action potential generation was prevented. 240Normally, the twitch response is not reduced unless more than 70% of the receptors are occupied by a nondepolarizing relaxant. 240These “extra” receptors insure the remarkable fidelity of neuromuscular transmission. Unlike the healthy patient, the patient with functional denervation may have preexisting clinical or subclinical weakness and a reduction in the margin of safety for neuromuscular transmission. In such a patient, inhibition of even 10% of postsynaptic nAChRs by a small dose of a nondepolarizing muscle relaxant may result in clinically detected weakness. Thus, despite up-regulation of fetal-type nAChRs and an associated reduction in the potency of nondepolarizing neuromuscular blockers, complete reversal of neuromuscular blockade is required for adequate respiratory function in these patients.

Spinal cord trauma and stroke are associated with muscle weakness or paralysis based on the dysfunction of central motor neurons. Degeneration of the α-motor neuron results from central malfunction, most likely because of a loss of trophic factors. 241Reduced exposure to acetylcholine results in up-regulation of the immature form of the nAChR. Up-regulation of extra junctional fetal nAChRs is associated with resistance to nondepolarizing neuromuscular blockers and increased sensitivity to succinylcholine and susceptibility to hyperkalemia. 242 

The period of vulnerability to succinylcholine-induced hyperkalemia has not been defined. Based on several case reports, the onset of the hyperkalemic response ranges from 1 week to several months. 243,244Because nAChR up-regulation occurs within 48 h after partial denervation, 233succinylcholine appears to be safe within the first 24 h after the insult. In one report, succinylcholine-induced hyperkalemia persisted for a period of up to 6 months in patients with upper motor neuron lesions. 245With recovery after stroke or cord section, the exaggerated response to succinylcholine is not likely to diminish until after resistance to nondepolarizing relaxants has reverted to normal. 246Up-regulation of the immature form of nAChRs may persist indefinitely, depending on the degree of denervation and renervation. Recent work demonstrated an increased hyperkalemic response to succinylcholine more than 1 yr after injury. 247Because the safe period varies depending on the degree of abnormal nAChR expression and other factors, it is best to avoid succinylcholine in these patients if possible.

Prolonged immobility, in which individuals are confined to wheelchairs or beds, is associated with muscle atrophy secondary to disuse. In contrast with upper or lower motor neuron disease, the nerves themselves are not damaged. The decreased protein synthesis, increased protein degradation, muscle atrophy, decreased glucose uptake, and apoptosis observed after muscle disuse or immobilization have been attributed to decreased insulin action and defective insulin signaling via  phosphatidylinositol 3-kinase. 248The latter is a key signaling molecule that is needed for the anabolic actions of insulin.

Despite the presence of an intact motor neuron, extrasynaptic nAChRs develop with some of the characteristics of immature nAChRs. 249There is resistance to nondepolarizing neuromuscular blockers and increased sensitivity to acetylcholine and succinylcholine. 250Resistance to nondepolarizing neuromuscular blockers was noted 4 days after complete immobilization in dogs. 231Administration of succinylcholine resulted in hyperkalemia and cardiac arrest and death. 251After remobilization, changes at the NMJ revert to normal within 20–50 days. 231 

Studies on the effect of single-limb immobilization in animals showed that the increased response to nondepolarizing neuromuscular blockers was not only noticeable in the immobilized limb, but also in the other unaffected limbs. 252The diaphragm, however, was not affected. 252Another interesting finding was that the potassium release after succinylcholine was significantly increased in beagles who had one limb immobilized by casting. 253This increase required 14–42 days to become apparent. 253 

In the case of total-body immobilization, the onset of vulnerability to succinylcholine-induced hyperkalemia has not been well defined. Death caused by hyperkalemic cardiac arrest after the administration of succinylcholine was reported in one patient 5 days after immobilization. 251For this reason, it is probably best to avoid succinylcholine when total-body immobilization exceeds 24 h. Data regarding single-limb immobilization are less conclusive, and the reported response to succinylcholine in the literature should be interpreted in relation to both the etiology and duration of the immobilization. 253 

Syndromes of weakness in critically ill patients are relatively common and likely polymorphic in origin. In a retrospective study of 92 critically ill patients with clinically diagnosed weakness, electromyographic studies indicated that 43% of the patients suffered from myopathy, wheresa 28% suffered from peripheral neuropathy. 254Weakness can lead to prolonged weaning from the ventilator and increased time for rehabilitation. 238,254Myopathy may be either caused by immobility discussed above or the catabolism associated with negative nitrogen balance. 238In addition, myasthenia-like syndromes are also seen in critically ill patients. Evidence for local immune activation by cytokine expression in the skeletal muscle was reported in patients with critical illness polyneuropathy and myopathy. 255Furthermore, the presence of antibodies to nAChR associated with decreased number of nAChRs and increased sensitivity to d-tubocurarine has been demonstrated in a rodent model of subacute or prolonged sepsis. 256Three main types have been identified: critical illness myopathy, myopathy with selective loss of myosin filaments, and acute necrotizing myopathy of intensive care.

The polyneuropathy seen in the critically ill has been termed “critical illness polyneuropathy.” Critical illness polyneuropathy is a diffuse axonal polyneuropathy and occurs in 50–70% of patients with multisystem organ failure and sepsis. 257Recovery from critical illness polyneuropathy can be rapid and complete when the patient survives the critical illness. 238,257There may be a role for humoral factors associated with multisystem organ failure, but the etiology is likely multifactorial. 238,257Prolonged use of neuromuscular blocking agents singly and in association with glucocorticoids 258may have toxic effects on motor axons, but the results of studies so far have been inconclusive.

It is likely that up-regulation of nAChRs induced by immobilization and chronic neuromuscular blockade contributed to the cardiac arrest associated with the use of succinylcholine in ICU patients. 251,259As most critically ill patients are immobilized, it is impossible to determine whether weakness is caused by immobility, polyneuropathy, or myopathy of critical illness without pathologic diagnosis. Nevertheless, as succinylcholine can cause hyperkalemia in any of these syndromes, it is best to avoid succinylcholine in ICU patients in whom total-body immobilization exceeds 24 h.

Several reports have implicated nondepolarizing neuromuscular blocking drugs to cause generalized weakness after their long-term administration (to ICU patients) that required recovery periods from 2 days to 6 months. 260However, it is not clear whether neuromuscular blockers were a precipitating factor since other possible contributing conditions were frequently present, e.g. , polyneuropathy of critical illness, disuse atrophy, renal failure, aminoglycoside, and steroid administration. 257,258,260,261A clinical impression has been reported that prolonged recovery from neuromuscular block occurs more frequently when steroidal neuromuscular relaxants are used. 260Prolonged neuromuscular block has been associated with renal failure and increased serum concentrations of the active metabolite of vecuronium, 3-desacetylvecuronium. 260Although corticosteroids are not thought to be a risk factor for polyneuropathy of critical illness, when administered with vecuronium, both in vivo  and in vitro , inhibition of nAChR activation is additive. 258Recovery of neuromuscular function after discontinuation of neuromuscular blocking drug infusion in ICU patients was found to be faster with cisatracurium than with vecuronium despite equivalent reduction in train-of-four suppression at baseline. 262Nevertheless, the variability in the time course and the etiology of alteration recovery of neuromuscular function demonstrates that routine neuromuscular monitoring alone is not sufficient in eliminating prolonged recovery and myopathy in ICU patients. 262 

Multiple sclerosis is a demyelinating disease resulting from an abnormal immune response to an antigen present in the myelin sheath within the central nervous system. It is common in young adults. Demyelination in multiple sclerosis follows a waxing and waning pattern and is thought to be inflammatory in origin. There is evidence for both genetic predispo-sition and previous exposure to an unknown causative agent. 263Demyelinating lesions may occur in any part of the brain and spinal cord and can result in sensory, motor, autonomic, or neuropsychological disability. In multiple sclerosis, mean firing rates of the motor unit action potentials are reduced, and firing variability is increased. 264 

There is some evidence that the stress of intercurrent illness–surgery–anesthesia may increase the rate of relapse in multiple sclerosis, but the interplay between these factors is unclear. There have been several case series published that do not demonstrate any association between the use of general anesthesia and an increased rate of relapse. 265,266The use of regional anesthesia in multiple sclerosis is more controversial. Both lumbar epidural and subarachnoid anesthesia have been reported in patients with multiple sclerosis without clear evidence for an increase in the relapse rate. 267There is some suggestion that higher concentrations of local anesthetic may be neurotoxic. In one study in which patients received either 0.5 or 0.25% bupivacaine for epidural anesthesia, relapses only occurred in patients receiving the higher dose of local anesthetic. 265Patients with multiple sclerosis may have exacerbations of their symptoms if they become hyperthermic. 266 

The use of neuromuscular blockers in patients with multiple sclerosis depends on the clinical syndrome. In patients with chronic motor weakness, central denervation is the probable cause. As with any patient with denervation or disuse, there may be up-regulation in nAChR numbers and increased sensitivity to depolarizing neuromuscular blockers. In this case, the patient is at risk for hyperkalemia after administration of succinylcholine. As discussed above, there are paradoxical reports of increased sensitivity to nondepolarizing neuromuscular blockers in patients with multiple sclerosis, probably because of reduced muscle mass or reduced margin of safety for neuromuscular transmission. 264It is significant to note that muscle denervation of any origin will cause muscle depolarization. 268As a consequence, the inactive state of sodium channels will be favored so that endplate potentials fail to generate action potentials. 269The denervation-induced decline of the resting potential will significantly contribute to muscle weakness.

The motor neuron diseases are a group of heterogenous disorders characterized by muscle weakness, atrophy, or spastic paralysis caused by involvement of lower or upper motor neurons, respectively. ALS is the most common motor neuron disease and involves both upper and lower motor neurons. Spinobulbar muscular atrophy (or Kennedy disease) affects lower motor neurons only. Hereditary spastic paraplegia, on the other hand, involves upper motor neurons. ALS, commonly known as Lou Gehrig disease, is a progressive disease characterized by degeneration of cortical, brainstem, and spinal motor neurons. 270Motor neuron degeneration results in denervation, muscle wasting, and eventual paralysis and death. Cognitive and sensory systems are left intact. The incidence of ALS is 2–4 in 100,000. The etiology of ALS is not known, although a role for oxidative stress has been suggested since mutations in the gene for Cu²⁺–Zn²⁺superoxide dismutase (SOD1) have been identified in familial ALS. Knockout of the SOD1  gene in mice results in a syndrome similar to ALS. 271Experimental data also suggest the presence of antibodies to voltage-gated Ca²⁺channels in ALS patients. 272These antibodies cause an increase in quantal release at the NMJ probably secondary to increased function of the presynaptic Ca²⁺channels. 273Increased Ca²⁺influx and intracellular Ca²⁺concentration may contribute to pathologic changes seen at the NMJ. 274In animals, long-term neuromuscular dysfunction is reproduced by passive transfer of ALS immunoglobulins. 275There is currently no cure for ALS, and treatment is aimed at symptomatic support and comfort. 270 

As in other patients with muscle wasting from states of functional denervation (multiple sclerosis, GBS), there is compensatory up-regulation of nAChRs that may be extrasynaptic. These patients are at risk for hyperkalemia after administration of succinylcholine. 276There may be a perceived hypersensitivity to nondepolarizing neuromuscular blockers because of weakness caused by muscle wasting. Patients, particularly in late stages of the disease, may be cachectic from inadequate nutrition and have reduced plasma protein binding for many of the anesthetic drugs. These patients have reduced respiratory muscle reserve, abnormal airway protective reflexes, and are at increased risk for respiratory depression and aspiration secondary to the use of sedative and anesthetic drugs. Epidural anesthesia has been used in ALS patients without untoward effects. 277 

Guillain-Barré syndrome is made up of a spectrum of diseases that commonly include a generalized polyradiculopathy, affecting the limbs proximally more than distally, and may also involve cranial and bulbar nerves. 278GBS is relatively common, with an incidence of 4 in 10,000 throughout the world. 278The diseases that underlie GBS have recently been reclassified. 278Acute inflammatory polyradiculoneuropathy is common in the white populations of North America and Europe. Lymphocytic invasion in the peripheral nervous system and primary macrophage penetration of apparently normal myelin are typical of acute inflammatory polyradiculoneuropathy. In contrast, in Central America, China, Japan, and India, GBS is caused by an axonopathy that affects both motor and sensory neurons. These syndromes are called acute motor axonal neuropathy and acute motor and sensory neuropathy depending on the presence of sensory involvement. The Fisher syndrome is an additional variant of GBS in which the patients have ophthalmoplegia, ataxia, and loss of tendon reflexes but no limb weakness.

There is strong evidence for an association between certain infections and GBS. The most prevalent infections and events associated with GBS are Campylobacter jejuni , Cytomegalovirus, Epstein-Barr virus, Mycoplasma pneumoniae , rabies, and the “Swine Flu” vaccines. Undoubtedly, very few patients infected or vaccinated with the above agents will develop GBS. A predisposition for GBS possibly requires a particular genetic background or specific strains of infective organisms. 278 

Patients with GBS commonly have high autoantibody titer to antiganglioside antibodies directed at the ganglioside GQ1b. Gangliosides are present in high concentrations in peripheral nerve axons and myelin, and several studies indicated that different gangliosides are present at nodes of Ranvier and at the NMJ. 279It is unclear whether these antibodies cause demyelination or are a secondary result of the disease. 278Neuromuscular weakness in the acute stage of GBS has been attributed, in part, to circulating antibodies that can block both presynaptic voltage-gated calcium channels 280and postsynaptic nAChR channels. 281Patients with GBS commonly have symptomatic improvement after plasmapheresis. 282The final common pathway in acute inflammatory polyradiculoneuropathy is invasion of the myelin sheath by macrophages. The macrophages displace and phagocytose the myelin from the axon, leaving cleanly demyelinated axons. 283Demyelination produces functional denervation of the muscle and up-regulation of nAChRs at the postsynaptic membrane.

Patients with GBS present to the anesthesiologist in the ICU with motor weakness, at which time tracheal intubation and ventilation is often necessary because of insufficient ability to generate inspiratory force or because of concurrent infection. These patients may need anesthetic intervention for surgery or for assisted delivery in pregnancy. Succinylcholine is contraindicated because of the risk of hyperkalemic cardiac arrest secondary to the proliferation of postsynaptic nAChRs. 284,285This risk may persist over a long period after recovering from the symptomatic neurologic deficit. 286Preexistent loss of motor units and presynaptic or postsynaptic nAChR channel blockade by antibodies 281may result in sensitivity to nondepolarizing neuromuscular blockers. 285,287Regional anesthesia is not contraindicated, although patients with GBS are sensitive to local anesthetics secondary to preexistent axonal conduction abnormalities. 285Patients with GBS have a high incidence of autonomic instability, and the slower onset of an epidural block may be preferable to the rapid onset of subarachnoid anesthesia. GBS has been reported in four patients 1–2 weeks after epidural anesthesia. It was postulated that local trauma to nerve roots may initiate a cascade of immunologic events that result in demyelinating neuropathy in these patients. 288 

Charcot-Marie-Tooth disease (CMTD; hereditary motor and sensory demyelinating polyneuropathy) is the most common genetic neuropathy, with an incidence of 1 in 2,600. 289CMTD has heterogeneous genetic (autosomal dominant, X-linked, or autosomal recessive) and clinical presentations. 290,291Three genes responsible for CMTD type 1 have been identified: peripheral myelin protein 22 and myelin protein zero for the autosomal dominant form and connexin 32 for the X-linked dominant variant. 292The latter variant encodes a gap junction protein. 291 

Gap junctions are aggregations of intercellular channels that provide a direct pathway for the exchange of nutrients, metabolites, ions, and small molecules up to approximately 1,000 Da between closely apposed cells. 293The channels are composed of connexins, a family of highly related proteins. 293In the nervous system, gap junctional channels play a key role in the propagation of signals between electrically excitable cells. 294Failure of the gap junctions may therefore lead to impaired Schwann cell function and subsequent demyelination. Electron microscopy shows gap junctions to be extremely rare between adjacent myelinating Schwann cells in genetic abnormalities in connexin 32 associated with CMTD. 292 

Charcot-Marie-Tooth disease can be divided into two distinct groups based on electrophysiologic studies. 295CMTD type 1 exhibits moderately to severely reduced motor nerve conduction velocities. 296The conduction deficit in CMTD type 1 is bilaterally symmetric, which suggests an intrinsic Schwann cell defect. 296In contrast, CMTD type 2 results from neuronal atrophy and degeneration and exhibits normal or only mildly reduced motor nerve conduction velocities with decreased amplitudes. 295 

Peroneal nerve atrophy leading to weakness in the anterior and lateral compartments is the most common clinical pattern in CMTD, but considerable variability exists in the pattern of atrophy. Abnormalities of feet and toes, including pes cavus, are usually present. Intrinsic atrophy of the calf musculature is a common finding in CMTD. The sensory disturbance is milder than the motor disturbance. Autonomic disturbances such as orthostatic hypotension and hypohidrosis are occasionally reported. 297Pregnancy may be associated with exacerbations of CMTD. 298Respiratory insufficiency has also been described in patients with CMTD. 299 

Loss of motor units and the resultant muscle weakness in CMTD might result in sensitivity to nondepolarizing neuromuscular blocking drugs. However, there is no evidence of prolonged response to atracurium and mivacurium in patients with CMTD. 300Succinylcholine and other malignant hyperthermia-triggering agents have been used in CMTD patients without untoward effects. 301Although there is no clear evidence that CMTD predisposes patients to an increased risk of malignant hyperthermia, incidents of malignant hyperthermia in patients with CMTD have been reported. 302Use of drugs known to trigger malignant hyperthermia must be carefully considered.

Muscular dystrophies are a group of heterogeneous, genetically determined disorders of skeletal muscle and, in some cases, cardiac muscle. These disorders have been classified on the basis of clinical symptomatology and genetic inheritance, but with the advent of molecular diagnosis, categories have shifted (table 2). Patients may present with symptoms of muscle weakness and atrophy at different stages of development. The time course and prognosis differs with each syndrome. Most symptomatology is a result of muscle weakness and related pulmonary complications and, in some cases, cardiac abnormalities. 303 

Duchenne muscular dystrophy is one of the most common genetic diseases in humans, with an incidence of 1 in 3,500 male births, whereas Becker muscular dystrophy is milder and affects 1 in 30,000 male births. 303Duchenne–Becker dystrophy is caused by an X-linked recessive mutation resulting in abnormal or absent dystrophin or related glycoproteins that link the extracellular matrix to the cytoskeleton (see also Signals from the Nerve). In Duchenne muscular dystrophy, dystrophin is usually absent, whereas in Becker muscular dystrophy, the protein is present but qualitatively and quantitatively abnormal. 304As a result of chronic muscle regeneration in patients with Duchenne dystrophy, there is coexpression of both fetal and adult nAChRs in the mature muscle membrane. 305In Duchenne dystrophy, weakness leading to inability to ambulate generally occurs before puberty, and patients typically develop nocturnal hypoventilation by their late 20s. 303Progressive cardiomyopathy develops in the midteens, and patients typically succumb to cardiac or pulmonary manifestations of their disease in their late teens or 20s. 303Cognitive impairment is also observed and has been attributed to an abnormality in the neuronal membrane caused by a lack of dystrophin. 306 

Becker muscular dystrophy results from abnormalities in the same gene as Duchenne dystrophy with similar symptomatology. However, it is milder and has slower progression. Onset in childhood may occur as late as 16 yr. Cardiac problems may be more severe than the skeletal muscle weakness. 303Limb-girdle dystrophy is similar to Duchenne dystrophy and is found most commonly in families in North Africa. Congenital muscular dystrophy has the worst prognosis. Affected infants present at birth with hypotonia, weakness, and respiratory and swallowing abnormalities. Mutations in the laminin α2 chain cause the most severe form of congenital muscular dystrophy. Muscle fiber deterioration in this disease is thought to be caused by impaired formation of the basement membrane and its inability to interact with the DGC or the integrins. Deficiency of laminin α2 is accompanied by up-regulation of the laminin α4 chain, giving rise to laminin-8, which binds poorly to DGC in the muscle fiber. Recently, it has been possible to rescue dystrophic symptoms in a mouse model for congenital muscular dystrophy by muscle-specific overexpression of an agrin minigene, which bound to laminin-8 and the DGC, 306Areplacing the missing link between the basement membrane and the muscle fiber. 306ATherefore, overexpression of an engineered molecule may become an exciting novel approach to devising new therapeutic tools to restore muscle function in human muscular dystrophies. Facioscapulohumeral muscular dystrophy usually presents in late childhood with facial and scapulohumeral weakness. There may also be weakness of the pelvic girdle with a lordotic posture, but there is usually no cardiac involvement. Patients may develop retinal vasculopathy and sensorineural hearing loss. 303 

There have been many reports of succinylcholine-induced hyperkalemia and cardiac arrest in patients with undiagnosed muscular dystrophies. 307,308This response has lead to a Food and Drug Administration–mandated warning against the use of succinylcholine in pediatric patients because of potential mortality in patients with clinically inapparent muscular dystrophies. Innervation is relatively normal in dystrophic muscle, but the postsynaptic nAChRs are expressed as a mixture of fetal- and mature-type receptors characteristic of chronic denervation. 305The expression of the fetal nAChR in the dystrophic muscle is not a characteristic of dystrophy but a consequence of muscle regeneration. 305 

Resistance to nondepolarizing neuromuscular blockers would be expected on the basis of the reduced sensitivity of fetal nAChRs to competitive antagonists. However, clinically the reverse is seen. 250,309Patients with myopathy are unusually sensitive to nondepolarizing neuromuscular blockers. 309There is an increase in the incidence of malignant hyperthermia in patients with myopathies, 310and there is an association of rhabdomyolysis with the use of volatile anesthetics. 310 

Myotonias are characterized by difficulty in initiating muscle movement with delayed muscle relaxation after voluntary contraction. Myotonic dystrophy is a progressive disease that manifests in late childhood or adulthood with muscle weakness and atrophy. There is associated frontal balding, cataracts, and testicular atrophy. Myotonic dystrophy occurs with an incidence of 1 in 8,000, making it one of the most common neuromuscular diseases. Myotonia may be precipitated by cold, shivering, diathermy, and succinylcholine. Mutations in the pore-forming subunits of sodium and chloride channels cause myotonia because of an alteration in the electrical excitability of the muscle fiber. 311,312Myotonic dystrophy is an autosomal dominant disorder associated with an expanded trinucleotide sequence at the 3′ untranslated end of the gene for myotonic dystrophy protein kinase (DMPK). 313DMPK is a serine–threonine protein kinase highly expressed in heart, brain, and skeletal muscle. 313In skeletal muscle, DMPK is located at the terminal cisternae of the sarcoplasmic reticulum, but its role in the pathophysiology of the disease is unclear. 314There is some evidence that DMPK is involved in cellular Ca²⁺homeostasis. 314Maturational-related abnormality or altered modulatory mechanisms of sarcoplasmic reticulum Ca²⁺transport have been noted in myotonic dystrophic slow-twitch muscle fibers. 315 

In normal muscle, depolarization of the postsynaptic membrane causes brief openings of sodium channels that occur within the first few milliseconds after membrane depolarization. The voltage-sensitive chloride channels then traffic chloride ion to return the muscle membrane potential to its resting level. 141Sodium channels harboring mutations causing myotonia exhibit an abnormal tendency to open later or more persistently after membrane depolarization. 141,316Residual sodium entry through these abnormal channels repeatedly reinitiates the cycle of membrane depolarization. 141Chloride channel mutations associated with myotonia reduce the amount of chloride ion that can enter the cell to repolarize the membrane, leading to oscillations. 141In patients with either abnormal sodium or chloride channels, the muscle becomes hyperexcitable. 146The increased excitability in the muscle results in the generation of repetitive action potentials after voluntary contractions.

Patients with myotonic dystrophy have increased mortality from respiratory complications of their muscle weakness as well as cardiac disease. Cardiac abnormalities include conduction block distal to the His bundle, ventricular arrhythmias, and an increased incidence of sudden death. The severity of the symptoms is somewhat related to the number to trinucleotide repeats in DMPK. 317The mechanism of muscle weakness is loss of contractile tissue, probably in combination with contractile dysfunction. 317 

Despite an apparently normal response to curare in a research setting, 239patients with myotonia have been reported to require reduced doses of nondepolarizing neuromuscular blockers in a clinical setting. 318This has been attributed to the underlying muscle wasting and reduced ability to produce contractile force. 319Anticholinesterase agents may precipitate myotonia 320because of increased sensitivity of the myotonic muscle to the effects of acetylcholine. The use of succinylcholine in patients with myotonic dystrophy, despite apparently normal nAChRs, is to be avoided. There are reports of extreme muscle rigidity and cardiac arrest after a dose of succinylcholine in patients with myotonic dystrophy. 319The cardiac arrest was assumed to be caused by increased serum potassium concentration; however, potassium concentrations were not verified before cardiopulmonary resuscitation. The cardiac arrest might have been caused by intrinsic cardiac abnormalities that are associated with myotonic dystrophy. In contrast, trauma patients with undiagnosed severe myotonic dystrophy were given succinylcholine without side effects. 321The association between myotonia and malignant hyperthermia is uncertain, and the difficulty in interpretation of the caffeine–halothane contracture test in myotonic patients further complicates the nature of the association. 322,323 

Patients with myotonic dystrophy may suffer respiratory compromise as a result of muscle weakness. There may be an increased risk of aspiration caused by velopalatal insufficiency. Children with myotonic dystrophy are at particular risk for the respiratory-depressant effects of general anesthetics and should be carefully monitored before discharge. 324Clinical deterioration may occur in pregnancy, probably because of hormonal changes, with exacerbation of the muscle weakness, myotonia, and muscle wasting. 325 

Muscle weakness and fatigability are pathognomonic of the myasthenic syndromes. In recent years it has become clear that the myasthenias represent a group of diseases. MG and the LEMS are both caused by autoimmune disease. MG is caused by autoantibody targeting of an extracellular portion of the muscle receptor for acetylcholine. 237Antibody targeting of this region results in cross-linking of two adjacent nAChRs, complement fixation, and focal lysis of the postsynaptic membrane. 326Antigenic modulation also results in an increased rate of internalization and degradation of nAChR on the muscle membrane. There are, in addition, antigenic T-cell epitopes throughout the α subunit. Interestingly, antibodies from MG patients do not cross-react with the α3 nAChR subunit that is found principally in the autonomic nervous system or α⁴β²nAChRs that occur in the central nervous system, perhaps explaining the lack of autonomic and central nervous system symptoms in typical MG. 327The net result of antigenic modulation and focal lysis is a reduced number and altered structure of the postsynaptic nAChRs, which impairs neuromuscular transmission and causes muscle weakness. 237Electron microscopic studies show that the postsynaptic membrane has abnormally sparse, shallow folds with markedly simplified geometric patterns. 328 

The cause of the induction of the immune response in MG is not well known. It is clear, however, that immunization with nAChRs from Electrophorus  electric organs can cause the induction of antibodies to the nAChR and a syndrome of muscular weakness that has become an animal model for MG. 329A small percentage of MG patients develop autoantibodies as part of a paraneoplastic syndrome (12% have thymoma). 237Thymic myoid cells express fetal nAChRs and other muscle proteins. Approximately 70% of MG patients have thymic lymphoid follicular hyperplasia and exhibit germinal centers that produce antibodies to nAChRs. 237Antibodies to nAChRs must also be produced in other locations. This is based on the evidence that thymectomy may be beneficial to the clinical course of MG, but it may not be curative. Fetal-type nAChRs may be immunogenic, as indicated by the common involvement of extraocular muscles in MG that selectively express fetal nAChRs in adult life. 237There is also some evidence to indicate that immune molecules created in response to microbial antigens may cross-react with nAChR. This may constitute initial triggers of MG in some patients. 330 

In chronic MG, the nAChR content is reduced to approximately 30%, and most of the remaining nAChRs are bound by antibody. 237Acetylcholine sensitivity is reduced, and decrementing response to repetitive stimulation occurs. There is no specific immunotherapy for MG as there are abnormalities in all arms of the immune response. Nonspecific immunosuppression with steroids and other drugs and plasmapheresis are often combined with thymectomy and symptomatic treatment with anticholinesterases.

Congenital myasthenic syndromes are heterogenous disorders that do not occur because of autoantibodies, but are caused by inherited mutations in the SVs, acetylcholinesterase, or nAChRs. 114,187,237,331These mutations result in a range of muscle weaknesses and fatigability that are characteristic of myasthenia. Mutations in the α, β, δ, and most frequently the ϵ subunit of nAChRs can cause congenital myasthenic syndromes (see also Subunit Mutations and the Myasthenic Syndromes and Acetylcholinesterase at the Neuromuscular Junction). The inheritance of congenital myasthenic syndromes is either autosomal dominant or autosomal recessive. In contrast to neonatal MG that is caused by passive transfer of anti-AChR antibodies to the fetus by a myasthenic mother, the mother of congenital myasthenic syndromes has no myasthenia.

The LEMS is a presynaptic disorder of neuromuscular transmission in which patients exhibit profound muscle weakness in response to nerve stimuli. LEMS is an autoimmune disease that is known to occur with, or precede, a variety of malignancies. Approximately 60% of LEMS patients exhibit a paraneoplastic response, often in conjunction with small-cell carcinoma of the lung. 237LEMS is caused by an autoantibody targeting the voltage-gated Ca²⁺channels that mediate acetylcholine release at the motor neuron terminals. 332Depolarization of the motor axon causes less Ca²⁺influx, and less acetylcholine is released. The acetylcholine content and acetyltransferase activity in diseased nerve endings are normal. In contrast to MG, there is an increase in contractile force on sustained muscle stimulation in LEMS. In fact, repetitive stimulation causes summation of presynaptic Ca²⁺signals and improved release. 333Exercise or tetanic stimulation improves rather than reduces muscle strength in LEMS. As discussed in Vesicle Mobilization and Docking, the interaction between synaptotagmin and the voltage-gated Ca²⁺channel plays an important role in docking synaptic vesicles at the plasma membrane before rapid neurotransmitter release. It has been suggested that an autoantibody binding to a synaptotagmin–Ca²⁺-channel complex may be involved in the etiology of LEMS. 163Assay of voltage-gated calcium channels antibody titer and electrophysiologic tests help to differentiate LEMS from other disorders of the NMJ. In contrast to MG, approximately 30% of patients with LEMS have autonomic dysfunction.

Treatment with 3,4-diaminopyridine results in significant improvement in symptoms and in the summated amplitude of compound muscle action potentials in patients with LEMS. 3343,4-Diaminopyridine selectively blocks potassium channels, preventing potassium efflux and causing increase in action potential duration. The latter results in prolonged activation of voltage-gated Ca²⁺channels and increases intracellular Ca²⁺concentrations in the nerve terminal with a concomitant increase in acetylcholine release.

Anesthesia for myasthenic patients has been reviewed by Baraka. 335Because of the decreased number of nAChRs or their functional blockade by antibodies, myasthenic patients are resistant to succinylcholine. 336On the other hand, butyrylcholinesterase activity may be decreased in myasthenic patients by preoperative plasmapheresis or administration of pyridostigmine, and this would result in potentiation of succinylcholine- (or mivacurium 337)-induced block. The interplay between these two factors (resistance to succinylcholine vs.  reduction in butyrylcholinesterase activity) should be considered when administering succinylcholine to patients with MG. Progression to phase II block is not uncommon in these patients. 335Succinylcholine should be avoided in patients with SCCMS because succinylcholine would be expected to worsen the existent state of excitotoxicity.

With the loss of 70–89% of the functional nAChRs and hence the margin of safety of neurotransmission, patients with MG are extremely sensitive to nondepolarizing neuromuscular blockers. The decrease in available nAChRs in MG means that even mildly symptomatic myasthenic patients are just at the border of the safety margin for neuromuscular transmission, as evidenced by their easy fatigability. 237The effective dose of vecuronium is 250% greater in control patients than in MG patients. 338Indeed, a case of congenital myasthenia with minor clinical signs has been diagnosed as the result of an exaggerated response to a small dose of a nondepolarizing neuromuscular blocker. 339However, with careful titration and with adequate monitoring of neuromuscular function, nondepolarizing agents have been used safely in myasthenic patients undergoing thymectomy. 340 

On the other hand, patients with LEMS are sensitive to both depolarizing and nondepolarizing neuromuscular blockers. 341In fact, patients with LEMS have a significantly greater sensitivity to nondepolarizing neuromuscular blockers when compared with those with MG. 239 

The mitochondrial myopathies are a clinically and biochemically heterogeneous group of disorders characterized by abnormalities of mitochondrial structure. Mitochondrial myopathies are often associated with abnormal proliferation of mitochondria, which accumulate beneath the sarcolemma and between muscle fibers. The massive proliferation of giant mitochondria is probably caused by up-regulation of both mitochondrial DNA and nuclear DNA transcripts, presumably in an effort to compensate for the bioenergetic defect caused by a mitochondrial DNA mutation. 342These collections of abnormal mitochondria stain purple or red with the modified Gomori trichrome stain, resulting in so called “ragged red fibers.” However, ragged red fibers are not pathognomonic of a mitochondrial DNA mutation, as they also appear in aged muscle and in other myopathies. 343In some cases, the fibers do not have a ragged appearance. Affected fibers also contain an excess of glycogen granules and increased numbers of fine neutral lipid droplets.

Mutations in mitochondrial DNA have been associated with mitochondrial myopathies. 344These mutations will cause impaired electron transport chain function. This, in turn, results in decreased ATP production and formation of damaging free radicals. These toxic events produce further mitochondrial damage, including oxidation of mitochondrial DNA, proteins, and lipids. Reactive oxygen species have also been implicated in mitochondrial myopathies. 345Normally, mammalian mitochondria generate most of the ATP for cells by the process of oxidative phosphorylation. During oxidative phosphorylation, between 0.4 and 4% of the oxygen consumed is reduced to form superoxide anion. 345During normal circumstances, superoxide is reduced to H²O²by the mitochondrial form of superoxide dismutase. Within the mitochondria, the H²O²is either converted to water by mitochondrial glutathione peroxidase or can participate in Fenton type chemistry, giving rise to further reactive oxygen species such as the hydroxyl radical. 345 

Both isolated myopathies and several multisystem syndromes have been identified. The syndromes, which are defined through characteristic clinical manifestations in addition to mitochondrial myopathy, are chronic progressive external ophthalmoplegia, including Kearns-Sayre syndrome, MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes) syndrome, MERRF (myoclonus epilepsy and ragged red fibers) syndrome, MNGIE (myopathy, external ophthalmoplegia, neuropathy, and gastrointestinal encephalopathy) syndrome, and NARP (neuropathy, ataxia, and retinitis pigmentosa) syndrome. Acquired mitochondrial myopathy has been associated with the use of zidovudine, an antiretroviral drug that depletes muscle mitochondrial DNA. 346 

There is evidence to support mitochondrial involvement in posttetanic potentiation of synaptic transmission at the NMJ. 347Electrophysiologic studies do not correlate with any specific biochemical or genetic defect, but are consistent with diagnosis in patients presenting with clinical signs of mitochondrial myopathy. 348 

Although it has been suggested that mitochondrial myopathy does not involve the NMJ, 349increased sensitivity to different nondepolarizing neuromuscular blockers has been demonstrated in patients with mitochondrial myopathies. 350This enhanced sensitivity is of a magnitude similar to that observed in MG. 350Increased sensitivity to succinylcholine was also noted in these patients. 351The association between malignant hyperthermia and mitochondrial myopathies is not clear, but published reports indicate a possible association. 352 

Cell membranes are composed of two lipid layers that are not permeable to ions. The channels are macromolecular protein complexes within the lipid membrane and are opened by ligands or voltage changes. They regulate the traffic of ions in and out of the cell causing depolarization and hyperpolarization of the cell. The channel structure is determined by different genes that encode each protein subunit in the channel. 353In skeletal muscle, disorders are associated with mutations in Na⁺, K⁺, Ca²⁺, Cl⁻, and nAChR channels. Some of these disorders (myotonia, CMTD, and congenital myasthenic syndromes) have been discussed previously (table 3). For more extensive accounts on ion channels and disease, see Ashcroft. 354 

Hyperkalemic periodic paralysis is an autosomal dominant disorder characterized by episodes of muscle weakness associated with hyperkalemia. Mutations in the gene encoding the human skeletal muscle Na⁺channel α subunit have been identified in hyperkalemic periodic paralysis. 355Muscle fibers from affected individuals exhibit sustained Na⁺currents that depolarize the sarcolemma and inactivate normal Na⁺channels. This inactivation disables the formation of action potentials during the attack of paralysis. 144,146,316Attacks usually begin in the second decade and vary both in frequency and duration. Respiration is rarely affected, and the disorder is considered benign. 146The attacks of paralysis are frequent, brief, and often precipitated by rest after exertion, stress, the ingestion of foods with high potassium content such as bananas, or the administration of potassium. Prophylactic treatment with potassium-wasting diuretics is often successful in reducing the frequency and severity of attacks by lowering serum potassium.

Depletion of potassium before surgery, prevention of carbohydrate depletion, avoidance of potassium-releasing anesthetic drugs, and maintenance of normothermia are the key steps of anesthetic management. 356Succinylcholine should be avoided because it will result in increases in serum potassium concentrations and can cause myotonic symptoms in these patients. The association between malignant hyperthermia and hyperkalemic periodic paralysis to the adult skeletal muscle sodium channel gene has been established. 357There is no evidence that patients with hyperkalemic periodic paralysis exhibit abnormal sensitivity to nondepolarizing neuromuscular relaxants. 356 

Hypokalemic periodic paralysis is an autosomal dominant muscle disease manifested by episodic weakness associated with hypokalemia during attacks. It is thought to also arise from the abnormal function of Ca²⁺channels. 358The causative gene was shown to encode the α1 subunit of the dihydropyridine receptor. Although it is the most common form of the periodic paralyzes in humans, it is still a rare disease, with a prevalence of only 1:100,000. 144The hypokalemia has been attributed to the stimulation of the sodium-potassium pump by insulin. Low potassium concentration may cause electrical destabilization of the cell membrane because the potassium equilibrium becomes very negative, and the potassium conductance approaches zero. 144 

This disorder differs from hyperkalemic periodic paralysis in several additional aspects: the attacks can be very severe in certain patients, women with the same mutation are much less severely affected than men, attacks are often triggered by high carbohydrate intake or insulin challenge, and this condition can lead to a progressive disabling myopathy. 146Symptomatic treatment of severe attacks entails ingestion of high levels of potassium. Prophylactic treatment with acetazolamide (a carbonic anhydrase inhibitor) is also successful, perhaps by producing metabolic acidosis. The latter decreases the urinary excretion of K⁺. 359 

Hypothermia, glucose and salt loading, or metabolic alkalosis can precipitate an attack. Therefore, careful intraoperative monitoring of body temperature, glucose, serum electrolytes, and acid-base status is important. Careful and frequent monitoring of plasma potassium concentrations is of greatest importance.

Despite the recommendation that relaxants be avoided, careful titration of short- or intermediate-acting nondepolarizing neuromuscular blockers with adequate monitoring of neuromuscular function is uneventful in patients not suffering from acute episodes of paralysis. 360In a review of 21 anesthetics administered to members of a family with hypokalemic periodic paralysis, seven patients suffered from mild or severe postoperative paralysis. 361Hypokalemia should be considered in the differential diagnosis of postoperative residual weakness. A normal response to succinylcholine was noted in these patients, 360but the association between hypokalemic periodic paralysis and malignant hyperthermia has been reported. 362 

Spinal and epidural anesthesia was reported to be safe alternatives to general anesthesia in these patients. 363It should be noted, however, that epidural nerve blocks lower serum potassium concentrations. 364Administration of epinephrine with the local anesthetic accounts for a proportion of this decline, but another unknown mechanism appears to contribute to the reduction in serum potassium in patients not receiving β-adrenergic agonists. 364 

Malignant hyperthermia (MH) is the genetic predisposition that responds to triggering agents such as inhalational anesthetics and depolarizing neuromuscular blockers with abnormalities in intracellular Ca²⁺homeostasis. These abnormalities, more common in patients with muscle disease, result in tetany, increased metabolism, rhabdomyolysis, hyperkalemia, acidosis, and, if untreated, death. 322,365Often inherited as an autosomal dominant trait, MH has linkage to 30 different mutations in the type-1 ryanodine receptor (RyR1 ) gene. The RyR  gene encodes a channel that mediates the release of Ca²⁺from the sarcoplasmic reticulum membrane to initiate contraction in skeletal muscle. The reverse of this is muscle relaxation, via  both inactivation (closure) of the channel and ATP-dependent pumping of calcium back into the sarcoplasmic reticulum. RyR1  mutations with linkage to MH are thought to cause an abnormal opening of the calcium-release channel, when it is exposed to certain anesthetic drugs. Mutations in this gene are considered to account for susceptibility to MH in more than 50% of cases. 230MH is a heterogeneous disorder and may, in some pedigrees, be caused by mutations in genes on chromosomes other than 19q. 365Another mutation in the Ca²⁺channel α²δ subunit has also been linked to MH. 366The molecular diagnosis of this disease is made more complicated because it is variably expressed, and there is incomplete penetrance of the clinical phenotype. Although multiple mutations likely exist, the final common result is abnormal Ca²⁺homeostasis in response to triggering agents commonly used in anesthesia. For a recent account on MH, see the review by Hopkins. 367 

In North America and Europe, the overall frequency is 1 in 15,000 anesthetics. If adult patients are considered only, the occurrence may be as low as 1 in 50,000 anesthetics. Mortality is more than 60% in untreated patients. Early administration of dantrolene (a lipid-soluble hydantoin analog) is invaluable in the treatment of MH crises, presumably by preventing Ca²⁺release from the sarcoplasmic reticulum. Although prompt recognition and appropriate treatment have markedly reduced the mortality rate in recent years, MH remains an important contributor to anesthetic-induced morbidity and mortality.

Central core disease (CCD) is also a dominantly inherited neuromuscular condition often associated with a susceptibility to malignant hyperthermia. CCD is linked to mutations in the gene encoding RyR1  and is thought to arise from “leaky” or “uncoupled” sarcoplasmic reticulum Ca²⁺-release channels. 368It has been widely assumed that CCD and a locus for MH may be allelic (i.e. , a single genetic defect is responsible for coinheritance of CCD and MH). However, not all individuals with CCD are susceptible to MH. 369It has also been suggested that all patients with CCD should be tested for MH susceptibility. 369 

Histologic examination of CCD muscles shows the presence of central areas (cores) mainly in type 1 muscle fibers. The core regions consist of unstructured myofibrils and a general lack (or absence) of mitochondria and oxidative enzymatic activity. 370Electron microscopic analysis of the central cores reveals a disintegration of the contractile apparatus and alterations in the structure and amount of sarcoplasmic reticulum and transverse tubule membranes. 370Because expression of the CCD phenotype is variable, the CCD diagnosis is based on histologic signs as well as on clinical expression of the disease. The phenotype may include fetal hypotonia (floppy infant syndrome), delayed motor development, and proximal muscle weakness. Exercise-induced muscle cramps are frequently reported. However, because the clinical course of CCD is slow or nonprogressive, many patients are not diagnosed until later in life. A small number of patients may be severely affected. Muscle atrophy is a frequent finding. Musculoskeletal deformities, including kyphoscoliosis, congenital hip dislocation, foot deformities, and joint contractures, are not uncommon. Cardiac abnormalities have rarely been reported in association with CCD. 371 

Surgical treatment may be required for some of the musculoskeletal deformities in patients with CCD. All patients with CCD should be considered at risk for MH unless in vitro  contracture tests show that the particular patient is free of the trait. 369,371 

Considerable information about the molecular mechanisms regulating formation of the NMJ is available, but the complete picture remains fragmentary. The molecular mechanisms by which polyneuronal innervation of perinatal muscle is reduced to a single motor neuron are largely unknown. Other major questions remaining unresolved concern the mechanisms by which agrin activates MuSK, as well as the signaling pathways downstream of MuSK. Neural agrin activation of MuSK requires additional muscle-specific activities, 31,34none of which are identified. In addition to aggregating AChRs and other components of the subsynaptic apparatus, phosphorylated MuSK also organizes a secondary NRG–ErbB receptor pathway activating the transcription of genes encoding AChR subunits and possibly other synaptic components. These two aspects of MuSK function are apparently mediated by different signaling pathways since agrin-induced AChR gene transcription but not AChR clustering depends on the binding of agrin to a substrate located in the BL. 63The intracellular signals involved in either of these pathways are still largely unknown. As a first step toward their identification, recent experiments have aimed at mapping the intracellular MuSK domain for the tyrosine residues phosphorylated by and involved in agrin-induced signaling.

Another problem requiring further investigation is whether synapse-specific expression of AChR genes depends on NRG-1 derived from motor neurons, or whether subsynaptic differentiation is controlled by neural agrin alone. The development of conditional gene knockout technology allowing the deletion of NRG-1 selectively in either skeletal muscle or in motor neurons will help to resolve this problem.

Studies of the SCCMS provide a view of ongoing evolution of the nAChR. The impact of gain- or loss-of-function mutations in these syndromes is becoming clearer, but the physiologic role, if any, of subunit changes during development remains a mystery. The fact that developmental subunit changes also occur for other ligand-gated ion channels suggests an influence on brain function. Receptor subunit composition clearly influences the pharmacologic sensitivity of central synapses.

General anesthetics may interact with specific amino acid residues of both muscle and neuronal nAChRs, 372but the modulatory role of lipids on these drug-receptor interactions is not clear. Understanding of such issues can be expected to lead to the development of drugs having a greater selectivity of action. Advances in research in the neurobiology of the NMJ are impressive. Future investigation will continue to impact on the practice of anesthesiology.