Medical Physiology, 3rd Edition

Synaptic Transmission at the Neuromuscular Junction

Neuromuscular junctions are specialized synapses between motor neurons and skeletal muscle

The chemical synapse between peripheral motor nerve terminals and skeletal muscle fibers is the most intensely studied synaptic connection in the nervous system. Even though the detailed morphology and the specific molecular components (e.g., neurotransmitters and receptors) differ considerably among different types of synapses, the basic electrophysiological principles of the neuromuscular junction are applicable to many other types of chemical synapses, including neuronal synaptic connections in the brain, to which we will return in Chapter 13. In this chapter, we focus on the neuromuscular junction in discussing the basic principles of synaptic transmission.

Motor neurons with cell bodies located in the ventral horn of the spinal cord have long axons that branch extensively near the point of contact with the target muscle (Fig. 8-4). Each axon process innervates a separate fiber of skeletal muscle. The whole assembly of muscle fibers innervated by the axon from one motor neuron is called a motor unit.


FIGURE 8-4 Vertebrate neuromuscular junction or motor end plate. A motor neuron, with its cell body in the ventral horn of the spinal cord, sends out an axon that progressively bifurcates to innervate several muscle fibers (a motor unit). The neuron contacts a muscle fiber at exactly one spot called a neuromuscular junction or motor end plate. The end plate consists of an arborization of the nerve into many presynaptic terminals, or boutons, as well as the specializations of the postsynaptic membrane. A high-magnification view of a bouton shows that the synaptic vesicles containing the neurotransmitter ACh cluster and line up at the active zone of the presynaptic membrane. The active zones on the presynaptic membrane are directly opposite the secondary postsynaptic clefts that are created by infoldings of the postsynaptic membrane (postjunctional folds). Depolarization of the bouton causes the vesicles to fuse with the presynaptic membrane and to release their contents into the synaptic cleft. The ACh molecules must diffuse at least 50 nm before reaching nicotinic AChRs. Note the high density of AChRs at the crests of the postjunctional folds. The activity of the released ACh is terminated mainly by an AChE. The bouton reloads its discharged synaptic vesicles by resynthesizing ACh and transporting this ACh into the vesicle via an ACh-H exchanger. CoA, coenzyme A.

Typically, an axon makes a single point of synaptic contact with a skeletal muscle fiber, midway along the length of the muscle fiber. This specialized synaptic region is called the neuromuscular junction or the end plate (see Fig. 8-4). An individual end plate consists of a small tree-like patch of unmyelinated nerve processes that are referred to as terminal arborizations. The bulb-shaped endings that finally contact the muscle fiber are called boutons. Schwann cells are intimately associated with the nerve terminal and form a cap over the face of the nerve membrane that is located away from the muscle membrane. The postsynaptic membrane of the skeletal muscle fiber lying directly under the nerve terminal is characterized by extensive invaginations known as postjunctional folds. These membrane infoldings greatly increase the surface area of the muscle plasma membrane in the postsynaptic region. The intervening space of the synaptic cleft, which is ~50 nm wide, is filled with a meshwork of proteins and proteoglycans that are part of the extracellular matrix. A particular region of the muscle basement membrane called the synaptic basal lamina contains various proteins (e.g., collagen, laminin, agrin) that mediate adhesion of the neuromuscular junction and play important roles in synapse development and regeneration. The synaptic basal lamina also contains a high concentration of the enzyme acetylcholinesterase (AChE), which ultimately terminates synaptic transmission by rapidly hydrolyzing free ACh to choline and acetate.

Electron micrographs of the bouton region demonstrate the presence of numerous spherical synaptic vesicles, each with a diameter of 50 to 60 nm. The cell bodies of motor neurons in the spinal cord produce these vesicles, and the microtubule-mediated process of fast axonal transport (see p. 25) translocates them to the nerve terminal. The quantal nature of transmitter release (described below in more detail) reflects the fusion of individual synaptic vesicles with the plasma membrane of the presynaptic terminal. Each synaptic vesicle contains 6000 to 10,000 molecules of ACh. The ACh concentration in synaptic vesicles is ~150 mM. ACh is synthesized in the nerve terminal—outside the vesicle—from choline and acetyl coenzyme A by the enzyme choline acetyltransferase. The ACh moves into the synaptic vesicle via a specific ACh-H exchanger, which couples the inward transport of ACh to the efflux of H+. Energetically, this process is driven by the vesicular proton electrochemical gradient (positive voltage and low pH inside), which in turn is produced by a vacuolar-type H pump fueled by ATP (see pp. 118–119). The nerve terminal also contains numerous mitochondria that produce the ATP required to fuel energy metabolism.

The process of fusion of synaptic vesicles and release of ACh occurs at differentiated regions of the presynaptic membrane called active zones. In electron micrographs, active zones appear as dense spots over which synaptic vesicles are closely clustered in apposition to the membrane. High-resolution images of active zones reveal a double linear array of synaptic vesicles and intramembranous particles. These zones are oriented directly over secondary postsynaptic clefts that lie between adjacent postjunctional folds. Molecular localization studies have shown that the density of ionotropic (nicotinic) AChRs is very high at the crests of postjunctional folds. Examination of the detailed microarchitecture of the neuromuscular synapse thus reveals a highly specialized structure for delivery of neurotransmitter molecules to a precise location on the postsynaptic membrane.

ACh activates nicotinic AChRs to produce an excitatory end-plate current

Electrophysiological experiments on muscle fibers have characterized the electrical nature of the postsynaptic response at the muscle end plate. Figure 8-5 illustrates results obtained from a classic experiment performed by Fatt and Katz in 1951. Their work is the first description of how stimulation of the motor nerve affects the membrane potential (Vm) at the postsynaptic region (i.e., muscle cell) of the neuromuscular junction. Nerve stimulation normally drives the Vm of the muscle above threshold and elicits an action potential (see p. 173). However, Fatt and Katz were interested not in seeing the action potential but in studying the small graded electrical responses that are produced as ACh binds to receptors on the muscle cell membrane. Therefore, Fatt and Katz greatly reduced the response of the AChRs by blocking most of them with a carefully selected concentration of d-tubocurarine, which we discuss below. imageN8-2 They inserted a KCl-filled microelectrode into the end-plate region of a frog sartorius muscle fiber. This arrangement allowed them to measure tiny changes in Vm at one location of the muscle cell called the neuromuscular junction or NMJ.


FIGURE 8-5 EPPs elicited at the frog neuromuscular junction by stimulation of the motor neuron. The magnitude of the excitatory postsynaptic potential is greatest near the end plate and decays farther away. (Data from Fatt P, Katz B: An analysis of the end-plate potential recorded with an intracellular electrode. J Physiol 115:320–370, 1951.)



For more information about tubocurarine, visit (accessed October 2014).

When Fatt and Katz electrically excited the motor nerve axon, they observed a transient depolarization in the muscle membrane after a delay of a few milliseconds. The delay represents the time required for the release of ACh, its diffusion across the synapse, and activation of postsynaptic AChRs. The positive voltage change follows a biphasic time course: Vm rapidly rises to a peak and then more slowly relaxes back to the resting value, consistent with an exponential time course. This signal, known as the end-plate potential (EPP), is an example of an excitatory postsynaptic potential. It is produced by the transient opening of AChR channels, which are selectively permeable to monovalent cations such as Na+ and K+. The increase in Na+ conductance drives Vm to a more positive value in the vicinity of the end-plate region. In this experiment, curare blockade allows only a small number of AChR channels to open, so that the EPP does not reach the threshold to produce an action potential. If the experiment is repeated by inserting the microelectrode at various distances from the end plate, the amplitude of the potential change is successively diminished and its peak is increasingly delayed. This decrement with distance occurs because the EPP originates at the end-plate region and spreads away from this site according to the passive cable properties (see pp. 201–203) of the muscle fiber. Thus, the EPP in Figure 8-5 is an example of a propagated graded response. However, without the curare blockade, more AChR channels would open and a larger EPP would ensue, which would drive Vm above threshold and consequently trigger a regenerating action potential (see p. 173).

What ions pass through the AChR channels during generation of the EPP? This question can be answered by the same voltage-clamp technique imageN7-4 that was used to study the basis of the action potential (see Fig. 7-5B). Figure 8-6A illustrates the experimental preparation for a two-electrode voltage-clamp experiment in which the motor nerve is stimulated while the muscle fiber in the region of its end plate is voltage-clamped to a chosen Vm. The recorded current, which is proportional to the conductance change at the muscle end plate, is called the end-plate current (EPC). The EPC has a characteristic time course that rises to a peak within 2 ms after stimulation of the motor nerve and falls exponentially back to zero (see Fig. 8-6B). The time course of the EPC corresponds to the opening and closing of a population of AChR channels, governed by the rapid binding and disappearance of ACh as it diffuses to the postsynaptic membrane and is hydrolyzed by AChE.


FIGURE 8-6 EPCs obtained at different membrane potentials in a voltage-clamp experiment. A, Two-electrode voltage clamp is used to measure the EPC in a frog muscle fiber. The tips of the two microelectrodes are in the muscle fiber. B, The six records represent EPCs that were obtained while the motor nerve was stimulated and the postsynaptic membrane was clamped to Vm values of −120, −91, −68, −37, +24, and +38 mV. Notice that the peak current reverses from inward to outward as the holding potential shifts from −37 to +24 mV. C, The reversal potential is near 0 mV because the nicotinic AChR has a poor selectivity for Na+ versus K+. (Data from Magleby KL, Stevens CF: The effect of voltage on the time course of end-plate current. J Physiol 223:151–171, 1975.)

As shown in Figure 8-6B, when the muscle fiber is clamped to a “holding potential” of −120 mV, we observe a large inward current (i.e., the EPC). This inward current decreases in magnitude as Vm is made more positive, and the current reverses direction to become an outward current at positive values of Vm. A plot of the peak current versus the clamped Vm shows that the reversal potential for the EPC is close to 0 mV (see Fig. 8-6C). Because the EPC specifically corresponds to current through AChR channels, this reversal potential reflects the ionic selectivity of these channels when extracellular Na+ and K+concentrations ([Na+]o and [K+]o) are normal.

By varying the concentrations of the extracellular ions while monitoring the shift in the reversal potential of the EPC, researchers found that the AChR channel is permeable to Na+, K+, and Ca2+ but not to anions such as Cl. Because of its low extracellular concentration, the current attributable to Ca2+ is small under physiological conditions and its contribution to reversal potential can be ignored. imageN8-3 By plugging the values for the various cations into the Goldman-Hodgkin-Katz voltage equation (see Equation 6-9), one can obtain the permeability of the AChR channel to various alkali monovalent ions relative to Na+ permeability. The result is the following sequence of relative permeability: 0.87 (Li+), 1.00 (Na+), 1.11 (K+), and 1.42 (Cs+). This weak ionic selectivity stands in marked contrast to typical voltage-gated Na+channels, which have PNa/PK ratios of ~20, and voltage-gated K+ channels, which have PK/PNa ratios of >100. On this basis, the ionotropic (nicotinic) AChR channel at the muscle end plate is often classified as a nonselective cation channel. Nevertheless, the weak ionic selectivity of the AChR is well suited to its basic function of raising Vm above the threshold of about −50 mV, which is necessary for firing of an action potential. When the nicotinic AChR channel at the muscle end plate opens, the normally high resting permeability of the muscle plasma membrane for K+ relative to Na+ falls so that Na+ and K+ become equally permeant and Vm shifts to a value between EK (approximately −80 mV) and ENa (approximately +50 mV).


Contribution of Ca2+ to the Resting Membrane Potential

Inspired by Jack Rose, Idaho State University

Contributed by Emile Boulpaep, Walter Boron

Equation 6-9 in the text (shown here as Equation NE 8-1) is the Goldman-Hodgkin-Katz (GHK) voltage equation:


Of course, we could insert additional terms for other cations besides K+ and Na+; for example, if we included Ca2+, the equation would look something like the following*:


A typical value for [Ca2+]i would be 10−7 M or 0.0001 mM, and a typical value for [Ca2+]o would be 1.2 mM. Thus, even though the concentration ratio for Ca2+ across the plasma membrane is large, this ratio per se has no bearing on the GHK equation. What counts here are the magnitudes of the product PCa[Ca2+], which are generally small compared to the other terms in both the numerator and denominator. Thus, Ca2+ makes very little contribution to Vm in the resting state. However, if we were to reduce the size of the other terms in either the numerator or denominator, the Ca2+ would begin to matter.

*The GHK equation has dropped the z (valence) term, as if all ions were monovalent. In order to insert Ca2+ into this simple equation, we treat the ion as if it were monovalent, which is clearly not the case. Thus, this equation merely serves to make the point that Ca2+ contributes very little to Vm because of the small magnitude of the product of permeability and concentration.

As we shall see in Chapter 13, which focuses on synaptic transmission in the CNS, similar principles hold for the generation of postsynaptic currents by other types of agonist-gated channels. For example, the receptor-gated channels for serotonin and glutamate are cation selective and give rise to depolarizing excitatory postsynaptic potentials. In contrast, the receptor-gated channels for glycine and GABA are anion selective and drive Vm in the hyperpolarizing direction, toward the equilibrium potential for Cl. These hyperpolarizing postsynaptic responses are called inhibitory postsynaptic potentials.

The nicotinic AChR is a member of the pentameric Cys-loop receptor family of ligand-gated ion channels

The molecular nature of the nicotinic AChR channel was revealed by studies that included protein purification, amino-acid sequencing of isolated subunits, molecular cloning, and cryoelectron microscopy. Purification of the receptor was aided by the recognition that the electric organs of certain fish are a particularly rich source of the nicotinic AChR. In the electric eel and torpedo ray, the electric organs are embryologically derived from skeletal muscle. The torpedo ray can deliver large electrical discharges by summating the simultaneous depolarizations of a stack of many disk-like cells called electrocytes. These cells have the skeletal muscle isoform of the nicotinic AChR, which is activated by ACh released from presynaptic terminals.

The purified torpedo AChR consists of four subunits (α, β, γ, and δ) in a pentameric stoichiometry of 2α:1β:1γ:1δ (Fig. 8-7). Each subunit has a molecular mass of ~50 kDa and is homologous to the other subunits. The primary sequences of nicotinic AChR subunits are ~90% identical between the torpedo ray and human.


FIGURE 8-7 Structure of the Torpedo nicotinic AChR. The nicotinic AChR receptor is a heteropentamer with the subunit composition of α2βγδ. These subunits are homologous to one another, and each has four membrane-spanning segments (M1 to M4). (For view from above, data from N Unwin: Refined structure of the nicotinic receptor at 4 Å resolution. J Mol Biol 346:968-989, 2005.) imageN8-14


Structure of the Nicotinic Acetylcholine Receptor

Contributed by Ed Moczydlowski

The nicotinic acetylcholine receptors (AChRs), which are all ligand-gated ion channels, come in two major subtypes, N1 and N2. The N1 nicotinic AChRs are at the neuromuscular junction, whereas the N2AChRs are found in the autonomic nervous system (on the postsynaptic membrane of the postganglionic sympathetic and parasympathetic neurons) and in the CNS. Both N1 and N2 are ligand-gated ion channels activated by ACh or nicotine. However, whereas the N1 receptors at the neuromuscular junction are stimulated by decamethonium and preferentially blocked by d-tubocurarine and α-bungarotoxin, the autonomic N2 receptors are stimulated by tetramethylammonium, blocked by hexamethonium, but resistant to α-bungarotoxin. When activated, N1 and N2 receptors are both permeable to Na+ and K+, with the entry of Na+ dominating. Thus, the nicotinic stimulation leads to rapid depolarization.

The nicotinic AChRs in skeletal muscle and autonomic ganglia are heteropentamers. That is, five nonidentical protein subunits surround a central pore, in a rosette fashion. imageN6-20 Because the five subunits are not identical, the structure exhibits pseudosymmetry, rather than true symmetry. There are at least ten α subunits (α1 to α10) and four β subunits (β1 to β4). As we will see below, the basis for these differences is a difference in subunit composition.

The N1 receptors have different subunit compositions depending upon location and developmental stage. The subunit composition of α2βγδ is found in fetal skeletal muscle, as well as the nonjunctional regions of denervated adult skeletal muscle. The electric organ of the electric eel (Torpedo), from which the channel was first purified, has the same subunit composition. The subunit composition of α2βεδ is found at the neuromuscular junction of adult skeletal muscle. Here, the ε subunit replaces the γ subunit. In both the α2βγδ and α2βεδ pentamers, the α subunits are of the α1 subtype and the β subunits are of the β1 subtype.

In the Torpedo N1 AChRs, the α, β, γ, and δ subunits have polypeptide lengths of 437 to 501 amino acids. eFigure 8-1 shows side and top views of this AChR.


EFIGURE 8-1 Three-dimensional view of the Torpedo or human fetal nicotinic AChR channel. (Data from Toyoshima C, Unwin N: Ion channel of acetylcholine receptor reconstructed from images of postsynaptic membranes. Nature 336:247–250, 1988.)

The N2 receptors in the nervous system, like those in muscle, are heteromers, probably heteropentamers. N2 receptors use α2 to α10 and β2 to β4.

Nicotinic Receptors

Receptor Type



N1 Nicotinic ACh

ACh (nicotine, decamethonium)

d-tubocurarine α-bungarotoxin

N2 Nicotinic ACh

ACh (nicotine, tetramethylammonium)



Unwin N. Refined structure of the nicotinic acetylcholine receptor at 4Å resolution. J Mol Biol. 2005;346:968–989.

Unwin N, Fujiyoshi Y. Gating movement of acetylcholine receptor caught by plunge-freezing. J Mol Biol. 2012;422:617–634.

The α, β, γ, and δ subunits each have four distinct hydrophobic regions known as M1 to M4, which correspond to membrane-spanning segments. For each of the subunits, the M2 transmembrane segment lines the aqueous pore through which Na+ and K+ cross the membrane.

The pentameric complex has two agonist binding sites. The two ACh binding sites are located at the extracellular α/γ interface of one α subunit and the α/δ interface of the other α subunit. imageN8-4


Ligand-Binding Sites of the Nicotinic AChRs

Contributed by Ed Moczydlowski

New insight into molecular details of the extracellular agonist-binding domain of AChR has been obtained from the x-ray crystal structure of an ACh-binding protein (AChBP) from Lymnaea stagnalis, a freshwater snail. AChBP is a soluble protein of 229 residues that is homologous to the amino-terminal region of nicotinic AChR and other members of the pentameric ligand-gated channel superfamily. As shown in Figure 3 of the paper by Brejc and colleagues, the crystal structure demonstrates that AChBP is formed as a radially symmetrical homopentamer of the monomer subunit with the agonist-binding site located between the five subunit interfaces. The tertiary structure of a single monomer subunit of AChBP features ten β strands folded into a β sandwich. The snail AChBP specifically binds many of the same agonist and antagonist molecules as AChR, including ACh, carbamylcholine, nicotine, d-tubocurarine, and α-bungarotoxin. AChBP serves as a particularly good homology model for the structure of nicotinic receptors in the mammalian nervous system that are formed as homopentamers of α subunits.


Brejc K, van Dijk WJ, Klaassen RV, et al. Crystal structure of an ACh-binding protein reveals the ligand-binding domain of nicotinic receptors. Nature. 2001;411:269–276.

Celie PHN, van Rossum-Fikkert SE, van Dijk WJ, et al. Nicotine and carbamylcholine binding to nicotinic acetylcholine receptors as studied in AChBP crystal structures. Neuron. 2004;41:907–914.

AChRs of normal adult muscle fibers are present in high density in the junctional folds of the postsynaptic membrane. However, in developing muscle fibers of the mammalian embryo and in denervated fibers of adult skeletal muscle, AChRs are also widely distributed in the membrane outside the end-plate region. The two types of AChRs, called junctional and nonjunctional receptors, have different functional properties. The unitary conductance of nonjunctional receptors is ~50% larger and the single channel lifetime is longer in duration than that of junctional receptors. The basis for this phenomenon is a difference in subunit composition. The nonjunctional (or fetal) receptors are a pentameric complex with a subunit composition of α2βγδ in mammals, just as in the electric organ of the torpedo ray. For the junctional AChR in adult skeletal muscle, substitution of an ε subunit for the fetal γ subunit results in a complex with the composition α2βεδ.

The functional properties of the two types of receptors have been studied in preparations of Xenopus oocytes that coexpress the cloned subunits. Figure 8-8A shows patch-clamp recordings of single ACh-activated channels in oocytes that had been injected with mRNA encoding either α, β, γ, δ or α, β, ε, δ. Measurements of currents at different voltages yielded single channel current-voltage (I-V) curves (see Fig. 8-8B) showing that the channel formed with the ε subunit had a unitary conductance of 59 picosiemens (pS), whereas that formed with the γ subunit had a conductance of 40 pS. The mean lifetime of single channel openings at 0 mV was 1.6 ms for ε-type and 4.4 ms for γ-type receptors, which closely corresponds to values found in native fetal and adult muscle, respectively. The different functional properties of junctional and nonjunctional nicotinic AChRs presumably reflect their specialized roles in synaptic transmission versus development and synapse formation.


FIGURE 8-8 Properties of fetal and adult AChRs from skeletal muscle. A, The results of patch-clamp experiments, with the patch pipettes in the outside-out configuration and the patch exposed to 0.5 µM ACh, are summarized. In the upper panel, recordings are from Xenopus oocytes that expressed the fetal AChR, which has the subunit composition α2βγδ. In the lower panel, the oocytes expressed the adult AChR, which has the subunit composition α2βεδ. Notice that the mean open times are greater for the fetal form, whereas the unitary currents are greater for the adult form. B, The two lines summarize data that are similar to those shown in A. The single channel conductance of the adult form (59 pS) is higher than that of the fetal form (40 pS). (Data from Mishina M, Takai T, Imoto K, et al: Molecular distinction between fetal and adult forms of muscle acetylcholine receptor. Nature 321:406–411, 1986.)

Humans have ten genes (CHRNA1 to CHRNA10) that encode homologous α subunits α1 to α10 of nicotinic ACh-activated receptors (see Fig. 6-20O). Humans also have four genes (CHRNB1 to CHRNB4) that encode β subunits β1 to β4, as well as separate genes encoding γ,δ, and ε subunits. The skeletal muscle receptor comprises two α1 subunits, one β1, one δ, and one ε in the adult receptor but one γ in the fetal receptor. Various heteromeric combinations of the other α subunits (α2 to α10) and β subunits (β2 to β4) produce a diverse array of functional receptor isoforms in neurons. Whereas muscle AChRs can be activated only by high concentrations of nicotine, the neuronal (α4)22)3 AChR isoform in the CNS and autonomic ganglia has the highest affinity for nicotine and is responsible for the behavioral and addictive effects of nicotine in tobacco. The α subunits α1, α7, and α9 bind a snake venom protein called α-bungarotoxin (see p. 226) from the Taiwanese banded krait.

The nicotinic AChRs belong to the pentameric Cys-loop receptor family of ligand-gated ion channels (see Table 6-2, family No. 11), so named because they contain a highly conserved pair of disulfide-bonded cysteine residues. This family also contains three other classes of agonist-activated channels, those activated by serotonin (5-HT3 receptor), glycine (GlyR), and GABA (GABAA receptor). As is the case for the AChRs, the 5-HT3 receptor channels are permeable to cations and thus produce excitatory currents. In contrast, glycine-activated and GABAA channels are permeable to anions such as Cl and produce inhibitory currents. Figure 8-9 shows examples of macroscopic and unitary Cl currents mediated by glycine-activated and GABAA channels. Cloned genes encoding subunits of these receptor channels encode proteins that are homologous to AChR subunits. Their primary amino-acid sequences share a common arrangement of M1, M2, M3, and M4 transmembrane segments, as described above for the nicotinic AChR (see Fig. 8-7). Sequence analysis of these genes indicates that they evolved from a common ancestor. The basis for cation versus anion selectivity appears to reside solely within the M2 segment. Mutation of only three residues within the M2 segment of a cation-selective α subunit of a neuronal nicotinic AChR is sufficient to convert it to an anion-selective channel activated by ACh.


FIGURE 8-9 Currents activated by glycine and GABA. A, These data are from experiments performed on cultured mouse spinal cord neurons using patch-clamp techniques. The left panel shows the macroscopic Cl current, which is measured in the whole-cell configuration and carried by glycine receptor (GlyR) channels when exposed to glycine. The right panel shows single channel currents that are recorded using the outside-out patch configuration. In both scenarios, the holding potential was −70 mV. B, The left panel shows the macroscopic Cl current that is carried by GABAA receptor channels when exposed to GABA. The right panel shows single channel currents. (Data from Bormann J, Hamill OP, Sakmann B: Mechanism of anion permeation through channels gated by glycine and γ-aminobutyric acid in mouse spinal neurones. J Physiol 385:243–286, 1987.)

Activation of AChR channels requires binding of two ACh molecules

The EPC is the sum of many single channel currents, each representing the opening of a single AChR channel at the neuromuscular junction. Above we described the random opening and closing of an idealized channel in a two-state model in which the channel could be either closed or open (see p. 181):


In the case of an agonist-activated channel, such as the AChR channel, binding of an agonist to the channel in its closed state favors channel opening. This gating process may be represented by the following kinetic model:



In this two-step scheme, the closed state (C) of the channel must bind one molecule of the agonist ACh to form a closed agonist-bound channel (AC) before it can convert to an open agonist-bound channel (AO). However, studies of the dependence of the probability of channel opening on agonist concentration indicate that binding of two molecules of ACh is required for channel opening. This feature of nicotinic receptor gating is described by the following modification of Equation 8-2:



The kinetics of channel opening are important for understanding the mechanism by which certain channel inhibitors work. For example, a competitive antagonist such as d-tubocurarine prevents the binding of the agonist ACh to each of its two sites. However, many noncompetitive antagonists of the AChR channel, including some local anesthetics, act by entering the lumen of the channel and blocking the flow of ionic current. Figure 8-10A shows the results of a patch-clamp experiment in which a single AChR channel opened and closed in response to its agonist, ACh. After the addition of QX-222, an analog of the local anesthetic agent lidocaine (see pp. 187–189), to the extracellular side, the channel exhibits a rapidly flickering behavior. This flickering represents a series of brief interruptions of the open state by numerous closures (see Fig. 8-10B). This type of flickering block is caused by rapid binding and unbinding of the anesthetic drug to a site in the mouth of the open channel. When the drug binds, it blocks the channel to the flow of ions (A2B). Conversely, when the drug dissociates, the channel becomes unblocked (A2O):



FIGURE 8-10 Effect of a local anesthetic on the AChR. A, Single channel recording of nicotinic AChR expressed in a Xenopus oocyte. The patch was in the outside-out configuration, and the holding potential was −150 mV. The continuous presence of 1 µM ACh caused brief channel openings. B, Recordings from an experiment similar to that in A except that, in addition to the ACh, the lidocaine analog QX-222 (20 µM) was present at the extracellular surface of the receptor channel. Note that the channel opening is accompanied by rapid flickering caused by many brief channel closures. The time scale of the lower panel is expanded 10-fold. (Data from Leonard RJ, Labarca CG, Charnet P, et al: Evidence that the M2 membrane-spanning region lines the ion channel pore of the nicotinic receptor. Science 242:1578–1581, 1988.)

Channel blockers are often used as molecular tools to study the mechanism of ion permeation. For example, studies with the blocker QX-222—in combination with site-directed mutagenesis—helped to localize the blocker binding site on the M2 transmembrane segment, thus identifying residues that line the aqueous pore.

Miniature EPPs reveal the quantal nature of transmitter release from the presynaptic terminals

Under physiological conditions, an action potential in a presynaptic motor nerve axon produces a depolarizing postsynaptic EPP that peaks at a level ~40 mV more positive than the resting Vm. This large signal results from the release of ACh from only about 200 synaptic vesicles, each containing 6000 to 10,000 molecules of ACh. The neuromuscular junction is clearly designed for excess capacity inasmuch as a single end plate is composed of numerous synaptic contacts (~1000 at the frog muscle end plate), each with an active zone that is lined with dozens of mature synaptic vesicles. Thus, a large inventory of ready vesicles (>104), together with the ability to synthesize ACh and to package it into new vesicles, allows the neuromuscular junction to maintain a high rate of successful transmission without significant loss of function as a result of presynaptic depletion of vesicles or ACh.

The original notion of a vesicular mode of transmitter delivery is based on classic observations of EPPs under conditions of reduced ACh release. In 1950, Fatt and Katz observed an interesting kind of electrophysiological “noise” in their continuous high-resolution recordings of Vm with a microelectrode inserted at the end-plate region of a frog muscle fiber. Their recordings from resting muscle fibers that were not subjected to nerve stimulation revealed the occurrence of tiny depolarizations of ~0.4 mV that appeared at random intervals. These small depolarizations were blocked by the AChR antagonist curare and they increased in size and duration with application of the AChE inhibitor neostigmine. Because the spontaneous Vm fluctuations also exhibited a time course similar to that of the normal EPP, they were named miniature end-plate potentials (also known as MEPPs or minis). These observations suggested that even in the absence of nerve stimulation, there is a certain low probability of transmitter release at the presynaptic terminal, resulting in the opening of a small number of AChRs in the postsynaptic membrane. An examination of the size of individual MEPPs suggested that they occur in discrete multiples of a unitary amplitude. This finding led to the notion that ACh release is quantized, with the quantal event corresponding to ACh release from one synaptic vesicle.

Another way of studying the quantal release of ACh is to stimulate the presynaptic motor neuron and to monitor Vm at the end plate under conditions in which the probability of ACh release is greatly decreased by lowering [Ca2+]o and increasing [Mg2+]o. Low [Ca2+]o decreases Ca2+ entry into the presynaptic terminal (see Fig. 8-2, step 3). High [Mg2+]o partially blocks the presynaptic Ca2+ channels and thus also decreases Ca2+ entry. Therefore, the consequence of either decreased [Ca2+]o or increased [Mg2+]o is a fall in [Ca2+]i in the presynaptic terminal, which reduces transmitter release and thus the amplitude of the EPP (Fig. 8-11). Del Castillo and Katz exploited this suppression of transmitter release under conditions of low [Ca2+]o and high [Mg2+]o to observe the Vm changes caused by the quantal release of transmitter. Figure 8-12A shows seven superimposed records of MEPPs that were recorded from a frog muscle fiber during seven repetitive trials of nerve stimulation under conditions of reduced [Ca2+]o and elevated [Mg2+]o. The records are aligned at the position of the nerve stimulus artifact. The amplitudes of the peak responses occur in discrete multiples of ~0.4 mV. Among the seven records were one “nonresponse,” two responses of ~0.4 mV, three responses of ~0.8 mV, and one response of ~1.2 mV. One of the recordings also revealed a spontaneous MEPP with a quantal amplitude of ~0.4 mV that appeared later in the trace. Del Castillo and Katz proposed that the macroscopic EPP is the sum of many unitary events, each having a magnitude of ~0.4 mV (see Fig. 8-12B). Microscopic observation of numerous vesicles in the synaptic terminal naturally led to the hypothesis that a single vesicle releases a relatively fixed amount of ACh and thereby produces a unitary MEPP. According to this view the quantized MEPPs thus correspond to the fusion of discrete numbers of synaptic vesicles: 0, 1, 2, 3, and so on. imageN8-5


FIGURE 8-11 Effect of extracellular Ca2+ and Mg2+ on EPPs. The data obtained by stimulating the motor neuron and monitoring the evoked subthreshold EPP show that the EPP is stimulated by increasing levels of Ca2+ but inhibited by increasing levels of Mg2+. (Data from Dodge FA Jr, Rahaminoff R: Cooperative action of calcium ions in transmitter release at the neuromuscular junction. J Physiol 193:419–432, 1967.)


FIGURE 8-12 Evoked and spontaneous MEPPs. A, The investigators recorded Vm in frog skeletal muscle fibers that were exposed to extracellular solutions having a [Ca2+] of 0.5 mM and [Mg2+] of 5 mM. These values minimize transmitter release, and therefore it was possible to resolve the smallest possible MEPP, which corresponds to the release of a single synaptic vesicle (i.e., 1 quantum). The investigators stimulated the motor neuron seven consecutive times and recorded the evoked MEPPs. In one trial, the stimulus evoked no response (0 quanta). In two trials, the peak MEPP was about 0.4 mV (1 quantum). In three others, the peak response was about 0.8 mV (2 quanta). Finally, in one, the peak was about 1.2 mV (3 quanta). In one case, a MEPP of the smallest magnitude appeared spontaneously. B, The histogram summarizes data from 198 trials on a cat neuromuscular junction in the presence of 12.5 mM extracellular Mg2+. The data are in bins with a width of 0.1 mV. The distribution has eight peaks. The first represents stimuli that evoked no responses. The other seven represent stimuli that evoked MEPPs that were roughly integral multiples of the smallest MEPP. The curve overlying each cluster of bins is a gaussian or “normal” function and facilitates calculation of the average MEPP for each cluster of bins. The peak values of these gaussians follow a Poisson distribution. (Data from Magleby KL: Neuromuscular transmission. In Engel AG, Franzini-Armstrong C [eds]: Myology: Basic and Clinical, 2nd ed. New York, McGraw-Hill, 1994, pp 442–463.)


Quantal Nature of Transmitter Release

Contributed by Ed Moczydlowski

The quantal nature of transmitter release can be expressed quantitatively by postulating that a nerve terminal contains a population of N quanta or vesicles and that each has a finite probability (P) of releasing under any given set of conditions. Thus, the mean number (m) of quanta released after any single nerve impulse is


Figure 8-12B in the text illustrates the results of an experiment very similar to that producing the data shown in Figure 8-12A, except that the investigators—Boyd and Martin—repeated the nerve stimulation 198 times, rather than 7 times as in Figure 8-12A. In each case, Boyd and Martin recorded the magnitude of the MEPP and placed it into a “bin” that was 0.1 mV wide. Thus, if they observed a MEPP of 1.23 mV, they placed it into the 1.2 bin. Figure 8-12B, a histogram summarizing the results of the 198 nerve-evoked responses, shows a series of peaks. The peak at 0 mV corresponds to the 18 trials in which the nerve stimulus evoked no end-plate potential. The peaks labeled I, II, III, and so on correspond to MEPPs that are multiples of the unit event—which is 0.4 mV—at amplitudes of 0.4 mV, 0.8 mV, 1.2 mV, and so forth. Thus, peak I corresponds to 1 quantum released, peak II corresponds to 2 quanta released, and so on.

If we sum up all the MEPPs in the 198 trials, we see that the total change in Vm was 184 mV. Dividing by 198 produces the mean amplitude of the MEPPs, 0.93 mV. If we assume a unitary response of 0.4 mV, 0.93 mV corresponds to 2.3 quanta, which is the m in Equation NE 8-3. Thus, on average, a nerve impulse produces a MEPP of 0.93 mV, which corresponds to the release of 2.3 quanta. However, in any given nerve impulse, the actual MEPP—if we could measure it with perfect accuracy—must correspond to an integral number of quanta released (x = 0, 1, 2, 3, …). Of course, because of noise and inaccuracies in the measuring system, Boyd and Martin also measured MEPPs that corresponded to nonintegral numbers of quanta. The y-axis in Figure 8-12B gives the number of times Boyd and Martin observed a given MEPP, out of the total of 198 observations. The seven bell-shaped or gaussian curves in Figure 8-12B represent the probability of releasing 1, 2, 3, 4, 5, 6, or 7 quanta.

Because each bin is 0.1 mV wide, and because the unitary MEPP is 0.4 mV, Boyd and Martin added up 0.4/0.1 or four consecutive bins to obtain the number of observations (nx) corresponding to the release of x quanta, out of the total of 198 observations (ntotal). For example, for x = 0 quanta, n0 was 18; for x = 1 quantum, n1 was 44; the second column in eTable 8-1 in this webnote gives the number of events observed (nx) for each number of quanta x (listed in the first column). The probability (px) that we saw x quanta being released after a single nerve impulse is

eTABLE 8-1

Poisson Distribution of Quanta Released During Nerve Stimulation

Number of Quanta (x)

Number of Events Observed (nx)

Probability Observed

Probability Predicted






































Thus, for x = 0, p0 would be 18/198 or 0.091; for x = 1, p1 would be 44/198 or 0.222; the other values are given in the third column of eTable 8-1.

How do these observed values agree with those predicted by probability theory? Probability theory predicts that px should follow a Poisson distribution:


Note that m in this equation is once again the mean number of quanta released per nerve impulse, 2.3 in our example. This theory assumes that the underlying probability of vesicle release (P in Equation NE 8-3) is very small and that the population of replenishable vesicles (N in Equation NE 8-3) is very large. The fourth column of eTable 8-1 shows that the px predicted by Equation NE 8-5 is very nearly the same as the observed px for each number of quanta.

We can also check the agreement of the data with the theory by testing whether the observed number of blank records (0-mV events) can predict the mean number (m) of quanta released after any single nerve impulse. According to Equation NE 8-5, when x = 0,


Because p0 is 18/198 or 0.091, the m value that we compute from Equation NE 8-6 is 2.4 quanta. This value is very close to the measured mean of 2.3 quanta. Findings such as these have provided strong support for the quantal theory of neurotransmitter release at the neuromuscular junction.


Boyd IA, Martin AR. The end-plate potential in mammalian muscle. J Physiol. 1956;132:74–91.

For elucidating the mechanism of synaptic transmission at the neuromuscular junction, Bernard Katz shared the 1970 Nobel Prize in Physiology or Medicine. imageN8-6


Sir Bernard Katz

For more information about Bernard Katz and the work that led to his Nobel Prize, visit (accessed October 2014).

Direct sensing of extracellular transmitter also shows quantal release of transmitter

Instead of using the postsynaptic AChR as a biological detector of quantum release, one can use a microscopic electrochemical sensor to measure neurotransmitter levels directly. Figure 8-13 shows results from an experiment in which a fine carbon-fiber electrode was placed very close to the presynaptic terminal membrane of a leech neuron that uses serotonin as its only neurotransmitter. The carbon fiber is an electrochemical detector of serotonin (see Fig. 8-13A); the current measured by this electrode corresponds to four electrons per serotonin molecule oxidized at the tip. Stimulation of the leech neuron to produce an action potential also elicits an oxidation current, as measured by the carbon fiber, that corresponds to the release of serotonin. At a [Ca2+]o of 5 mM, the current is large and composed of many small spikes (see Fig. 8-13B, top). On the other hand, reducing [Ca2+]o to 1 mM—presumably reducing Ca2+ influx into the nerve terminal and thus reducing the number of quanta released—reveals individual spikes of serotonin release. The release spikes come in two sizes, small and large (see Fig. 8-13B, bottom), corresponding to two separate classes of synaptic vesicles that are evident on electron micrographs. Injection of the cell with tetanus toxin, which blocks the release of synaptic vesicles, abolishes the serotonin release spikes. Thus, the spikes represent genuine events of synaptic exocytosis.


FIGURE 8-13 Detection of serotonin that is released from synaptic vesicles. A, The serotonin that is released from a synaptic terminal of a leech neuron can be detected electrochemically by use of a carbon-fiber microelectrode. The current carried by the carbon fiber increases with the amount of serotonin that is released, reflecting the oxidation of serotonin molecules on the surface of the carbon fiber. B, The top panel shows the action potential recorded from the stimulated motor neuron. The middle panel shows the evoked serotonin release (measured as a current) at both a [Ca2+]o of 5 mM (high level of serotonin release) and a [Ca2+]o of 1 mM (lower level or release). The bottom panel shows results of five consecutive trials at a [Ca2+]o of 1 mM and illustrates that the release of serotonin can occur in either small quanta or very large quanta. These two sizes of quanta correspond to small clear vesicles and large dense-core vesicles, both of which can be observed by electron microscopy. (Data from Bruns D, Jahn R: Real-time measurement of transmitter release from single synaptic vesicles. Nature 377:62–65, 1995.)

The nearly immediate appearance of the small release spikes after electrical stimulation of the cell shows that this type of vesicular release is extremely rapid. From the height and duration of the small and large spikes in Figure 8-13B, one can estimate the amount of electrical charge and thus the number of serotonin molecules oxidized at the carbon fiber per spike. A unitary small event corresponds to the release of ~4700 serotonin molecules, whereas a unitary large event corresponds to the release of 15,000 to 300,000 serotonin molecules. Thus, the amount of serotonin released by the small synaptic vesicles of the leech neuron is about half the number of ACh molecules contained in a synaptic vesicle at the frog neuromuscular junction. These and other observations of the synaptic function of nerve-muscle and nerve-nerve synapses have led to the conclusion that chemical neurotransmission operates by a fundamentally similar mechanism at many types of synapses in different animal species (see Chapter 13).

Short-term or long-term changes in the relative efficiency of neurotransmitter release can increase or decrease the strength of a particular synapse and thereby give rise to an alteration in behavior. Three types of synaptic modulation occur at the neuromuscular junction, and they differ in how they affect the quantal release of neurotransmitter: imageN8-7


Modulation of Quantal Release

Contributed by Ed Moczydlowski

As discussed in imageN8-5, the quantal nature of transmitter release can be expressed quantitatively by postulating that a nerve terminal contains a population of N quanta or vesicles and that each has a finite probability (P) of releasing under any given set of conditions. Thus, the mean number (m) of quanta released after any single nerve impulse is


As noted in the text, facilitation is a short-lived enhancement of the postsynaptic potential in response to a brief increase in the frequency of nerve stimulation. One way facilitation may occur is by a transient increase in the mean number of quanta per nerve stimulus, corresponding to an increase in the m parameter of Equation NE 8-7.

Potentiation is a long-lived and pronounced increase in transmitter release that occurs after a long period of high-frequency nerve stimulation. This effect can last for minutes after the conditioning stimulus. Potentiation may be caused by a period of intense nerve firing, which increases [Ca2+]i in the presynaptic terminal and thus increases the probability of exocytosis (the P parameter in Equation NE 8-7).

Synaptic depression is a transient decrease in the efficiency of transmitter release—and, consequently, a reduction in the postsynaptic potential—in response to a period of frequent nerve stimulation. Depression may result from a temporary depletion of transmitter-loaded vesicles in the presynaptic terminal—that is, a reduction in the number of available quanta, corresponding to the parameter N in Equation NE 8-7.

Thus, these three temporal changes in synaptic strength and efficiency appear to reflect changes at different steps of synaptic transmission. Similar modulation of synaptic strength in the CNS provides a mechanistic paradigm for understanding how individual nerve terminals may “learn.”

Facilitation is a short-lived enhancement of the EPP in response to a brief increase in the frequency of nerve stimulation. One way that facilitation may occur is by a transient increase in the mean number of quanta per nerve stimulus.

Potentiation (or post-tetanic potentiation) is a long-lived and pronounced increase in transmitter release that occurs after a long period of high-frequency nerve stimulation. This effect can last for minutes after the conditioning stimulus. Potentiation may be caused by a period of intense nerve firing, which increases [Ca2+]i in the presynaptic terminal and thus increases the probability of exocytosis.

Synaptic depression is a transient decrease in the efficiency of transmitter release and, consequently, a reduction in the EPP in response to a period of frequent nerve stimulation. Depression may result from temporary depletion of transmitter-loaded vesicles from the presynaptic terminal, that is, a reduction in the number of available quanta. Thus, these three temporal changes in synaptic strength and efficiency appear to reflect changes at different steps of synaptic transmission. Similar modulation of synaptic strength in the CNS provides a mechanistic paradigm to understand how individual nerve terminals may “learn” and “remember” (see p. 328).

Synaptic vesicles package, store, and deliver neurotransmitters

The physiology of synaptic vesicles in the nervous system is a variation on the universal theme used by endocrine or secretory cells in animals from the most primitive invertebrates up to mammals (see pp. 35–37). Many proteins involved in synaptic vesicle movement and turnover are related to those involved in intracellular membrane trafficking in almost all eukaryotic cells. This trafficking involves vesicular translocation from the endoplasmic reticulum to the Golgi network and fusion with the plasma membrane. Genetic analysis of the yeast secretory pathway has identified various gene products that are homologous to those associated with synaptic vesicles of higher vertebrates. Thus, the processes underlying synaptic function are inherently quite similar to cellular exocytosis and endocytosis.

As shown in Figure 8-14nascent synaptic vesicles are produced in the neuronal cell body by a process similar to that in the secretory pathway (see pp. 34–35). The membrane proteins of synaptic vesicles are synthesized in the rough endoplasmic reticulum and are then directed to the Golgi network, where processing, maturation, and sorting occur. Nascent synaptic vesicles—which are, in fact, secretory vesicles—are then transported to the nerve terminal by fast axonal transport (see p. 25) mediated by the microtubule system, which also carries mitochondria to the terminal.


FIGURE 8-14 Synthesis and recycling of synaptic vesicles and their content.

Vesicles destined to contain peptide neurotransmitters travel down the axon with the presynthesized peptides or peptide precursors already inside. On arrival at the nerve terminal (see Fig. 8-14), the vesicles—now called dense-core secretory granules (100 to 200 nm in diameter)—become randomly distributed in the cytoplasm of the terminal (see p. 310).

Vesicles destined to contain nonpeptide neurotransmitters (e.g., ACh) travel down the axon with no transmitter inside. On arrival at the nerve terminal (see Fig. 8-14), the vesicles take up the nonpeptide neurotransmitter, which is synthesized locally in the nerve terminal. These nonpeptide clear synaptic vesicles (40 to 50 nm in diameter) then attach to the actin-based cytoskeletal network. At this point, the mature clear synaptic vesicles are functionally ready for Ca2+-dependent transmitter release and become docked at specific release sites in the active zones of the presynaptic membrane. After exocytotic fusion of the clear synaptic vesicles, endocytosis via clathrin-coated vesicles (see pp. 40–42) recovers membrane components and recycles them to an endosome compartment in the terminal. Synaptic vesicles may then be re-formed within the terminal for reuse in neurotransmission, or they may be transported back to the cell body for turnover and degradation.

The concentrative uptake of nonpeptide neurotransmitters into clear synaptic vesicles is accomplished by the combination of a vacuolar-type H-ATPase and various neurotransmitter transport proteins (Fig. 8-15, top). The vacuolar-type H pump (see pp. 118–119) is a large, multisubunit complex that catalyzes the inward movement of H+ into the vesicle, coupled to the hydrolysis of cytosolic ATP to ADP and inorganic phosphate. The resulting pH and voltage gradients across the vesicle membrane energize the uptake of neurotransmitters into the vesicle via three families of neurotransmitter transport proteins that mediate the exchange of neurotransmitters in the cytosol for H+ in the vesicle (see Table 5-4). The SLC18 family includes members specific for monoamines (e.g., epinephrine, norepinephrine, dopamine, serotonin, histamine) and ACh. Members of the SLC17 family transport glutamate, and the SLC32 family handles GABA and glycine.


FIGURE 8-15 Model of fusion and exocytosis of synaptic vesicles. imageN8-15 SNARE, SNAreceptor.


Proteins Involved in Fusion and Endocytosis of Synaptic Vesicles

Contributed by Ed Moczydlowski

SNAP stands for soluble NSF attachment protein. imageN8-8

SNARE proteins are so-named because they act as receptors for SNAP. Thus, the acronym SNARE is a concatenation of SNAP and REceptor.

NSF stands for N-ethylmaleimide sensitive factor. It is a homohexameric ATPase enzyme that functions in the dissociation and recycling of SNARE complexes after vesicle fusion has occurred.

Synaptobrevin is also known as VAMP (vesicle-associated membrane protein).

Many proteins are involved in the fusion and recycling of synaptic vesicles (see Fig. 8-15). The SNARE proteins (see p. 37), which also participate in the secretory pathway, comprise the force-generating molecular machinery for membrane-membrane fusion. One SNARE protein—named synaptobrevin or VAMP—is a “v” SNARE because it is in the vesicle membrane. The two other SNARE proteins—one called syntaxin-1 and the other, SNAP-25 (synaptosome-associated protein, 25 kDa)—are “t” SNAREs because they are in the target (i.e., presynaptic) membrane. Synaptobrevin of the vesicle membrane and syntaxin-1 of the presynaptic membrane are anchored in the presynaptic membrane by single membrane-spanning segments. On the other hand, the presynaptic SNAP-25—which exists as a dual-helix bundle—is tethered to the presynaptic membrane by palmitoyl lipid chains. In the next section, we discuss how the SNAREs produce fusion.

Synaptotagmin is a synaptic vesicle protein that has, at its cytosolic end, two repetitive domains that are homologous to the C2 domain of protein kinase C. It is the Ca2+-sensor for exocytosis. Rab3 is a member of a large family of low-molecular-weight GTP-binding proteins that appears to be universally involved in cellular membrane trafficking (see p. 37) via the binding and hydrolysis of GTP, and it also regulates synaptic release.

Neurotransmitter release occurs by exocytosis of synaptic vesicles

Although the mechanism by which synaptic vesicles fuse with the plasma membrane and release their contents is still not completely understood, we have working models (see Fig. 8-15) for the function of various key components and steps involved in synaptic vesicle release. These models are based on a variety of in vitro experiments. The use of specific toxins that act at nerve synapses and elegant functional studies of genetic mutants in Drosophila, Caenorhabditis elegans, and gene knockout mice have provided important information on the role of various components.

We have already introduced key proteins located in the synaptic vesicle, including the v-SNAREs synaptobrevin and the Ca2+ sensor synaptotagmin. We also have introduced the two t-SNAREs syntaxin-1 and SNAP-25 (see p. 37) in the target area of the presynaptic membrane. These are also essential for the fusion process. As discussed below on pages 224–225, tetanus toxin and certain botulinum toxins are endoproteinases that cleave synaptobrevin, whereas other botulinum endoproteinases cleave syntaxin-1 and SNAP-25. These toxins thus block the fusion of synaptic vesicles. imageN8-8


“SNAP” Nomenclature

Contributed by Emile Boulpaep, Walter Boron

Unfortunately, SNAP means different things to different people: “SNAP” in SNAP-25 means “synaptosome-associated protein, 25 kDa,” and “SNAP” in α-SNAP means “soluble NSF-attachment protein.”

As summarized in Figure 8-15, after docking of the vesicle near the presynaptic membrane, Sec-1/Munc18, Munc13, and RIM catalyze assembly of the partial SNARE complex. The free helical ends of synaptobrevin, syntaxin, and SNAP-25 begin to coil around each other to form a four-helix bundle—the trans-SNARE complex—formed by four ~70–amino-acid SNARE helix motifs, one from synaptobrevin, one from syntaxin-1, and two from SNAP-25. The result, called priming stage 1, is a ternary SNARE complex with an extraordinarily stable rod-shaped structure of intertwined α helices. As the energetically favorable coiling of the three SNAREs continues in a zipper-like process, the vesicle membrane is pulled ever closer to the presynaptic membrane. Next, a cytosolic protein called complexin inserts into the trans-SNARE complex, preventing spontaneous fusion. The result is priming stage 2.

As Ca2+ enters through voltage-gated Ca2+ channels (see Table 7-2)—located in register with the active zone of the presynaptic membrane—it binds to multiple sites on the C2 domains of synaptotagmin. These Ca2+-bound C2 domains promote the binding of synaptotagmin to acidic phospholipids in the presynaptic membrane and also displace the complexin, thereby reversing the block to fusion. These events trigger the actual membrane fusion event, accompanied by fusion-pore opening and the beginning of transmitter release. The role of the synaptotagmin as the Ca2+ sensor is supported by experiments with knockout mice and Drosophila mutants showing that the absence of the appropriate isoform of this protein results in impaired Ca2+-dependent transmitter release.

Following fusion completion, as the plasma-membrane Ca-ATPase (PMCA) extrudes Ca2+ across the plasma membrane and as mitochondria take up Ca2+, [Ca2+]i rapidly falls, causing synaptotagmin to dissociate from the tightly wound SNARE complex. The soluble α-SNAP (soluble NSF attachment protein) imageN8-8 binds to the SNARE complex and promotes the binding of NSF (N-ethylmaleimide–sensitive factor, a homohexameric ATPase), which uses the energy of ATP hydrolysis to disassemble the SNAREs. The now-free synaptobrevin presumably undergoes recycling endocytosis, whereas the syntaxin and SNAP-25 on the presynaptic membrane are available for the next round of vesicle fusion.

The model just presented leaves unanswered some important questions. For example, what is the exact structure of the fusion pore detected by electrophysiological measurements as a primary event in membrane fusion? Also, precisely how do the many regulatory and scaffolding proteins control the numerous conformational changes that accompany the membrane fusion event? Physiologists are very interested in the details of synaptic vesicle fusion because regulation of the exocytotic process is a target for control of the strength of synaptic transmission and is undoubtedly involved in synaptic plasticity phenomena responsible for changes in animal behavior. For their work on vesicle trafficking, including fusion of synaptic vesicle, James Rothman, Randy Schekman, and Thomas Südhof shared the Nobel Prize in Physiology or Medicine in 2013. imageN8-16


James Rothman, Randy Schekman, and Thomas Südhof

For more information about James Rothman, Randy Schekman, and Thomas Südhof and the work that led to their Nobel Prize, visit (accessed October 2015).

Re-uptake or cleavage of the neurotransmitter terminates its action

Effective transmission across chemical synapses requires not only release of the neurotransmitter and activation of the receptor on the postsynaptic membrane but also rapid and efficient mechanisms for removal of the transmitter. At synapses where ACh is released, this removal is accomplished by enzymatic destruction of the neurotransmitter. However, the more general mechanism in the nervous system involves re-uptake of the neurotransmitter mediated by specific high-affinity transport systems located in the presynaptic plasma membrane and surrounding glial cells. These secondary active transport systems use the normal ionic gradients of Na+, K+, H+, or Cl to achieve concentrative uptake of transmitter. Vertebrates have two distinct families of neurotransmitter transport proteins. The first family is characterized by a common motif of 12 membrane-spanning segments and includes transporters with specificity for catecholamines, serotonin, GABA, glycine, and choline. Energy coupling of transport in this class of systems is generally based on cotransport of the substrate with Na+ and Cl. The second family is represented by transporters for the excitatory amino acids glutamate and aspartate; in these systems, substrate transport generally couples to cotransport of Na+ and H+ and to exchange of K+.

At the neuromuscular junction and other cholinergic synapses, immediate termination of the action of ACh is accomplished enzymatically by the action of AChE. Although AChE is primarily found at the neuromuscular junction, AChE activity can be detected throughout the nervous system. The enzyme occurs in a variety of physical forms. The globular or G forms exist as monomers, dimers, or tetramers of a common ~72-kDa glycoprotein catalytic subunit. These molecules can be found either in soluble form or bound to cell membranes via a GPI linkage (see p. 13) in which a post-translational modification attaches the C terminus of the protein to a glycolipid moiety. The asymmetric or A forms consist of one to three tetramers of the globular enzyme coupled via disulfide bond linkage to a collagen-like structural protein. The largest asymmetric form, which has 12 catalytic subunits attached to the collagen-like tail, is the major species located at the neuromuscular junction. The triple-helical, collagen-like tail attaches the asymmetric AChE complex to extracellular matrix components of the synaptic basal lamina. The various physical forms of AChE are a result of the alternative splicing that occurs in the transcription of a single AChE gene. imageN8-9



Contributed by Ed Moczydlowski

The acetylcholinesterase (AChE) enzyme is an ellipsoidal globular protein, approximately 4.5 nm × 6.0 nm × 6.5 nm. It includes a central 12-stranded β sheet surrounded by 14 α-helical segments. The active site of the enzyme is composed of three residues (Ser200, His440, and Glu327) located on different loops. These residues are analogous to the Ser-His-Asp catalytic triad of serine proteases such as trypsin and chymotrypsin. This similarity is an example of convergent evolution, inasmuch as there is little structural similarity between the two types of enzymes. A unique aspect of the structure of AChE is that the active site of ACh hydrolysis is located at the bottom of a 2.0-nm-deep gorge (the active site gorge) that the substrate must enter by diffusion from the surface of the protein. The three-dimensional structure of the catalytic subunit of AChE from the electric ray, Torpedo californica, has been solved by x-ray crystallography.

In the first step of the enzymatic reaction (see Equation 8-5), the H from the hydroxyl of Ser200 becomes attached to the oxygen in the ester linkage of ACh, which results in the formation of choline and a tetrahedral acyl-enzyme intermediate at Ser200.

In the second step of Equation 8-5, the hydrolysis of the acyl-enzyme yields acetate and the free enzyme.

The enzyme AChE rapidly hydrolyzes ACh to choline and acetate in a two-step process:



In the first step of the reaction, the enzyme cleaves choline from ACh, which results in the formation of a transient intermediate in which the acetate group is covalently coupled to a serine group on the enzyme. The second step is the hydrolysis and release of this acetate as well as the recovery of free enzyme. The nerve terminal recovers the extracellular choline via a high-affinity Na+-coupled uptake system and uses it for the synthesis of ACh (Box 8-1).

Box 8-1

Diseases of Neuromuscular Transmission

The term myasthenia means muscle weakness (from the Greek mys + asthenia) and is generally used clinically to denote weakness in the absence of a CNS disorder, neuropathy, or primary muscle disease. Thus, myasthenia can be due to any one of a wide range of aberrations of neuromuscular transmission.

Myasthenia Gravis

Myasthenia gravis, one specific type of myasthenia and the most common adult form, affects 25 to 125 of every 1 million people. It can occur at any age but has a bimodal distribution, with peak incidences occurring among people in their 20s and 60s. Those affected at an early age tend to be women with hyperplasia of the thymus. Those who are older are more likely to be men with coexisting cancer of the thymus gland. The cells of the thymus possess nicotinic AChRs, and the disease arises as a result of antibodies directed against these receptors. The antibodies then lead to skeletal muscle weakness caused in part by competitive antagonism of AChRs. Symptoms include fatigue and weakness of skeletal muscle. Two major forms of the disease are recognized: one that involves weakness of only the extraocular muscles and another that results in generalized weakness of all skeletal muscles. In either case, myasthenia gravis is typified by fluctuating symptoms, with weakness greatest toward the end of the day or after exertion. In severe cases, paralysis of the respiratory muscles can lead to death. Treatment directed at enhancing cholinergic transmission, alone or combined with thymectomy or immunosuppression, is highly effective in most patients.

Progress toward achieving an understanding of the cause of myasthenia gravis was first made when electrophysiological analysis of involved muscle revealed that the amplitude of the MEPP was decreased, although the frequency of quantal events was normal. This finding suggested either a defect in the postsynaptic response to ACh or a reduced concentration of ACh in the synaptic vesicles. A major breakthrough occurred in 1973, when Patrick and Lindstrom found that symptoms similar to those of humans with myasthenia developed in rabbits immunized with AChR protein purified from the electric eel. This finding was shortly followed by the demonstration of anti-AChR antibodies in human patients with myasthenia gravis and a severe reduction in the surface density of AChR in the junctional folds. The anti-AChR antibodies are directed against one or more subunits of the receptor, where they bind and activate complement and accelerate destruction of the receptors. The most common target of these antibodies is a region of the AChR α subunit called MIR (main immunogenic region).

Myasthenia gravis is now recognized to be an acquired autoimmune disorder in which the spontaneous production of anti-AChR antibodies results in progressive loss of muscle AChRs and degeneration of postjunctional folds. Treatment is aimed at either reducing the potency of the immunological attack or enhancing cholinergic activity within the synapse. Reduction of the potency of the immunological attack is achieved by the use of immunosuppressants (most commonly corticosteroids) or plasmapheresis (removal of antibodies from the patient's serum). Some patients with myasthenia gravis have a thymoma (a tumor of the thymus gland) that is often readily seen on routine chest radiographs. In these patients, removal of the thymoma leads to clinical improvement in nearly 75% of the cases. Enhancement of cholinergic activity is achieved via the use of AChE inhibitors; pyridostigmine is the most widely used agent. The dosage of these drugs must be carefully monitored to prevent overexposure of the remaining AChRs to ACh. Overexposure can lead to overstimulation of the postsynaptic receptors, prolonged depolarization of the postsynaptic membrane, inactivation of neighboring Na+ channels, and thus synaptic blockade.

Congenital Myasthenic Syndrome

Congenital myasthenic syndrome (CMS) refers to a variety of inherited disorders, present at birth, that affect neuromuscular transmission in a variety of ways. Because specific cases can involve abnormal presynaptic release of ACh, AChE deficiency, or defective AChR function (without the presence of antireceptor antibodies), the signs and symptoms can also vary widely.

In 1995, an unusual example of a CMS disorder was traced to a mutation in the ε subunit of the human AChR. Single channel recordings from biopsy samples of muscle fibers of a young myasthenic patient revealed a profound alteration in AChR kinetics. The burst duration of AChR openings was greatly prolonged in comparison with that of normal human AChR channels. The patient had a mutation of threonine to proline at position 264 in the adult ε subunit of the AChR. This amino-acid residue corresponds to an evolutionarily conserved position in the M2 membrane-spanning segment, which is involved in formation of the channel pore. Thus, a human mutation in the pore region of the AChR protein results in failure of the channel to close normally, thereby causing excessive depolarization and pathological consequences at the muscle end plate.

Many human mutations in α, β, δ, ε, and γ subunits of the nicotinic muscle AChR have since been characterized. These mutations have many different effects on AChR kinetics including slow and fast channel syndromes depending on where they occur in the receptor channel protein.

Lambert-Eaton Myasthenic Syndrome

Another condition characterized by progressive muscle weakness and fatigue is Lambert-Eaton myasthenic syndrome (LEMS; also called Lambert-Eaton syndrome), an impairment of presynaptic Ca2+channels at motor nerve terminals. LEMS is an autoimmune disorder, most often seen in patients with certain types of cancer, such as small-cell lung carcinoma. In LEMS, antibodies attack Cav2.2 (CACNA1B), which reduces Ca2+ entry during the presynaptic action potential and thus reduces ACh release. LEMS differs from myasthenia gravis as follows: (1) LEMS primarily attacks the limb muscles, not the ocular and bulbar muscles. (2) In LEMS, repetitive stimulation of a particular muscle (which leads to a progressive rise in [Ca2+]i) causes a gradual increase in the amplitude of the compound motor action potential (CMAP) in the stimulated muscle—as measured using electromyography (EMG)imageN8-10. In patients with myasthenia, repetitive stimulation leads to progressive lessening of the CMAP. Thus, repeated muscle stimulation leads to increasing contractile strength in patients with LEMS and to decreasing strength in patients with myasthenia.


The Electromyogram and the Compound Motor Action Potential

Contributed by Walter Boron

An electromyogram (EMG) is a record obtained using an instrument called an electromyograph. The technique—called electromyography—is used to assess the electrical activity generated by excited skeletal muscle cells.

compound motor action potential (CMAP) is the summation of many nearly simultaneous action potentials from many skeletal muscle fibers that are in the same vicinity. One would typically observe CMAPs during electromyography.


Wikipedia. s.v. Compound muscle action potential.

Wikipedia. s.v. Electromyography.