Ganong’s Review of Medical Physiology, 24th Edition

CHAPTER 6 Synaptic & Junctional Transmission


After studying this chapter, you should be able to:

image Describe the main morphologic features of synapses.

image Distinguish between chemical and electrical transmission at synapses.

image Describe fast and slow excitatory and inhibitory postsynaptic potentials, outline the ionic fluxes that underlie them, and explain how the potentials interact to generate action potentials.

image Define and give examples of direct inhibition, indirect inhibition, presynaptic inhibition, and postsynaptic inhibition.

image Describe the neuromuscular junction, and explain how action potentials in the motor neuron at the junction lead to contraction of the skeletal muscle.

image Define denervation hypersensitivity.


The “all-or-none” type of conduction seen in axons and skeletal muscle has been discussed in Chapters 4 and 5. Impulses are transmitted from one nerve cell to another cell at synapses. These are the junctions where the axon or some other portion of one cell (the presynaptic cell) terminates on the dendrites, soma, or axon of another neuron (Figure 6–1) or, in some cases, a muscle or gland cell (the postsynaptic cell). Cell-to-cell communication occurs across either a chemical or electrical synapse. At chemical synapses, a synaptic cleft separates the terminal of the presynaptic cell from the postsynaptic cell. An impulse in the presynaptic axon causes secretion of a chemical that diffuses across the synaptic cleft and binds to receptors on the surface of the postsynaptic cell. This triggers events that open or close channels in the membrane of the postsynaptic cell. In electrical synapses, the membranes of the presynaptic and postsynaptic neurons come close together, and gap junctions form between the cells (see Chapter 2). Like the intercellular junctions in other tissues, these junctions form low-resistance bridges through which ions can pass with relative ease. There are also a few conjoint synapses in which transmission is both electrical and chemical.


FIGURE 6–1 Synapses on a typical motor neuron. The neuron has dendrites (1), an axon (2), and a prominent nucleus (3). Note that rough endoplasmic reticulum extends into the dendrites but not into the axon. Many different axons converge on the neuron, and their terminal boutons form axodendritic (4) and axosomatic (5) synapses. (6) Myelin sheath. (Reproduced with permission from Krstic RV: Ultrastructure of the Mammalian Cell. Springer, 1979.)

Regardless of the type of synapse, transmission is not a simple transmission of an action potential from the presynaptic to the postsynaptic cell. The effects of discharge at individual synaptic endings can be excitatory or inhibitory, and when the postsynaptic cell is a neuron, the summation of all the excitatory and inhibitory effects determines whether an action potential is generated. Thus, synaptic transmission is a complex process that permits the grading and adjustment of neural activity necessary for normal function. Because most synaptic transmission is chemical, consideration in this chapter is limited to chemical transmission unless otherwise specified.

Transmission from nerve to muscle resembles chemical synaptic transmission from one neuron to another. The neuromuscular junction, the specialized area where a motor nerve terminates on a skeletal muscle fiber, is the site of a stereotyped transmission process. The contacts between autonomic neurons and smooth and cardiac muscle are less specialized, and transmission in these locations is a more diffuse process. These forms of transmission are also considered in this chapter.


The anatomic structure of synapses varies considerably in the different parts of the mammalian nervous system. The ends of the presynaptic fibers are generally enlarged to form terminal boutons or synaptic knobs (Figure 6–2). In the cerebral and cerebellar cortex, endings are commonly located on dendrites and frequently on dendritic spines, which are small knobs projecting from dendrites (Figure 6–3). In some instances, the terminal branches of the axon of the presynaptic neuron form a basket or net around the soma of the postsynaptic cell (eg, basket cells of the cerebellum). In other locations, they intertwine with the dendrites of the postsynaptic cell (eg, climbing fibers of the cerebellum) or end on the dendrites directly (eg, apical dendrites of cortical pyramidal cells). Some end on axons of postsynaptic neurons (axoaxonal endings). On average, each neuron divides to form over 2000 synaptic endings, and because the human central nervous system (CNS) has 1011 neurons, it follows that there are about 2 × 1014 synapses. Obviously, therefore, communication between neurons is extremely complex. Synapses are dynamic structures, increasing and decreasing in complexity and number with use and experience.


FIGURE 6–2 Electronmicrograph of synaptic knob (S) ending on the shaft of a dendrite (D) in the central nervous system. P, postsynaptic density; M, mitochondrion. (×56,000). (Courtesy of DM McDonald.)


FIGURE 6–3 Axodendritic, axoaxonal, and axosomatic synapses. Many presynaptic neurons terminate on dendritic spines, as shown at the top, but some also end directly on the shafts of dendrites. Note the presence of clear and granulated synaptic vesicles in endings and clustering of clear vesicles at active zones.

It has been calculated that in the cerebral cortex, 98% of the synapses are on dendrites and only 2% are on cell bodies. In the spinal cord, the proportion of endings on dendrites is less; there are about 8000 endings on the dendrites of a typical spinal neuron and about 2000 on the cell body, making the soma appear encrusted with endings.


Each presynaptic terminal of a chemical synapse is separated from the postsynaptic structure by a synaptic cleft that is 20–40 nm wide. Across the synaptic cleft are many neurotransmitter receptors in the postsynaptic membrane, and usually a postsynaptic thickening called the postsynaptic density (Figures 6–2 and 6–3). The postsynaptic density is an ordered complex of specific receptors, binding proteins, and enzymes induced by postsynaptic effects.

Inside the presynaptic terminal are many mitochondria, as well as many membrane-enclosed vesicles, which contain neurotransmitters. There are three kinds of synaptic vesicles: small, clear synaptic vesicles that contain acetylcholine, glycine, GABA, or glutamate; small vesicles with a dense core that contain catecholamines; and large vesicles with a dense core that contain neuropeptides. The vesicles and the proteins contained in their walls are synthesized in the neuronal cell body and transported along the axon to the endings by fast axoplasmic transport. The neuropeptides in the large dense-core vesicles must also be produced by the protein-synthesizing machinery in the cell body. However, the small clear vesicles and the small dense-core vesicles recycle in the nerve ending. These vesicles fuse with the cell membrane and release transmitters through exocytosis and are then recovered by endocytosis to be refilled locally. In some instances, they enter endosomes and are budded off the endosome and refilled, starting the cycle over again. The steps involved are shown in Figure 6–4. More commonly, however, the synaptic vesicle discharges its contents through a small hole in the cell membrane, then the opening reseals rapidly and the main vesicle stays inside the cell (kiss-and-run discharge). In this way, the full endocytotic process is short-circuited.


FIGURE 6–4 Small synaptic vesicle cycle in presynaptic nerve terminals. Vesicles bud off the early endosome and then fill with neurotransmitter (NT; top left). They then move to the plasma membrane, dock, and become primed. Upon arrival of an action potential at the ending, Ca2+ influx triggers fusion and exocytosis of the granule contents to the synaptic cleft. The vesicle wall is then coated with clathrin and taken up by endocytosis. In the cytoplasm, it fuses with the early endosome, and the cycle is ready to repeat. (Reproduced with permission from Südhof TC: The synaptic vesicle cycle: A cascade of protein-protein interactions. Nature 1995;375:645.)

The large dense-core vesicles are located throughout the presynaptic terminals that contain them and release their neuropeptide contents by exocytosis from all parts of the terminal. On the other hand, the small vesicles are located near the synaptic cleft and fuse to the membrane, discharging their contents very rapidly into the cleft at areas of membrane thickening called active zones (Figure 6–3). The active zones contain many proteins and rows of Ca2+channels.

The Ca2+ that triggers exocytosis of transmitters enters the presynaptic neurons, and transmitter release starts within 200 μs. Therefore, it is not surprising that the voltage-gated Ca2+ channels are very close to the release sites at the active zones. In addition, the transmitter must be released close to the postsynaptic receptors to be effective on the postsynaptic neuron. This orderly organization of the synapse depends in part on neurexins, proteins bound to the membrane of the presynaptic neuron that bind neurexin receptors in the membrane of the postsynaptic neuron. In many vertebrates, neurexins are produced by a single gene that codes for the α isoform. However, in mice and humans they are encoded by three genes, and both α and β isoforms are produced. Each of the genes has two regulatory regions and extensive alternative splicing of their mRNAs. In this way, over 1000 different neurexins are produced. This raises the possibility that the neurexins not only hold synapses together, but also provide a mechanism for the production of synaptic specificity.

As noted in Chapter 2, vesicle budding, fusion, and discharge of contents with subsequent retrieval of vesicle membrane are fundamental processes occurring in most, if not all, cells. Thus, neurotransmitter secretion at synapses and the accompanying membrane retrieval are specialized forms of the general processes of exocytosis and endocytosis. The details of the processes by which synaptic vesicles fuse with the cell membrane are still being worked out. They involve the v-snare protein synaptobrevin in the vesicle membrane locking with the t-snare protein syntaxin in the cell membrane; a multiprotein complex regulated by small GTPases such as Rab3 is also involved in the process (Figure 6–5). The one-way gate at the synapses is necessary for orderly neural function.


FIGURE 6–5 Main proteins that interact to produce synaptic vesicle docking and fusion in nerve endings. The processes by which synaptic vesicles fuse with the cell involve the v-snare protein synaptobrevin in the vesicle membrane locking with the t-snare protein syntaxin in the cell membrane; a multiprotein complex regulated by small GTPases such as Rab3 is also involved in the process. (Reproduced with permission from Ferro-Novick S, John R: Vesicle fusion from yeast to man. Nature 1994;370:191.)

Several deadly toxins that block neurotransmitter release are zinc endopeptidases that cleave and hence inactivate proteins in the fusion-exocytosis complex. Clinical Box 6–1 describes how neurotoxins from bacteria called Clostridium tetani and Clostridium botulinum can disrupt neurotransmitter release in either the CNS or at the neuromuscular junction.


Botulinum and Tetanus Toxins

Clostridia are gram-positive bacteria. Two varieties, Clostridium tetani and Clostridium botulinum, produce some of the most potent biological toxins (tetanus toxin and botulinum toxin) known to affect humans. These neurotoxins act by preventing the release of neurotransmitters in the CNS and at the neuromuscular junction. Tetanus toxin binds irreversibly to the presynaptic membrane of the neuromuscular junction and uses retrograde axonal transport to travel to the cell body of the motor neuron in the spinal cord. From there it is picked up by the terminals of presynaptic inhibitory interneurons. The toxin attaches to gangliosides in these terminals and blocks the release of glycine and GABA. As a result, the activity of motor neurons is markedly increased. Clinically, tetanus toxin causes spastic paralysis; the characteristic symptom of “lockjaw” involves spasms of the masseter muscle. Botulism can result from ingestion of contaminated food, colonization of the gastrointestinal tract in an infant, or wound infection. Botulinum toxins are actually a family of seven neurotoxins, but it is mainly botulinum toxins A, B, and E that are toxic to humans. Botulinum toxins A and E cleave synaptosome-associated protein (SNAP-25). This is a presynaptic membrane protein needed for fusion of synaptic vesicles containing acetylcholine to the terminal membrane, an important step in transmitter release. Botulinum toxin B cleaves synaptobrevin, a vesicle-associated membrane protein (VAMP). By blocking acetylcholine release at the neuromuscular junction, these toxins cause flaccid paralysis. Symptoms can include ptosis, diplopia, dysarthria, dysphonia, and dysphagia.


Tetanus can be prevented by treatment with tetanus toxoid vaccine. The widespread use of this vaccine in the U.S. beginning in the mid 1940s has led to a marked decline in the incidence of tetanus toxicity. The incidence of botulinum toxicity is also low (about 100 cases per year in the U.S.), but in those individuals that are affected, the fatality rate is 5–10%. An antitoxin is available for treatment, and those who are at risk for respiratory failure are placed on a ventilator. On the positive side, local injection of small doses of botulinum toxin (botox) has proven to be effective in the treatment of a wide variety of conditions characterized by muscle hyperactivity. Examples include injection into the lower esophageal sphincter to relieve achalasia and injection into facial muscles to remove wrinkles.



Penetration of an α-motor neuron is a good example of a technique used to study postsynaptic electrical activity. It is achieved by advancing a microelectrode through the ventral portion of the spinal cord. Puncture of a cell membrane is signaled by the appearance of a steady 70 mV potential difference between the microelectrode and an electrode outside the cell. The cell can be identified as a spinal motor neuron by stimulating the appropriate ventral root and observing the electrical activity of the cell. Such stimulation initiates an antidromic impulse (see Chapter 4) that is conducted to the soma and stops at that point. Therefore, the presence of an action potential in the cell after antidromic stimulation indicates that the cell that has been penetrated is an α-motor neuron. Stimulation of a dorsal root afferent (sensory neuron) can be used to study both excitatory and inhibitory events in α-motor neurons (Figure 6–6).



FIGURE 6–6 Excitatory and inhibitory synaptic connections mediating the stretch reflex provide an example of typical circuits within the CNS. A) The stretch receptor sensory neuron of the quadriceps muscle makes an excitatory connection with the extensor motor neuron of the same muscle and an inhibitory interneuron projecting to flexor motor neurons supplying the antagonistic hamstring muscle. B) Experimental setup to study excitation and inhibition of the extensor motor neuron. Top panel shows two approaches to elicit an excitatory (depolarizing) postsynaptic potential or EPSP in the extensor motor neuron–electrical stimulation of the whole Ia afferent nerve using extracellular electrodes and intracellular current passing through an electrode inserted into the cell body of a sensory neuron. Bottom panel shows that current passing through an inhibitory interneuron elicits an inhibitory (hyperpolarizing) postsynaptic potential or IPSP in the flexor motor neuron. (From Kandel ER, Schwartz JH, Jessell TM [editors]: Principles of Neural Science, 4th ed. McGraw-Hill, 2000.)

Once an impulse reaches the presynaptic terminals, a response can be obtained in the postsynaptic neuron after a synaptic delay. The delay is due to the time it takes for the synaptic mediator to be released and to act on the receptors on the membrane of the postsynaptic cell. Because of it, conduction along a chain of neurons is slower if there are many synapses compared to if there are only a few synapses. Because the minimum time for transmission across one synapse is 0.5 ms, it is also possible to determine whether a given reflex pathway is monosynaptic or polysynaptic (contains more than one synapse) by measuring the synaptic delay.

A single stimulus applied to the sensory nerves characteristically does not lead to the formation of a propagated action potential in the postsynaptic neuron. Instead, the stimulation produces either a transient partial depolarization or a transient hyperpolarization. The initial depolarizing response produced by a single stimulus to the proper input begins about 0.5 ms after the afferent impulse enters the spinal cord. It reaches its peak 11.5 ms later and then declines exponentially. During this potential, the excitability of the neuron to other stimuli is increased, and consequently the potential is called an excitatory postsynaptic potential (EPSP) (Figure 6–6).

The EPSP is produced by depolarization of the postsynaptic cell membrane immediately under the presynaptic ending. The excitatory transmitter opens Na+ or Ca2+ channels in the postsynaptic membrane, producing an inward current. The area of current flow thus created is so small that it does not drain off enough positive charge to depolarize the whole membrane. Instead, an EPSP is inscribed. The EPSP due to activity in one synaptic knob is small, but the depolarizations produced by each of the active knobs summate.

EPSPs are produced by stimulation of some inputs, but stimulation of other inputs produces hyperpolarizing responses. Like the EPSPs, they peak 11.5 ms after the stimulus and decrease exponentially. During this potential, the excitability of the neuron to other stimuli is decreased; consequently, it is called an inhibitory postsynaptic potential (IPSP) (Figure 6–6).

An IPSP can be produced by a localized increase in Cl transport. When an inhibitory synaptic knob becomes active, the released transmitter triggers the opening of Cl channels in the area of the postsynaptic cell membrane under the knob. Cl moves down its concentration gradient. The net effect is the transfer of negative charge into the cell, so that the membrane potential increases.

The decreased excitability of the nerve cell during the IPSP is due to movement of the membrane potential away from the firing level. Consequently, more excitatory (depolarizing) activity is necessary to reach the firing level. The fact that an IPSP is mediated by Cl can be demonstrated by repeating the stimulus while varying the resting membrane potential of the postsynaptic cell. When the membrane potential is at the equilibrium potential for chloride (ECl), the postsynaptic potential disappears (Figure 6–7), and at more negative membrane potentials, it becomes positive (reversal potential).


FIGURE 6–7 IPSP is due to increased Cl influx during stimulation. This can be demonstrated by repeating the stimulus while varying the resting membrane potential (RMP) of the postsynaptic cell. When the membrane potential is at ECl, the potential disappears, and at more negative membrane potentials (eg, EK and below), it becomes positive (reversal potential).

Because IPSPs are net hyperpolarizations, they can be produced by alterations in other ion channels in the neuron. For example, they can be produced by opening of K+ channels, with movement of K+ out of the postsynaptic cell, or by closure of Na+ or Ca2+ channels.


In addition to the EPSPs and IPSPs described previously, slow EPSPs and IPSPs have been described in autonomic ganglia, cardiac and smooth muscle, and cortical neurons. These postsynaptic potentials have a latency of 100–500 ms and last several seconds. The slow EPSPs are generally due to decreases in K+ conductance, and the slow IPSPs are due to increases in K+ conductance.


At synaptic junctions where transmission is electrical, the impulse reaching the presynaptic terminal generates an EPSP in the postsynaptic cell that, because of the low-resistance bridge between the two, has a much shorter latency than the EPSP at a synapse where transmission is chemical. In conjoint synapses, both a short-latency response and a longer-latency, chemically mediated postsynaptic response can occur.


The constant interplay of excitatory and inhibitory activity on the postsynaptic neuron produces a fluctuating membrane potential that is the algebraic sum of the hyperpolarizing and depolarizing activities. The soma of the neuron thus acts as an integrator. When the level of depolarization reaches the threshold voltage, a propagated action potential will occur. However, the discharge of the neuron is slightly more complicated than this. In motor neurons, the portion of the cell with the lowest threshold for the production of an action potential is the initial segment, the portion of the axon at and just beyond the axon hillock. This unmyelinated segment is depolarized or hyperpolarized electrotonically by the current sinks and sources under the excitatory and inhibitory synaptic knobs. It is the first part of the neuron to fire, and its discharge is propagated in two directions: down the axon and back into the soma. Retrograde firing of the soma in this fashion probably has value in wiping the slate clean for subsequent renewal of the interplay of excitatory and inhibitory activity on the cell.


Two passive membrane properties of a neuron affect the ability of postsynaptic potentials to summate to elicit an action potential (Figure 6–8). The time constant of a neuron determines the time course of the synaptic potential, and the length constant of a neuron determines the degree to which a depolarizing current is reduced as it spreads passively. Figure 6–8 also shows how the time constant of the postsynaptic neuron can affect the amplitude of the depolarization caused by consecutive EPSPs produced by a single presynaptic neuron. The longer the time constant, the greater is the chance for two potentials to summate to induce an action potential. If a second EPSP is elicited before the first EPSP decays, the two potentials summate and, as in this example, their additive effects are sufficient to induce an action potential in the postsynaptic neuron (temporal summation). Figure 6–8 also shows how the length constant of a postsynaptic neuron can affect the amplitude of two EPSPs produced by different presynaptic neurons in a process called spatial summation. If a neuron has a long length constant, the membrane depolarization induced by input arriving at two points on the neuron can spread to the trigger zone of the neuron with minimal decrement. The two potentials can summate and induce an action potential.


FIGURE 6–8 Central neurons integrate a variety of synaptic inputs through temporal and spatial summation. A) The time constant of the postsynaptic neuron affects the amplitude of the depolarization caused by consecutive EPSPs produced by a single presynaptic neuron. In cases of a long time constant, if a second EPSP is elicited before the first EPSP decays, the two potentials summate to induce an action potential. B) The length constant of a postsynaptic cell affects the amplitude of two EPSPs produced by two presynaptic neurons, A and B. If the length constant is long, the depolarization induced at two points on the neuron can spread to the trigger zone with minimal decrement so that the two potentials summate and an action potential is elicited. (From Kandel ER, Schwartz JH, Jessell TM [editors]: Principles of Neural Science, 4th ed. McGraw-Hill, 2000.)


For many years, the standard view was that dendrites were simply the sites of current sources or sinks that electrotonically change the membrane potential at the initial segment; that is, they were regarded merely as extensions of the soma that expand the area available for integration. When the dendritic tree of a neuron is extensive and has multiple presynaptic knobs ending on it, there is room for a great interplay of inhibitory and excitatory activity.

It is now well established that dendrites contribute to neural function in more complex ways. Action potentials can be recorded in dendrites. In many instances, these are initiated in the initial segment and conducted in a retrograde fashion, but propagated action potentials are initiated in some dendrites. Further research has demonstrated the malleability of dendritic spines. Dendritic spines appear, change, and even disappear over a time scale of minutes and hours, not days and months. Also, although protein synthesis occurs mainly in the soma with its nucleus, strands of mRNA migrate into the dendrites. There, each can become associated with a single ribosome in a dendritic spine and produce proteins, which alters the effects of input from individual synapses on the spine. These changes in dendritic spines have been implicated in motivation, learning, and long-term memory.


Inhibition in the CNS can be postsynaptic or presynaptic. The neurons responsible for postsynaptic and presynaptic inhibition are compared in Figure 6–9Postsynaptic inhibition during the course of an IPSP is called direct inhibition because it is not a consequence of previous discharges of the postsynaptic neuron. There are various forms of indirect inhibition, which is inhibition due to the effects of previous postsynaptic neuron discharge. For example, the postsynaptic cell can be refractory to excitation because it has just fired and is in its refractory period. During after-hyperpolarization it is also less excitable. In spinal neurons, especially after repeated firing, this after-hyperpolarization may be large and prolonged.


FIGURE 6–9 Comparison of neurons producing presynaptic and postsynaptic inhibition. Presynaptic inhibition is a process mediated by neurons whose terminals are on excitatory nerve endings, forming axoaxonal synapses, and reducing transmitter release form the excitatory neuron. Postsynaptic inhibition occurs when an inhibitory transmitter (eg, glycine, GABA) is released from a presynaptic nerve terminal onto the postsynaptic neuron.


Postsynaptic inhibition occurs when an inhibitory transmitter (eg, glycine, GABA) is released from a presynaptic nerve terminal onto the postsynaptic neuron. Various pathways in the nervous system are known to mediate postsynaptic inhibition, and one illustrative example is presented here. Afferent fibers from the muscle spindles (stretch receptors) in skeletal muscle project directly to the spinal motor neurons of the motor units supplying the same muscle (Figure 6–6). Impulses in this afferent fiber cause EPSPs and, with summation, propagated responses in the postsynaptic motor neurons. At the same time, IPSPs are produced in motor neurons supplying the antagonistic muscles which have an inhibitory interneuron interposed between the afferent fiber and the motor neuron. Therefore, activity in the afferent fibers from the muscle spindles excites the motor neurons supplying the muscle from which the impulses come, and inhibits the motor neurons supplying its antagonists (reciprocal innervation). These reflexes are considered in more detail in Chapter 12.


Another type of inhibition occurring in the CNS is presynaptic inhibition, a process mediated by neurons whose terminals are on excitatory endings, forming axoaxonal synapses (Figure 6–3). Three mechanisms of presynaptic inhibition have been described. First, activation of the presynaptic receptors increases Cl conductance, and this has been shown to decrease the size of the action potentials reaching the excitatory ending (Figure 6–10). This in turn reduces Ca2+ entry and consequently the amount of excitatory transmitter released. Voltage-gated K+ channels are also opened, and the resulting K+ efflux also causes a decrease in Ca2+ influx. Finally, there is evidence for direct inhibition of transmitter release independent of Ca2+ influx into the excitatory ending.


FIGURE 6–10 Effects of presynaptic inhibition and facilitation on the action potential and the Ca2+ current in the presynaptic neuron and the EPSP in the postsynaptic neuron. In each case, the solid lines are the controls and the dashed lines the records obtained during inhibition or facilitation. Presynaptic inhibition occurs when activation of presynaptic receptors increases Cl conductance which decreases the size of the action potential. This reduces Ca2+ entry and thus the amount of excitatory transmitter released. Presynaptic facilitation is produced when the action potential is prolonged and the Ca2+ channels are open for a longer duration. (Modified from Kandel ER, Schwartz JH, Jessell TM [editors]: Principles of Neural Science, 4th ed. McGraw-Hill, 2000.)

The first transmitter shown to produce presynaptic inhibition was GABA. Acting via GABAA receptors, GABA increases Cl conductance. GABAB receptors are also present in the spinal cord and appear to mediate presynaptic inhibition via a G protein that produces an increase in K+ conductance. Baclofen, a GABAB agonist, is effective in the treatment of the spasticity of spinal cord injury and multiple sclerosis, particularly when administered intrathecally via an implanted pump. Other transmitters also mediate presynaptic inhibition by G protein-mediated effects on Ca2+ channels and K+ channels.

Conversely, presynaptic facilitation is produced when the action potential is prolonged (Figure 6–10) and the Ca2+ channels are open for a longer period. The molecular events responsible for the production of presynaptic facilitation mediated by serotonin in the sea snail Aplysia have been worked out in detail. Serotonin released at an axoaxonal ending increases intraneuronal cAMP levels, and the resulting phosphorylation of one group of K+channels closes the channels, slowing repolarization and prolonging the action potential.


Presynaptic inhibition and postsynaptic inhibition are usually produced by stimulation of certain systems converging on a given postsynaptic neuron. Neurons may also inhibit themselves in a negative feedback fashion (negative feedback inhibition). For instance, a spinal motor neuron emits a recurrent collateral that synapses with an inhibitory interneuron, which then terminates on the cell body of the spinal neuron and other spinal motor neurons (Figure 6–11). This particular inhibitory neuron is sometimes called a Renshaw cell after its discoverer. Impulses generated in the motor neuron activate the inhibitory interneuron to secrete the inhibitory neurotransmitter glycine, and this reduces or stops the discharge of the motor neuron. Similar inhibition via recurrent collaterals is seen in the cerebral cortex and limbic system. Presynaptic inhibition due to descending pathways that terminate on afferent pathways in the dorsal horn may be involved in the gating of pain transmission.


FIGURE 6–11 Negative feedback inhibition of a spinal motor neuron via an inhibitory interneuron. The axon of a spinal motor neuron has a recurrent collateral that synapses on an inhibitory interneuron that terminates on the cell body of the same and other motor neurons. The inhibitory interneuron is called a Renshaw cell and its neurotransmitter is glycine.

Another type of inhibition is seen in the cerebellum. In this part of the brain, stimulation of basket cells produces IPSPs in the Purkinje cells. However, the basket cells and the Purkinje cells are excited by the same parallel-fiber excitatory input (see Chapter 12). This arrangement, which has been called feed-forward inhibition, presumably limits the duration of the excitation produced by any given afferent volley.



As the axon supplying a skeletal muscle fiber approaches its termination, it loses its myelin sheath and divides into a number of terminal boutons (Figure 6–12). The terminal contains many small, clear vesicles that contain acetylcholine, the transmitter at these junctions. The endings fit into junctional folds, which are depressions in the motor end plate, the thickened portion of the muscle membrane at the junction. The space between the nerve and the thickened muscle membrane is comparable to the synaptic cleft at neuron-to-neuron synapses. The whole structure is known as the neuromuscular junction. Only one nerve fiber ends on each end plate, with no convergence of multiple inputs.


FIGURE 6–12 The neuromuscular junction. A) Scanning electronmicrograph showing branching of motor axons with terminals embedded in grooves in the muscle fiber’s surface. B) Structure of a neuromuscular junction. (From Widmaier EP, Raff H, Strang KT: Vanders Human Physiology. McGraw-Hill, 2008.)


The events occurring during transmission of impulses from the motor nerve to the muscle are somewhat similar to those occurring at neuron-to-neuron synapses (Figure 6–13). The impulse arriving in the end of the motor neuron increases the permeability of its endings to Ca2+. Ca2+ enters the endings and triggers a marked increase in exocytosis of the acetylcholine-containing synaptic vesicles. The acetylcholine diffuses to nicotinic cholinergic (NM) receptors that are concentrated at the tops of the junctional folds of the membrane of the motor end plate. Binding of acetylcholine to these receptors increases the Na+ and K+ conductance, and the resultant influx of Na+ produces a depolarizing potential, the end plate potential. The current sink created by this local potential depolarizes the adjacent muscle membrane to its firing level. Action potentials are generated on either side of the end plate and are conducted away from the end plate in both directions along the muscle fiber. The muscle action potential, in turn, initiates muscle contraction, as described in Chapter 5. Acetylcholine is then removed from the synaptic cleft by acetylcholinesterase, which is present in high concentration at the neuromuscular junction.


FIGURE 6–13 Events at the neuromuscular junction that lead to an action potential in the muscle fiber plasma membrane. The impulse arriving in the end of the motor neuron increases the permeability of its endings to Ca2+which enters the endings and triggers exocytosis of the acetylcholine (ACh)-containing synaptic vesicles. ACh diffuses and binds to nicotinic cholinergic (NM) receptors in the motor end plate which increases Na+ and K+conductance. The resultant influx of Na+ produces the end plate potential. The current sink created by this local potential depolarizes the adjacent muscle membrane to its firing level. Action potentials are generated on either side of the end plate and are conducted away from the end plate in both directions along the muscle fiber and the muscle contracts. ACh is then removed from the synaptic cleft by acetylcholinesterase. (From Widmaier EP, Raff H, Strang KT: Vanders Human Physiology. McGraw-Hill, 2008.)

An average human end plate contains about 15–40 million acetylcholine receptors. Each nerve impulse releases acetylcholine from about 60 synaptic vesicles, and each vesicle contains about 10,000 molecules of the neurotransmitter. This amount is enough to activate about 10 times the number of NM receptors needed to produce a full end plate potential. Therefore, a propagated action potential in the muscle is regularly produced, and this large response obscures the end plate potential. However, the end plate potential can be seen if the 10-fold safety factor is overcome and the potential is reduced to a size that is insufficient to activate the adjacent muscle membrane. This can be accomplished by administration of small doses of curare, a drug that competes with acetylcholine for binding to NM receptors. The response is then recorded only at the end plate region and decreases exponentially away from it. Under these conditions, end plate potentials can be shown to undergo temporal summation.


Small quanta (packets) of acetylcholine are released randomly from the nerve cell membrane at rest. Each produces a minute depolarizing spike called a miniature end plate potential, which is about 0.5 mV in amplitude. The size of the quanta of acetylcholine released in this way varies directly with the Ca2+ concentration and inversely with the Mg2+ concentration at the end plate. When a nerve impulse reaches the ending, the number of quanta released increases by several orders of magnitude, and the result is the large end plate potential that exceeds the firing level of the muscle fiber. Quantal release of acetylcholine similar to that seen at the myoneural junction has been observed at other cholinergic synapses, and quantal release of other transmitters occurs at noradrenergic, glutamatergic, and other synaptic junctions. Two diseases of the neuromuscular junction, myasthenia gravis and Lambert-Eaton syndrome, are described in Clinical Box 6–2 and Clinical Box 6–3, respectively.


Myasthenia Gravis

Myasthenia gravis is a serious and sometimes fatal disease in which skeletal muscles are weak and tire easily. It occurs in 25 to 125 of every 1 million people worldwide and can occur at any age but seems to have a bimodal distribution, with peak occurrences in individuals in their 20s (mainly women) and 60s (mainly men). It is caused by the formation of circulating antibodies to the muscle type of nicotinic cholinergic receptors. These antibodies destroy some of the receptors and bind others to neighboring receptors, triggering their removal by endocytosis. Normally, the number of quanta released from the motor nerve terminal declines with successive repetitive stimuli. In myasthenia gravis, neuromuscular transmission fails at these low levels of quantal release. This leads to the major clinical feature of the disease, muscle fatigue with sustained or repeated activity. There are two major forms of the disease. In one form, the extraocular muscles are primarily affected. In the second form, there is a generalized skeletal muscle weakness. In severe cases, all muscles, including the diaphragm, can become weak and respiratory failure and death can ensue. The major structural abnormality in myasthenia gravis is the appearance of sparse, shallow, and abnormally wide or absent synaptic clefts in the motor end plate. Studies show that the postsynaptic membrane has a reduced response to acetylcholine and a 70–90% decrease in the number of receptors per end plate in affected muscles. Patients with mysathenia gravis have a greater than normal tendency to also have rheumatoid arthritis, systemic lupus erythematosus, and polymyositis. About 30% of mysathenia gravis patients have a maternal relative with an autoimmune disorder. These associations suggest that individuals with myasthenia gravis share a genetic predisposition to autoimmune disease. The thymus may play a role in the pathogenesis of the disease by supplying helper T cells sensitized against thymic proteins that cross-react with acetylcholine receptors. In most patients, the thymus is hyperplastic; and 10–15% have a thymoma.


Muscle weakness due to myasthenia gravis improves after a period of rest or after administration of an acetylcholinesterase inhibitor such as neostigmine or pyridostigmine. Cholinesterase inhibitors prevent metabolism of acetylcholine and can thus compensate for the normal decline in released neurotransmitters during repeated stimulation. Immunosuppressive drugs (eg, prednisone, azathioprine, or cyclosporine) can suppress antibody production and have been shown to improve muscle strength in some patients with myasthenia gravis. Thymectomy is indicated especially if a thymoma is suspected in the development of myasthenia gravis. Even in those without thymoma, thymectomy induces remission in 35% and improves symptoms in another 45% of patients.


Lambert–Eaton Syndrome

In a relatively rare condition called Lambert–Eaton Syndrome (LEMS), muscle weakness is caused by an autoimmune attack against one of the voltage-gated Ca2+ channels in the nerve endings at the neuromuscular junction. This decreases the normal Ca2+ influx that causes acetylcholine release. The incidence of LEMS in the U.S. is about 1 case per 100,000 people; it is usually an adult-onset disease that appears to have a similar occurrence in men and women. Proximal muscles of the lower extremities are primarily affected, producing a waddling gait and difficulty raising the arms. Repetitive stimulation of the motor nerve facilitates accumulation of Ca2+ in the nerve terminal and increases acetylcholine release, leading to an increase in muscle strength. This is in contrast to myasthenia gravis in which symptoms are exacerbated by repetitive stimulation. About 40% of patients with LEMS also have cancer, especially small cell cancer of the lung. One theory is that antibodies that have been produced to attack the cancer cells may also attack Ca2+ channels, leading to LEMS. LEMS has also been associated with lymphosarcoma, malignant thymoma, and cancer of the breast, stomach, colon, prostate, bladder, kidney, or gall bladder. Clinical signs usually precede the diagnosis of cancer. A syndrome similar to LEMS can occur after the use of aminoglycoside antibiotics, which also impair Ca2+ channel function.


Since there is a high comorbidity with small cell lung cancer, the first treatment strategy is to determine whether the individual also has cancer and, if so, to treat that appropriately. In patients without cancer, immunotherapy is initiated. Prednisone administration, plasmapheresis, and intravenous immunoglobulin are some examples of effective therapies for LEMS. Also, the use of aminopyridines facilitates the release of acetylcholine in the neuromuscular junction and can improve muscle strength in LEMS patients. This class of drugs causes blockade of presynaptic K+ channels and promote activation of voltage-gated Ca2+ channels. Acetylcholinesterase inhibitors can be used but often do not ameliorate the symptoms of LEMS.


The postganglionic neurons in the various smooth muscles that have been studied in detail branch extensively and come in close contact with the muscle cells (Figure 6–14). Some of these nerve fibers contain clear vesicles and are cholinergic, whereas others contain the characteristic dense-core vesicles that contain norepinephrine. There are no recognizable end plates or other postsynaptic specializations. The nerve fibers run along the membranes of the muscle cells and sometimes groove their surfaces. The multiple branches of the noradrenergic and, presumably, the cholinergic neurons are beaded with enlargements (varicosities) and contain synaptic vesicles (Figure 6–14). In noradrenergic neurons, the varicosities are about 5 μm apart, with up to 20,000 varicosities per neuron. Transmitter is apparently liberated at each varicosity, that is, at many locations along each axon. This arrangement permits one neuron to innervate many effector cells. The type of contact in which a neuron forms a synapse on the surface of another neuron or a smooth muscle cell and then passes on to make similar contacts with other cells is called a synapse en passant.


FIGURE 6–14 Endings of postganglionic autonomic neurons on smooth muscle. The nerve fibers run along the membranes of the smooth muscle cells and sometimes groove their surfaces. The multiple branches of postganglionic neurons are beaded with enlargements (varicosities) and contain synaptic vesicles. Neurotransmitter is released from the varicosities and diffuses to receptors on smooth muscle cell plasma membranes. (From Widmaier EP, Raff H, Strang KT: Vanders Human Physiology. McGraw-Hill, 2008.)

In the heart, cholinergic and noradrenergic nerve fibers end on the sinoatrial node, the atrioventricular node, and the bundle of His (see Chapter 29). Noradrenergic fibers also innervate the ventricular muscle. The exact nature of the endings on nodal tissue is not known. In the ventricle, the contacts between the noradrenergic fibers and the cardiac muscle fibers resemble those found in smooth muscle.


In smooth muscles in which noradrenergic discharge is excitatory, stimulation of the noradrenergic nerves produces discrete partial depolarizations that look like small end plate potentials and are called excitatory junction potentials (EJPs). These potentials summate with repeated stimuli. Similar EJPs are seen in tissues excited by cholinergic discharges. In tissues inhibited by noradrenergic stimuli, hyperpolarizing inhibitory junction potentials (IJPs) are produced by stimulation of the noradrenergic nerves. Junctional potentials spread electrotonically.


When the motor nerve to skeletal muscle is cut and allowed to degenerate, the muscle gradually becomes extremely sensitive to acetylcholine. This is called denervation hypersensitivity or supersensitivity. Normally nicotinic receptors are located only in the vicinity of the motor end plate where the axon of the motor nerve terminates. When the motor nerve is severed, there is a marked proliferation of nicotinic receptors over a wide region of the neuromuscular junction. Denervation supersensitivity also occurs at autonomic junctions. Smooth muscle, unlike skeletal muscle, does not atrophy when denervated, but it becomes hyperresponsive to the chemical mediator that normally activates it. This hyperresponsiveness can be demonstrated by using pharmacological tools rather than actual nerve section. Prolonged use of a drug such as reserpine can be used to deplete transmitter stores and prevent the target organ from being exposed to norepinephrine for an extended period. Once the drug usage is stopped, smooth muscle and cardiac muscle will be supersensitive to subsequent release of the neurotransmitter.

The reactions triggered by section of an axon are summarized in Figure 6–15. Hypersensitivity of the postsynaptic structure to the transmitter previously secreted by the axon endings is a general phenomenon, largely due to the synthesis or activation of more receptors. Both orthograde degeneration (wallerian degeneration) and retrograde degeneration of the axon stump to the nearest collateral (sustaining collateral) will occur. There are a series of changes in the cell body that leads to a decrease in Nissl substance (chromatolysis). The nerve then starts to regrow, with multiple small branches projecting along the path the axon previously followed (regenerative sprouting). Axons sometimes grow back to their original targets, especially in locations like the neuromuscular junction. However, nerve regeneration is generally limited because axons often become entangled in the area of tissue damage at the site where they were disrupted. This difficulty has been reduced by administration of neurotrophins (see Chapter 4).


FIGURE 6–15 Summary of changes occurring in a neuron and the structure it innervates when its axon is crushed or cut at the point marked X. Hypersensitivity of the postsynaptic structure to the transmitter previously secreted by the axon occurs largely due to the synthesis or activation of more receptors. There is both orthograde (wallerian) degeneration from the point of damage to the terminal and retrograde degeneration of the axon stump to the nearest collateral (sustaining collateral). Changes also occur in the cell body, including chromatolysis. The nerve starts to regrow, with multiple small branches projecting along the path the axon previously followed (regenerative sprouting).

Denervation hypersensitivity has multiple causes. As noted in Chapter 2, a deficiency of a given chemical messenger generally produces an upregulation of its receptors. Another factor is a lack of reuptake of secreted neurotransmitters.


image The terminals of the presynaptic fibers have enlargements called terminal boutons or synaptic knobs. The presynaptic terminal is separated from the postsynaptic structure by a synaptic cleft. The postsynaptic membrane contains neurotransmitter receptors and usually a postsynaptic thickening called the postsynaptic density.

image At chemical synapses, an impulse in the presynaptic axon causes secretion of a neurotransmitter that diffuses across the synaptic cleft and binds to postsynaptic receptors, triggering events that open or close channels in the membrane of the postsynaptic cell. At electrical synapses, the membranes of the presynaptic and postsynaptic neurons come close together, and gap junctions form low-resistance bridges through which ions pass with relative ease from one neuron to the next.

image An EPSP is produced by depolarization of the postsynaptic cell after a latency of 0.5 ms; the excitatory transmitter opens Na+ or Ca2+ ion channels in the postsynaptic membrane, producing an inward current. An IPSP is produced by a hyperpolarization of the postsynaptic cell; it can be produced by a localized increase in Cl transport. Slow EPSPs and IPSPs occur after a latency of 100–500 ms in autonomic ganglia, cardiac, and smooth muscle, and cortical neurons. The slow EPSPs are due to decreases in K+ conductance, and the slow IPSPs are due to increases in K+ conductance.

image Postsynaptic inhibition during the course of an IPSP is called direct inhibition. Indirect inhibition is due to the effects of previous postsynaptic neuron discharge; for example, the postsynaptic cell cannot be activated during its refractory period. Presynaptic inhibition is a process mediated by neurons whose terminals are on excitatory endings, forming axoaxonal synapses; in response to activation of the presynaptic terminal. Activation of the presynaptic receptors can increase Cl conductance, decreasing the size of the action potentials reaching the excitatory ending, and reducing Ca2+ entry and the amount of excitatory transmitter released.

image The axon terminal of motor neurons synapses on the motor end plate on the skeletal muscle membrane to form the neuromuscular junction. The impulse arriving in the motor nerve terminal leads to the entry of Ca2+ which triggers the exocytosis of the acetylcholine-containing synaptic vesicles. The acetylcholine diffuses and binds to nicotinic cholinergic receptors on the motor end plate, causing an increase in Na+ and K+ conductance; the influx of Na+induces the end plate potential and subsequent depolarization of the adjacent muscle membrane. Action potentials are generated and conducted along the muscle fiber, leading in turn to muscle contraction.

image When a nerve is damaged and then degenerates, the postsynaptic structure gradually becomes extremely sensitive to the transmitter released by the nerve. This is called denervation hypersensitivity or supersensitivity.


For all questions, select the single best answer unless otherwise directed.

1. Which of the following electrophysiological events is correctly paired with the change in ionic currents causing the event?

A. Fast inhibitory postsynaptic potentials (IPSPs) and closing of Cl channels.

B. Fast excitatory postsynaptic potentials (EPSPs) and an increase in Ca2+ conductance.

C. End plate potential and an increase in Na+ conductance.

D. Presynaptic inhibition and closure of voltage-gated K+ channels.

E. Slow EPSPs and an increase in K+ conductance.

2. Which of the following physiological processes is not correctly paired with a structure?

A. Electrical transmission : gap junction

B. Negative feedback inhibition : Renshaw cell

C. Synaptic vesicle docking and fusion : presynaptic nerve terminal

D. End plate potential : muscarinic cholinergic receptor

E. Action potential generation : initial segment

3. Initiation of an action potential in skeletal muscle

A. requires spatial facilitation.

B. requires temporal facilitation.

C. is inhibited by a high concentration of Ca2+ at the neuromuscular junction.

D. requires the release of norepinephrine.

E. requires the release of acetylcholine.

4. A 35-year-old woman sees her physician to report muscle weakness in the extraocular eye muscles and muscles of the extremities. She states that she feels fine when she gets up in the morning, but the weakness begins soon after she becomes active. The weakness is improved by rest. Sensation appears normal. The physician treats her with an anticholinesterase inhibitor, and she notes immediate return of muscle strength. Her physician diagnoses her with

A. Lambert–Eaton syndrome.

B. myasthenia gravis.

C. multiple sclerosis.

D. Parkinson disease.

E. muscular dystrophy.

5. A 55-year-old female had an autonomic neuropathy which disrupted the sympathetic nerve supply to the pupillary dilator muscle of her right eye. While having her eyes examined, the ophthalmologist placed phenylephrine in her eyes. The right eye became much more dilated than the left eye. This suggests that

A. the sympathetic nerve to the right eye had regenerated.

B. the parasympathetic nerve supply to the right eye remained intact and compensated for the loss of the sympathetic nerve.

C. phenylephrine blocked the pupillary constrictor muscle of the right eye.

D. denervation supersensitivity had developed.

E. the left eye also had nerve damage and so was not responding as expected.

6. A 47-year-old female was admitted to the hospital after reporting that she had been experiencing nausea and vomiting for about two days and then developed severe muscle weakness and neurological symptoms including ptosis and dysphagia. She indicated she had eaten at a restaurant the evening before the symptoms began. Lab tests were positive for Clostridium botulinum. Neurotoxins

A. block the reuptake of neurotransmitters into presynaptic terminals.

B. such as tetanus toxin bind reversibly to the presynaptic membrane at the neuromuscular junction.

C. reach the cell body of the motor neuron by diffusion into the spinal cord.

D. exert all of their adverse effects by acting centrally rather than peripherally.

E. such as botulinum toxin prevent the release of acetylcholine from motor neurons due to cleavage of either synaptosome-associated proteins or vesicle-associated membrane proteins.


Di Maoi V: Regulation of information passing by synaptic transmission: A short review. Brain Res 2008;1225:26.

Hille B: Ionic Channels of Excitable Membranes, 3rd ed. Sinauer Associates, 2001.

Magee JC: Dendritic integration of excitatory synaptic input. Nature Rev Neurosci 2000;1:181.

Sabatini B, Regehr WG: Timing of synaptic transmission. Annu Rev Physiol 1999;61:521.

Van der Kloot W, Molg J: Quantal acetylcholine release at the vertebrate neuromuscular junction. Physiol Rev 1994;74:899.

WuH, Xiong WC, Mei L: To build a synapse: signaling pathways in neuromuscular junction assembly. Development 2010;137:1017.