Neurotransmitters Are Synthesized in Presynaptic Endings and in Neuronal Cell Bodies

Nearly all known or suspected neurotransmitters fall into one of two general categories: some are small molecules such as amines or amino acids, and the others are peptides (neuropeptides). Small-molecule transmitters are synthesized in presynaptic cytoplasm, using locally available substrates (e.g., acetate and choline) and soluble enzymes that arrive by slow axonal transport (Fig. 8-4A). Making peptide transmitters, in contrast, requires the protein synthesis machinery that resides in the neuronal cell body. Neuropeptides start out there as larger precursor proteins that are packed into vesicles and dispatched by fast axonal transport to synaptic endings. Along the way, the precursors are cleaved and processed into the final, biologically active neuropeptides (Fig. 8-4B).

Figure 8-4 Life cycles of small-molecule neurotransmitters (A) and neuropeptides (B). A, Enzymes required for the synthesis and packaging of small-molecule neurotransmitters are themselves synthesized in the cell body (1), released from the Golgi apparatus (2), and conveyed to presynaptic terminals by slow axonal transport (3). The neurotransmitters are then synthesized from substrates transported into the terminals (4) and concentrated in vesicles (5) that were either recycled from the presynaptic membrane (6) or assembled from components transported from the cell body. B, Precursors of neuropeptides are synthesized in the cell body (1) and packaged in the Golgi apparatus into vesicles (2), which are conveyed to presynaptic terminals by fast axonal transport along microtubules (3). Either in the cell body or during the journey down the axon, the precursor proteins are modified and become neuropeptides (4). After exocytosis, the membranes of large, dense-core vesicles are returned to the cell body for recycling (5).

Neurotransmitters Are Packaged into Synaptic Vesicles before Release

Synaptic vesicles are the units of currency at chemical synapses, the sites where neurotransmitters are packaged, concentrated, and protected from catabolism while awaiting release. All presynaptic endings contain small vesicles (about 40 nm), and many also contain less numerous but larger vesicles (≥100 nm). Depending on the preparation conditions used for electron microscopy, some of the small vesicles may appear dark, and others may look clear and either round or flattened; each large vesicle contains a dark core (Fig. 8-5). These differences in appearance correspond to differences in neurotransmitter content. For example, endings with clear, flattened vesicles are usually inhibitory.

Figure 8-5 Types of synaptic vesicles in axon terminals (At) making synaptic contact with transversely sectioned dendrites (D) in the anterior horn of rat spinal cord. Two of the terminals (At1) contain clear, round vesicles; one (At2) contains clear, flattened vesicles; and one (At3) contains a mixture of small, round vesicles and large, dense-cored vesicles. The actual size of the dendrite in the center of the micrograph is about 2.5 × 1.5 μm.

(From Pannese E: Neurocytology: fine structure of neurons, nerve processes, and neuroglial cells, New York, 1994, Thieme Medical Publishers.)

Small synaptic vesicles contain small-molecule transmitters, concentrated to levels far beyond that found in the cytoplasm by specific transporters in the vesicle walls. Large dense-core vesicles contain neuropeptides at lower concentrations and may contain one or more small-molecule transmitters as well. The two kinds of vesicle, like the two kinds of neurotransmitter, are manufactured and recycled differently (Fig. 8-4). Small-molecule transmitters are synthesized by cytoplasmic enzymes and can therefore be manufactured and packaged for release in individual synaptic endings. Small synaptic vesicles can therefore be recycled entirely within a presynaptic ending, whereas large dense-core vesicles must be created anew in the cell body.

Presynaptic Endings Release Neurotransmitters into the Synaptic Cleft

Release of neurotransmitter is a Ca²⁺-mediated secretory process. Each presynaptic density, or active zone, contains an abundance of voltage-gated Ca²⁺ channels, together with anchoring sites for a cluster of small vesicles that are held there (“docked”) by a Ca²⁺-sensitive system of proteins. Depolarization of the presynaptic terminal, such as by propagation of an action potential down the axon and subsequent electrotonic spread into the terminal, causes opening of the voltage-gated Ca²⁺ channels. Because the free intracellular Ca²⁺ concentration is only about 10⁻⁷ M, Ca²⁺ ions flow in through these channels and momentarily elevate the Ca²⁺ concentration near the active zone by as much as 1000-fold. During the brief period before the excess Ca²⁺ diffuses away or is sequestered, a vesicle docked nearby may fuse with the presynaptic membrane and discharge its contents into the synaptic cleft in a process called exocytosis (Fig. 8-6A). The whole process, from the arrival of an action potential to the release of a small vesicle’s contents, can take less than 100 μsec. Because the large, dense-cored vesicles are not located adjacent to active zones, repetitive action potentials, additional Ca²⁺ entry, and more time (10s of milliseconds) are typically required for their exocytosis (Fig. 8-6B). This exocytotic addition of membrane clearly cannot continue for very long, or else presynaptic endings would expand continuously. In fact, vesicle membranes are taken back up (by endocytosis), and those used for small-molecule transmitters can be recycled in less than a minute (Fig. 8-4A).

Figure 8-6 Ca²⁺-triggered release of neurotransmitter. A, Invasion of a presynaptic terminal by a single action potential (1) causes brief opening of voltage-gated Ca²⁺ channels (2) and a local increase in Ca²⁺ concentration. One or more nearby small vesicles may fuse with the presynaptic membrane (3) and discharge their contents into the synaptic cleft; they are then recycled (4) in the synaptic terminal. B, Repetitive action potentials (1) cause opening of more voltage-gated Ca²⁺ channels for longer periods (2) and a correspondingly more widespread increase in Ca²⁺ concentration. This causes fusion not only of small vesicles docked at active zones (3) but also of larger vesicles away from active zones (4).

Neurotransmitters Diffuse across the Synaptic Cleft and Bind to Postsynaptic Receptors

Neurotransmitters released from small vesicles find postsynaptic receptor molecules waiting for them directly across the synaptic cleft (Fig. 8-7). For this reason, it takes the contents of such vesicles very little time to exert their effects. The total synaptic delay from presynaptic action potential to postsynaptic effect can be less than 200 μsec. The contents of large vesicles, in contrast, take longer to be released and often diffuse to receptors relatively far away (see Fig. 8-9), so their effects develop more slowly.

Figure 8-7 Juxtaposition of presynaptic voltage-gated Ca²⁺ channels and postsynaptic neurotransmitter receptors, shown at a neuromuscular junction by using toxins coupled to fluorescent dyes. The area of a frog neuromuscular junction outlined in A is enlarged in B and C. B, A marine snail toxin (ω-conotoxin) that binds selectively to voltagegated Ca²⁺ channels demonstrates these channels’ locations in the presynaptic terminal. C, Staining the same area with a snake toxin (α-bungarotoxin) that binds to nicotinic acetylcholine receptors (the type found at neuromuscular junctions) demonstrates that these receptors have an almost exactly parallel distribution in the postjunctional muscle membrane.

(From Robitaille R, Adler EM, Charlton MP: Neuron 5:773, 1990.)

Figure 8-9 Endocytotic removal of a neuropeptide (substance P) and its receptor by postsynaptic neurons, demonstrated by staining substance P receptors using a fluorescent antibody technique. In the absence of stimulation, substance P receptors coat the somatic and dendritic membranes of pain-sensitive second-order neurons in the spinal cord posterior horn (A, red and yellow fluorescence), and the distal dendrites of these cells have a uniform diameter (B). Five minutes after a painful stimulus on the ipsilateral side, most of the substance P receptor has left the surface membranes and is found intracellularly (C), and distal dendrites have a beaded appearance, with varicosities filled with endosomes containing substance P receptor (D). Over about the next hour, the internalized substance P receptor molecules are recycled to the cell surface. The scale mark in A (also applies to C) is 20 μm. The scale mark in B (also applies to D) is 10 μm.

(From Mantyh PW et al: Science 268:1629, 1995.)

Neurotransmitter Action Is Terminated by Uptake, Degradation, or Diffusion

Neurotransmitter molecules need to be removed quickly once they have had a chance to bind to receptors, so that the postsynaptic membrane will be prepared for subsequent releases of transmitter. This is accomplished by virtually every means imaginable (Fig. 8-8). Binding of neurotransmitter and receptor is a reversible event, so receptors and removal mechanisms compete for transmitter. Some transmitter simply diffuses away, but this is too slow a process to be the principal mechanism. In most synapses, transmitter is reabsorbed into the presynaptic ending or taken up by neighboring glial cells or even by the postsynaptic process. In others, enzymes in the synaptic cleft degrade free transmitter. Reabsorbed transmitters or their metabolic products, like vesicle membranes, are often recycled for use in a subsequent synaptic event.

Figure 8-8 Neurotransmitter binds to postsynaptic receptors (1) while other mechanisms compete for the transmitter, trying to remove it from the synaptic cleft. Although it is unlikely that all these mechanisms would be used at a single synapse, different synapses use various combinations of reuptake of neurotransmitter by the presynaptic terminal (2) or nearby glial cells (3), enzymatic inactivation of neurotransmitter (4), and uptake by the postsynaptic terminal (5). Finally, some neurotransmitter simply diffuses out of the synaptic cleft (6).

Different kinds of neurotransmitters have different preferred mechanisms of removal. For example, serotonin, norepinephrine, and dopamine are transported rapidly back into the presynaptic ending for repackaging in synaptic vesicles. Acetylcholine, in contrast, is split into acetate and choline by acetylcholinesterase in the synaptic cleft; the choline is then transported back into the presynaptic ending and used for the synthesis of more acetylcholine. Neuropeptides are either degraded by extracellular peptidases or swallowed up by the postsynaptic cell while still attached to their receptors (Fig. 8-9).

Rapid Synaptic Transmission Involves Transmitter-Gated Ion Channels

Early studies of synaptic transmission demonstrated that a single action potential in the motor nerve ending at a neuromuscular junction causes a large but brief EPSP,* sufficient to cause an action potential and contraction in the muscle fiber (Fig. 8-10B). Closer analysis revealed small, brief depolarizing events in the postjunctional muscle membrane during the periods between action potentials (Fig. 8-10A), each corresponding to a brief period of channel opening and depolarizing current flow. We now know that each of these small depolarizing events is the response of the postsynaptic membrane to the spontaneous release of one synaptic vesicle. Each vesicle contains about 10,000 acetylcholine molecules, which diffuse across the synaptic cleft and bind briefly to acetylcholine receptors, which themselves are ligand-gated ion channels permeable to both Na⁺ and K⁺ ions. Acetylcholinesterase competes with the receptors for the released acetylcholine, so the channels stay open for only a millisecond or two. During this time, they allow a current flow that tries to move the membrane potential to a value around 0 mv (between the Na⁺ and K⁺ equilibrium potentials; see Appendix 7B, Equation B-7). Because 2 msec is considerably less than the time constant of the muscle membrane, the membrane potential never reaches 0 mv. Instead, it rises rapidly at the beginning of the postsynaptic current flow and then decays slowly at the end of the current flow (Fig. 8-10A). Entry of depolarizing current is localized to the site of the receptors, so the postsynaptic potential decays electrotonically (Fig. 8-10C and D). However, the presynaptic ending at each neuromuscular junction contains upwards of 1000 active zones (Figs. 8-7 and 8-11), so a single action potential in the motor axon causes the nearly simultaneous release of hundreds of vesicles full of acetylcholine at closely spaced sites, which in turn causes the large EPSP normally seen (Fig. 8-9B).

The basic elements of neuromuscular transmission have proved to be generally true of fast synaptic transmission throughout the nervous system (Fig. 8-12A). Depolarization-induced Ca²⁺ entry causes the release of one or more packets of neurotransmitter, called quanta, each of which is generally assumed to be the contents of a single vesicle docked at an active zone. The neurotransmitter then diffuses across the synaptic cleft and binds transiently to a transmitter-gated (i.e., ligand-gated) ion channel. The selectivity of the channel determines the postsynaptic effect. Opening channels selective for monovalent cations, as in the example of the neuromuscular junction, causes an EPSP. Opening channels selective for Cl⁻, in contrast, moves the membrane potential toward the Cl⁻ equilibrium potential; this is the most common basis of fast IPSPs in the central nervous system (CNS). The selective concentration of receptors on the postsynaptic membrane ensures that fast postsynaptic potentials are localized spatially. Transmitter-gated ion channels are also referred to as ionotropic receptors (from the same Greek word that gave rise to tropism, a movement toward or away from a stimulus, as in the heliotropic growth of plants in the direction of sunlight).

Figure 8-12 Basic events in typical fast (A) and slow (B) synaptic transmission. A, At rest, ligand-gated ion channels (ionotropic receptors) in the postsynaptic membrane are closed (1), and vesicles are docked in the presynaptic terminal (2) awaiting release. Depolarization of the terminal (3) causes Ca²⁺ influx (4), transmitter exocytosis, binding of transmitter to postsynaptic ligand-gated ion channels (5), and opening of the channels. In this example, open channels allow Na⁺ influx (6) and K⁺ efflux (7), depolarizing the postsynaptic membrane. B, Most slow synaptic responses involve metabotropic receptors (1) coupled at rest to a three-subunit G protein (2), as well as transmitters in large, dense-core vesicles (3) or in certain small vesicles. Prolonged or repetitive depolarization (4) of the presynaptic terminal causes widespread Ca²⁺ influx (5), exocytosis of transmitter from large, dense-core vesicles (6), and binding to G protein–coupled receptors (some of which may be located some distance from the presynaptic terminal; see Fig. 8-9). This binding causes dissociation of the G protein into subunits (7), which can have a variety of effects. They may bind to ion channels and alter their conductance (8); activate an enzyme (9), which in turn alters the concentration of a second messenger (e.g., cyclic adenosine monophosphate, inositol triphosphate, arachidonic acid metabolites); or have even more complex effects (10), such as altering gene expression.

Slow Synaptic Transmission Usually Involves Postsynaptic Receptors Linked to G Proteins

Slow synaptic potentials are also caused by changes in current flow through membrane ion channels, but in this case a multistep process is involved in which binding of neurotransmitter leads to altered concentrations of second messengers, which in turn modulate channel conductance. Because of the intermediate metabolic events, the receptors are referred to as metabotropic receptors. Most metabotropic receptors are membrane-spanning proteins linked to adjacent guanine nucleotide-binding proteins (G proteins). Transmitter binding by a G protein–coupled receptor causes dissociation of subunits of the G protein, which are then able to move laterally in the postsynaptic membrane and exert other effects (Fig. 8-12B).

In the simplest scenario, a G protein subunit directly influences the state of ion channels, for example, causing K⁺ or Ca²⁺ channels to open; in this instance, the G protein subunit itself is the second messenger. In most cases, however, the G protein subunit increases or decreases the activity of an enzyme, which in turn causes changes in the concentration of something else.

Multistep pathways such as this are slow and seem unwieldy, but they actually have considerable advantages. Synaptic inputs can be amplified, because a single receptor can cause the dissociation of multiple G proteins, and changing the activity of an enzyme can cause the synthesis or degradation of thousands of second messenger molecules. In addition, because there are many types of G protein–coupled receptors and many types of G proteins, a wide array of postsynaptic effects can result. Some, as indicated previously, are as simple as opening or closing an ion channel. Others may actually be electrically silent, causing things such as changes in sensitivity to other transmitters or changes in gene transcription. This kind of strategy is so versatile that it is not restricted to synapses. G proteins and second messengers are the basis for the receptor potentials generated by retinal rods and cones, olfactory sensory neurons, and some other receptor cells (see Chapter 9).