Neurotransmission: The Autonomic and Somatic Motor Nervous Systems

The autonomic nervous system, as delineated by Langley over a century ago (Langley, 1898), also is called the visceral, vegetative, or involuntary nervous system. In the periphery, its representation consists of nerves, ganglia, and plexuses that provide the innervation to the heart, blood vessels, glands, other visceral organs, and smooth muscle in various tissues. It is therefore widely distributed throughout the body and regulates autonomic functions, which occur without conscious control.

Differences between Autonomic and Somatic Nerves

The efferent nerves of the involuntary system supply all innervated structures of the body except skeletal muscle, which is served by somatic nerves. The most distal synaptic junctions in the autonomic reflex arc occur in ganglia that are entirely outside the cerebrospinal axis. These ganglia are small but complex structures that contain axodendritic synapses between preganglionic and postganglionic neurons. Somatic nerves contain no peripheral ganglia, and the synapses are located entirely within the cerebrospinal axis. Many autonomic nerves form extensive peripheral plexuses, but such networks are absent from the somatic system. Whereas motor nerves to skeletal muscles are myelinated, postganglionic autonomic nerves generally are nonmyelinated. When the spinal efferent nerves are interrupted, the skeletal muscles they innervate lack myogenic tone, are paralyzed, and atrophy, whereas smooth muscles and glands generally show some level of spontaneous activity independent of intact innervation.

Visceral Afferent Fibers

The afferent fibers from visceral structures are the first link in the reflex arcs of the autonomic system. With certain exceptions, such as local axon reflexes, most visceral reflexes are mediated through the central nervous system (CNS). The afferent fibers are, for the most part, nonmyelinated and are carried into the cerebrospinal axis by the vagus, pelvic, splanchnic, and other autonomic nerves. For example, about four-fifths of the fibers in the vagus are sensory. Other autonomic afferents from blood vessels in skeletal muscles and from certain integumental structures are carried in somatic nerves. The cell bodies of visceral afferent fibers lie in the dorsal root ganglia of the spinal nerves and in the corresponding sensory ganglia of certain cranial nerves, such as the nodose ganglion of the vagus. The efferent link of the autonomic reflex arc is discussed in the following sections.

The autonomic afferent fibers are concerned with the mediation of visceral sensation (including pain and referred pain); with vasomotor, respiratory, and viscerosomatic reflexes; and with the regulation of interrelated visceral activities. An example of an autonomic afferent system is that arising from the pressoreceptive endings in the carotid sinus and the aortic arch and from the chemoreceptor cells in the carotid and aortic bodies; this system is important in the reflex control of blood pressure, heart rate, and respiration, and its afferent fibers pass in the glossopharyngeal and vagus nerves to the medulla oblongata in the brainstem.

The neurotransmitters that mediate transmission from sensory fibers have not been unequivocally characterized. However, substance P is present in afferent sensory fibers, in the dorsal root ganglia, and in the dorsal horn of the spinal cord, and this peptide is a leading candidate for the neurotransmitter that functions in the passage of nociceptive stimuli from the periphery to the spinal cord and higher structures. Other neuroactive peptides, including somatostatin, vasoactive intestinal polypeptide (VIP), and cholecystokinin, also have been found in sensory neurons (Lundburg, 1996; Hkfelt et al., 2000), and one or more such peptides may play a role in the transmission of afferent impulses from autonomic structures. Enkephalins, present in interneurons in the dorsal spinal cord (within an area termed the substantia gelatinosa), have antinociceptive effects that appear to be brought about by presynaptic and postsynaptic actions to inhibit the release of substance P and diminish the activity of cells that project from the spinal cord to higher centers in the CNS. The excitatory amino acids, glutamate and aspartate, also play major roles in transmission of sensory responses to the spinal cord.

Central Autonomic Connections

There probably are no purely autonomic or somatic centers of integration, and extensive overlap occurs. Somatic responses always are accompanied by visceral responses and vice versa. Autonomic reflexes can be elicited at the level of the spinal cord. They clearly are demonstrable in the spinal animal, including human beings, and are manifested by sweating, blood pressure alterations, vasomotor responses to temperature changes, and reflex emptying of the urinary bladder, rectum, and seminal vesicles. Extensive central ramifications of the autonomic nervous system exist above the level of the spinal cord. For example, the integration of the control of respiration in the medulla oblongata is well known. The hypothalamus and the nucleus of the solitary tract (nucleus tractus solitarius) generally are regarded as principal loci of integration of autonomic nervous system functions, which include regulation of body temperature, water balance, carbohydrate and fat metabolism, blood pressure, emotions, sleep, respiration, and sexual responses. Signals are received through ascending spinobulbar pathways. Also, these areas receive input from the limbic system, neostriatum, cortex, and, to a lesser extent, other higher brain centers. Stimulation of the nucleus of the solitary tract and the hypothalamus activates bulbospinal pathways and hormonal output to mediate autonomic and motor responses in the organism (Andresen and Kunze, 1994; Loewy and Spyer, 1990; see also Chapter 12: Neurotransmission and the Central Nervous System). The hypothalamic nuclei that lie posteriorly and laterally are sympathetic in their main connections, while parasympathetic functions evidently are integrated by the midline nuclei in the region of the tuber cinereum and by nuclei lying anteriorly.

Divisions of the Peripheral Autonomic System

On the efferent side, the autonomic nervous system consists of two large divisions: (1) the sympathetic or thoracolumbar outflow and (2) the parasympathetic or craniosacral outflow. A brief outline of those anatomical features necessary for an understanding of the actions of autonomic drugs is given here.

The arrangement of the principal parts of the peripheral autonomic nervous system is presented schematically in Figure 61. As discussed below, the neurotransmitter of all preganglionic autonomic fibers, all postganglionic parasympathetic fibers, and a few postganglionic sympathetic fibers is acetylcholine (ACh); these so-called cholinergic fibers are depicted in blue. The adrenergic fibers, shown in red, compose the majority of the postganglionic sympathetic fibers; here the transmitter is norepinephrine (noradrenaline, levarterenol). The terms cholinergic and adrenergic were proposed originally by Dale (1954) to describe neurons that liberate ACh and norepinephrine, respectively. As noted above, all of the transmitter(s) of the primary afferent fibers, shown in green, have not been identified conclusively. Substance P and glutamate are thought to mediate many afferent impulses; both are present in high concentrations in the dorsal regions of the spinal cord.

Sympathetic Nervous System

The cells that give rise to the preganglionic fibers of this division lie mainly in the intermediolateral columns of the spinal cord and extend from the first thoracic to the second or third lumbar segment. The axons from these cells are carried in the anterior (ventral) nerve roots and synapse with neurons lying in sympathetic ganglia outside the cerebrospinal axis. The sympathetic ganglia are found in three locations: paravertebral, prevertebral, and terminal.

The paravertebral sympathetic ganglia consist of 22 pairs that lie on either side of the vertebral column to form the lateral chains. The ganglia are connected to each other by nerve trunks and to the spinal nerves by rami communicantes. The white rami are restricted to the segments of the thoracolumbar outflow; they carry the preganglionic myelinated fibers that exit from the spinal cord by way of the anterior spinal roots. The gray rami arise from the ganglia and carry postganglionic fibers back to the spinal nerves for distribution to sweat glands and pilomotor muscles and to blood vessels of skeletal muscle and skin. The prevertebral ganglia lie in the abdomen and the pelvis near the ventral surface of the bony vertebral column and consist mainly of the celiac (solar), superior mesenteric, aorticorenal, and inferior mesenteric ganglia. The terminal ganglia are few in number, lie near the organs they innervate, and include ganglia connected with the urinary bladder and rectum and the cervical ganglia in the region of the neck. In addition, there are small intermediate ganglia, especially in the thoracolumbar region, that lie outside the conventional vertebral chain. They are variable in number and location but usually are in close proximity to the communicating rami and to the anterior spinal nerve roots.

Preganglionic fibers issuing from the spinal cord may synapse with the neurons of more than one sympathetic ganglion. Their principal ganglia of termination need not correspond to the original level from which the preganglionic fiber exits the spinal cord. Many of the preganglionic fibers from the fifth to the last thoracic segment pass through the paravertebral ganglia to form the splanchnic nerves. Most of the splanchnic nerve fibers do not synapse until they reach the celiac ganglion; others directly innervate the adrenal medulla (see below).

Postganglionic fibers arising from sympathetic ganglia innervate visceral structures of the thorax, abdomen, head, and neck. The trunk and the limbs are supplied by means of sympathetic fibers in spinal nerves, as previously described. The prevertebral ganglia contain cell bodies, the axons of which innervate the glands and the smooth muscles of the abdominal and the pelvic viscera. Many of the upper thoracic sympathetic fibers from the vertebral ganglia form terminal plexuses, such as the cardiac, esophageal, and pulmonary plexuses. The sympathetic distribution to the head and the neck (vasomotor, pupillodilator, secretory, and pilomotor) is by way of the cervical sympathetic chain and its three ganglia. All postganglionic fibers in this chain arise from cell bodies located in these three ganglia; all preganglionic fibers arise from the upper thoracic segments of the spinal cord, there being no sympathetic fibers that leave the CNS above the first thoracic level.

The adrenal medulla and other chromaffin tissue are embryologically and anatomically similar to sympathetic ganglia; all are derived from the neural crest. The adrenal medulla differs from sympathetic ganglia in that the principal catecholamine that is released in human beings and many other species is epinephrine (adrenaline), whereas norepinephrine is released from postganglionic sympathetic fibers. The chromaffin cells in the adrenal medulla are innervated by typical preganglionic fibers that release acetylcholine.

Parasympathetic Nervous System

The parasympathetic nervous system consists of preganglionic fibers that originate in three areas of the CNS and their postganglionic connections. The regions of central origin are the midbrain, the medulla oblongata, and the sacral part of the spinal cord. The midbrain, or tectal, outflow consists of fibers arising from the Edinger-Westphal nucleus of the third cranial nerve and going to the ciliary ganglion in the orbit. The medullary outflow consists of the parasympathetic components of the seventh, ninth, and tenth cranial nerves. The fibers in the seventh cranial, or facial, nerve form the chorda tympani, which innervates the ganglia lying on the submaxillary and sublingual glands. They also form the greater superficial petrosal nerve, which innervates the sphenopalatine ganglion. The ninth cranial, or glossopharyngeal, autonomic components innervate the otic ganglion. Postganglionic parasympathetic fibers from these ganglia supply the sphincter of the iris (pupillae constrictor muscle), the ciliary muscle, the salivary and lacrimal glands, and the mucous glands of the nose, mouth, and pharynx. These fibers also include vasodilator nerves to the organs mentioned. The tenth cranial, or vagus, nerve arises in the medulla and contains preganglionic fibers, most of which do not synapse until they reach the many small ganglia lying directly on or in the viscera of the thorax and abdomen. In the intestinal wall, the vagal fibers terminate around ganglion cells in the plexuses of Auerbach and Meissner. Preganglionic fibers are thus very long, whereas postganglionic fibers are very short. The vagus nerve, in addition, carries a far greater number of afferent fibers (but apparently no pain fibers) from the viscera into the medulla; the cell bodies of these fibers lie mainly in the nodose ganglion.

The parasympathetic sacral outflow consists of axons that arise from cells in the second, third, and fourth segments of the sacral cord and proceed as preganglionic fibers to form the pelvic nerves (nervi erigentes). They synapse in terminal ganglia lying near or within the bladder, rectum, and sexual organs. The vagal and sacral outflows provide motor and secretory fibers to thoracic, abdominal, and pelvic organs, as indicated in Figure 61.

Enteric Nervous System

Stimulation of particular vagal nuclei in the medulla oblongata or certain fibers in the vagal trunk was known for some time to elicit muscle relaxation in certain regions of the stomach or intestine, such as sphincters, instead of the expected and more common contractile response. In the mid-1960s, it became evident that relaxation of the gastrointestinal tract and other visceral organs was not necessarily mediated by adrenergic stimulation; rather, release of other putative transmitters from enteric neurons, located in Auerbach's and Meissner's plexuses, gave rise to hyperpolarization and relaxation of the smooth muscle (Figure 61). Over the succeeding years, certain peptides (i.e., VIP), nucleotides (ATP), and nitric oxide (NO) were found to be inhibitory transmitters in the gastrointestinal tract and other visceral organs (see Bennett, 1997). Inhibition is achieved either through guanylyl cyclase activation by nitric oxide or hyperpolarization through the activation of K⁺ channels. Specific K⁺ channel inhibitors such as apamin or inhibitors of nitric oxide synthase can distinguish the inhibitory events and their durations. Noncholinergic excitatory transmitters such as tachykinins (e.g., substance P) also are found to be released in regions of the enteric plexus. Substance P is a transmitter of the sensory afferent system, which is released locally or from afferent nerve branches that link to intramural ganglia. The enteric system does not have a unique connection to the CNS. While under the influence of parasympathetic preganglionic nerves, release of transmitters usually is dominated by local control. Coordination of contraction and relaxation at a local level would be expected for regulation of peristaltic waves in the intestine.

Differences among Sympathetic, Parasympathetic, and Motor Nerves

The sympathetic system is distributed to effectors throughout the body, whereas parasympathetic distribution is much more limited. Furthermore, the sympathetic fibers ramify to a much greater extent. A preganglionic sympathetic fiber may traverse a considerable distance of the sympathetic chain and pass through several ganglia before it finally synapses with a postganglionic neuron; also, its terminals make contact with a large number of postganglionic neurons. In some ganglia, the ratio of preganglionic axons to ganglion cells may be 1:20 or more. In this manner, a diffuse discharge of the sympathetic system is possible. In addition, synaptic innervation overlaps, so that one ganglion cell may be supplied by several preganglionic fibers.

The parasympathetic system, in contrast, has its terminal ganglia very near to or within the organs innervated and thus is more circumscribed in its influences. In some organs a 1:1 relationship between the number of preganglionic and postganglionic fibers has been suggested, but the ratio of preganglionic vagal fibers to ganglion cells in Auerbach's plexus has been estimated as 1:8000. Hence, this distinction between the two systems does not apply to all sites.

The cell bodies of somatic motor neurons are in the ventral horn of the spinal cord; the axon divides into many branches, each of which innervates a single muscle fiber, so that more than 100 muscle fibers may be supplied by one motor neuron to form a motor unit. At each neuromuscular junction, the axonal terminal loses its myelin sheath and forms a terminal arborization that lies in apposition to a specialized surface of the muscle membrane, termed the motor end-plate. Mitochondria and a collection of synaptic vesicles are concentrated at the nerve terminal. Through trophic influences of the nerve, those cell nuclei in the multinucleated skeletal muscle cell lying in close apposition to the synapse acquire the capacity to activate specific genes which express synapse-localized proteins (Hall and Sanes, 1993; Sanes and Lichtman, 1999).

Details of Innervation

The terminations of the postganglionic autonomic fibers in smooth muscle and glands form a rich plexus, or terminal reticulum. The terminal reticulum (sometimes called the autonomic ground plexus) consists of the final ramifications of the postganglionic sympathetic (adrenergic), parasympathetic (cholinergic), and visceral afferent fibers, all of which are enclosed within a frequently interrupted sheath of satellite or Schwann cells. At these interruptions, varicosities packed with vesicles are seen in the efferent fibers. Such varicosities occur repeatedly but at variable distances along the course of the ramifications of the axon.

'Protoplasmic bridges' occur between the smooth muscle fibers themselves at points of contact between their plasma membranes. They are believed to permit the direct conduction of impulses from cell to cell without the need for chemical transmission. These structures have been termed nexuses or tight junctions, and they enable the smooth muscle fibers to function as a unit or syncytium.

Sympathetic ganglia are extremely complex, both anatomically and pharmacologically (see Chapter 9: Agents Acting at the Neuromuscular Junction and Autonomic Ganglia). The preganglionic fibers lose their myelin sheaths and divide repeatedly into a vast number of end fibers with diameters ranging from 0.1 to 0.3 m; except at points of synaptic contact, they retain their satellite-cell sheaths. The vast majority of synapses are axodendritic. Apparently, a given axonal terminal may synapse with one or more dendritic processes.

Responses of Effector Organs to Autonomic Nerve Impulses

From the responses of the various effector organs to autonomic nerve impulses and the knowledge of the intrinsic autonomic tone, one can predict the actions of drugs that mimic or inhibit the actions of these nerves. In most instances, the sympathetic and parasympathetic neurotransmitters can be viewed as physiological or functional antagonists. If one neurotransmitter inhibits a certain function, the other usually augments that function. Most viscera are innervated by both divisions of the autonomic nervous system, and the level of activity at any one moment represents the integration of influences of the two components. Despite the conventional concept of antagonism between the two portions of the autonomic nervous system, their activities on specific structures may be either discrete and independent or integrated and interdependent. For example, the effects of sympathetic and parasympathetic stimulation of the heart and the iris show a pattern of functional antagonism in controlling heart rate and pupillary aperture, respectively. Their actions on male sexual organs are complementary and are integrated to promote sexual function. The control of peripheral vascular resistance is primarily, but not exclusively, due to sympathetic control of arteriolar resistance. The effects of stimulating the sympathetic (adrenergic) and parasympathetic (cholinergic) nerves to various organs, visceral structures, and effector cells are summarized in Table 61.

General Functions of the Autonomic Nervous System

The integrating action of the autonomic nervous system is of vital importance for the well-being of the organism. In general, the autonomic nervous system regulates the activities of structures that are not under voluntary control and that function below the level of consciousness. Thus, respiration, circulation, digestion, body temperature, metabolism, sweating, and the secretions of certain endocrine glands are regulated, in part or entirely, by the autonomic nervous system. As Claude Bernard (18781879), J.N. Langley (1898, 1901), and Walter Cannon (1929, 1932) emphasized, the constancy of the internal environment of the organism is to a large extent controlled by the vegetative, or autonomic, nervous system.

The sympathetic system and its associated adrenal medulla are not essential to life in a controlled environment. Under circumstances of stress, however, the lack of the sympathoadrenal functions becomes evident. Body temperature cannot be regulated when environmental temperature varies; the concentration of glucose in blood does not rise in response to urgent need; compensatory vascular responses to hemorrhage, oxygen deprivation, excitement, and exercise are lacking; resistance to fatigue is lessened; sympathetic components of instinctive reactions to the external environment are lost; and other serious deficiencies in the protective forces of the body are discernible.

The sympathetic system normally is continuously active; the degree of activity varies from moment to moment and from organ to organ. In this manner, adjustments to a constantly changing environment are accomplished. The sympathoadrenal system also can discharge as a unit. This occurs particularly during rage and fright, when sympathetically innervated structures over the entire body are affected simultaneously. Heart rate is accelerated; blood pressure rises; red blood cells are poured into the circulation from the spleen (in certain species); blood flow is shifted from the skin and splanchnic region to the skeletal muscles; blood glucose rises; the bronchioles and pupils dilate; and, on the whole, the organism is better prepared for 'fight or flight.' Many of these effects result primarily from, or are reinforced by, the actions of epinephrine, secreted by the adrenal medulla (see below). In addition, signals are received in higher brain centers to facilitate purposeful responses or to imprint the event in memory.

The parasympathetic system is organized mainly for discrete and localized discharge. Although it is concerned primarily with conservation of energy and maintenance of organ function during periods of minimal activity, its elimination is not compatible with life. Sectioning the vagus, for example, soon gives rise to pulmonary infection because of the inability of cilia to remove irritant substances from the respiratory tract. The parasympathetic system slows the heart rate, lowers the blood pressure, stimulates gastrointestinal movements and secretions, aids absorption of nutrients, protects the retina from excessive light, and empties the urinary bladder and rectum. Many parasympathetic responses are rapid and reflexive in nature.

Nerve impulses elicit responses in smooth, cardiac, and skeletal muscles, exocrine glands, and postsynaptic neurons through liberation of specific chemical neurotransmitters. The steps involved and the evidence for them are presented in some detail because the concept of chemical mediation of nerve impulses profoundly affects our knowledge of the mechanism of action of drugs at these sites.

Historical Aspects

The earliest concrete proposal of a neurohumoral mechanism was made shortly after the turn of the twentieth century. Lewandowsky (1898) and Langley (1901) noted independently the similarity between the effects of injection of extracts of the adrenal gland and stimulation of sympathetic nerves. A few years later, in 1905, T.R. Elliott, while a student with Langley at Cambridge, England, extended these observations and postulated that sympathetic nerve impulses release minute amounts of an epinephrine-like substance in immediate contact with effector cells. He considered this substance to be the chemical step in the process of transmission. He also noted that, long after sympathetic nerves had degenerated, the effector organs still responded characteristically to the hormone of the adrenal medulla. In 1905, Langley suggested that effector cells have excitatory and inhibitory 'receptive substances' and that the response to epinephrine depended on which type of substance was present. In 1907, Dixon was so impressed by the correspondence between the effects of the alkaloid muscarine and the responses to vagal stimulation that he advanced the important idea that the vagus nerve liberated a muscarine-like substance that acted as a chemical transmitter of its impulses. In the same year, Reid Hunt described the actions of ACh and other choline esters. In 1914, Dale thoroughly investigated the pharmacological properties of ACh along with other esters of choline and distinguished its nicotine-like and muscarine-like actions. He was so intrigued with the remarkable fidelity with which this drug reproduced the responses to stimulation of parasympathetic nerves that he introduced the term parasympathomimetic to characterize its effects. Dale also noted the brief duration of the action of this chemical and proposed that an esterase in the tissues rapidly splits ACh to acetic acid and choline, thereby terminating its action.

The studies of Otto Loewi, begun in 1921, provided the first direct evidence for the chemical mediation of nerve impulses by the release of specific chemical agents. Loewi stimulated the vagus nerve of a perfused (donor) frog heart and allowed the perfusion fluid to come in contact with a second (recipient) frog heart used as a test object. The recipient frog heart was found to respond, after a short lag, in the same way as did the donor heart. It was thus evident that a substance was liberated from the first organ that slowed the rate of the second. Loewi referred to this chemical substance as Vagusstoff ('vagus substance'; parasympathin); subsequently, Loewi and Navratil (1926) presented evidence to identify it as ACh. Loewi also discovered that an accelerator substance similar to epinephrine and called Acceleranstoff was liberated into the perfusion fluid in summer, when the action of the sympathetic fibers in the frog's vagus, a mixed nerve, predominated over that of the inhibitory fibers. Loewi's discoveries eventually were confirmed and became universally accepted. Evidence that the cardiac vagus-substance also is ACh in mammals was obtained in 1933 by Feldberg and Krayer.

In addition to the role of ACh as the transmitter of all postganglionic parasympathetic fibers and of a few postganglionic sympathetic fibers, this substance has been shown to have transmitter function in three additional classes of nerves: preganglionic fibers of both the sympathetic and the parasympathetic systems, motor nerves to skeletal muscle, and certain neurons within the CNS.

In the same year as Loewi's discovery, Cannon and Uridil (1921) reported that stimulation of the sympathetic hepatic nerves resulted in the release of an epinephrine-like substance that increased blood pressure and heart rate. Subsequent experiments firmly established that this substance is the chemical mediator liberated by sympathetic nerve impulses at neuroeffector junctions. Cannon called this substance 'sympathin.' In many of its pharmacological and chemical properties, 'sympathin' closely resembled epinephrine, but also differed in certain important respects. As early as 1910, Barger and Dale noted that the effects of sympathetic nerve stimulation were more closely reproduced by the injection of sympathomimetic primary amines than by that of epinephrine or other secondary amines. The possibility that demethylated epinephrine (norepinephrine) might be 'sympathin' had been repeatedly advanced, but definitive evidence for its being the sympathetic nerve mediator was not obtained until specific assays were developed for the determination of sympathomimetic amines in extracts of tissues and body fluids. von Euler in 1946 found that the sympathomimetic substance in highly purified extracts of bovine splenic nerve resembled norepinephrine by all criteria used. Norepinephrine is the predominant sympathomimetic substance in the postganglionic sympathetic nerves of mammals and is the adrenergic mediator liberated by their stimulation (see von Euler, 1972). Norepinephrine, its immediate precursor, dopamine, and epinephrine also are neurotransmitters in the CNS (see Chapter 12: Neurotransmission and the Central Nervous System).

Evidence for Neurohumoral Transmission

The concept of neurohumoral transmission or chemical neurotransmission was developed primarily to explain observations relating to the transmission of impulses from postganglionic autonomic fibers to effector cells. The general lines of evidence to support the concept have included (1) demonstration of the presence of a physiologically active compound and its biosynthetic enzymes at appropriate sites; (2) recovery of the compound from the perfusate of an innervated structure during periods of nerve stimulation but not (or in greatly reduced amounts) in the absence of stimulation; (3) demonstration that the compound is capable of producing responses identical with responses to nerve stimulation; and (4) demonstration that the responses to nerve stimulation and to the administered compound are modified in the same manner by various drugs, usually competitive antagonists.

Chemical, rather than electrogenic, transmission at autonomic ganglia and the neuromuscular junction of skeletal muscle was not generally accepted for a considerable period, because techniques were limited in time and chemical resolution. Techniques of intracellular recording and microiontophoretic application of drugs as well as sensitive analytical assays have overcome these limitations.

Neurotransmission in the peripheral and central nervous systems once was believed to conform to the hypothesis that each neuron contains only one transmitter substance. However, peptides, such as enkephalin, substance P, neuropeptide Y, VIP, and somatostatin; purines such as ATP or adenosine; and small molecules such as nitric oxide, have been found in nerve endings. These substances can depolarize or hyperpolarize nerve terminals or postsynaptic cells. Furthermore, results of histochemical, immunocytochemical, and autoradiographic studies have demonstrated that one or more of these substances is present in the same neurons that contain one of the classical biogenic amine neurotransmitters (Bartfai et al., 1988; Lundberg, 1996). For example, enkephalins are found in postganglionic sympathetic neurons and adrenal medullary chromaffin cells. VIP is localized selectively in peripheral cholinergic neurons that innervate exocrine glands, and neuropeptide Y is found in sympathetic nerve endings. These observations suggest that in many instances synaptic transmission may be mediated by the release of more than one neurotransmitter (see below).

Steps Involved in Neurotransmission

The sequence of events involved in neurotransmission is of particular importance pharmacologically, since the actions of most drugs modulate the individual steps. The term conduction is reserved for the passage of an impulse along an axon or muscle fiber; transmission refers to the passage of an impulse across a synaptic or neuroeffector junction. With the exception of the local anesthetics, very few drugs modify axonal conduction in the doses employed therapeutically. Hence, this process is described only briefly.

Axonal Conduction

Current knowledge of axonal conduction stems largely from the investigative work of Hodgkin and Huxley (1952).

At rest, the interior of the typical mammalian axon is approximately 70 mV negative to the exterior. The resting potential is essentially a diffusion potential, based chiefly on the fortyfold higher concentration of K⁺ in the axoplasm as compared with the extracellular fluid and the relatively high permeability of the resting axonal membrane to this ion. Na⁺ and Cl are present in higher concentrations in the extracellular fluid than in the axoplasm, but the axonal membrane at rest is considerably less permeable to these ions; hence their contribution to the resting potential is small. These ionic gradients are maintained by an energy-dependent active transport or pump mechanism, which involves an adenosine triphosphatase (ATPase) activated by Na⁺ at the inner and by K⁺ at the outer surface of the membrane (see Hille, 1992; Hille et al., 1999a).

In response to depolarization to a threshold level, an action potential or nerve impulse is initiated at a local region of the membrane. The action potential consists of two phases. Following a small gating current resulting from depolarization inducing an open conformation of the channel, the initial phase is caused by a rapid increase in the permeability of Na⁺ through voltage-sensitive Na⁺ channels. The result is inward movement of Na⁺ and a rapid depolarization from the resting potential, which continues to a positive overshoot. The second phase results from the rapid inactivation of the Na⁺ channel and the delayed opening of a K⁺ channel, which permits outward movement of K⁺ to terminate the depolarization. Inactivation appears to involve a voltage-sensitive conformational change in which a hydrophobic peptide loop physically occludes the open channel at the cytoplasmic side. Although not important in axonal conduction, Ca²⁺ channels in other tissues (e.g., heart) contribute to the action potential by prolonging depolarization by an inward movement of Ca²⁺. This influx of Ca²⁺ also serves as a stimulus to initiate intracellular events (Hille, 1992; Catterall, 2000).

The transmembrane ionic currents produce local circuit currents around the axon. As a result of such localized changes in membrane potential, adjacent resting channels in the axon are activated, and excitation of an adjacent portion of the axonal membrane occurs. This brings about the propagation of the action potential without decrement along the axon. The region that has undergone depolarization remains momentarily in a refractory state. In myelinated fibers, permeability changes occur only at the nodes of Ranvier, thus causing a rapidly progressing type of jumping, or saltatory, conduction. The puffer fish poison, tetrodotoxin, and a close congener found in some shellfish, saxitoxin, selectively block axonal conduction; they do so by blocking the voltage-sensitive Na⁺ channel and preventing the increase in permeability to Na⁺ associated with the rising phase of the action potential. In contrast, batrachotoxin, an extremely potent steroidal alkaloid secreted by a South American frog, produces paralysis through a selective increase in permeability of the Na⁺ channel to Na⁺, which induces a persistent depolarization. Scorpion toxins are peptides that also cause persistent depolarization, but they do so by inhibiting the inactivation process (see Catterall, 2000). Na⁺ and Ca²⁺ channels are discussed in more detail in Chapters 15: Local Anesthetics, 32, and 35.

Junctional Transmission

The arrival of the action potential at the axonal terminals initiates a series of events that trigger transmission of an excitatory or inhibitory impulse across the synapse or neuroeffector junction. These events, diagrammed in Figure 62, are as follows.

Cholinergic Transmission

Two enzymes, choline acetyltransferase and AChE, are involved in the synthesis and degradation, respectively, of ACh.

Choline Acetyltransferase

Choline acetyltransferase catalyzes the final step in the synthesis of AChthe acetylation of choline with acetyl coenzyme A (CoA; see Wu and Hersh, 1994; Parsons et al., 1993). The primary structure of choline acetyltransferase is known from molecular cloning, and its immunocytochemical localization has proven useful for identification of cholinergic axons and nerve cell bodies.

Acetyl CoA for this reaction is derived from pyruvate via the multistep pyruvate dehydrogenase reaction or is synthesized by acetate thiokinase, which catalyzes the reaction of acetate with adenosine triphosphate (ATP) to form an enzyme-bound acyladenylate (acetyl AMP). In the presence of CoA, transacetylation and synthesis of acetyl CoA proceed.

Choline acetyltransferase, like other protein constituents of the neuron, is synthesized within the perikaryon and then is transported along the length of the axon to its terminal. Axonal terminals contain a large number of mitochondria, where acetyl CoA is synthesized. Choline is taken up from the extracellular fluid into the axoplasm by active transport. The final step in the synthesis occurs within the cytoplasm, following which most of the ACh is sequestered within the synaptic vesicles. Although moderately potent inhibitors of choline acetyltransferase exist, they have no therapeutic utility, in part because the uptake of choline is the rate-limiting step in the biosynthesis of ACh.

Choline Transport

Transport of choline from the plasma into neurons is accomplished by distinct high- and low-affinity transport systems. The high-affinity system (K_m= 1 to 5 M) is unique to cholinergic neurons, is dependent on extracellular Na⁺, and is inhibited by hemicholinium. Plasma concentrations of choline approximate 10 M; thus, the concentration of choline does not limit its availability to cholinergic neurons. Much of the choline formed from AChE-catalyzed hydrolysis of ACh is recycled back into the nerve terminal. The recent cloning of the high-affinity choline transporter found in presynaptic terminals reveals a sequence and structure differing from those of other neurotransmitter transporters, but similar to that of the Na⁺-dependent glucose transporter family (Okuda et al., 2000).

Upon acetylation of choline, ACh is transported into and packaged in the synaptic vesicle. The vesicular transporter relies on a proton gradient to drive amine uptake. Vesamicol blocks ACh vesicular transport at micromolar concentrations. The genes for choline acetyltransferase and the vesicular transporter are found at the same locus, with the transporter gene positioned in the first intron of the transferase gene. Hence, a common promoter regulates the expression of both genes (Eiden, 1998).

Acetylcholinesterase (AChE)

For ACh to serve as a neurotransmitter in the motor system and certain neuronal synapses, it must be removed or inactivated within the time limits imposed by the response characteristics of the synapse. At the neuromuscular junction, immediate removal is required to prevent lateral diffusion and sequential activation of receptorswith 'flashlike suddenness,' as Dale expressed it. Modern biophysical methods have revealed that the time required for hydrolysis of ACh is less than a millisecond at the neuromuscular junction. Choline has only 10³ to 10⁵ of the potency of ACh at the neuromuscular junction.

While AChE is found in cholinergic neurons (dendrites, perikarya, and axons), it is more widely distributed than cholinergic synapses. It is highly concentrated at the postsynaptic end-plate of the neuromuscular junction. Butyrylcholinesterase (BuChE; also known as pseudocholinesterase) is present in low abundance in glial or satellite cells but is virtually absent in neuronal elements of the central and peripheral nervous systems. BuChE is synthesized primarily in the liver and is found in liver and plasma; its likely vestigial physiological function is the hydrolysis of ingested esters from plant sources. AChE and BuChE typically are distinguished by the relative rates of ACh and butyrylcholine hydrolysis and by effects of selective inhibitors (see Chapter 8: Anticholinesterase Agents). Almost all the pharmacological effects of the anti-ChE agents are due to the inhibition of AChE, with the consequent accumulation of endogenous ACh in the vicinity of the nerve terminal. Distinct, but single, genes encode AChE and BuChE in mammals; the diversity of molecular structures of AChE arise from alternative mRNA processing (Taylor et al., 2000).

Storage and Release of Acetylcholine

Fatt and Katz (1952) recorded at the motor end-plate of skeletal muscle and observed the random occurrence of small (0.1 to 3.0 mV), spontaneous depolarizations at a frequency of approximately one per second. The magnitude of these miniature end-plate potentials (mepps) is considerably below the threshold required to fire a muscle AP; that they are due to the release of ACh is indicated by their enhancement by neostigmine (an anti-ChE agent) and their blockade by d-tubocurarine (a competitive antagonist that acts at nicotinic receptors). These results led to the hypothesis that ACh is released from motor-nerve endings in constant amounts, or quanta. The likely morphological counterpart of quantal release was discovered shortly thereafter in the form of synaptic vesicles (De Robertis and Bennett, 1955). Most of the storage and release properties of ACh originally investigated in motor end-plates apply to other fast-responding synapses. When an action potential arrives at the motor-nerve terminal, there is a synchronous release of 100 or more quanta (or vesicles) of ACh (Katz and Miledi, 1965).

Estimates of the ACh content of synaptic vesicles range from 1000 to over 50,000 molecules per vesicle, and it has been calculated that a single motor-nerve terminal contains 300,000 or more vesicles. In addition, an uncertain but possibly significant amount of ACh is present in the extravesicular cytoplasm. Recording the electrical events associated with the opening of single channels at the motor end-plate during continuous application of ACh has permitted estimation of the potential change induced by a single molecule of ACh (3 x 10⁷ V); from such calculations, it is evident that even the lower estimate of the ACh content per vesicle (1000 molecules) is sufficient to account for the magnitude of the mepps (Katz and Miledi, 1972).

The release of ACh and other neurotransmitters by exocytosis through the prejunctional membrane is inhibited by botulinum and tetanus toxins from Clostridium. Some of the most potent toxins known are produced by these spore-forming anaerobic bacteria (Shapiro et al., 1998). The Clostridium toxins, consisting of disulfide-linked heavy and light chains, bind to an as-yet-unidentified receptor on the membrane of the cholinergic nerve terminal. Through endocytosis, they are transported into the cytosol. The light chain is a Zn²⁺-dependent protease that becomes activated and hydrolyzes components of the core or SNARE complex involved in exocytosis. The various serotypes of botulinum toxin proteolyse selective proteins in the plasma membrane (syntaxin and SNAP-25) and the synaptic vesicle (synaptobrevin). Therapeutic uses of botulinum toxin are described in Chapters 9: Agents Acting at the Neuromuscular Junction and Autonomic Ganglia and 66: Ocular Pharmacology.

By contrast, tetanus toxin primarily has a central action, since it is transported in retrograde fashion up the motor neuron to its soma in the spinal cord. From there, the toxin migrates to inhibitory neurons that synapse with the motor neuron and blocks exocytosis in the inhibitory neuron. The block of release of inhibitory transmitter gives rise to tetanus or spastic paralysis. The toxin from the venom of black widow spiders (-latrotoxin) binds to neurexins, transmembrane proteins that reside on the nerve terminal membrane. This gives rise to massive synaptic vesicle exocytosis (Schiavo et al., 2000).

Characteristics of Cholinergic Transmission at Various Sites

From the comparisons noted above, it is obvious that there are marked differences among various sites of cholinergic transmission with respect to architecture and fine structure, the distributions of AChE and receptors, and the temporal factors involved in normal functioning. For example, in skeletal muscle the junctional sites occupy a small, discrete portion of the surface of the individual fibers and are relatively isolated from those of adjacent fibers; in the superior cervical ganglion, approximately 100,000 ganglion cells are packed within a volume of a few cubic millimeters, and both the presynaptic and postsynaptic neuronal processes form complex networks.

Skeletal Muscle

Stimulation of a motor nerve results in the release of ACh from perfused muscle; close intraarterial injection of ACh produces muscular contraction similar to that elicited by stimulation of the motor nerve. The amount of ACh (10¹⁷mol) required to elicit an EPP following its microiontophoretic application to the motor end-plate of a rat diaphragm muscle fiber is equivalent to that recovered from each fiber following stimulation of the phrenic nerve (Krnjević and Mitchell, 1961).

The combination of ACh with nicotinic acetylcholine receptors at the external surface of the postjunctional membrane induces an immediate, marked increase in permeability to cations. Upon activation of the receptor by ACh, its intrinsic channel opens for about 1 millisecond; during this interval about 50,000 Na⁺ ions traverse the channel (Katz and Miledi, 1972). The channel opening process is the basis for the localized depolarizing EPP within the end-plate, which triggers the muscle action potential. The latter, in turn, leads to contraction. Further details concerning these events and their modification by neuromuscular blocking agents are presented in Chapter 9: Agents Acting at the Neuromuscular Junction and Autonomic Ganglia.

Following section and degeneration of the motor nerve to skeletal muscle or of the postganglionic fibers to autonomic effectors, there is a marked reduction in the threshold doses of the transmitters and of certain other drugs required to elicit a response, i.e., denervation supersensitivity has occurred. In skeletal muscle, this change is accompanied by a spread of the receptor molecules from the end-plate region to the adjacent portions of the sarcoplasmic membrane, which eventually involves the entire muscle surface. Embryonic muscle also exhibits this uniform sensitivity to ACh prior to innervation. Hence, innervation represses the expression of the receptor gene by the nuclei that lie in extrajunctional regions of the muscle fiber and directs the subsynaptic nuclei to the expression of the structural and functional proteins of the synapse (Sanes and Lichtman, 1999).

Autonomic Effectors

Stimulation or inhibition of autonomic effector cells occurs upon activation of muscarinic acetylcholine receptors (see below). In this case the effector is coupled to the receptor by a G protein (see Chapter 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship Between Drug Concentration and Effect). In contrast to skeletal muscle and neurons, smooth muscle and the cardiac conduction system (SA node, atrium, AV node, and the His-Purkinje system) normally exhibit intrinsic activity, both electrical and mechanical, that is modulated but not initiated by nerve impulses. In the basal condition, unitary smooth muscle exhibits waves of depolarization and/or spikes that are propagated from cell to cell at rates considerably slower than the AP of axons or skeletal muscle. The spikes apparently are initiated by rhythmic fluctuations in the membrane resting potential. In intestinal smooth muscle, the site of the pacemaker activity continually shifts, but in the heart, spontaneous depolarizations normally arise from the SA node; however, when activity of the SA node is repressed or under pathological conditions, they can arise from any part of the conduction system (see Chapter 35: Antiarrhythmic Drugs).

Application of ACh (0.1 to 1 M) to isolated intestinal muscle causes a decrease in the resting potential (i.e., the membrane potential becomes less negative) and an increase in the frequency of spike production accompanied by a rise in tension. A primary action of ACh in initiating these effects through muscarinic receptors is probably the partial depolarization of the cell membrane, brought about by an increase in Na⁺ and, in some instances, Ca²⁺ conductance. ACh also can produce contraction of some smooth muscles when the membrane has been completely depolarized by high concentrations of K⁺, provided Ca²⁺ is present. Hence, ACh stimulates ion fluxes across membranes and/or mobilizes intracellular Ca²⁺ to cause contraction.

In the cardiac conduction system, particularly in the SA and the AV nodes, stimulation of the cholinergic innervation or the direct application of ACh causes inhibition, associated with hyperpolarization of the membrane and a marked decrease in the rate of depolarization. These effects are due, at least in part, to a selective increase in permeability to K⁺ (Hille, 1992).

Autonomic Ganglia

The primary pathway of cholinergic transmission in autonomic ganglia is similar to that at the neuromuscular junction of skeletal muscle. Ganglion cells can be discharged by injecting very small amounts of ACh into the ganglion. The initial depolarization is the result of activation of nicotinic ACh receptors, which are ligand-gated cation channels with properties similar to those found at the neuromuscular junction. Several secondary transmitters or modulators either enhance or diminish the sensitivity of the postganglionic cell to ACh. This sensitivity appears to be related to the membrane potential of the postsynaptic nerve cell body or its dendritic branches. Ganglionic transmission is discussed in more detail in Chapter 9: Agents Acting at the Neuromuscular Junction and Autonomic Ganglia.

Actions of Acetylcholine at Prejunctional Sites

Considerable attention has been focused on the possible involvement of prejunctional cholinoceptive sites in both cholinergic and noncholinergic transmission and in the actions of various drugs. Although cholinergic innervation of blood vessels is limited, prejunctional muscarinic receptors appear to be present on sympathetic vasoconstrictor nerves (Steinsland et al., 1973). The physiological role of these receptors is not clear, but their activation causes inhibition of neurally mediated release of norepinephrine (see Chapter 7: Muscarinic Receptor Agonists and Antagonists). Because ACh is rapidly hydrolyzed by tissue-localized and circulating esterases, it is unlikely that it plays a role as a circulating hormone analogous to that of epinephrine.

Dilation of blood vessels in response to administered choline esters involves several sites of action, including prejunctional inhibitory synapses on sympathetic fibers and inhibitory cholinergic receptors in the vasculature that are not innervated. The vasodilator effect of ACh on isolated blood vessels requires an intact endothelium. Activation of muscarinic receptors results in the liberation of a vasodilator substance (endothelium-derived relaxing factor or nitric oxide) that diffuses from the endothelium to the adjoining smooth muscle and causes relaxation (see below and Chapter 7: Muscarinic Receptor Agonists and Antagonists; see also Furchgott, 1999).

Cholinergic Receptors and Signal Transduction

Sir Henry Dale noted that the various esters of choline elicited responses that were similar to those of either nicotine or muscarine, depending on the pharmacological preparation (Dale, 1914). A similarity in response also was noted between muscarine and nerve stimulation in those organs innervated by the craniosacral divisions of the autonomic nervous system. Thus, Dale suggested that ACh or another ester of choline was a neurotransmitter in the autonomic nervous system; he also stated that the compound had dual actions, which he termed a nicotine action (nicotinic) and a muscarine action (muscarinic).

The capacities of tubocurarine and atropine to block nicotinic and muscarinic effects of ACh, respectively, provided further support for the proposal of two distinct types of cholinergic receptors. Although Dale had access only to crude plant alkaloids of then-unknown structure from Amanita muscaria and Nicotiana tabacum, this classification remains as the primary subdivision of cholinergic receptors. Its utility has survived the discovery of several distinct subtypes of nicotinic and muscarinic cholinergic receptors.

Although ACh and certain other compounds stimulate both muscarinic and nicotinic receptors, several other agonists and antagonists are selective for one of the two major types of receptor. ACh itself is a flexible molecule, and indirect evidence suggests that the conformations of the neurotransmitter are distinct when it is bound to nicotinic or muscarinic receptors.

Nicotinic receptors are ligand-gated ion channels, and their activation always causes a rapid (millisecond) increase in cellular permeability to Na⁺ and Ca²⁺, depolarization, and excitation. By contrast, muscarinic receptors belong to the class of G proteincoupled receptors. Responses to muscarinic agonists are slower; they may be either excitatory or inhibitory, and they are not necessarily linked to changes in ion permeability.

The primary structures of various species of nicotinic receptors (Numa et al., 1983; Changeux and Edelstein, 1998) and muscarinic receptors (Bonner, 1989; Caulfield and Birdsall, 1998) have been deduced from the sequences of their respective genes. That these two types of receptor belong to distinct families of proteins is not surprising, retrospectively, in view of their distinct differences in chemical specificity and function.

The nicotinic receptors exist as pentameric arrangements of one to four distinct subunits that are homologous in sequence; the individual subunits are arranged to surround an internal channel. One of the subunits, designated , is present in at least two copies, and the multiple binding sites for ACh are formed at one of the interfaces of the -subunit with the neighboring subunit. One -helical membrane-spanning sequence from each subunit forms the channel boundary (Changeux and Edelstein, 1998; see Chapters 9: Agents Acting at the Neuromuscular Junction and Autonomic Ganglia and 12: Neurotransmission and the Central Nervous System). The general properties of muscarinic receptor coupling to G proteins and the characteristics of the muscarinic ligand-binding site are described in Chapters 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship between Drug Concentration and Effect and 7: Muscarinic Receptor Agonists and Antagonists.

Subtypes of Nicotinic Receptors

Based on the distinct actions of certain agonists and antagonists that interact with nicotinic receptors from skeletal muscle and ganglia, it long has been evident that not all nicotinic receptors are identical. Heterogeneity of this type of receptor was further revealed by molecular cloning. For example, the muscle nicotinic receptor contains four distinct subunits in a pentameric complex ( or ). Receptors in embryonic or denervated muscle contain a subunit, whereas an subunit replaces the in adult innervated muscle. This change in expression of the genes encoding the and subunits gives rise to small differences in ligand selectivity, but the switch may be more important for dictating rates of turnover of the receptors or their tissue localization. Nicotinic receptors in the CNS also exist as pentamers. Because of the diversity of neuronal nicotinic receptor subunits, they have been designated as the and subtypes. There are eight subtypes of 9) and three subtypes of 4) in the mammalian nervous system. Although not all combinations of and are functional, the number of permutations of and that yield functional receptors is sufficiently large to preclude a pharmacological classification of all subtypes. Homooligomeric pentamers of 8, and 9 subunits form functional receptors. Distinctions between nicotinic receptors are listed in Table 62; the structure, function, distribution, and subtypes of nicotinic receptors are described in more detail in Chapter 9: Agents Acting at the Neuromuscular Junction and Autonomic Ganglia.

Subtypes of Muscarinic Receptors

Five subtypes of muscarinic ACh receptors have been detected by molecular cloning. Similar to the different forms of nicotinic receptors, these variants have distinct anatomical localizations and chemical specificities. The muscarinic receptors all act through G-protein signaling systems (see discussion below and Table 62).

Of the large number of muscarinic antagonists studied over many decades, only pirenzepine, found in the 1970s, showed the unique property of blocking gastric acid secretion at concentrations that did not affect several other responses to muscarinic agonists. These observations and subsequent study of other agonists and antagonists, followed by rapid advances in the cloning of cDNAs that encode muscarinic receptors, led to the identification of five subtypes of muscarinic receptors. They have been designated as M₁ through M₅ based on pharmacological specificity (Bonner, 1989; see also Chapter 7: Muscarinic Receptor Agonists and Antagonists).

M₁ receptors are found in ganglia and in some secretory glands; M₂ receptors predominate in the myocardium and also appear to be found in smooth muscle; and M₃ and M₄ receptors are located in smooth muscle and secretory glands. All five subtypes are found in the CNS. Various tissues may contain several subtypes of muscarinic receptors; parasympathetic ganglia in the tissue also contain muscarinic receptors.

The basic functions of muscarinic receptors are mediated by interactions with members of the family of G proteins and thus by G proteininduced changes in the functions of distinct membrane-bound effector molecules. The M₁, M₃, and M₅ subtypes activate a G protein, termed G_q/11, that is responsible for stimulation of phospholipase C activity; the immediate result is hydrolysis of phosphatidylinositol polyphosphates (which are components of the plasma membrane) to form inositol polyphosphates. Some of the inositol phosphate isomers (chiefly inositol-1,4,5-trisphosphate) cause release of intracellular Ca²⁺ from stores in the endoplasmic reticulum. Thus, these receptors mediate such Ca²⁺-dependent phenomena as contraction of smooth muscle and secretion (see Chapter 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship Between Drug Concentration and Effect; see also Berridge, 1993). The second product of the phospholipase C reaction, diacylglycerol, activates protein kinase C (in conjunction with Ca²⁺). This arm of the pathway plays a role in modulation of function and in the later phases of the functional response (Dempsey et al., 2000).

A second pathway for mediation of responses to muscarinic agonists is evoked by activation of M₂ and M₄ receptors. These receptors interact with a distinct group of G proteins (in particular those termed G_i and G_o) with resultant inhibition of adenylyl cyclase, activation of receptor-operated K⁺ channels (in the heart, for example), and suppression of the activity of voltage-gated Ca²⁺ channels in certain cell types. The functional consequences of these effects are most clear in the myocardium, where inhibition of adenylyl cyclase and activation of K⁺ conductances can account for both the negative chronotropic and inotropic effects of ACh.

Other cellular events such as the release of arachidonic acid and the activation of guanylyl cyclase also can result from activation of muscarinic receptors; typically these responses are secondary to the production of other mediators.

Adrenergic Transmission

Under this general heading are included norepinephrine, the transmitter of most sympathetic postganglionic fibers and of certain tracts in the CNS, and dopamine, the predominant transmitter of the mammalian extrapyramidal system and of several mesocortical and mesolimbic neuronal pathways, as well as epinephrine, the major hormone of the adrenal medulla.

A tremendous amount of information about catecholamines and related compounds has accumulated in recent years, partly because of the importance of interactions between the endogenous catecholamines and many of the drugs used in the treatment of hypertension, mental disorders, and a variety of other conditions. The details of these interactions and of the pharmacology of the sympathomimetic amines themselves will be found in subsequent chapters. The basic physiological, biochemical, and pharmacological features are presented here.

Synthesis, Storage, and Release of Catecholamines

Synthesis

The synthesis of epinephrine from tyrosine, by the steps shown in Figure 63, was proposed by Blaschko in 1939. The enzymes involved have been identified, cloned, and characterized (Nagatsu, 1991). It is important to note that these enzymes are not completely specific; consequently, other endogenous substances as well as certain drugs are similarly acted upon at the various steps. For example, 5-hydroxytryptamine (5-HT, serotonin) can be produced by aromatic L-amino acid decarboxylase (or dopa decarboxylase) from 5-hydroxy-L-tryptophan. Dopa decarboxylase also converts dopa into dopamine, and methyldopa is converted to -methyldopamine, which, in turn, is converted by dopamine -hydroxylase to the 'false transmitter,'-methylnorepinephrine.

The hydroxylation of tyrosine generally is regarded as the rate-limiting step in the biosynthesis of catecholamines (Zigmond et al., 1989), and tyrosine hydroxylase is activated following stimulation of adrenergic nerves or the adrenal medulla. The enzyme is a substrate for cyclic AMPdependent and Ca²⁺-calmodulin-sensitive protein kinase and protein kinase C; kinase-catalyzed phosphorylation may be associated with increased hydroxylase activity (Zigmond et al., 1989; Daubner et al., 1992). This is an important acute mechanism for increasing catecholamine synthesis in response to increased nerve stimulation. In addition, there is a delayed increase in tyrosine hydroxylase gene expression after nerve stimulation. There is evidence suggesting that this increased expression can occur at multiple levels of regulation, including transcription, RNA processing, regulation of RNA stability, translation, and enzyme stability (Kumer and Vrana, 1996). These mechanisms serve to maintain the content of catecholamines in response to increased release of these transmitters. In addition, tyrosine hydroxylase is subject to feedback inhibition by catechol compounds, an allosteric modulation of enzyme activity. Patients with mutations in the tyrosine hydroxylase gene have been described (Wevers et al., 1999).

Current knowledge concerning the cellular sites and mechanisms of synthesis, storage, and release of catecholamines has been derived from studies of both adrenergically innervated organs and of adrenal medullary tissue. Nearly all the norepinephrine content of the former is confined to the postganglionic sympathetic fibers; it disappears within a few days after section of the nerves. In the adrenal medulla, catecholamines are stored in chromaffin granules (Winkler, 1997; Aunis, 1998). These vesicles contain extremely high concentrations of catecholamines (approximately 21% dry weight), ascorbic acid, and ATP, as well as specific proteins such as chromogranins, the enzyme dopamine -hydroxylase (DBH), and peptides including enkephalin and neuropeptide Y. Interestingly, vasostatin-1, the N-terminal fragment of chromogranin A, has been found to have antibacterial and antifungal activity (Lugardon et al., 2000). Two types of storage vesicles are found in sympathetic nerve terminals: large dense-core vesicles corresponding to chromaffin granules and small dense-core vesicles containing norepinephrine, ATP, and membrane-bound dopamine -hydroxylase.

The main features of the mechanisms of synthesis, storage, and release of catecholamines and their modifications by drugs are summarized in Figure 64. In the case of adrenergic neurons, the enzymes that participate in the formation of norepinephrine are synthesized in the cell bodies of the neurons and are then transported along the axons to their terminals. In the course of synthesis (see Figure 63), the hydroxylation of tyrosine to dopa and the decarboxylation of dopa to dopamine take place in the cytoplasm. About half the dopamine formed in the cytoplasm then is actively transported into the DBH-containing storage vesicles, where it is converted to norepinephrine; the remainder, which escaped capture by the vesicles, is deaminated to 3,4-dihydroxyphenylacetic acid (DOPAC) and subsequently O-methylated to homovanillic acid (HVA). The adrenal medulla has two distinct catecholamine-containing cell types: those with norepinephrine and those with primarily epinephrine. The latter cell population contains the enzyme phenylethanolamine-N-methyltransferase. In these cells, the norepinephrine formed in the granules leaves these structures, presumably by diffusion, and is methylated in the cytoplasm to epinephrine. Epinephrine then reenters the chromaffin granules, where it is stored until released. In adults, epinephrine accounts for approximately 80% of the catecholamines of the adrenal medulla, with norepinephrine making up most of the remainder (von Euler, 1972).

A major factor that controls the rate of synthesis of epinephrine, and hence the size of the store available for release from the adrenal medulla, is the level of glucocorticoids secreted by the adrenal cortex. The latter hormones are carried in high concentration, by the intraadrenal portal vascular system, directly to the adrenal medullary chromaffin cells, where they induce the synthesis of phenylethanolamine-N-methyltransferase (see Figure 63). The activities of both tyrosine hydroxylase and DBH also are increased in the adrenal medulla when the secretion of glucocorticoids is stimulated (Carroll et al., 1991; Viskupic et al., 1994). Thus, any stress that persists sufficiently to evoke an enhanced secretion of corticotropin mobilizes the appropriate hormones of both the adrenal cortex (predominantly cortisol) and medulla (epinephrine).

This remarkable relationship is present only in certain mammals, including human beings, for which the adrenal chromaffin cells are enveloped entirely by steroid-secreting cortical cells. In the dogfish, for example, where the chromaffin cells and steroid-secreting cells are located in independent, noncontiguous glands, epinephrine is not formed. Nonetheless, there is evidence indicating that phenylethanolamine-N-methyltransferase is expressed in mammalian tissues such as brain, heart, and lung, leading to extra-adrenal epinephrine synthesis (Kennedy and Ziegler, 1991; Kennedy et al., 1993).

In addition to its de novo synthesis, outlined above, there is a second major mechanism for replenishment of the norepinephrine stored in the terminal portions of the adrenergic fibersnamely, recapture by active transport of norepinephrine previously released to the extracellular fluid. This process is responsible for the termination of the effects of adrenergic impulses in most organs. In blood vessels and in tissues with wide synaptic gaps, recapture of released norepinephrine is less important. At such sites, a relatively large fraction of the released neurotransmitter is inactivated by a combination of extraneuronal uptake (see below) and enzymatic breakdown and diffusion. To effect the reuptake of norepinephrine into adrenergic nerve terminals and to maintain the concentration gradient of norepinephrine within the vesicles, at least two distinct carrier-mediated transport systems are involved: one across the axoplasmic membrane from the extracellular fluid to the cytoplasm and the other from the cytoplasm into the storage vesicles.

Storage of Catecholamines

Catecholamines are stored in vesicles to ensure their regulated release; this storage decreases intraneuronal metabolism of these transmitters as well as their leakage outside the cell. The amine transporter has been extensively characterized (Schuldiner, 1994). Uptake of catecholamine and ATP into isolated chromaffin granules appears to be driven by pH and potential gradients that are established by an ATP-dependent proton translocase. For every molecule of amine taken up, two H⁺ ions are extruded (Brownstein and Hoffman, 1994). Monoamine transporters are relatively promiscuous, capable of transporting dopamine, norepinephrine, epinephrine, and serotonin, for example. Also, meta-iodobenzylguanidine, used clinically to image chromaffin-cell tumors, is a substrate for this transport system (Schuldiner, 1994). Reserpine is a drug that inhibits monoamine transport into these vesicles which ultimately leads to depletion of catecholamine from sympathetic nerve endings and in the brain. Several vesicular transport cDNAs have been identified with molecular cloning techniques; these cDNAs reveal open reading frames predictive of proteins with 12 putative transmembrane domains with structural homology to other transport proteins such as bacterial drug-resistance transporters (Schuldiner, 1994). Regulation of the expression of these various transporters may be important in the regulation of synaptic transmission (Varoqui and Erickson, 1997).

When catecholamines such as norepinephrine are injected into the blood of experimental animals, they are rapidly accumulated in tissues with extensive sympathetic innervation, such as heart and spleen; labeled catecholamines are concentrated in sympathetic nerve endings, and tissue uptake disappears after denervation (reviewed in Brownstein and Hoffman, 1994). This and other evidence suggested the presence of transporters on the plasma membrane of sympathetic neurons that could take up catecholamines. The amine transport system across the axoplasmic membranes is Na⁺-dependent and is blocked selectively by a number of drugs, including cocaine and the tricyclic antidepressants, such as imipramine. This transporter has a high affinity for norepinephrine and a somewhat lower affinity for epinephrine; the synthetic -adrenergic receptor agonist isoproterenol is not a substrate for this system. The neuronal uptake process has been termed uptake-1 (Iversen, 1975). A number of highly specific neurotransmitter transporters have been identified by protein purification or expression cloning techniques. High-affinity transporters have been identified, for example, for dopamine, norepinephrine, serotonin, and a amino acid transmitters (Amara and Kuhar, 1993; Brownstein and Hoffman, 1994; Masson et al., 1999). These transporters are members of an extended family sharing common structural motifs, particularly the putative 12 transmembrane helices. These plasma membrane transporters appear to have greater substrate specificity than do vesicular transporters. Indeed, these transport systems may be viewed as targets ('receptors') for specific drugs such as cocaine (dopamine transporter) or fluoxetine (serotonin transporter).

Certain sympathomimetic drugs (e.g., ephedrine, tyramine) produce some of their effects indirectly, chiefly by displacing norepinephrine from the nerve-ending binding sites to the extracellular fluid, where the released endogenous transmitter then acts at receptor sites of the effector cells. The mechanisms by which indirect-acting sympathomimetic amines release norepinephrine from nerve endings are complex. All such agents are substrates for uptake-1. As a result of their transport across the neuronal membrane and release into the axoplasm, they make carrier available at the inner surface of the membrane for the outward transport of norepinephrine ('facilitated exchange diffusion'). In addition, these amines are able to mobilize norepinephrine stored in the vesicles by competing for the vesicular uptake process. Reserpine, which depletes vesicular stores of norepinephrine, also inhibits this uptake mechanism, but, in contrast with the indirect-acting sympathomimetic amines, it enters the adrenergic nerve ending by passive diffusion across the axonal membrane (Bnisch and Trendelenburg, 1988).

These actions of indirect-acting sympathomimetic amines are associated with the phenomenon of tachyphylaxis. For example, repeated administration of tyramine results in rapidly decreasing effectiveness, whereas repeated administration of norepinephrine does not reduce effectiveness and, in fact, reverses the tachyphylaxis to tyramine. Although these phenomena have not been explained fully, several hypotheses have been proposed. One possible explanation of tachyphylaxis to tyramine and similarly acting sympathomimetic agents is that the pool of neurotransmitter available for displacement by these drugs is small relative to the total amount stored in the sympathetic nerve ending. This pool is presumed to reside in close proximity to the plasma membrane, and the norepinephrine of such vesicles may be replaced by the less potent amine following repeated administration of the latter substance. In any case, neurotransmitter release by displacement is not associated with the release of dopamine -hydroxylase and does not require extracellular Ca²⁺; thus, it is presumed not to involve exocytosis.

There also is an extraneuronal amine transport system, termed uptake-2, which exhibits a low affinity for norepinephrine, a somewhat higher affinity for epinephrine, and a still higher affinity for isoproterenol. This uptake process is ubiquitous and is present in glial, hepatic, myocardial, and other cells. Uptake-2 is not inhibited by imipramine or cocaine. It probably is of relatively little physiological importance unless the neuronal uptake mechanism is blocked (Iversen, 1975; Trendelenburg, 1980). It may be of greater importance in the disposition of circulating catecholamines than in the removal of amines that have been released from adrenergic nerve terminals.

Release of Catecholamines

The full sequence of steps by which the nerve impulse effects the release of norepinephrine from adrenergic fibers is not known. In the adrenal medulla, the triggering event is the liberation of ACh by the preganglionic fibers and its interaction with nicotinic receptors on chromaffin cells to produce a localized depolarization; a succeeding step is the entrance of Ca²⁺ into these cells, which results in the extrusion by exocytosis of the granular contents, including epinephrine, ATP, some neuroactive peptides or their precursors, chromogranins, and DBH. Influx of Ca²⁺ likewise plays an essential role in coupling the nerve impulse, membrane depolarization, and opening of voltage-gated Ca²⁺ channels with the release of norepinephrine at adrenergic nerve terminals. Blockade of N-type Ca²⁺ channels leads to hypotension, likely due to inhibition of norepinephrine release (Bowersox et al., 1992). Ca²⁺-triggered secretion involves interaction of highly conserved molecular scaffolding proteins leading to docking of granules at the plasma membrane, ultimately leading to secretion (Aunis, 1998). Enhanced activity of the sympathetic nervous system is accompanied by an increased concentration of both DBH and chromogranins in the circulation, supporting the argument that the process of release following adrenergic nerve stimulation also involves exocytosis.

Adrenergic fibers can sustain the output of norepinephrine during prolonged periods of stimulation without exhausting their reserve supply, provided synthesis and uptake of the transmitter are unimpaired. To meet increased needs for norepinephrine, acute regulatory mechanisms come into play that involve activation of tyrosine hydroxylase and dopamine -hydroxylase (see above).

Termination of the Actions of Catecholamines

The actions of norepinephrine and epinephrine are terminated by (1) reuptake into nerve terminals; (2) dilution by diffusion out of the junctional cleft and uptake at extraneuronal sites; and (3) metabolic transformation. Two enzymes are important in the initial steps of metabolic transformation of catecholaminesmonoamine oxidase (MAO) and catechol-O-methyltransferase (COMT; see Axelrod, 1966; Kopin, 1972). In addition, catecholamines are metabolized by sulfotransferases (Dooley, 1998). However, it is evident that a powerful degradative enzymatic pathway, such as that provided by AChE, is absent from the adrenergic nervous system. The importance of neuronal reuptake of catecholamines is shown by observations that inhibitors of this process (e.g., cocaine, imipramine) potentiate the effects of the neurotransmitter; inhibitors of MAO and COMT have relatively little effect. However, transmitter that is released within the nerve terminal is metabolized by MAO. COMT, particularly in the liver, plays a major role in the metabolism of endogenous circulating and administered catecholamines.

Both MAO and COMT are widely distributed throughout the body, including the brain; the highest concentrations of each are in the liver and the kidney. However, little or no COMT is found in adrenergic neurons. There are distinct differences in the cytological locations of the two enzymes; whereas MAO is associated chiefly with the outer surface of mitochondria, including those within the terminals of adrenergic fibers, COMT is located largely in the cytoplasm. These factors are of importance both in determining the primary metabolic pathways followed by catecholamines in various circumstances and in explaining the effects of certain drugs. Two different isozymes of MAO (MAO-A and MAO-B) are found in widely varying proportions in different cells in the CNS and in peripheral tissues. Selective inhibitors of these two isozymes are available (see Chapter 19: Drugs and the Treatment of Psychiatric Disorders: Depression and Anxiety Disorders). Irreversible antagonists of MAO-A enhance the bioavailability of tyramine contained in many foods; tyramine-induced norepinephrine release from sympathetic neurons may lead to markedly increased blood pressure. Selective MAO-B inhibitors (e.g., selegiline) or reversible MAO-A-selective inhibitors (moclobemide) are less likely to cause this potential interaction (Volz and Geiter, 1998; Wouters, 1998). MAO inhibitors are useful in the treatment of Parkinson's disease and mental depression (see Chapters 19: Drugs and the Treatment of Psychiatric Disorders: Depression and Anxiety Disorders and 22: Treatment of Central Nervous System Degenerative Disorders).

Most of the epinephrine and norepinephrine that enter the circulationfrom the adrenal medulla or following administration or that is released by exocytosis from adrenergic fibersis methylated by COMT to metanephrine or normetanephrine, respectively (Figure 65). Norepinephrine that is released intraneuronally by drugs such as reserpine is initially deaminated by MAO to 3,4-dihydroxyphenylglycolaldehyde (DOPGAL; see Figure 65). The aldehyde is reduced by aldehyde reductase to the glycol, 3,4-dihydroxyphenylethylene glycol (DOPEG), or is oxidized by aldehyde dehydrogenase to form 3,4-dihydroxymandelic acid (DOMA). 3-Methoxy-4-hydroxymandelic acid [generally but incorrectly called vanillylmandelic acid (VMA)] is the major metabolite of catecholamines excreted in the urine. The corresponding product of the metabolic degradation of dopamine, which contains no hydroxyl group in the side chain, is homovanillic acid (HVA). Other metabolic reactions are described in Figure 65. Measurement of the concentrations of catecholamines and their metabolites in blood and urine is useful in the diagnosis of pheochromocytoma, a catecholamine-secreting tumor of the adrenal medulla.

Inhibitors of MAO (e.g., pargyline, nialamide) can cause an increase in the concentration of norepinephrine, dopamine, and 5-HT in the brain and other tissues accompanied by a variety of pharmacological effects. No striking pharmacological action in the periphery can be attributed to the inhibition of COMT. However, a COMT inhibitor, entacapone, has been found to be efficacious in the therapy of Parkinson's disease (Chong and Mersfelder, 2000; see also Chapter 22: Treatment of Central Nervous System Degenerative Disorders).

Classification of Adrenergic Receptors

Crucial to understanding the remarkably diverse effects of the catecholamines and related sympathomimetic agents is an understanding of the classification and properties of the different types of adrenergic receptors (or adrenoceptors). Elucidation of the characteristics of these receptors and the biochemical and physiological pathways they regulate has increased our understanding of the seemingly contradictory and variable effects of catecholamines on various organ systems. Although structurally related (see below), different adrenergic receptors regulate distinct physiological processes by controlling the synthesis or release of a variety of second messengers (see Tables 63 and 64).

Ahlquist (1948) first proposed the existence of more than one adrenergic receptor; he based his hypothesis on a study of the abilities of epinephrine, norepinephrine, and other related agonists to regulate various physiological processes. It was known that these drugs could cause either contraction or relaxation of smooth muscle, depending on the site, the dose, and the agent chosen. For example, norepinephrine was known to have potent excitatory effects on smooth muscle and correspondingly low activity as an inhibitor; isoproterenol displayed the opposite pattern of activity. Epinephrine could both excite and inhibit smooth muscle. Thus, Ahlquist proposed the designations and for receptors on smooth muscle where catecholamines produce excitatory and inhibitory responses, respectively. An exception is the gut, which generally is relaxed by activation of either - or -adrenergic receptors. The rank order of potency of agonists is isoproterenol > epinephrine norepinephrine for -adrenergic receptors and epinephrine norepinephrine >> isoproterenol for -adrenergic receptors (see Table 63). This initial classification of adrenergic receptors was corroborated by the finding that certain antagonists produce selective blockade of the effects of adrenergic nerve impulses and sympathomimetic agents at -adrenergic receptors (e.g., phenoxybenzamine), whereas others produce selective -adrenergic receptor blockade (e.g., propranolol).

-Adrenergic receptors later were subdivided into (e.g., those in the myocardium) and (smooth muscle and most other sites), because epinephrine and norepinephrine essentially are equipotent at the former sites, whereas epinephrine is 10- to 50-fold more potent than norepinephrine at the latter (Lands et al., 1967). Antagonists that discriminate between - and -adrenergic receptors subsequently were developed (see Chapter 10: Catecholamines, Sympathomimetic Drugs, and Adrenergic Receptor Antagonists). A human gene that encodes a third -adrenergic receptor (designated ) has been isolated (Emorine et al., 1989; Granneman et al., 1993). Since the receptor is about tenfold more sensitive to norepinephrine than to epinephrine and is relatively resistant to blockade by antagonists such as propranolol, it may mediate responses to catecholamine at sites with 'atypical' pharmacological characteristics (e.g., adipose tissue). However, the role of this receptor in regulating lipolysis in human beings remains uncertain (Rosenbaum et al., 1993; Krief et al., 1993; Lnnqvist et al., 1993). It has been suggested that polymorphisms in the receptor gene may be related to risk of obesity or type 2 diabetes in some populations (Arner and Hoffstedt, 1999). Also, there has been interest in the possibility that -receptor-selective agonists might have benefit in treating these disorders (Weyer et al., 1999).

The heterogeneity of -adrenergic receptors also is now appreciated. The initial distinction was based on functional and anatomical considerations when it was realized that norepinephrine and other -adrenergic agonists could profoundly inhibit the release of norepinephrine from neurons (see Starke, 1987; see also Figure 64). Indeed, when sympathetic nerves are stimulated in the presence of certain -adrenergic antagonists, the amount of norepinephrine liberated by each nerve impulse increases markedly. This feedback inhibitory effect of norepinephrine on its release from nerve terminals is mediated by receptors that are pharmacologically distinct from the classical postsynaptic receptors. Accordingly, these presynaptic receptors were designated , whereas the postsynaptic 'excitatory' receptors were designated (see Langer, 1997). Compounds such as clonidine are more potent agonists at than at receptors; by contrast, phenylephrine and methoxamine selectively activate postsynaptic receptors. Although there is little evidence to suggest that -adrenergic receptors function presynaptically in the autonomic nervous system, it now is clear that -adrenergic receptors also are present at postjunctional or nonjunctional sites in several tissues. For example, stimulation of postjunctional receptors in the brain is associated with reduced sympathetic outflow from the CNS and appears to be responsible for a significant component of the antihypertensive effect of drugs such as clonidine (see Chapter 10: Catecholamines, Sympathomimetic Drugs, and Adrenergic Receptor Antagonists). Thus, the anatomical concept of prejunctional - and postjunctional -adrenergic receptors has been abandoned in favor of a pharmacological and functional classification (see Table 63).

Cloning revealed additional heterogeneity of both - and -adrenergic receptors (Bylund, 1992). There are three pharmacologically defined -adrenergic receptors (_{1A 1B}, and _1D; see Table 65), with distinct sequences and tissue distributions. Nonetheless, unique functional properties of the different -adrenergic receptor subtypes, for the most part, have not been elucidated. There are three cloned subtypes of -adrenergic receptors (_{2A 2B 2C}; Table 65). Distinct patterns of distribution of these subtypes exist in the brain, and it is likely that at least the _2A subtype can serve as a presynaptic autoreceptor (Aantaa et al., 1995; Lakhlani et al., 1997).

Molecular Basis of Adrenergic Receptor Function

The responses that follow activation of all types of adrenergic receptors appear to result from G proteinmediated effects on the generation of second messengers and on the activity of ion channels. As discussed in Chapter 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship Between Drug Concentration and Effect, these systems involve three interacting proteinsthe receptor, the coupling G protein, and effector enzymes or ion channels. The pathways overlap broadly with those discussed for muscarinic receptors and are summarized in Table 64.

Structure of Adrenergic Receptors

The adrenergic receptors constitute a family of closely related proteins. They also are related both structurally and functionally to receptors for a wide variety of other hormones and neurotransmitters that are coupled to G proteins (Lefkowitz, 2000). This wider family of receptors includes the muscarinic acetylcholine receptors and even the visual 'photon receptor,' rhodopsin (see Chapter 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship Between Drug Concentration and Effect). Ligand binding, site-directed labeling, and mutagenesis have revealed that the conserved membrane-spanning regions are crucially involved in the binding of ligands (Strader et al., 1994; Hutchins, 1994). These regions appear to create a ligand-binding pocket analogous to that formed by the membrane-spanning regions of rhodopsin to accommodate the covalently attached chromophore, retinal, with molecular models placing catecholamines either horizontally (Strader et al., 1994) or perpendicularly (Hutchins, 1994) in the bilayer. The crystal structure of mammalian rhodopsin has been determined and confirms a number of predictions about the structure of G proteincoupled receptors (Palczewski et al., 2000).

-Adrenergic Receptors

The three -adrenergic receptors share approximately 60% amino acid sequence identity within the presumed membrane-spanning domains, where the ligand-binding pocket for epinephrine and norepinephrine is found. Based on results of site-directed mutagenesis, individual amino acids in the -adrenergic receptor that interact with each of the functional groups on the catecholamine agonist molecule have been identified.

All -adrenergic receptors stimulate adenylyl cyclase via interaction with G_s (see Chapter 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship Between Drug Concentration and Effect; see also Taussig and Gilman, 1995). Stimulation of the receptor leads to the accumulation of cyclic AMP, activation of the cyclic AMPdependent protein kinase, and altered function of numerous cellular proteins as a result of their phosphorylation (see below). In addition, G_s can enhance directly the activation of voltage-sensitive Ca²⁺ channels in the plasma membrane of skeletal and cardiac muscle; this action provides an additional means of regulating the function of these tissues.

The cyclic AMPdependent protein kinase (protein kinase A) generally is considered the major intracellular receptor for cyclic AMP. It exists as a tetramer (R₂C₂), consisting of two regulatory (R) and two catalytic (C) subunits. Binding of cyclic AMP causes dissociation of the regulatory subunits, as a result of a 10,000- to 100,000-fold decrease in affinity of R for C, with resultant activation of the catalytic subunits (Francis and Corbin, 1994; Smith et al., 1999). Phosphorylation of various cellular proteins then causes responses that are characteristic of those produced by -adrenergic agonists. When the stimulus is removed, dephosphorylation of the various protein substrates is catalyzed by phosphoprotein phosphatases. Compartmentalization of protein kinase A is a key determinant in the specificity of responses mediated by this kinase. Localization of protein kinase A is fostered by so-called A-kinase anchoring proteins (AKAPs; Edwards and Scott, 2000).

A well-defined example of these mechanisms is the activation of hepatic glycogen phosphorylase, the enzyme that promotes the rate-limiting step in glycogenolysis, the conversion of glycogen to glucose-1-phosphate. This activation is itself the result of a cascade of phosphorylation reactions. Protein kinase A catalyzes the phosphorylation of phosphorylase kinase, thereby activating it; phosphorylase kinase then phosphorylates and activates phosphorylase. This sequence of successive phosphorylations permits considerable amplification of the initial signal. Thus, only a few receptors need to be stimulated to activate a large number of phosphorylase molecules in a very brief period of time.

Concurrent with the activation of hepatic glycogen phosphorylase, protein kinase A also catalyzes the phosphorylation and inactivation of another enzyme, glycogen synthase, which catalyzes the transfer of glycosyl units from UDP-glucose to glycogen. Phosphorylation decreases the net rate of synthesis of glycogen from glucose. The dual effects of cyclic AMP to enhance conversion of glycogen to glucose and to decrease the synthesis of glycogen from glucose summate to increase the output of glucose from the liver.

Similar types of reactions result in the activation of triglyceride lipase in adipose tissue, with resultant increased release of free fatty acids. The lipase is activated when it is phosphorylated by protein kinase A. Catecholamines provide an increased supply of substrate for oxidative metabolism by this mechanism.

In the heart, stimulation of -adrenergic receptors leads to positive inotropic and chronotropic responses. Increased intracellular concentrations of cyclic AMP and enhanced phosphorylation of proteins such as troponin and phospholamban are detected after -adrenergic stimulation. Although these phosphorylation events appear to influence both the actions and the disposition of cellular Ca²⁺, other events also may contribute to the inotropic response such as direct activation of voltage-sensitive Ca²⁺ channels by G_s.

-Adrenergic Receptors

The deduced amino acid sequences from the six -adrenergic receptor genes, three genes (_{1A 1B 1D}; Zhong and Minneman, 1999) and three genes (_{2A 2B 2C}; Bylund, 1992), conform to the well-established paradigm of the seven membrane-spanning G proteincoupled receptors. While not as thoroughly investigated as -adrenergic receptors, the general structural features and their relation to the functions of ligand binding and G protein activation appear to agree with those set forth in Chapter 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship Between Drug Concentration and Effect and above for the -adrenergic receptors. Within the membrane-spanning domains, the three -adrenergic receptors share approximately 75% identity in amino acid residues, as do the three -adrenergic receptors, but the and subtypes are no more similar than are the and subtypes (approximately 30% to 40%).

-Adrenergic Receptors

As shown in Table 64, -adrenergic receptors couple to a variety of effectors (Aantaa et al., 1995; Bylund, 1992). Inhibition of adenylyl cyclase activity was the earliest effect observed, but in some systems the enzyme actually is stimulated by receptors either by G_i subunits or by weak direct stimulation of G_s. The physiological significance of these latter processes is not currently clear. -Adrenergic receptors activate G proteingated K⁺ channels, resulting in membrane hyperpolarization. In some cases (e.g., cholinergic neurons in the myenteric plexus), this may be Ca²⁺ dependent, whereas in others (e.g., muscarinic acetylcholine receptors in atrial myocytes), it is not and results from direct G proteinmediated coupling of the receptors to the K⁺ channels. -Adrenergic receptors also are capable of inhibiting voltage-gated Ca²⁺ channels; this is mediated by G_o proteins. Other second-messenger systems linked to -adrenergic receptor activation include acceleration of Na⁺/H⁺ exchange, stimulation of phospholipase C activity and arachidonic acid mobilization, increased polyphosphoinositide hydrolysis, and increased intracellular availability of Ca²⁺. The latter is involved in the smooth musclecontracting effect of receptor agonists. In addition, it now is clear that adrenergic receptors activate mitogen-activated protein kinases (MAPKs), likely via subunits released from pertussis toxin-sensitive G proteins (Della Rocca et al., 1997; Richman and Regan, 1998). This and related pathways lead to activation of a variety of tyrosine kinasemediated downstream events. These pathways are reminiscent of pathways activated by peptide tyrosine kinase receptors. Although -adrenergic receptors may activate several different signaling pathways, the exact contribution of each to many physiological processes is not clear. The _2A-adrenergic receptor plays a major role in inhibiting norepinephrine release from sympathetic nerve endings and suppressing sympathetic outflow from the brain leading to hypotension (MacMillan et al., 1996; Docherty, 1998; Kable et al., 2000). In addition, the _2A-adrenergic receptor contributes to sedative and anesthetic-sparing effects of -selective agonists (Lakhlani et al., 1997).

-Adrenergic Receptors

Stimulation of -adrenergic receptors results in the regulation of multiple effector systems. A primary mode of signal transduction involves the mobilization of intracellular Ca²⁺ from endoplasmic stores. This increase in intracellular Ca²⁺ currently is thought to result from activation of phospholipase C isoforms through the G_q family of G proteins. The hydrolysis of membrane-bound polyphosphoinositides via phospholipase C results in the generation of two second messengersdiacylglycerol (DAG) and inositol-1,4,5-trisphosphate (IP₃). IP₃ stimulates the release of Ca²⁺ from intracellular stores via a specific receptor-mediated process, while DAG is a potent activator of protein kinase C (see Berridge, 1993). A major component of the responses that follow receptor activation involves regulation of several protein kinases. In addition to protein kinase C, which is activated by Ca²⁺ and diacylglycerol, these include a group of Ca²⁺- and calmodulin-sensitive protein kinases (Dempsey et al., 2000; Braun and Schulman, 1995). For example, -adrenergic receptors regulate hepatic glycogenolysis in some animal species; this effect results from the activation of phosphorylase kinase by the mobilized Ca²⁺, aided by the inhibition of glycogen synthase caused by protein kinase Cmediated phosphorylation. Protein kinase C phosphorylates many substrates, including membrane proteins such as channels, pumps, and ion-exchange proteins (e.g., Ca²⁺-transport ATPase). These effects presumably lead to regulation of various ion conductances.

-Adrenergic stimulation of phospholipase A₂ leads to the release of free arachidonate, which is then metabolized via the cyclooxygenase and lipoxygenase pathways to the bioactive prostaglandins and leukotrienes, respectively (see Chapter 26: Lipid-Derived Autacoids: Eicosanoids and Platelet-Activating Factor). Stimulation of phospholipase A₂ activity by various agonists (including epinephrine acting at -adrenergic receptors) is found in many tissues and cell lines, suggesting that this effector is physiologically important. Phospholipase D hydrolyzes phosphatidylcholine to yield phosphatidic acid (PA). Although PA itself may act as a second messenger by releasing Ca²⁺ from intracellular stores, it also is metabolized to the second messenger DAG. Recent studies have demonstrated that phospholipase D is an effector for ADP-ribosylating factor (ARF), suggesting that phospholipase D may play a role in membrane trafficking. Finally, some evidence in vascular smooth muscle suggests that -adrenergic receptors are capable of regulating a Ca²⁺ channel via a G protein.

In most smooth muscles, the increased concentrations of intracellular Ca²⁺ ultimately cause contraction as a result of activation of Ca²⁺-sensitive protein kinases such as the calmodulin-dependent myosin light chain kinase; phosphorylation of the light chain of myosin is associated with the development of tension (Stull et al., 1990). In contrast, the increased concentrations of intracellular Ca²⁺ that result from stimulation of -adrenergic receptors in gastrointestinal smooth muscle cause hyperpolarization and relaxation by activation of Ca²⁺-dependent K⁺ channels (see McDonald et al., 1994).

As with receptors described above, there is considerable evidence demonstrating that receptors activate MAPKs and other kinases such as PI 3-kinase leading to important effects on cell growth and proliferation (Dorn and Brown, 1999; Gutkind, 1998). For example, prolonged stimulation of receptors promotes growth of cardiac myocytes and vascular smooth muscle cells.

Localization of Adrenergic Receptors

Presynaptically located - and -adrenergic receptors fulfill important roles in the regulation of neurotransmitter release from sympathetic nerve endings. Presynaptic -adrenergic receptors also may mediate inhibition of release of neurotransmitters other than norepinephrine in the central and peripheral nervous systems. Both - and -adrenergic receptors also are located at postsynaptic sites, e.g., on many types of neurons in the brain. In peripheral tissues, postsynaptic -adrenergic receptors are found in vascular and other smooth muscle cells (where they mediate contraction), adipocytes, and many types of secretory epithelial cells (intestinal, renal, endocrine). Postsynaptic -adrenergic receptors can be found in the myocardium (where they mediate contraction) and also on vascular and other smooth muscle cells (where they mediate relaxation). Both - and -adrenergic receptors may be situated at sites that are relatively remote from nerve terminals releasing norepinephrine. Such extrajunctional receptors typically are found on vascular smooth muscle cells and blood elements (platelets and leukocytes), and may be activated preferentially by circulating catecholamines, particularly epinephrine.

In contrast, - and -adrenergic receptors appear to be located mainly in the immediate vicinity of adrenergic nerve terminals in peripheral target organs, strategically placed to be activated during stimulation of these nerves. These receptors also are widely distributed in the mammalian brain.

The cellular distributions of the three - and three -adrenergic receptor subtypes still are incompletely understood. In situ hybridization of receptor mRNA and receptor subtype-specific antibodies indicate that _2A-adrenergic receptors in the brain may be both pre- and postsynaptic. These findings and other studies indicate that this receptor subtype functions as a presynaptic autoreceptor in central noradrenergic neurons (Aantaa et al., 1995; Lakhlani et al., 1997). Using similar approaches, _1A receptor mRNA was found to be the dominant subtype message expressed in prostatic smooth muscle (Walden et al., 1997).

Refractoriness to Catecholamines

Exposure of catecholamine-sensitive cells and tissues to adrenergic agonists causes a progressive diminution in their capacity to respond to such agents. This phenomenon is variously termed refractoriness, desensitization, or tachyphylaxis, and it significantly limits the therapeutic efficacy and duration of action of catecholamines and other agents (see Chapter 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship Between Drug Concentration and Effect). Although descriptions of such adaptive changes are common, mechanisms are incompletely understood. They have been studied most extensively in cells that synthesize cyclic AMP in response to -adrenergic receptor agonists.

There is evidence for multiple points of regulation of responsiveness, including the receptors, G proteins, adenylyl cyclase, and cyclic nucleotide phosphodiesterase. The pattern of refractoriness varies according to the extent to which these different components are modified. In some cases, especially when the receptors themselves are altered, desensitization may be limited to the actions of -adrenergic agents. This often is termed homologous desensitization. In other cases, stimulation by a -adrenergic agonist can cause diminished responsiveness to a wide variety of receptor-mediated stimulators of cyclic AMP synthesis. Although such heterologous desensitization may result from changes in receptors, it also may involve perturbations of more distal elements in the signaling pathway.

One of the most important mechanisms for rapidly regulating -adrenergic receptor function is agonist stimulation of receptor phosphorylation, which leads to decreased sensitivity to further catecholamine stimulation. The receptors may be phosphorylated by several different protein kinases, but in all cases the end result is the same, decreased coupling to G_s and decreased stimulation of adenylyl cyclase.

Mechanisms for Heterologous Desensitization

One protein kinase that phosphorylates G_s-coupled receptors is protein kinase A, which is stimulated by -adrenergic receptor-mediated activation of adenylyl cyclase and subsequent elevation of intracellular cyclic AMP levels. This kinase thus enables completion of a negative feedback regulatory loop, phosphorylating and desensitizing the receptor responsible for its stimulation (Hausdorff et al., 1990). The sites of protein kinase A phosphorylation on the -adrenergic receptors have been mapped to the distal portion of the third cytoplasmic loop and the proximal part of the carboxy terminal cytoplasm tail of the receptor, and heterologous desensitization is paralleled by phosphorylation of the residue in the third cytoplasmic loop (Clark et al., 1989; denoted P₂, see Figure 66). The phosphorylation presumably changes the conformation of the receptor, thereby impairing the coupling to G_s (see Figure 66).

Mechanisms for Homologous Desensitization

A receptor-directed protein kinase, termed the -adrenergic receptor kinase (ARK), phosphorylates the receptors only when they are occupied by an agonist (Benovic et al., 1986). It subsequently was discovered that ARK is a member of a family of at least six G proteincoupled receptor kinases (GRKs) that phosphorylate and regulate a wide variety of G proteincoupled receptors. Because only the agonist-occupied 'activated' forms of the -adrenergic and other receptors are substrates for GRKs, these enzymes provide a mechanism for achieving homologous or agonist-specific desensitization. GRKs share a similar structural organization (Krupnick and Benovic, 1998; Pitcher et al., 1998). For example, the function of the visual light 'receptor' rhodopsin is regulated by a related enzyme, rhodopsin kinase; now called GRK1. While GRK1 is preferentially expressed in retinal rods and cones, GRK2 is broadly expressed in diverse cell types. With the exception of rhodopsin kinase, it is not known with certainty which GRKs regulate which receptors. When -adrenergic receptors are activated by agonists, they interact with G_s, dissociating it into _s and subunits (see Chapter 2: Pharmacodynamics: Mechanisms of Drug Action and the Relationship Between Drug Concentration and Effect). The G-subunit complex, which is attached to the plasma membrane by a lipid group (geranylgeranyl), appears to foster or stabilize ARK (GRK2) association with the plasma membrane, facilitating phosphorylation of the agonist-occupied and activated receptor on multiple serine residues located close to the carboxy terminus of the cytoplasmic tail of the receptor (Figure 66).

GRK3 also contains a binding domain. GRK4 and GRK6 are palmitoylated, and GRK5 contains 2 basic phospholipid-binding domains (Krupnick and Benovic, 1998). GRKs also have been implicated in phosphorylating many other G proteincoupled receptors, including _1B and _2A adrenergic receptors, and receptors for thrombin, angiotensin II, and many other agents. Inhibitors of GRK activity may reduce the development of desensitization. Overexpression of GRK2 in heart cells attenuates -adrenergic responses in these cells (Koch et al., 1995). Interestingly, there is evidence for increased expression of GRK in the myocardium in patients with congestive heart failure who often have blunted responses to -adrenergic agonists (Ungerer et al., 1993).

Unlike the situation with protein kinase A-mediated receptor phosphorylation, covalent modification of the G proteincoupled receptor by GRKs alone is not sufficient to fully desensitize receptor function. Rather, a second reaction must occur in which an 'arresting' protein binds to the phosphorylated receptor and presumably sterically inhibits its functional coupling to G_s. This protein, called -arrestin, is one of a family of proteins that fulfills this function in different receptor systems (Krupnick and Benovic, 1998; Lefkowitz, 1998). The homologous protein in the visual system is called arrestin. The arrestin protein binds much more rapidly to the GRK-phosphorylated forms of the receptors than to the nonphosphorylated forms. The binding of an arrestin to the phosphorylated receptor plays a critical role in attenuating receptor signaling in response to agonists.

Agonists also promote a rapid (minutes) and reversible sequestration (internalization) of their receptors and a slower (hours) 'down-regulation' of the receptors in which the actual number of receptors in the cell declines. The function of receptor sequestration is not fully understood. Interestingly, there is evidence suggesting that receptor internalization is important for some (Daaka et al., 1998) but not all G proteincoupled receptor-mediated activation of MAP kinase (Schramm and Limbird, 1999; Pierce et al., 2000). Quantitatively, sequestration may not contribute significantly to the mechanisms underlying desensitization, particularly because there is a high degree of amplification between -adrenergic receptor occupancy and ultimate cyclic AMPmediated responses in many cells. Nonetheless, some evidence suggests that it may lead to receptor dephosphorylation and resensitization. In addition down-regulation of receptor number contributes to longer-term desensitization of receptor function and undoubtedly is mediated by several distinct processes. These include changes in the rate of receptor turnover, receptor gene transcription, and receptor mRNA turnover. These processes are complex and poorly understood at present (Collins et al., 1992).

There is evidence that sequestration, internalization, and down-regulation may occur for receptors, although there are important differences among the various subtypes (Saunders and Limbird, 1999; Heck and Bylund, 1998). In addition, some studies have demonstrated phosphorylation and internalization of receptors after activation by an agonist (Wang et al., 1997; Diviani et al., 1997; Garcia-Sainz et al., 2000).