Chapter 12:  Catecholamines

Tyrosine hydroxylase is the rate-limiting enzyme for the biosynthesis of catecholamines

Tyrosine hydroxylase (TH) is found in all cells that synthesize catecholamines and is a mixed-function oxidase that uses molecular oxygen and tyrosine as its substrates and biopterin as its cofactor [3]. TH is a homotetramer, each subunit of which has a molecular weight of approximately 60,000. It catalyzes the addition of a hydroxyl group to the meta position of tyrosine, thus forming 3,4-dihydroxy-l-phenylalanine (l-DOPA). TH can also hydroxylate phenylalanine to form tyrosine, which is then converted to l-DOPA; this alternative synthetic route may be of significance in patients affected with phenylketonuria, a condition in which phenylalanine hydroxylase activity is depressed (see Chap. 44). TH has a K_m for tyrosine in the micromolar range. As a result, it is virtually saturated by the high tissue concentrations of endogenous tyrosine. The cofactor, biopterin, may be at subsaturating concentrations within catecholamine-containing neurons and, thus, may play an important role in regulating NE biosynthesis. TH is primarily a soluble enzyme; however, interactions with membrane constituents, such as phosphatidylserine, or with polyanions, such as heparin sulfate, have been shown to alter its kinetic characteristics. Analogs of tyrosine, such as α-methyl-p-tyrosine (AMPT), are competitive inhibitors of TH. Sequence analysis [4] reveals consensus sequences for phosphorylation primarily in the N-terminal portion of the molecule. The gene reveals considerable sequence homology with phenylalanine hydroxylase and tryptophan hydroxylase.

DOPA decarboxylase catalyzes the removal of the carboxyl group from DOPA to form dopamine

DOPA decarboxylase (DDC) is a pyridoxine-dependent enzyme that has a low K_m and a high V_max with respect to l-DOPA; thus, endogenous l-DOPA is efficiently converted to DA [5]. DDC can also decarboxylate 5-hydroxytryptophan, the precursor of serotonin, as well as other aromatic amino acids; accordingly, it has also been called aromatic amino acid decarboxylase (AADC). DDC is widely distributed throughout the body, where it is found both in catecholamine- and serotonin-containing neurons and in non-neuronal tissues, such as kidney and blood vessels. In DA-containing neurons, this enzyme is the final step in the pathway. α-Methyldopa inhibits DDC in vitro and leads to a reduction in blood pressure after being converted to the false transmitter α-methylnorepinephrine in vivo.

For neurons that synthesize epinephrine or norepinephrine, dopamine β-hydroxylase is the next step in the biosynthetic pathway

Like TH, dopamine β-hydroxylase (DBH) is a mixed-function oxidase that uses molecular oxygen to form the hydroxyl group added to the β carbon on the side chain of DA [6]. Ascorbate, reduced to dihydroascorbate during the reaction, provides a source of electrons. DBH contains Cu²⁺, which is involved in electron transfer in the reaction; accordingly, copper chelators, such as diethyldithiocarbamate, are potent inhibitors of the enzyme. DBH is a tetrameric glycoprotein containing subunits of 77 and 73 kDa, as determined by sodium dodecyl sulfate (SDS) gel electrophoresis. A full-length clone encodes a polypeptide chain of 578 amino acids [7]. The enzyme is concentrated within the vesicles that store catecholamines; most of the DBH is bound to the inner vesicular membrane, but some is free within the vesicles. DBH is released along with catecholamines from nerves and from the adrenal gland and is found in plasma.

In cells that synthesize epinephrine, the final step in the pathway is catalyzed by the enzyme phenylethanolamine N-methyltransferase

This enzyme is found in a small group of neurons in the brainstem that utilize epinephrine as their neurotransmitter and in the adrenal medullary cells, for which epinephrine is the primary neurohormone. Phenylethanolamine N-methyltransferase (PNMT) transfers a methyl group from S-adenosylmethionine to the nitrogen of NE, forming a secondary amine [8]. The coding sequence of bovine PNMT is contained in a single open reading frame encoding a protein of 284 amino acids [9]. PNMT activity is regulated by corticosteroids.

Catecholamines are concentrated in storage vesicles that are present at high density within nerve terminals

Ordinarily, low concentrations of catecholamines are free in the cytosol, where they may be metabolized by enzymes including monoamine oxidase (MAO). Thus, conversion of tyrosine to l-DOPA and l-DOPA to DA occurs in the cytosol; DA then is taken up into the storage vesicles. In NE-containing neurons, the final β hydroxylation occurs within the vesicles. In the adrenal gland, NE is N-methylated by PNMT in the cytoplasm. Epinephrine is then transported back into chromaffin granules for storage.

cDNA clones encoding vesicular amine transporters have been obtained. The sequence suggests that the proteins have 12 transmembrane domains and are homologous to a family of bacterial drug-resistance transporters. The expressed protein, referred to as vesicular membrane transporter 2 (VMAT2), has a high affinity for reserpine, which blocks vesicular uptake in vivo [10]. The mechanism that concentrates catecholamines within the vesicles is an ATP-dependent process linked to a proton pump. The intravesicular concentration of catecholamines is approximately 0.5 M, and they exist in a complex with ATP and acidic proteins known as chromogranins. The vesicular uptake process has broad substrate specificity and can transport a variety of biogenic amines, including tryptamine, tyramine and amphetamines; these amines may compete with endogenous catecholamines for vesicular storage sites. Reserpine is a specific, irreversible inhibitor of the vesicular amine pump that blocks the ability of the vesicles to concentrate the amines. Treatment with reserpine causes a profound depletion of endogenous catecholamines in neurons. The effect of reserpine is to inhibit the uptake of DA and other catecholamines into vesicles. Knockout mice lacking VMAT2 are not viable (Table 12-1).

The vesicles play a dual role: they maintain a ready supply of catecholamines at the terminal available for release, and they mediate the process of release. When an action potential reaches the nerve terminal, Ca²⁺ channels open, allowing an influx of the cation into the terminal; increased intracellular Ca²⁺ promotes the fusion of vesicles with the neuronal membrane (see Chap. 9). The vesicles then discharge their soluble contents, including NE, ATP and DBH, into the extraneuronal space [11]. The demonstration that DBH is released concurrently and proportionately with NE established that release occurs by the process of exocytosis since proteins would not be expected to diffuse across cell membranes. Exocytotic release from sympathetic neurons may be the source of some of the DBH found in the plasma and cerebrospinal fluid (CSF) of animals and humans. Indirectly acting sympathomimetics, like tyramine and amphetamine, release catecholamines by a mechanism that is neither dependent on Ca²⁺ nor associated with release of DBH. These drugs displace catecholamines from storage vesicles, resulting in leakage of neurotransmitter from the nerve terminals.

The concentration of catecholamines within nerve terminals remains relatively constant

Despite the marked fluctuations in the activity of catecholamine-containing neurons, efficient regulatory mechanisms modulate the rate of synthesis of catecholamines [12]. A long-term process affecting catecholamine synthesis involves alterations in the amounts of TH and DBH present in nerve terminals [1]. When sympathetic neuronal activity is increased for a prolonged period of time, the amounts of mRNA coding for TH and DBH are increased in the neuronal perikarya. DDC does not appear to be modulated by this process. The newly synthesized enzyme molecules are then transported down the axon to the nerve terminals.

Alteration in the rate of synthesis of TH and DBH provides a mechanism to modulate synthesis of catecholamines in response to persistent changes in neuronal activity. In addition, two mechanisms operative at the level of the nerve terminal play important roles in the short-term modulation of catecholamine synthesis and are responsive to momentary changes in neuronal activity [13]. TH, the rate-limiting enzyme in the synthesis pathway, is modulated by end-product inhibition [12]. Thus, free intraneuronal catecholamines inhibit the further activity of TH by competing at the site that binds the pterin cofactor; conversely, neuronal activity results in the release of catecholamines, a decrease in cytoplasmic concentrations and disinhibition of the enzyme. An additional and probably more important effect of depolarization of catecholaminergic terminals is activation of TH. The kinetic characteristics of the enzyme change so that it has a higher affinity for the pterin cofactor and is less sensitive to end-product inhibition. Activation of the enzyme is associated with reversible phosphorylation of the enzyme (Fig. 12-2) [14]. Protein kinase C (PKC), cAMP-dependent protein kinase (PKA) and Ca²⁺/calmodulin-dependent protein kinases (CaMKs) are all capable of inducing phosphorylation of the enzyme, leading to an increase in activity (see Chaps. 21 and 22).

Monoamine oxidase and catechol-O-methyltransferase are primarily responsible for the inactivation of catecholamines

MAO and catechol-O-methyltransferase (COMT) are widely distributed throughout the body (Fig. 12-3). MAO is a flavin-containing enzyme located on the outer membrane of the mitochondria [15]. This enzyme oxidatively deaminates catecholamines to their corresponding aldehydes; these can be converted, in turn, by aldehyde dehydrogenase to acids or by aldehyde reductase to form glycols. Because of its intracellular localization, MAO plays a strategic role in inactivating catecholamines that are free within the nerve terminal and not protected by storage vesicles. Accordingly, drugs that interfere with vesicular storage, such as reserpine, or indirectly acting sympathomimetics, such as amphetamines, which displace catecholamines from vesicles, cause a marked increase in deaminated metabolites. Isozymes of MAO with differential substrate specificities have been identified: MAO-A preferentially deaminates NE and serotonin and is selectively inhibited by clorgyline, whereas MAO-B acts on a broad spectrum of phenylethylamines, including β-phenylethylamine. MAO-B is selectively inhibited by deprenyl. MAO in the gastrointestinal tract and liver plays an important protective role by preventing access to the general circulation of ingested, indirectly acting amines, such as tyramine and phenylethylamine, that are contained in food; however, patients being treated for depression or hypertension with MAO inhibitors are not afforded this protection and can suffer severe hypertensive crises after ingesting foods that contain large amounts of tyramine. Such foods include port wine, Stilton cheese and herring. A methyl substituent on the α carbon of the phenylethylamine side chain protects against deamination by MAO; the prolonged action of amphetamine and related indirectly acting stimulants is in part a consequence of the presence of an α-methyl group, which prevents their inactivation by MAO.

COMT is found in nearly all cells, including erythrocytes [16]; thus, the enzyme can act on extraneuronal catecholamines. Most studies of COMT are carried out with enzyme purified from homogenates of liver. The enzyme, which requires Mg²⁺, transfers a methyl group from the cosubstrate S-adenosylmethionine to the 3-hydroxy group on the catecholamine ring. This enzyme has broad substrate specificity, methylating virtually any catechol regardless of the side-chain constituents; for this reason, competitive inhibitors of the enzyme that are of pharmacological significance have not been developed.

Measurement of catecholamine metabolites can provide insight into the rate of release or turnover of catecholamines in the brain. In clinical studies, metabolites of catecholamines are generally assayed in the CSF because the large quantities derived from the peripheral sympathomedullary system obscure the small contribution from the brain to urinary concentrations. However, acid metabolites are actively excreted from the CSF; more reliable estimates of turnover in the brain are obtained when this transport process is blocked by pretreatment with the drug probenecid.

4-Hydroxy-3-methoxy-phenylacetic acid, more commonly known as homovanillic acid (HVA), is a major metabolite of DA. Spinal fluid concentrations of HVA provide insight into the turnover of DA in the striatum. Concentrations of HVA are decreased, for example, in CSF of patients with Parkinson's disease (see Chap. 45). A metabolite of NE formed relatively selectively in the brain is 3-methoxy-4-hydroxyphenylglycol (MHPG). Because this is a minor metabolite of the much larger amounts of NE metabolized in the periphery, it is estimated that between 30 and 50% of the MHPG excreted in urine is derived from the brain. MHPG has been measured in CSF and in urine to provide an index of NE turnover in the brain and concentrations of MHPG have been shown to be decreased in certain forms of depression (see Chap. 52).

The action of catecholamines released at the synapse is terminated by diffusion and reuptake into presynaptic nerve terminals

Catecholamines diffuse from the synaptic cleft and are taken up or transported back into the nerve terminal. Some of the catecholamine molecules may be catabolized by MAO and COMT. The catecholamine-reuptake process was originally described by Axelrod [17]. He observed that when radioactive NE was injected intravenously, it accumulated in tissues in direct proportion to the density of the sympathetic innervation in the tissue. The amine taken up into the tissues was protected from catabolic degradation, and studies of the subcellular distribution of catecholamines showed that they are localized in synaptic vesicles. Ablation of the sympathetic input to organs abolished the ability of vesicles to accumulate and store radioactive NE. Subsequent studies demonstrated that this Na⁺ and Cl⁻-dependent uptake process is a characteristic feature of catecholamine-containing neurons in both the periphery and the brain; the transport process has been extensively studied in sheared-off nerve terminals or synaptosomes isolated from the brain (Table 12-2).

The uptake process is mediated by a carrier or transporter located on the outer membrane of the catecholaminergic neurons. It is saturable and obeys Michaelis-Menten kinetics. A transport process selective for NE is found only in noradrenergic neurons, whereas a transporter with different specificity is found in DA-containing neurons. Cloning of genes for transporters responsible for uptake of NE and DA has been accomplished, revealing proteins with conserved structural features [18]. The presence of 11 to 13 transmembrane domains is a recurrent theme (Fig. 12-4, Table 12-2). Transmembrane domains 1, 2 and 4–8 show the highest degree of sequence identity. The transporters are part of a larger family of neurotransporters (Chap. 5) [18]. They have consensus sites for phosphorylation, although the importance of phosphorylation has not been elucidated. The uptake process is energy-dependent since it can be inhibited by incubation at a low temperature or by metabolic inhibitors. The energy requirements reflect a coupling of the uptake process with the Na⁺ gradient across the neuronal membrane. The process is also Cl⁻-dependent. Drugs such as ouabain, which inhibits Na,K-ATPase, or veratridine, which opens Na⁺ channels, inhibit the uptake process. The linkage of uptake to the Na⁺ gradient may be of physiological significance since transport temporarily ceases at the time of depolarization-induced release of catecholamines. The transport of catecholamines can be inhibited selectively by such drugs as tricyclic antidepressants and cocaine. In addition, a variety of phenylethylamines, such as amphetamine, are substrates for carrier; thus, they can be concentrated within catecholamine-containing neurons and can compete with the catecholamines for transport. Neurotoxins such as mercaptopyrazide pyrimidine (MPP⁺) and 6-hydroxydopamine are also taken up by transporters, and this is required for the neurotoxic effect. Mice have been prepared with their transporter genes “knocked out” (see Chap. 40). Extensive studies with these mice confirm the important role of transporters (see Table 12-1).

Once an amine has been taken up across the neuronal membrane, it can be taken up by adrenergic storage vesicles. Neuronal uptake is Na⁺-dependent and is not affected by drugs like reserpine; uptake across the vesicle membrane requires H⁺ and is inhibited by reserpine (Table 12-2). Once a compound is taken up into the vesicles, it can be released in place of NE. Such substances are called false transmitters.

Our understanding of the function of catecholamine-containing neurons has been aided by neuroanatomical methods of visualizing these neurons

Nearly two decades ago, Falck and Hillarp took advantage of the fact that in the presence of formaldehyde catecholamines cyclize to form intensely fluorescent products [19]. With a fluorescence microscope, neurons containing catecholamines could be visualized in thin sections obtained from tissue previously exposed to formaldehyde vapor. A modification of the method uses glyoxylic acid and has resulted in enhanced sensitivity and a more stable fluorophor for even better visualization of the fine axons and terminals.

Once the enzymes that synthesize catecholamines were purified, it was possible to elicit antisera against each enzyme and to localize the enzyme by immunocytochemistry. Thin sections of tissue can be incubated with antibody against a particular enzyme, for example, rabbit anti-DBH, and then incubated with a second antibody linked to a marker, such as fluorescein or horseradish peroxidase. These markers can be readily visualized and examined with a microscope. By using this technique, the PNMT-containing neurons that synthesize epinephrine can be distinguished from the noradrenergic neurons that are devoid of PNMT; similarly, noradrenergic neurons that contain DBH can be separated from the DA-containing neurons that do not possess this enzyme. Cloning the genes that encode for catecholaminergic biosynthetic enzymes makes it possible to use in situ hybridization to localize mRNAs within particular neurons.

Finally, experimental advantage has been taken of the highly selective uptake process for catecholamines. Thus, after incubation with radioactive NE, noradrenergic axons can be demonstrated at the ultrastructural level by autoradiographic techniques. Alternatively, after administration of the congener 5-hydroxydopamine, which is taken up actively and stored within the vesicles, catecholamine-containing terminals can be distinguished by the presence of dense precipitates of 5-hydroxydopamine within their vesicles. Also, antibodies against transporters have been used in immunocytochemistry and their mRNAs mapped by in situ hybridization. These studies generally confirm the findings of earlier ones.

Cell bodies of noradrenergic neurons are clustered in the medulla oblongata, pons and midbrain and are considered to be anatomically part of the reticular formation

On the basis of their major axonal projections, noradrenergic fibers can be divided into two major pathways: the dorsal and ventral bundles (Fig. 12-5). The cell bodies of origin for the dorsal bundle are contained in a dense nucleus known as the locus ceruleus, located on the lateral aspect of the fourth ventricle. Axons of neurons in the locus ceruleus have endings in the spinal cord and cerebellum and course anteriorly through the medial forebrain bundle to innervate the entire cerebral cortex and hippocampus. The ventrally located cell bodies send fibers that innervate the brainstem and hypothalamus. As demonstrated by immunocytochemical techniques, in the ventral portion of the pons and medulla there are a small number of neurons that contain PNMT; the axons of these epinephrine-containing neurons terminate primarily in the brainstem and hypothalamus.

Large numbers of cell bodies of dopamine-containing neurons are located in the midbrain

Some of the DA-containing neurons can be divided into three groups: nigrostriatal, mesocortical and tuberohypophysial. A major dopaminergic tract in brain originates in the zona compacta of the substantia nigra and sends axons that provide a dense innervation to the caudate nucleus and putamen of the corpus striatum; nearly 80% of all DA in the brain is found in the corpus striatum. In Parkinson's disease, the nigrostriatal tract degenerates (see Chap. 45). This accounts for a profound depletion of DA from the striatum and for the symptoms of this disorder.

DA-containing cell bodies that lie medial to the substantia nigra in the ventral tegmental area provide a diffuse, but modest, innervation to the forebrain, including the frontal and cingulate cortex, septum, nucleus accumbens and olfactory tubercle. It has been hypothesized that these neurons are critical for the action of antipsychotic drugs, antihyperactivity drugs and psychostimulant drugs.

DA-containing cell bodies in the arcuate and periventricular nuclei of the hypothalamus send axons that innervate the intermediate lobe of the pituitary and the median eminence. These neurons play an important role in regulating the release of pituitary hormones, especially prolactin (see Chap. 18). In addition to these major pathways, DA-containing interneurons have been found in the olfactory bulb and in the neural retina.

The brain contains multiple classes of receptors for catecholamines

Effects of DA are mediated through interaction with D1-like (D1 and D5) and D2-like (D2, D3 and D4) receptors (Table 12-3), while effects of NE and epinephrine are mediated through α₁- and α₂-adrenergic receptors (Tables 12-4, 12-5) and through β-adrenergic receptors (Table 12-6). As of the present time, three subtypes of α₁-, three subtypes of α₂- and three subtypes of β-adrenergic receptors have been identified (Tables 12-4–12-6). Mice lacking these receptors can have significant physiological deficits (Table 12-1). The postsynaptic receptors on any given neuron receive information from transmitters released by another neuron. Typically, postsynaptic receptors are located on dendrites or cell bodies of neurons, but they may also occur on axons or nerve terminals; in the latter case, an axoaxonic synaptic relationship may cause presynaptic inhibition or excitation. In contrast, autoreceptors are situated on a given neuron and respond to transmitter molecules released from the same neuron. Autoreceptors may be widely distributed on the surface of the neuron. At the nerve terminal, they respond to transmitter molecules released into the synaptic cleft; on the cell body, they may respond to transmitter molecules released by dendrites. Functionally, most autoreceptors appear to regulate transmitter release in such a way that the released transmitter, acting on autoreceptors, regulates additional release. Autoreceptors have been identified for NE-, DA-, serotonin- and GABA-containing neurons. A major type of inhibitory autoreceptor described in both the sympathetic PNS and the brain has pharmacological properties resembling those of the α₂-adrenergic receptor [20].

In the sympathetic PNS, autoreceptors of the β-adrenergic type have also been described. These differ from most other known autoreceptors in that NE acting on these receptors facilitates transmitter release and thus amplifies the effects of neuronal firing. This effect contrasts with the inhibitory action of α-adrenergic and DA autoreceptors, which exert negative feedback control on transmitter release. The DA autoreceptors in the nigrostriatal pathway appear to be of the D2 subtype.

Multiple dopamine receptor subtypes exist

Two subtypes of DA receptor were initially identified on the basis of pharmacological and biochemical criteria. D1 receptors were shown to couple to stimulation of adenylyl cyclase activity, while D2 receptors inhibited enzyme activity (Fig. 12-6). More recently, multiple D1-like and D2-like receptors have been identified (Table 12-3) [21,22] and amino acid sequences determined. The known subtypes of DA receptor are members of the G protein-linked receptor family with seven hydrophobic domains, an extracellular N terminus and an intracellular C terminus (Fig. 12-4). Consensus sequences for phosphorylation are found in the second (i₂) and third intracellular (i₃) loops and the C-terminal tail. The D1-like receptors have relatively small i₃ loops and long C-terminal tails, while the D2-like receptors have large i₃ loops and short C-terminal tails.

The D1-like receptors include the D1 and D5 receptors [21,22]. The D1-like receptors have a high affinity for benzazepines like SCH-23390 and a low affinity for benzamides and are coupled to stimulation of adenylyl cyclase activity. The most striking pharmacological difference among them is the high affinity of D5 receptors for DA.

Molecular genetic studies have demonstrated the presence of two forms of mRNA coding for D2 receptors, designated D2L and D2S. These two forms differ by 87 bases, corresponding to a 29-amino-acid insert in the i₃ loop of the receptor (Fig. 12-4). The two species of D2 receptor mRNA appear to arise through alternative splicing. Both D2L and D2S receptors are coupled to inhibition of adenylyl cyclase activity, although D2S stimulation causes a greater inhibition. The D3 receptor, a second member of the D2-like receptor family, has been cloned and expressed in COS-7 cells. D3 receptor mRNA is found in limbic areas of the brain, including the nucleus accumbens. Comparison of the properties of D2 and D3 receptors shows that the D3 receptor has a relatively high affinity for atypical neuroleptics and for DA autoreceptor inhibitors, including (+)-UH232 and (+)-AJ76. Cloning of the D4 receptor has introduced an additional level of complexity to the study of DA receptors. Of particular interest is the high affinity of D4 receptors for the atypical neuroleptic clozapine. D4 receptor mRNA has been detected in the frontal cortex, midbrain, amygdala and medulla, with lower concentrations detected in the basal ganglia. The use of molecular approaches in the study of the D4 receptor has been hampered by the high G/C content of its coding sequences. D3 and D4 receptors inhibit adenylyl cyclase, and D4 receptors are positively coupled to K⁺ channels.

The number of D1 and D2 receptors can be modulated by antagonists or neurotoxins

The density of D2 receptors in rat striatum is increased following lesions with the neurotoxin 6-hydroxydopamine or by administration of antagonists. Similar results for D1 receptors were obtained following chronic administration of the D1-selective antagonist SCH-23390. Subtypes of DA receptor may be coregulated since the D2 antagonist sulpiride attenuated the ability of SCH-23390 to increase the density of D1 receptors. The increase in the density of D2 receptors following chronic administration of antagonists may be responsible for the development of a movement disorder called tardive dyskinesia (see Chap. 45).

Available behavioral data suggest that either acute or repeated administration of agonists acting at DA receptors results in augmentation of the behavioral effects of the drugs. This phenomenon, known as reverse tolerance or sensitization, is characterized by a selective increase in the intensity or duration or a shift to an earlier time of onset of stereotypical behaviors such as locomotion, sniffing, rearing, licking and gnawing. Sensitization to indirect DA agonists like amphetamine or cocaine also occurs. The mechanisms underlying behavioral sensitization are likely to be complex. It is known, for example, that stereotypical behavior and locomotor hyperactivity are critically dependent on activation of both D1 and D2 receptors.

Direct and indirect agonists at dopamine receptors, including amphetamine, bromocriptine and lisuride, have been shown to induce psychotic episodes

A strong correlation exists between the clinical doses of neuroleptics and their affinity for brain D2 receptors. This has led to the hypothesis that psychotic disorders result from overstimulation of D2 receptors. Long-term administration of neuroleptics to humans or experimental animals can result in an increase in the density of striatal D2 receptors and in the appearance of extrapyramidal side effects, including parkinsonian movement disorders and tardive dyskinesia. A panel of antipsychotic drugs referred to as atypical neuroleptics, including clozapine, melperone and fluperlapine, have been reported to produce fewer extrapyramidal side effects and have been useful in the treatment of patients with schizophrenia who respond poorly to typical antipsychotics such as haloperidol. The relative affinities of D2, D3 and D4 receptors for typical and atypical neuroleptics together with the selective expression of D3 receptor mRNA in limbic areas of the brain have led to the hypothesis that the clinical utility of neuroleptics in the treatment of psychiatric illness may be due, at least in part, to their ability to antagonize stimulation of D3 or D4 receptors, while the motor dysfunction observed following chronic treatment with typical neuroleptics could be due to alterations in the density of D2 receptors in the striatum.

The pharmacological responses to catecholamines were ascribed to effects of α- and β-adrenergic receptors in the late 1940s

NE and epinephrine act at both α and β receptors, but isoproterenol, a synthetic agonist, acts only at β receptors (Tables 12-4–12-6). Numerous antagonists also differentiate between α and β receptors. The prototypic β-adrenergic receptor antagonist propranolol is essentially inactive at α receptors; the α-adrenergic receptor antagonist phentolamine is very weak at β receptors.

Distinct subtypes of β-adrenergic receptor exist and have important pharmacological consequences. β₁-Adrenergic receptors predominate in the heart and in the cerebral cortex, whereas β₂-adrenergic receptors predominate in the lung and cerebellum. However, in many cases, β₁- and β₂-adrenergic receptors coexist in the same tissue, sometimes mediating the same physiological effect. A major side effect of β₂-selective agonists like metaproterenol, used to treat bronchial asthma, is cardiac acceleration. This is due to the coexistence of β₁- and β₂-adrenergic receptors in the heart. Both classes of receptor are coupled to the electrophysiological effects of catecholamines in the heart.

The brain contains both β₁ and β₂ receptors, which cannot be differentiated in terms of their physiological functions. Moreover, radioactive drugs that bind exclusively to one or the other type of β receptor are not yet available. However, one can label all of the β-adrenergic receptors in a given tissue with a nonselective radioligand and then selectively inhibit binding to one of the β-receptor subtypes with increasing concentrations of β₁- or β₂-selective agents [21]. ICI 89,406 and ICI 118,551 are highly selective antagonists at β₁- and β₂-adrenergic receptors, respectively. A similar approach can be used to define the anatomical localization of β₁- and β₂-adrenergic receptors using the technique of quantitative autoradiography. The density of β₁ receptors varies in different brain areas to a greater extent than does that of β₂ receptors. It has been suggested that this is due to the presence of β₂-adrenergic receptors on glia or blood vessels.

A third subtype of β-adrenergic receptor has been identified. This receptor has pharmacological properties distinct from those of β₁- and β₂-adrenergic receptors. Agonists that are selective for β₃ receptors exist and cause nonshivering thermogenesis in rodents. The β₃ receptor in humans has been linked to hereditary obesity, control of lipid metabolism and the development of diabetes. mRNA for β₃-adrenergic receptors is selectively expressed in brown adipose tissue present in rodents and in newborn humans. Message can be detected in white adipose tissue, but expression is very low.

The amino acid sequences of β-adrenergic receptors in brain and various tissues have been determined

A striking structural feature of the β-adrenergic receptors that have been cloned and sequenced, from turkey erythrocytes, hamster lung and human placenta and brain, and of the other members of the G protein-linked receptor family is their topographical orientation with respect to the membrane [23,24] (see Fig. 12-4). Hydropathicity analysis suggests that there are seven hydrophobic regions, each of 20 to 25 amino acids. These are potentially membrane-spanning. Other structural features of β-adrenergic receptors include a long C-terminal hydrophilic sequence thought to be intracellular, a somewhat shorter N-terminal hydrophilic sequence thought to be extracellular and a long cytoplasmic loop between presumptive transmembrane segments V and VI. Sites for N-linked glycosylation are found in the N-terminal extracellular portion of the molecule, while numerous sites that may be phosphorylated are found in the C-terminal portion of the molecule and on the i₂ and i₃ loops (see Chap. 22). Evidence from studies involving limited proteolysis and site-directed mutagenesis has led to the conclusion that the hydrophobic transmembrane helices are involved in the formation of the binding site for catecholamines, and the i₃ loop together with the C terminus may play a role in the interaction of the receptor with GTP-binding proteins (see Chap. 20). A conserved aspartate residue in transmembrane 3 and a pair of serines in transmembrane 5 are thought to provide counter-ions for the amino and catechol hydroxyl groups, respectively [24].

Multiple serine and threonine residues on the i₃ loop and C terminus and consensus sequences for cAMP-dependent phosphorylation may be important in explaining processes including agonist-induced receptor sequestration and desensitization (see also below). cAMP- and non-cAMP-dependent phosphorylations of β-adrenergic receptors have been observed. β-adrenergic receptor-stimulated synthesis of cAMP results in activation of PKA. The phosphorylated receptor is functionally uncoupled. Other receptors coupled to activation of adenylyl cyclase can also cause what is known as heterologous desensitization. In addition, occupancy of β-adrenergic receptors by agonists results in the activation of β-adrenergic receptor kinase (βARK), which leads to phosphorylation of the receptor. The uncoupling of the receptor from G_s also appears to involve a protein called β-arrestin, which is similar to a 48-kDa protein in the retina (see Chap. 47).

The proposed structure of the β-adrenergic receptor is strikingly similar in sequence and topography to that of bacterial rhodopsin (see Chap. 47) and the other members of the G protein-linked receptor family whose cDNAs have recently been cloned (see Chap. 20). Although these proteins mediate widely disparate biological effects, they show a high degree of homology. This is almost certainly related to the fact that, in each case, the immediate consequence of receptor activation is to promote an interaction between the receptor and a GTP-binding protein. The homologies between the members of the extended family of proteins are most evident within the presumed membrane-spanning helices.

Two families of α-adrenergic receptors exist

Radiolabeled agonists and antagonists have been used to label α receptors in both the brain and the peripheral tissues. As with β receptors, the binding properties of α receptors are essentially the same in the brain and the periphery. Some tissues possess only postsynaptic α₁ receptors, others postsynaptic α₂ receptors, and some organs have a mixture of both. Results of pharmacological and physiological studies have led to the suggestion that there are multiple types of α₁ and α₂ receptors. Of particular clinical importance are differences in the properties of junctional and extrajunctional α receptors. The proportions of α₁ and α₂ receptors also vary in different brain regions [25,26]. The physiological consequences of the two types of α receptor in the brain are unclear at the present time. It is striking that the drug specificity of postsynaptic α₂ receptors closely resembles that of adrenergic autoreceptors, which are therefore also referred to as α₂ receptors.

Studies involving the binding of radioligands are also consistent with the suggestion that there are subtypes of both α₁- and α₂-adrenergic receptors (Tables 12-4 and 12-5). The suggestion that there are subtypes of α₁-adrenergic receptor was initially based on a comparison of the properties of [³H]prazosin and [³H]WB4-101 binding to α₁-adrenergic receptors in rat brain and uterus. Heterogeneity of α₂-adrenergic receptors was initially based on a comparison between the binding of [³H]clonidine and [³H]yohimbine in a variety of tissues and species. The observation that prazosin is more potent in neonatal rat lung [25,26] and cerebral cortex than in the human platelet, the prototypic tissue for the study of α₂ receptors, was interpreted as indicating a heterogeneity in the pharmacological characteristics of α₂-adrenergic receptors. Cloning and sequence analysis suggest that there are three subtypes of α₁-adrenergic receptor and three subtypes of α₂-adrenergic receptor [26,27]. In some instances, the α_1D receptor has been linked to activation of Ca²+ channels, while the α_1B receptor has been shown to activate phosphoinositide-specific phospholipase C (PI-PLC), resulting in liberation of diacylglycerol (DAG) and inositol trisphosphate (IP₃) (see Chap. 20). Prototypic tissues expressing each of the subtypes of α₂ receptor have been identified. All three of the known subtypes of α₂-adrenergic receptor are linked to inhibition of adenylyl cyclase activity. As seen with other receptors linked to inhibition of adenylyl cyclase activity, the α₂-adrenergic receptors have long i₃ loops and relatively short C-terminal tails.

Not surprisingly, the known subtypes of α-adrenergic receptor share structural features with DA receptors (Fig. 12-4) and the other members of the G protein-linked receptor family. The degree of sequence identity is greater when the subtypes of α₁ or α₂ receptors are compared with each other than when α₁ and α₂ receptors are compared. The sequences of α₁ and α₂ receptors are not more closely related to each other than either is to the three known members of the β-adrenergic receptor family. Nomenclature notwithstanding, it is appropriate to think of three families of adrenergic receptor called α₁, α₂ and β. Sequence similarities both within and between families of adrenergic receptors are greater when the sequences of the putative transmembrane helices are compared than when one looks at overall sequence identity. It is sometimes difficult to distinguish between receptor subtypes and species homologs of the same receptor. Small differences in amino acid sequence can sometimes lead to large changes in the pharmacological specificity of an expressed receptor. The possibility that additional catecholamine receptors remain to be identified clearly exists.

Changes in the number of receptors appear to be associated with altered synaptic activity

Neurotransmitter receptors are not static entities; in both the sympathetic PNS and the brain, destruction of catecholamine-containing nerves is associated with functional supersensitivity of postsynaptic sites. Conversely, administration of tricyclic antidepressants or inhibitors of MAO leads to functional subsensitivity. These changes appear to be a compensatory response involving changes in the density of β-adrenergic receptors (see below). Destruction of the DA-containing nigrostriatal pathway also has well-described behavioral consequences. Unilateral nigrostriatal lesion causes asymmetry in DA innervation between the two cerebral hemispheres. Behavioral studies demonstrate that the DA receptors in the denervated corpus striatum are supersensitive. Apomorphine, a DA agonist that stimulates DA receptors selectively, causes rotational behavior in rats with unilateral lesions in a direction contralateral to the lesion. In contrast, indirect agonists such as cocaine cause rotation in the opposite direction.

The extent of receptor supersensitivity can be quantified by measuring the amount of rotational behavior. After selective nigrostriatal lesions have been produced in rats by injections of 6-hydroxydopamine in the substantia nigra, the number of DA receptors in the ipsilateral corpus striatum increases markedly, and the increase in the number of receptors may correlate with the extent of behavioral supersensitivity as monitored by rotational behavior [28]. Thus, the increase in receptor density appears to play a role in the behavioral supersensitivity of these animals.

Changes in the number of dopamine receptors may also be involved in pharmacological actions of neuroleptic drugs

One of the most serious side effects of the neuroleptic drugs is tardive dyskinesia, a disfiguring, excessive motor activity of the tongue, face, arms and legs in patients treated chronically with large doses of the drugs (see Chap. 45). Paradoxically, reduction of the dosage worsens the symptoms, whereas increasing the dosage alleviates the symptoms. It has been suggested that tardive dyskinesia reflects a supersensitivity of DA receptors that have been chronically blocked. This hypothesis gains support from the direct demonstration that chronic treatment with neuroleptics leads to an increase in the number of DA receptors in the corpus striatum. Moreover, the ability of neuroleptics to elicit this increase correlates with their ability to block DA receptors.

The number of α₁ and α₂ receptors increases after the noradrenergic neurons in the brain have been destroyed by injections of 6-hydroxydopamine. It is interesting that after this induced destruction of NE-containing neurons the number of β₁ receptors increases markedly but no changes occur in the number of β₂ receptors [23]. This may be a consequence of the fact that β₂ receptors have a low affinity for NE and that the concentration of epinephrine in the brain is relatively low. Similarly, chronic administration of tricyclic antidepressants, which block the reuptake of NE, leads to a selective decrease in the density of β₁-adrenergic receptors in the cerebral cortex. This suggests that the β₁-adrenergic receptors in the cortex are functionally innervated.

Exposure of cells to agonists results in diminished responsiveness, referred to as desensitization

The phenomenon of desensitization has been most extensively explored for β₂-adrenergic receptors. Within minutes of exposure to an agonist, β receptors are phosphorylated by βARK. This promotes the binding of another protein, β-arrestin, to the receptor, which uncouples the β receptors from the G protein. The uncoupling stops signaling and is followed by a sequestration of cell-surface β receptors into intracellular compartments. There is some evidence that β receptors are then recycled back to the surface of the cell. It is not clear if this regulatory process involving βARKs and β-arrestins [29–31] applies to all catecholamine receptors, but it is a reasonable hypothesis.

Other mechanisms of receptor regulation have been most thoroughly investigated on transformed and transfected cell lines expressing subtypes of β-adrenergic receptor. Transcriptional, post-transcriptional and post-translational regulatory phenomena have been described. Exposure of such cells to an agonist like isoproterenol results in an unexpected increase in mRNA levels. This is thought to be a consequence of the presence of a cAMP-response element (CRE) located approximately 50 bases upstream from the initiation codon. Exposure of cells to isoproterenol results in increases in cAMP and activation of PKA. A CRE-binding protein is then phosphorylated, resulting in activation of the CRE. The resulting increase in mRNA levels is transient and does not have an obvious effect on the synthesis of receptor protein. Over a somewhat longer time scale, post-transcriptional regulatory mechanisms are activated. In particular, mRNA concentrations decline, apparently as a consequence of a decrease in mRNA stability. The mechanism underlying this change in message stability has not been elucidated. Transcriptional and post-transcriptional types of regulation of β-adrenergic receptor synthesis are superimposed on the post-translational phenomena described above (see also Chap. 22).