NORADRENALINE: FATE AND CONTROL OF ITS BIOSYNTHESIS bY JULIUS AXELROD National Institute of Mental Health Bethesda, Maryland, U.S.A. Nobel Lecture, December 12, 1970 When I joined the National Institute of Mental Health in 1955, I began to think of an appropriate problem on which to work. In reading the literature I was surprised to learn that very little was known about the metabolism of noradrenaline and adrenaline. In 1946 von Euler (1) isolated and identified noradrenaline in the sympathetic nervous system and was later to develop sensitive methods for measuring this catecholamine in tissues (2). In 1954 I had been working on the in vivo (3) and in vitro (4) metabolism of am- phetamines and compounds related in structure to catecholamines. Because of this background, I decided to work on the metabolism of noradrenaline and adrenaline. Just as this work was begun, Armstrong et al. (5) identified 3-methoxy-4- hydroxymandelic acid in the urine of subjects with adrenaline-forming tumors. This observation immediately suggested that catecholamines might undergo an 0-methylation reaction. Cantoni had shown that S-adenosylmethionine Table 1. Enzymatic 0-Methylation of Catecholamines Substrate 0-Methoxy Product Formed in mfimoles l-Adrenaline 1 -Adrenaline (AMe omitted) l-Adrenaline (MgCl, omitted) dl-N-Methyl-4-hydroxy phenylethanolamine d-Adrenaline 1 -Noradrenaline dl-Octopamine Dopamine Tyramine Dopa 59 0 4 0 62 63 0 60 0 63 Catecholamines or other amines (0.3 pmole) were incubated at 37" C with partially purified catechol-0-methyltransferase from rat liver, 50 pmoles pH 7.8 phosphate buffer, 150 pmoles S-adenosylmethionine (AMe) 10 pmoles magnesium chloride in a final volume of 1 ml for 30 minutes. When adrenaline, noradrenaline or dopamine were used as substrates the O-me- thoxy derivatives were measured. With other substrates their disappearance was measured (9). 189 formed enzymatically from ATP and methionine can donate its methyl group to the nitrogen of nicotinamide (6) and it appeared possible that S-adenosyl- methionine could donate its methyl group to one of the hydroxy groups of catecholamines. In the initial experiment, a rat liver fraction was incubated with ATP, methionine and Mg" and adrenaline, and the disappearance of the catecholamine was measured (7). When these cofactors were added there was a marked disappearance of the catecholamine. When either cofactor was omitted no metabolism took place. The requirement for both ATP and methionine suggested that the liver extract was making S-adenosylmethionine. With S-adenosylmethionine instead of ATP and methionine, even greater metabolism of adrenaline occurred (Table 1) . The 0-methylated metabolite was isolated by solvent extraction and identified as 3-methoxy-4-hydroxy- phenyl-2-methylamino ethanol (metanephrine) . Metanephrine and normetane- phrine were synthesized within two days after isolation by Senoh and Witkop at the NIH. Rat urine and tissues were then examined by solvent extraction and paper chromatography for the normal occurrence of normetanephrine, metanephrine and 3-methoxy tyramine. All these compounds were present in brain, spleen Adrsnelitw OH $4 Dihydmxymandelic Acid t MAO 3.Mathoxy-4 Hydmxy- / 1Methoxy.4 3-&thaxy-4 Mondelic Acid Hydroxy- Hydroxy-i'henyl- Mondelic Glycol MAO Aldrhyds I OH OH ' Norrdnnaline Fig. 1. Nomwtmephrine Metabolism of noradrenaline and adrenaline. COMT is catechol-0-methyltransferase; MAO is monoamine oxidase. 190 and adrenal gland (8). Later, another 0-methylated metabolite, 3-methoxy-4- hydroxyphenylglycol, was identified. The administration of noradrenaline, adrenaline or dopamine resulted in an elevated excretion of 0-methyated amines, acid and alcohol metabolites. As a result of these experiments the scheme shown in Fig. 1 was proposed for the metabolism of noradrenaline and adrenaline. Dopamine undergoes an analogous pathway. CATECHOL-0-METHYLTRANSFERASE (COMT) The enzyme that 0-methylates catecholamines was partially purified from rat liver and its properties studied (9). It requires Mg" (Table 1 ), but other divalent ions such as Mn", Co++, Zn++, Fe++, Cd++ and Ni++ could be substituted. S-adenosylmethionine is necessary as the methyl donor. All catechols examined were 0-methylated by the enzyme, including adrenaline, noradrenaline, dopa- mine, dopa (Table l), 3,4-dihydroxy mandelic acid, 3,4-dihydroxy phenyl- acetic acid, 3-hydroxy estradiol and ascorbic acid. Foreign catechols such as 3,4-dihydroxy ephedrine, 3,4-dihydroxy amphetamine and many substituted catechols and polyphenols can serve as substrates for COMT. Monophenols are not 0-methylated (Table 1). 0-methylation occurs mainly on the meta position. However, 0-methylation in vitro occurs on both meta and para positions depending on the pH of the reaction mixture and the nature of the aromatic substrate (10). The purified enzyme has a molecular weight of approximately 24,000 ( 11). At least two separate forms of the enzyme have been identified on starch block electrophoresis (12). The enzyme can be inhibited by polyphenols (13), 3- hydroxy estradiol (14) and tropolone (15). The administration in vivo of COMT inhibitors results in a small, but definite, prolongation of the physio- logical effects of noradrenaline ( 16). COMT is present in all mammalian species examined (9) and exists also in some plants ( 17). Of all animal tissues, the liver and kidney exhibit highest activity. Unequally distributed in different regions of the brain, the enzyme's highest activity is present in the area postrema, and lowest activity is in the cerebell'ar cortex ( 18). Catechol-0-methyltransferase occurs mainly in the soiuble fraction of the cell, but sm%l amounts are present in fat cell mem- branes ( 19) and in microsomes (20). COMT acts on catecholamines mainly outside the neurone, whereas monoamine oxidase, the other major enzyme for catecholamine metabolism, is localized mainly within the neurone. How- ever, small amounts of COMT are present in the sympathetic nerves of the nictitating membrane and the vas deferens (22). COMT is involved mainly in the metabolism of catecholamines released into the circulation (23) and in the inactivation of noradrenaline in tissues with sparse adrenergic innervation (24). It also appears to be associated with an extraneural uptake mechanism (25). Recently we have observed that COMT is present within mammalian erythrocytes. This provided an easily available tissue to examine this enzyme in man. The activity of COMT in erythrocytes is reduced in women with primary affective disorders (27). 191 The discovery of COMT led to the description of other methyltransferases involved in biogenic amine metabolism: histamine-N-methyltransferase (28)) hydroxyindole-0-methyltransferase (29), phenylethanolamine-N-methyltrans- ferase (30)) and a nonspecific methyltransferase (3 1) . UPTAKE OF NORADRENALINE BY SYMPATHETIC NERVES Soon after the work on 0-methylation was begun, the distribution of H3-adren- aline in animal tissues was investigated. Fortunately, Seymour Kety arranged for the synthesis of tritiated noradrenaline and adrenaline of high specific activity labeled on a 7 position. This made possible the administration of physiological amounts of the neurotransmitter and a study of the localization and metabo- lism of the circulating catecholamine. In collaboration with Weil-Malherbe and Whitby, specific methods for the measurement of adrenaline, noradrenaline and its 0-methylated metabolites in tissues were developed. After the intra- venous injection of H3-adrenaline (32) or H3-noradrenaline (33) in cats, these catecholamines were rapidly and unequally distributed in tissues. The amines were selectively taken up in tissues heavily innervated with sympathetic nerves (heart, spleen). Since negligible amounts of H3-catecholamines were present in the brain, a blood brain barrier to these compounds was indicated. 0-methylated metabolites, H3-metanephrine and H3-normetanephrine, also occurred in tissues. When tissues were examined two hours following the administration of the catecholamines, long after the physiological effects had disappeared, they were found to have almost the same levels of H3-adrenaline and H3-noradrenaline as those found after two minutes. These experiments suggested that noradrenaline and adrenaline were taken up and retained in tissues in a physiologically inactive form. The selective binding of the cate- cholamines by tissues with a high adrenergic innervation pointed to the sympathetic nerves as, the sites of retention. To examine this possibility the superior cervical ganglia of cats were removed umlaterally and sufficient time (7 days) w'as allowed to elapse for complete degeneration of the sympathetic Table 2. Lack of Uptake df l-I"-Noradrenaline after Chronic Denervation of the Sympathetic Nerves Chronic Dencrvation Acute Dencrvation Denervated Innervated Denervated Innervated Salivary Gland 5 42 76 89 Lachrymal Gland 3 45 - - Retractor Muscle 2 11 13 13 Ocular Muscle 6 48 25 26 Right superior cervical ganglia were removed from 6 cats. After 7 days cats were given 25 fig/kg H*-noradrenaline and the Hs-catecholamine assayed in innervated and denervated structures one hour later. In the acute denervation experiments right superior ganglia were removed 15 minutes before the administration of Ha-noradrenaline. Results are expressed as mpg H'-noradrenaline per gm tissue (34). 192 nerve fibers. H3-noradrenaline was then given intravenously and the animals were killed one hour later and the H3-catecholamine content was examined in structures innervated by the sympathetic cervical ganglia (34). There was a sharp reduction in the uptake of H3-noradrenaline in the chronically de- nervated structures (Table 2). These results made it apparent that sympathetic nerve endings take up and retain the circulating catecholamine. To localize the intraneuronal site of the noradrenaline retention, combined electron microscopy and autoradiography were carried out. Hs-noradrenaline was injected; 30 minutes later the pineal was prepared for autoradiography and electron microscopy (35). The pineal gland was chosen because of its rich sympathetic innervation. Electron miscroscopy showed a striking localiza- tion of photographic grains overlying non-myelinated axons which contained granulated vesicles of about 500 A. With Potter attempts were made to isolate the dense core vesicles asso- ciated with the H3-noradrenaline (36, 37). Previously von Euler and Hillarp had isolated a high speed noradrenaline-containing granular fraction from bovine splenic nerves (38). Again, H3-noradrenaline was injected in rats and subcellular fractions of the heart and other tissues were separated in a continuous sucrose gradient (36). The predominant peak of the H3-nor- adrenaline together with the endogenous catecholamine coincided with the COUNTS/MlNtiTE (.) "M ICROSOME SUPER NATANT MITOCHONOR `IA , HE MOGLOBlN MUSCLE DEBRIS NUCLEI MICROGRAMS CATECHOLAMlNE(0; Fig. 2. Subcellular distribution of noradrenaline in the rat heart. Sprague-Dawley male rats were given 50 ,uc H3-noradrenaline and were killed 30 minutes later. The hearts were rapidly removed and homogenized in isotonic sucrose. A portion was layered on an exponential sucrose gradient and centrifuged in a Spinco preparative centrifuge using an SW39 rotor for 30 minutes (36). Drops were collected with a needle through the bottom of the tube and assayed for H3 and endogenous noradrenaline. 193 "microsomal band" (Fig. 2). The noradrenaline containing particles had no pressor action unless they were lysed in dilute acid, suggesting that the cate- cholamine was bound. In addition to H3-noradrenaline, the microsomal peak also contained large amounts of dopamine-&oxidase (37)) the enzyme that converts dopamine to noradrenaline. Further attempts to purify noradrenaline containing vesicles were unsuccessful. `The ability to take up and store H3-noradrenaline enabled Hertting and me to label the neurotransmitter in the nerve endings of tissues and to study its fate on liberation from sympathetic nerves (39). Cats were given H3-nor- adrenaline; the spleen, containing nerve endings labelled with H3-noradren- aline, was perfused; the splenic nerve was stimulated, as described by Brown and Gillespie (40) ; and the radioactive catecholamine and is metabolites were measured in the venous outflow. After each series of stimulations a marked increase occurred in the concentration `of H3-noradrenaline in the venous outflow. There was also a small but measurable elevation of the O- methylated metabolite, normetanephrine, but no increase in deaminated me- tabolites. From these experiments we concluded that noradrenaline liberated from the nerve terminals was inactivated by several mechanisms. Part is dis- charged into the bloodstream; part is 0-methylated by COMT, and ,part is taken up by the nerve terminals. Reuptake of noradrenaline by sympathetic neurones was examined in experiments performed with Rose11 and Kopin using the vascular bed of the dog gracilis muscle in situ (41). The sympathetic nerves of the gracilis muscle were labeled by an infusion of H3-noradrenaline, and the discharge of H3-noradrenaline measured after nerve stimulation. When the vasomotor nerves were stimulated, an initial reduction in the outflow of H3-noradrenaline was followed by a rise in outflow of the radioactive cate- cholamine (Fig. 3). The lag in the outflow was due to an increase in vascular ARTERIAL PERFUSION PRESSURE 200 100 0 240 I- mm Hg 4 V -- Fig. 3. Uptake and release of H3-noradrenaline in dog gracilis muscle. Dog gracilis muscle was perfused with Hs-noradrenaline as described by Rosell, Kopin and Axelrod (41). Peripheral resistance and venous outflow of Hs-noradrenaline was measured during sympathetic nerve stimulation, before and after treatment with phenoxybenzamine. 194 resistance. This observation indicates a reduced capacity of the vascular bed to carry away the released noradrenaline. After the stimulus was ended, decline in H3-noradrenaline outflow and return of the peripheral resistance were parallel. To block the constriction of the vascular bed, dogs were pretreated with phenoxybenzamine, an adrenergic blocking agent shown to inhibit reuptake of noradrenaline. Vasomotor stimulation resulted in an im- mediate and larger increase in noradrenaline outflow. The larger and im- mediate outflow of noradrenaline was due to a blockade of noradrenaline reuptake by phenoxybenzamine. It was concluded from this and other ex- periments that reuptake by sympathetic nerves was a major mechanism for terminating the actions of the neurotransmitter noradrenaline. Subsequent work by several investigators, particularly Iversen (42, 43) described the properties of the neuronal uptake mechanism. It obeys saturation kinetics of the Michaelis-Menton type: it is stereospecific for the l-isomer of noradren- aline and requires Na'. Many other amines structurally related to noradrenaline can be taken up and stored in sympathetic nerves by a neuronal uptake proc- ess. The fate of noradrenaline at the sympathetic nerve terminal and circula- tion is shown in Fig. 4. Noradrenaline can also be taken up by an extraneuronal process (44, 45) which has been shown to be similar to Iversen's Uptake 2 (46): This uptake Fig. 4. Fate of noradrenaline (NA) at a varicosity of the sympathetic nerve catechol-0-methyltransferase; MAO is monoamine oxidase. terminal. CUMl 15 195 is inhibited by adrenergic blocking agents and normetanephrine (25). Com- pounds such as isoproterenol which have a low affinity for intraneural uptake and a high affinity for extraneural uptake may be inactivated by the later process. Extraneural uptake operates at all concentrations of catecholamines (47) and serves to transport amines into non-neuronal tissues in which they are subsequently metabolized. EFFECT OF DRUGS ON NEURONAL UPTAKE The ability of the sympathetic nerves to take up H3-catecholamines provided a relatively simple technique for studying the effect of a variety of drugs acting on the sympathetic nervous system. In early experiments of this kind, mice were treated with a variety of drugs and the rate of disappearance of H3-adrenaline was measured (48). A wide variety of drugs (imipramine, chlorpromazine, cocaine, reserpine, amphetamine, tyramine) increased the rate of disappearance of the catecholamine. Such experiments suggested that these drugs might increase the rate of metabolism by interfering with the binding and/or uptake of the catecholamines, thus exposing them to enzymatic attack by COMT or monoamine oxidase. This suggestion was supported by the observation that the catechol quercitrin markedly slowed catecholamine metabolism in vivo, presumably by inhibiting COMT. The experiments that followed were more direct. Cats were treated with cocaine, and then H3-noradrenaline was injected intravenously. One hour later, heart, spleen and adrenal gland were examined for H3-noradrenaline (49). Cats pretreated with cocaine showed a dramatic decrease in tissue I-Is- noradrenaline. In addition, there was a sharp elevation in plasma levels of H3-noradrenaline in cocaine-treated animals. This experiment revealed that cocaine markedly reduces the uptake of noradrenaline in tissues, presumably the sympathetic neurone. The inhibition of uptake by cocaine thus raised the extraneural concentration of noradrenaline (as reflected by the elevated level in plasma catecholamine). By blocking uptake into the nerves, cocaine caused an elevated concentration of'noradrenaline to reach Che receptor (Fig. 5)) and this explains the effect of the drug and denervation of sympathetic nerves in producing supersensitivity. Experiments similar to those described with cocaine were carried out with COCAINE, NORMALUPTAKE SYMPATHOMIMETIC AMINES, IMIPRAMINE Fig. 5. Effect of drugs on uptake of noradrenaline at the sympathetic nerve terminal. 196 AMPHETAMINE Fig. 6. IMIPRAMINE GANGLIA BLOCKER cl CONTROL tza H'NA BEFORE lllmn" NA AFTER Effect of drugs in uptake and release of H3-noradrenaline in the rat heart. Rats were given 15 ,UC H3-noradrenaline 30 minutes before or after the administration of drugs and killed 24 hours later. The hearts were examined for H3-noradrenaline remaining (52, 54). The ganglia blocker was chlorisondamine. other drugs (50, 51) . The following compounds lowered the concentration of H3-nor-adrenaline in tissues: imipramine, chlorpromazine, tyramine, ampheta- mine, guanethedine, reserpine and phenoxybenzamine. All of these drugs also elevated the initial blood level of the HS catecholamine. Such observations indicate that these drugs also interfere with the uptake of noradrenaline into the adrenergic neurone. In addition to blocking uptake, these drugs could also prevent the storage or release of the bound H3-nor-adrenaline. If a drug prevents noradrenaline uptake, it should lower the tissue levels of H3-noradrenaline only when given before the H3 catecholamine. If it reduces the concentration when given after H3-noradrenaline, when the neurotransmitter is bound to tissue, then it realeases the catecholamine. To distinguish between these two possibilities rats were given drugs before or after the intravenous injection of H3-noradrenaline, and the amount of the H3-catecholamine in the heart was measured (52). Reserpiner amphetamine, and tyramine reduced the H3-catecholamine after it was bound. Pretreatment witl? imipramine (Fig. 6) or chlorpromazine lowered the concentration of cardiac H3-noradrenaline only when given before H3-noradrenaline, indicating that these drugs blocked uptake but did not release the amine. Amphetamine caused a greater reduction when given before H3-noradrenaline than after (Fig. 6), indicating that it not only blocks uptake but also relsases the qatecholamine. Many of thes'e observations were confirmed and extended .through `direct visualization of the sympathetic neurone by histofluorescent techniques (53). . H3-noradrenaline was also used to measure the effect of drugs in blocking the spontaneous release of the neurotransmitter. Long-lasting ganglionic block- ing agents (chlorisondamine; Fig. 6) and bretylium inhibited the spontaneous 197 release of H3-noradrenaline from the rat heart (54, 55). Decentralization of the superior cervical ganglion also slowed spontaneous release of H3-nor- adrenaline, again demonstrating that nerve impulses cause a release of the H3-noradrenaline (54). UPTAKE, STORAGE, RELEASE AND METABOLISM OF ~~~~~~~~~~~~~~~ IN THE &T BRAIN In 1954 Vogt demonstrated the presence of noradrenaline in the brain and showed that it was unequally distributed (56). Drugs such as amphetamine and reserpine (56, 57, 58) lowered the tissue concentration of endogenous noradrenaline, whereas monoaminoxidase inhibitors elevated the level of the catecholamine (58). In our earlier work we were unable to study the dis- position of noradrenaline in the brain because of its inability to cross the blood-brain barrier (33). In 1964 Jacques Glowinski devised a technique for introducing H3-noradrenaline into the rat brain via the lateral ventricle (59). This provided a means of labeling the brain stores of noradrenaline and en- abled us to study the fate of this compound in the brain and examine the effect of drugs. The initial concern was whether this exogenously administered H3- noradrenaline mixed with the endogenous pool of brain catecholamines. We first examined the distribution of the H3-noradrenaline in various brain areas. After an intraventricular injection, H3-noradrenaline was selectively distributed in areas which contained high concentrations of catecholamines, the highest levels occurred in the hypothalamus and the lowest in the cerebral cortex and cerebellum (60). However, considerable amounts of H3-noradrenaline were present in the corpus striatum, which normally contains high levels of dop- amine and little endogenous noradrenaline. In autographic studies intense labeling was also found in the periventricular and ventromedial nuclei of the hypothalamus, medial forebrain bundle, in specific tracts of the spinal cord and in the apical dendritic layer of the hypocampus. Subcellular distribution studies, using continuous sucrose gradients, showed the H"-noradrenaline, after its intraventricular administration in the brain, was present in the synaptosomal laye,r (pinched off nerve endings) tog&her with endogenous noradrenaline (61) . These observations indicated that H3-noradrenaline mixed to a considerable degree with the endogenous brain stores of the catecholamine. The H3-noradrenaline persisted in the brain for long periods of time, indicating that it was stored and protected from metabolism. The radioactive metabolites formed were normetanephrine and 0-methylated deaminated metabolites. The major product in the brain was H3-3-yethoxy-4-hydroxyphenylglycol. Labeling of the brain stores of noradrenaline provided an opportunity to study the effects of drugs on the uptake, storage, release and metabolism of noradrenalinte in the brain (62). It was previously shown that imipramine and chlorpromazine blocked the uptake of H3-noradrenaline in intact periph- eral tissues (51) and brain slices (42). In the intact rat brain, imipramine reduced the accumulation of H3-noradrenaline after its intraventricular in- jection (63), while chlorpromazine did not (Table 3). Other antidepressant 198 Table 3. Antidepressant Drugs and the Inhibition of Uptake of Ha-Noradrenaline in the Rat Brain Treatment Clinical H"-Noradrenaline Antidepressant gm/brain Action cpm X 1000 None Imipramine Desmethylimipramine Amitryptyline Compound 2 Compound 3 Chlorpromazine Yes Yes Yes No No No 30 f 2.0 19 * 1.0' 19 f- 1.11 23 f 2.1* 30 f 1.6 28 & 1.2 32 * 3.1 `P < .OOl `P < .05 Groups of 6 rats were given drugs (20 mg/kg) intraperitoneally 1 hour before the admini- stration of 0.07 pg of H"-noradrenaline into the lateral ventricle of the brain. Rats were killed two hours later and assayed for Ha-n&adrenaline. Compound 2 had the same struc- ture as imipramine except that a dimethyl isopropyl side chain was substituted for a dime- thylaminopropyl side chain. Compound 3 had the same structure as chlorpromazine except that a dimethylaminoethyl ether side chain was substituted for a dimethylaminopropyl side chain (63). drugs such as desmethylimipramine and amitriptyline reduced the accumula- tion of H3-noradrenaline in the brain, but structurally related derivatives of imipramine which are clinically inactive as antidepressants had no effect. Both monoamine oxidase inhibitors and imipramine are antidepressant drugs and cause an increased amount of physiologically active noradrenaline to react with the adrenergic receptors in the brain. Each of these compounds makes more noradrenaline available in the brain by different mechanisms. Imipramine and other tricyclic antidepressant drugs slow inactivation by reuptake into the neurone, and monoamine oxidase inhibitors prevent me- tabolism of the catecholamine. Amphetamine has multiple actions on the disposition of the catecholamine in the brain (,62). Like tricyclic antidepres- sants, it blocks uptake into the neurone, causes release of the catecholamine from its storage site, and inhibit4 monoamine oxidase. Amphetamine ad- ministration results in an increased formation of H3-normetanephrine in brain, whereas reserpine causes an increase in deaminated metabolites. These metabolic changes reflect a release from the neurone of physiologically active noradrenaline by amphetamine and a release of inactive metabolites by re- set-pine. Glowinski and Iversen performed-a study ,on metabolism of noradrenaline in different brain regions. They found that all .areas of the brain except the striatum can convert dopamine to,noradrenaline (64). Amphetamine blocked the reuptake of noradrenaline in all brain areas, whereas desmethylimipramine inhibited uptake in cerebellum, medulla oblongata and hypothalamus, but not in the corpus striatum (62). Rates of turnover of brain noradrenaline were 199 also examined by such different experimental approaches as measuring rates of disappearance of endogenous noradrenaline after inhibiting catecholamine biosynthesis, estimating rates of disappearance of H3-noradrenaline formed from H3-dopamine, and determining rates of disappearance of H3-noradren- aline after its intraventricular injection, These methods produced results in close agreement with one another. Cerebellum had the fastest turnover and the medulla oblongata and hypothalamus had the slowest turnover (65). With these techniques subsequent work has established that turnover of brain noradrenaline is altered by a variety of stresses, temperature changes and sleep. REGULATION OF THE BIOSYNTHESIS OF CATECHOLAMINES The catecholamines are synthesized as shown in Fig. 7. This biosynthetic pathway was first proposed by Blaschko in 1939 (66) and finally established by Udenfriend and his coworkers (67). The first step is catalyzed by the enzyme tyrosine hydroxylase (67), the second by dopa decarboxylase (68), and the third by dopamine-Soxidase (69). These reactions occur within the sympathetic nerve terminal. The final step is catalyzed by phenylethanolamine- N-methyltransferase and occurs almost exclusively in the adrenal medulla (30). In the adrenal gland the biosynthetic enzymes tyrosine hydroxylase, dopamine- /l-oxidase and phenylethanolamine-N-methyltransferase are confined almost entirely to the adrenal medulla. Noradrenaline in sympathetic nerves and catecholamines (noradrenaline and adrenaline) in the adrenal medulla are in constant flux. They are con- tinuously being released, metabolized, and synthesized, yet they maintain a remarkably constant levei in tissues. Recent work jn our laboratory and those of others revealed several mechanisms that regulate the biosynthesis of cate- cholamines, involving long-term hormonal controls as well as short and long- term neural regulation . tl ON TYROSINE DWA D~PAMINE PHT 4 C-C-N-CH, c AA PNMT OH NORADRENALINE ADRENALINE Fig. 7. The biosynthesis of catecholamines. PNMT is phenylethanolamine-N-methyltransferasc. 200 HORMONAL CONTROL In species such as dogfish, where the chromaffin tissue is located outside the adrenal gland, little or no adrenaline occurs (70). In species where the medulla is completely contiguous with the cortex (human and rat) almost all of the catecholamine content is adrenaline. This suggested to Wurtman and me (71) that the adrenal cortex might affect the activity of the adrenaline forming enzyme phenylethanolamine-N-metyhyltransferase. I had been work- ing on the properties of this enzyme and developed a sensitive and specific assay for its measurement (30). In the initial experiment we measured the effect of hypophysectomy on the phenylethanolamine-N-methyltransferase in the adrenal gland (71). The hypophysectomized rats showed steady fall of the adrenaline-forming enzyme until about 20 percent of the initial con- centration remained (Fig. 8). The daily administration of either ACTH (Fig. 8) or dexamethasone for 21 days restored enzyme activity to normal levels in hypophysectomized rats. To examine whether the corticoid-induced rise in PNMT was due to increased synthesis of new enzyme protein, dexa- methasone was given to rats whose RNA dependent protein synthesis had been inhibited by puromycin or actinomycin D. Both inhibitors of protein synthesis prevented the rise of enzyme activity caused by dexamethasone. However, repeated administration of ACTH or dexamethasone to intact rats failed to elevate adrenal PNMT activity above normal levels. In view of the effect of hypophysectomy on the adrenal PNMT activity, the effect on other catecholamine biosynthetic enzymes was examined. After hypophysectomy there was a fall of adrenal gland tyrosine hydroxylase (72). Fig. 8. PNMT TY.ROSINE HYDROXYLASE - cl NORMAL El HYPOX m HYPOX t ACTH AMINE j9-0XiD~sE Control of enzymatic synthesis of adrenaline in the adrenal medulla by ACTH. Phenyl- ethanolamine-N-methyltransferase (PNMT) (71) and dopamine-B-oxidase (73) were measured 21 days after hypophysectomy, and tyrosine hydroxylase 5 days after hypo- physectomy (72). ACTH was given, after hypophysectomy, daily for 6 days. 201 (Fig. 8). Enzyme activity was reduced 25 percent in 5 days (Fig. 8) and to about half in 10 days. Repeated administration of ACTH restorted tyrosine hydroxylase activity to normal values in hypophysectomized rats (Fig. 8). In contrast to PNMT, dexamethasone did not elevate tyrosine hydroxylase activ- ity in hypophysectomized rats. Again repeated doses of large amounts of ACTH did not increase adrenal tyrosine hydroxylase in normal rats. Dopamine-/Loxidase (the enzyme that converts dopamine to noradrenaline) activity was also examined in hypophysectomized rats (73). This enzyme decreases to about 30 percent of normal values after 21 days (Fig. 8). Ad- ministration of ACTH for 5 days caused dopamine-/?-oxidase activity to in- crease, but full activity was not reached in this period of time. These observa- tions indicate that the normal maintenance of the catecholamine biosynthetic enzymes in the adrenal glands requires ACTH. NEURAL REGULATION The biosynthesis of catecholamines in the sympathetic nerves and the adrenal gland is under precise control by nervous mechanisms, one of which is rapid and the other slower. After prolonged stimulation of the splanchnic nerve the sum of the amount of catecholamines released together with the amount remaining in the gland is greater than that initially present in the gland (74). This indicated that nerve impulses increases the biosynthesis of cate- cholamines. Weiner and his coworkers using an isolated preparation of the hypogastric nerve of the vas deferens showed that stimulation resulted in an increased synthesis of C?`-noradenaline from Cl*-tyrosine, but not from C4-dopa (75). They also found that addition of noradrenaline prevented an increase in Cl4 catecholamine formation from C14-tyrosine. However, stimula- tion of the vas deferens did not change the total amount of noradrenaline or tyrosine hydroxylase in vitro. The fact that noradrenaline is capable of inhibiting the conversion of Cl"- tyrosine to noradrenaline indicated a rapid feedback inhibition at the tyrqine hydroxylase step. Another type of regulation of catecholamine biosymhesis was uncovered in an unexpected manner. Tranzer and Thoenen reported that Ghydroxydop- amine selectively destroyed sympatheti? .ner%e terminals ( 76). Thoenen decided to spend a sabbatical year in my laboratory, and together with Mueller we examined the effect of chemical destruction of sympathetic nerve terminals by 6-hydroxydopamine on the biosynthetic enzyme tyrosine hydroxylase. As expected, the enzymes completely disappeared within two days after the administration of 6-hydr6xydopamjne ( 77). However, when the adrenal gland was examined a marked increase in `tyrosine hydroxylase was observed. Since 6-hydroxydopamine lowers blood pressure, the increase in enzyme activity caused by this compound might' be due to a reflex increase in sym- pathetic adrenal activity. Consequently we examined the effect of reserpine, which is known to reduce blood pressure and increase preganglionic neuronal activity. Reserpine produced a marked increase in tyrosine hydroxylase activity over several days in the adrenal gland of the rat and several other species, in 202 200 1 J- DOPAMINE IS-0x1~~s~ El DECENTRALIZED TYROSINE HYDROXYLASE O NORMAL CONTROL RESERPINE CONTROL RESERPINE Fig. 9. Transsynaptic induction of noradrenaline biosynthetic enzymes. Right or left superior cervical ganglion was decentralized by transection of the preganglionic trunk from 2 to 6 days before reserpine treatment. Reserpine (5 mg/kg) was given 24 hours before tyrosine hydroxylase was assayed in innervated and decentralized ganglia (78). In the case of dopamine-fi-oxidase, reserpine (2.5 mg/kg) was given on 3 alternated days and enzyme examined on the 7th day after decentralization (83). the superior cervical ganglion (Fig. 9) and in the brainstem of the rabbit (78, 79). The adrenergic blocking agent phenoxybenzamine also caused a reflex increase in sympathetic adrenal activity. And again the administration of this compound resulted in an elevation in tyrosine hydroxylase activity in the adrenal gland. To examine whether the increased enzyme activity is due to the formation of new enzyme molecules, protein synthesis was inhibited prior to the administration of the drugs. Inhibition of protein synthesis with either cycloheximide or actinomycin D prevented the drug-induced increase of tyrosine hydroxylase in the adrenal gland and ganglia (80). The most likely mechanisms for the increase in enzyme activily might be a blood-borne factor, as in th,e induction of PNMT by ACTH, or an increase in the activity of the preganglionic neurones. To examine .the latter possibility, we cut unilaterally the splanchnic nerve supplying the adrenal gland (81) and preganglionic fibers to the superior cervical ganglion and then administered reserpine (78). This drug caused the expected rise in tyrosine hydroxylase in the innervated side but the increase in tyrosine hydroxylase on the denetvated side was completely prevented (Fig. 9). These results indicate that the increase in tyrosine hydroxylase is due `to a transsynaptic induction of the enzyme. Studies on the molecular mechanisms that cause this induction across nerves have thus far proved unsuccessful. The neuronally-mediated induction of tyrosine hydroxylase after reserpine is also observed in the nerve terminals as well as the cell body. However, the increase in tyrosine hydroxylase in the nerve terminals lags behind the ganglia by two or three days (82). Experiments with inhibitors of protein synthesis point to a local formation of induced 203 tyrosine hydroxylase in the nerve terminals rather than the peripheral move- ment .of the completed enzyme. Similar studies on the induction of dopamine-/?-oxidase, an enzyme present in the noradrenaline storage granule, were undertaken with the collaboration of Molinoff, Weinshilboum and Brimijoin (83). These experiments were made possible because a very sensitive assay for dopamine-Soxidase was developed. This enzyme could be measured where it never has been found before. Re- peated administration of reserpine caused a marked rise in dopamine-& oxidase in rat stellate and superior cervical ganglia (Fig. 9) in the nerve ter- minals as well as in the adrenal medulla. The elevation of dopamine+ oxidase in sympathetic ganglia was blocked by protein synthesis inhibitors or by surgical decentralization (Fig. 9). Recently we have found that dopamine- /?-oxidase is present in the plasma of man and other mammalian species (84). Preliminary experiments indicate that the circulating dopamine-/?-oxidase comes from sympathetic nerve terminals. The activity of PNMT in the adrenal gland also increased after reserpine and this elevation in enzyme was blocked by interrupting the splanchnic nerve (85). Because of the increasing implications of catecholamines in behavioral changes we examined the effects of psychosocial deprivation and stimulation on the biosynthetic enzymes tyrosine hydroxylase, and PNMT (86). One group of mice was isolated to prevent visual contact and another was exposed to increased social stimulation by a specially designed cage system. After six months a marked decrease in adrenal tyrosine hydroxylase and PNMT activity occurred in the deprived mice and an increase in both these enzymes was found in the stimulated mice. In related experiments in Kopin's laboratory it was also observed that prolonged forced immobilization in rats also produced a rise in adrenal tyrosine hydroxylase and PNMT activities, and this elevation was abolished by interrupting the splanchnic nerve to the adrenal (87). All these results suggest that the increase in the catecholamine-forming enzymes in sustained stress may be neuronally mediated and that this response is not immediate, as in the case of a sudden discharge of noradrenaline and adren- aline in states of anger, fear or aggression. NORADRENALINE AS A NEUROCHEMICAL TRANSDUC~~R IN THE PINEAL GLAND The pineal gland is exceedingly rich in sympathetic nerve fibers which originate in the superior cervical ganglia (88). This organ has the unique capacity to synthesize the hormone melatonin (!%methoxy-N-acetyltryptamine) as follows: tryptophan + 5-hydroxytryptophan + serotonin + N-acetylserotonin + mel- atonin (89). A year before the discovery'of melantonin by Lerner (90) we found an 0-methylating enzyme, COMT (7). Th is stimulated a search for the en- zyme that 0-methylates indoles to form melatonin. Such an enzyme was found in the pineal gland and named hydroxyindole-0-methyltransferase (29). The enzyme 0-methylates N-acetylserotonin to form melatonin, S-adenosylmehio- nine serving as the methyl donor. At about the same time the enzyme (N- acetyltransferase) that acetylates serotonin to N-acetylserotonin was described 204 MELATONIN cl WITHOUT NORADRENALINE E2 WITH NORADRENALINE iEROTONlN Stimulation of melatonin synthesis in pineal gland by noradrenaline. Culture tubes of pineal gland of rats were incubated with C14-tryptophan in the absence or presence of noradrenaline (3 x 15-4 M). After 24 hours the pineal cultures were assayed for CPQerotonin and C14-melatonin (97). (91). The latter enzyme subsequently proved to be critical in the control of melatonin by the adrenergic nervous system. That environmental lighting had something to do with the pineal was suggested by Fiske in 1961 (92) when she found that continuous light changed the weight of the organ. Consequently rats were placed in continuous darkness or light and activity of the melatonin forming enzyme hydroxy indole-0-methyltransferase in the pineal was meas- ured (93). In constant darkness the hydroxyindole-0-methyltransferase activ- ity in the pineal was more than twice as great as that in constant light. Removal of the superior cervical ganglia abolished this difference in enzyme activity (94). Since noradrenaline is the neurotransmitter of the sympathetic nerves, it might be the agent involved in controlling melatonin synthesis. This possibility was reinforced by the demonstration that levels of noradren- aline in the pineal are markedly influenced by environmental lighting (95). A possible approach to an examination of the mechanism whereby the neuro- transmitter could influence the synthesis of `melatonin (which occurs outside the ne&one) was to use pineah in organ culture. Mainly through the efforts of Shein, we succeeded in growing pineal gland in organ culture. The pineal in organ culture was capable of carrying out all the steps in the formation of melatonin from tryptophan (96). Inhibition of protein synthesis completely stopped the conversion of tryptophan to melantonin, indicating that new enzyme protein was being formed. Addition of noradrenaline to the culture medium resulted in a shar$ increase of melatonin, but not serotonin formation from tryptophan over a period of 24 hours. (97). (Fig. 10). However, nor- adrenaline had only a marginal effect on the hydroxyindole-O-methyltrans- ferase activity. Klein et al. examined the enzyme that converts serotonin into N-acetylserotonin in pineal organ culture. He found that noradrenaline causes remarkable increase in the activity of this enzyme (98). When protein syn- thesis was blocked, noradrenaline no longer stimulated the N-acetyltransferase activity. These results show that noradrenaline released from sympathetic 205 nerves stimulates the formation of the pineal hormone melatonin by specific- ally increasing the synthesis of new N-acetyltransferase molecules. CONCLUDING REMARKS Since the demonstration by Otto Loewi (99) that sympathetic nerves exert their effects by the release of a chemical substance, numerous advances have occurred. The neurotransmitter has been identified as noradrenaline and its biosynthesis, metabolism and inactivation elucidated. Although the complexities of the storage, release and regulation of noradrenaline and adrenaline have been partially unravelled, much remains to be done. Our understanding of central adrenergic mechanisms is still at the early stages but shows great promise for rapid development. Drugs therapeutically effective in the treat- ment of affective disorders and neurological and cardiovascular diseases have also been shown to influence the uptake, storage, release, formation and metabolism of catecholamines. These findings implicating the peripheral and central sympathetic nervous system have provided insight into the causes and treatment of mental depression (loo), Parkinson's disease (101.) and hyper- tension ( 102). ACKNOWLEDGEMENTS For their incisive help in several of these investigations, I thank the many research associates and visiting scientists who worked with me. I am grateful to the National Institute of Mental Health and the National Institutes of Health, Bethesda, Maryland, for providing generous support, superb resources and a stimulating environment. REFERENCES 1. U. S. von Euler, Acta .Physiol. Stand. 22, 73 (1946). 2. U. S. von Euler and I. Floding, Acta Physiol. Stand. 33, Suppl. 118, 57 (1955). 3. J. Axelrod, J. Pharmacol. Exptl. Therap. 110, 315 (1954). 4. J. Axelrod, J. Biol. Chem. 214, 753 (1955). 5. M. D. Armstrong, A...McMillan and K. N. F. Shaw, Biochem. Biophys. Acta 25, 422 (1957): 6. G. L. Cantoni, J. Biol. Chem. 189, 203 (1951). 7. J. Axelrod, Science 126, 400 (1957). 8. J. Axelrod, S. Senoh and B. Witkop, j. Biol. Chem. 233, 697 (1958). 9. J. Axelrod and R. Tomchick, J. Biol. Chem. 233, 702 (1958). 10. S. Senoh, J. Daly, J. Axelrod and B. Witkop, J. Amer. Chem. Sot. 81, 6240 (1959). 11. M. Assicot and C. Bohuon, Eur. J. Biochem. 22, 490 (1970). 12. J. Axelrod and E. S. Vesell, Mol. Pharmacol. 6, 78 (1970). 13. J. Axelrod and M. J. LaRoche, Science 130, 800 (1959). 14. V. R. Knuppen, M. HGller, D. Tilmann and H. Breuer, Hoppe-Seyler's Z. Physiol. Chem. 350, 1301 (1969). 15. B. Belleau and J. Burba, J. Med. Chem. 6, 755 (1963). 206 16. D. W. Wylie, S. Archer and A. Arnold, J. Pharmacol. Exptl. Therap. 130, 239 (1961) 17. J. D. Mann, H. M. Fales and S. H. Mudd, J. Biol. Chem. 238, 3820 ( 1963). 18. J. Axelrod, W. Albers and C. D. Clemente, J. Neurochem. 5, 68 (1959). 19. G. J. Traiger and D. N. Calvert, Biochem. Pharmacol. 18, 109, 1969. 20. J. K. Inscoe, J. Daly and J. Axelrod, Biochem. Pharmacol. II, 1257 (1965) 21. J. Axelrod, Pharmacol. Rev. 18, 95 ( 1966). 22. B. Jarrot and L. Iversen, J. Neurochem. (in press). 23. I. J. Kopin, Pharmacol. Rev. 16, 179 (1964). 24. J. A. Levin and R. F. Furchgott, J. Pharmacol. Exptl. Therap. 272, 310 (1970). 25. A. J. Eisenfeld, L. Landsberg and J. Axelrod, J. Pharmacol. Exptl. Therap. 158, 378 (1967). 26. J. Axelrod and C. K. Cohn, J. Pharmacol. Exptl. Therap. (in press). 27. C. K. Cohn, D. Dunner and J. Axelrod, Science 170, 1323 (1970). 28. D. D. Brown, R. Tomchick and J. Axelrod, J. Biol. Chem. 234, 2948 (1959). 29. J. Axelrod and H. Weissbach, J. Biol. Chem. 236, 211 ( 1961). 30. J. Axelrod, J. Biol. Chem. 237, 1657 (1962). 31. J. Axelrod, J. Pharmacol. Exptl. Therap. 138, 28 (1962). 32. J. Axelrod, H. Weil-Malherbe and R. Tomchick, J. Pharmacol. Exptl. Therap. 127, 251 (1959). 33. L. G. Whitby, J. Axelrod and H. We&Malherbe, J. Pharmacol. Exptl. Therap. 132, 193 (1961). 34. G. Hertting, J. Axehod, I. J. Kopin and L. G. Whitby, Nature 189, 66 (1961). 35. D. E. Wolfe, L. T. Potter, K. C. Richardson and J. Axelrod, Science 138, 440 (1962) 36. L. T. Potter and J. Axelrod, J. Pharmacol. Exptl. Therap. 240, 199 ( 1963). 37. L. T. Potter and J. Axelrod, J. Pharmacol. Exptl. Therap. 142, 291 (1963). 38. U. S. von Euler and N.-A. Hillarp, Nature 177,44 (1956). 39. G. Hertting and J. Axelrod, Nature 192, 172 ( 1961). 40. G. L. Brown and J. S. Gillespie, J. Physiol. London 138, 81 (1957). 41. S. Rose& I. J. Kopin and J. Axelrod, Am. J. Physiol. 205, 3 17 ( 1963). 42. H. J. Dengler, I. A. Michaelson, H. E. Spiegel and E. Titus, Intern. J. Neuro- pharmacol. 1, 23 (1962). 43. L. L. Iversen, The Uptake and Storage of Noradrenaline in Sympathetic Nerves, Cambridge University Press, 1967. 44. A. J. Eisenfeld, J: Axelrod and L. R. Krakoff, J. Pharmacol. Exptl. Therap. 156, 107 (1967). 45. J. S. Gillespie, D. N. H. Hamilton and R. J. A. Hosie, J. Physiol. London 206, 563 (1970). 46. L. L. Iversen, Brit. J. Pharmacol. 25, 18 ( 1965). 47. S. L. Lightman.and L. L. Iversen, Brit. J. Pharmacol. 37, 638 (1969). 48. J. Axelrod and R. Tomchick, Nature 284, 2027 (1959). 49. L. G. Whitby, G. Hertting and J. Axelrod, Nature 187, 604 (1960). 50. J. Axelrod, L. G. Whitby and G. Hertting, Science 133, 383 (1961). 51. G. Hertting, J,,Axelrod and L. G. Whitby, J. Pharmacol. Exptl. Therap. 134, 146 (1961). : 52. J. Axelrod, G. Hertting and L. Potter, Nature' 194, 297 (1962). 53. T. Malmfors, Acta Physiol. Stand. 64, SuppL'248, 1 (1965). 54. G. Hertting, L. T. Potter and J. Axelrod, J. Pharmacol. Exptl. Therap. 136, 289 (1962). 55. G. Hertting, J. Axelrod and R. W. Patrick, Brit. J. Pharmacol. 28, 161 (1962). 56. M. Vogt, J. Physiol. London 123, 451 (1954). 57. A. Carlsson, E. Rosengren, A. Bertler and J. Nilsson, In: Psychotropic Drugs, Garratini and Ghetti, 363, Elsevier, Amsterdam, 1957. 58. B. B. Brodie, S. Spector and P. A. Shore, Ann. N.Y. Acad. Sci. 80, 609 (1959). 207 59. J. Glowinski, I. J. Kopin and J. Axelrod, J. Neurochern. 12, 25 (1965). 60. J. Glowinski and J. Axelrod, Pharmacol. Rev. 18, 775 (1966). 61. J. Glowinski, S. H. Snyder and J. Axelrod, J. Pharmacol. Exptl. Therap. 2.52, 282 (1966). 62. J. Glowinski, J. Axelrod and L. L. Iversen, J. Pharmacol. Exptl. Therap. 153, 30 (1966). 63. J. Glowinski and J. Axelrod, Nature 204, 1318 (1964). 64. J. Glowinski and L. L. Iversen, J. Neurochem. 13, 655 (1966). 65. L. L. Iversen and J. Glowinski, J. Neurochem. 13, 671 f, 1966). 66. H. Blaschko, J. Physiol. London 96, 50P (1939). 67. T. Nagatsu, M. Levitt and S. Udenfriend, J. Biol. Chem. 239, 2910 (1964). 68. P. Holtz, R. Heise and K. Ludtke, Arch. Exp. Pathol. Pharmakol. 191, 87 (1938). 69. S. Kaufman and S. Friedman, Pharmacol. Rev. 27, 71 (1965). 70. R. E. Coupland, J. Endocrinology 9, 194 (1953). 71. R. J. Wurtman and J. Axelrod, J. Biol. Chem. 242, 2301 (1966). 72. R. A. Mueller, H. Thoenen and J. Axelrod, Endocrinology 86, 751 (1970). 73. R. Weinshilboum and J. Axelrod, Endocrinology $7, 894 ( 1970). 74. S. Bygdeman and U. S. von Euler, Acta Physiol. Stand. 44, 375 (1958). 75. N. Weiner and M. Rabadjija, J. Pharmacol. Exptl. Therap. 160, 61 (1968). 76. J. P. Tranzer and H. Thoenen, Expedienta 24, 115 (1968). 77. R. A. Mueller, H. Thoenen and J. Axelrod, Science 163, 468 (1969). 78. H. Thoenen, R. A. Mueller and J. Axelrod, Nature 221, 1264 ( 1969). 79. R. A. Mueller, H. Thoenen and J. Axelrod, J. Pharmacol. Exptl. Therap. 169, 74 (1969). 80. R. A. Mueller, H. Thoenen and J. Axelrod, Mol. Pharmacol. 5, 463 (1969). 81. H. Thoenen, R. -4. Mueller and J. Axelrod, J. Pharmacol. Exptl. Therap. 169, 249 (1969). 82. H. Thoenen, R. A. Mueller and J. Axelrod, Proc. Natl. Acad. Sci. 65, 58 (1970). 83. P. B. Molinoff, W. S. Brimijoin, R. M. Weinshilboum and J. Axelrod, Proc. Natl. Acad. Sci. 66, 453 (1970). 84. R. M. Weinshilboum and J. Axelrod, The Pharmacologist 12, 214 (1970). 85. H. Thoenen,, R. A. Mueller and J. Axelrod, Biochem. Pharmacol. 29, 669 (1970). 86. J. Axelrod, R. A. Mueller, J. P. Henry and P. M. Stephens, Nature 22, 1059 (1970). 87. R. Kvetnansky, V. K. Weise and I. J. Kopin, Endocrinology 87, 744 (1970). 88. J. A. Kappers, Z. Zellforsch. Mikroskop. Anat. 52, 163 (1960). 89. R. J. Wurtman, J, Axelrod and D.-E. Kelly, The Pineal, A cademic Press, New York, 1968. 90. A. B. Lerner, J. D. Case, R. V. Heinzelman, T. Amer. Chem. Sot. 81, 6084 1959. 91. H. Weissbach, B. G. Redfield and J. Axelrod, Biochem. Biophys. Acta 54, 190 (1961). 92. V. M. Fiske, K. Bryant and J. Putnam, Endocrinology 66, 489 (1960). 93. R. J. Wurtman, J. Axelrod and L. S. Phillips, Science 142, 1071 ( 1963 ). 94. R. J. Wurtman, J. Axelrod, 8. W. Ch u and J. E. Fischer, Endocrinology 75, 266 (1964). : 95. R. J. Wurtman, J. Axelrod, G. Sedvall and R. Y. Moore; J. Pharmacol. Exptl. Therap. 157, 487 (1967). 96. R. J. Wurtman, F. Larin, J. Axelrod and H. M. Shein, Nature 217, 953 (1968). 97. J. Axelrod, H. M. Shein and R. J. Wurtman, Proc. Natl. Acad. Sci. 62, 544 (1969). 98. D. C. Klein and J. Weller, Federation Proc. 29, 615 (1970). 99. 0. Loewi, Pfliiger's Arch. Ges. Physiol. 289, 239 (1921). 100. J. J. Schildkraut and S. S. Kety, Science 256, 21 (1967). 101. 0. Hornykiewicz, Pharrnacol. Rev. 18, 929 (1966). 102. J. de Champlain, L. R. Krakoff and J. Axelrod, Circulation Res. 24, Suppl. 1, 75 (1969). 208