Proc. Nati. Acad. Sci. USAVol. 73, No. 10, pp. 3398-3402,' October 1976Biochemistry

Catecholamine-stimulated cyclic GMP accumulation in the ratpineal: Apparent presynaptic site of action

(I-norepinephrine/nerve ending/cyclic AMP/denervation/a-receptor)

ROBERT F. O'DEA AND MARTIN ZATZSection on Pharmacology, Laboratory of Clinical Science, National Institute of Mental Health, Bethesda, Maryland 20014

Communicated by Julius Axelrod, July 9, 1.976

ABSTRACT Guanosine 3':5'-cyclic monophosphate (cGMP)increased 7-fold in rat pineal glands incubated in the presenceof I-norepinephrine. This response consisted of two compo-nents-one was stereospecific and inhibited by a-adrenergicantagonists while the other was not stereospecific and notreadily inhibited by antagonists. Although 1-isoproterenol wasmore potent than I-norepinephrine it had less intrinsic activityand its action was not stereospecific. The increase in cGMPcaused by these catecholamines, unlike that of adenosine 3':5'-cyclic monophosphate (cAMP), was dependent upon extra-cellular calcium. Ouabain and high levels of potassium pro-duced a marked, calcium-dependent increase in pineal cGMP,without affecting cAMP. There was no effect of cholinergicagonists on cGMP. Surgical denervation markedly reduced thecGMP response to stimulation by I-norepinephrine, potassium,or ouabain. This was in contrast to the enhanced response ofcAMP in denervated glands. The nonspecific increase in cGMPcaused by I-isoproterenol, however, was not affected by den-ervation. These data demonstrate the existence of a calcium-dependent presynaptic mechanism for the generation of cGMPwhich may be mediated by an a-adrenergic-like receptor. Inaddition, the mechanisms regulating pineal cGMP appear tobe physiologically distinct from those regulating cAMP.

The physiologic role of guanosine 3':5'-cyclic monophosphate(cGMP) has remained elusive despite numerous observationsthat several biologically active agents can increase its concen-tration in a variety of cell types (1). Adenosine 3':5'-cyclicmonophosphate (cAMP) is a second messenger in the action ofseveral hormones (2), and a similar role has been sought forcGMP. It has been proposed that cGMP functions as an inde-pendent biological mediator whose effects oppose those ofcAMP (3). Contrasting and often antagonistic regulatory in-fluences of cGMP and cAMP have been reported in severalbiological systems (3). Since the initial demonstration thatacetylcholine raises myocardial cGMP in the perfused, isolatedrat heart (4), cholinergic stimulation of cGMP accumulationhas been observed in uterine myometrium (5), brain tissue(6-8), and superior cervical ganglia (9). More recently the ad-renergic neurotransmitter, norepinephrine, has been shown toelevate cGMP in vas deferens (10) and cerebellar tissue (11).The pineal gland has proven to be a useful organ in the study

of adrenergic neurotransmission and its effects upon responsivecells (12). Located between the cerebral hemispheres, the ratpineal gland is innervated exclusively by noradrenergic nerveswhose cell bodies lie in the superior cervical ganglia (13). Betaadrenergic stimulation of the pineal leads to the induction ofserotonin N-acetyltransferase (arylamine acetyltransferase;acetyl CoA:arylamine N-acetyltransferase, EC 2.3.1.5) via acAMP-mediated mechanism (14, 15), and ultimately to thesynthesis and secretion of the hormone melatonin (14, 16).Decreased release of norepinephrine from nerve terminals

caused by prolonged exposure of the animals to light or bysurgical denervation of the gland leads to a heightened (i.e.,supersensitive) response to subsequent ,B-adrenergic stimula-tion-in terms of adenylate cyclase activation, cAMP accu-mulation, protein kinase activity, and N-acetyltransferase in-duction (17-19, M. Zatz and R. F. O'Dea, submitted for pub-lication).

Because tissue cGMP is elevated by norepinephrine (10, 11),the regulation of this cyclic nucleotide was examined in intactand denervated rat pineal glands. Our experiments demonstratethe existence of an apparent presynaptic mechanism for thegeneration of cGMP, which may be mediated by an a-adren-ergic-like receptor.

MATERIALS AND METHODSChemicals. Tritium-labeled cGMP (specific activity 10 Ci/

mmol) was purchased from Schwarz/Mann. The [3H]cGMPwas purified free of a guanosine-like contaminant by applica-tion to a Dowex AG 1X-10 (200-400 mesh) formate column.The Dowex 1-formate column was then washed and the cGMPin the eluate was eluted as previously described (20). The formicacid was removed during lyophilization and the [3H]cGMP wasdissolved in water.

Cyclic GMP and cyclic AMP were obtained from Sigma.Standard solutions of cGMP were prepared from material pu-rified as above and their concentrations were determinedspectrophotometrically utilizing an extinction coefficient of13.7 X 103 at 252 nm. Antisera to cGMP and 125I-labeled cGMPderivative (succinyl-cGMP methyltyrosine) were purchasedfrom Schwarz/Mann.

1- and d-Norepinephrine-d-bitartrate, 1- and d-isoprotere-nol-d-bitartrate, and phenylephrine hydrochloride were pro-vided by Sterling-Winthrop Laboratories. Methoxamine hy-drochloride was obtained from Burroughs-Wellcome Inc. d-and l-Propranolol hydrochloride and dl-sotalol hydrochloridewere provided by Ayerst Labs and Regis Chemical Co., re-spectively. Phenoxybenzamine hydrochloride was obtainedfrom Smith, Kline and French. Phentolamine hydrochloridewas provided by Ciba. Other drugs and chemicals were pur-chased from commercial sources.

Animals. Male Sprague-Dawley rats (150-175 g) werepurchased from Zivic-Miller Laboratories (Allison Park, Pa.).All animals were housed under diurnal lighting conditions forat least 6 days before each experiment. Animals were killedafter exposure to light for 24 hr or to darkness for 12 hr as pre-viously described (21). Rats were killed by decapitation andtheir pineal glands were removed immediately (within 15 secafter decapitation) and either homogenized in perchloric acidor placed into organ culture. Rats with denervated pineal glandswere prepared by bilateral superior cervical ganglionectomyat least 1 month prior to use. Surgically prepared and their re-

3398

Abbreviations: cGMP, guanosine 3':5'-cyclic monophosphate; cAMP,adenosine 3':5'-cyclic monophosphate

Proc. Nati. Acad. Sci. USA 73 (1976) 3399

Table 1. Specificity of the actylation radioimmunoassay for pineal cGMP

cGMP cGMP Column cGMPStandard (fmol/assay) Tissue (fmol/assay) purification (fmol/pineal)

Control cGMP 60 Control pineal extract 70 Unpurified extract 502cGMP + 25,000 fmol of cAMP 77 Pineal extract + 25,000 fmol of cAMP 86 Purified extract 581PD-treated control 1 PD-treated pineal extract 4

cGMP was assayed by acetylation radioimmunoassay in neutralized, buffered perchloric acid containing either cGMP standard or tissueextract from l-norepinephrine-stimulated glands. Both standard and tissue extract, where shown, received a 400-fold excess of cAMP toassess the crossreactivity with this cyclic nucleotide. In addition, both standard and tissue extract were preincubated with 5 sgg of purified beefheart phosphodiesterase (PD) for 60 min at 300 prior to the assay of cGMP. The values shown are the average of triplicate determinations in arepresentative experiment. cGMP was assayed also in unpurified or Dowex 1-formate column-purified pooled extract from gland tissue afterstimulation with l-norepinephrine. The values shown are the means of six separate column purifications.

spective sham-operated animals were also provided by Zivic-Miller Laboratories.

Pineal Explant Cultures. Pineal glands were placed intoorgan culture in plastic petri dishes (Falcon, 60mm diameter)containing 2.5 ml of BGJb Fitton-Jackson medium (GrandIsland Biological Co.) containing 1 mM CaCl2, supplementedwith ascorbic acid (0.1 mg/ml), glutamine (2.0 mM), strepto-mycin (100 sg/ml), penicillin (100 units/ml), and 1-norepi-nephrine or other drugs as indicated. Six pineals were incubatedin each dish at 360 under 95% 02-5% CO2.Determination of Pineal cGMP and cAMP. Individual

pineal glands (approximately 1 mg of tissue, wet weight) werehomogenized in 500 41 of ice-cold 1.5 M perchloric acid. Thehomogenates were centrifuged at 12,000 X g for 15 min andmeasured aliquots of supernatant were neutralized with 2 MKHCO3. The neutralized extract was then assayed for cGMPby the radioimmunoassay described by Steiner et al. (22)employing the acetylation modification recently described byHarper and Brooker (23). cGMP was assayed in triplicate ali-quots of the perchloric acid extract from each pineal gland.cGMP determinations were reliably within the linear portionof the standard curve (10-200 fmol) when aliquots representingapproximately Ao of a pineal gland were assayed. A logit plotanalysis was used for the samples with an interassay variationof less than +5%. Cyclic AMP was assayed according to themethod of Gilman (24).

Validation and Specificity of the cGMP Assay. The acet-ylation radioimmunoassay for cGMP accurately measures

cGMP between 10 and 200 fmol per aliquot. Under the con-

ditions used, pineal aliquots contained 8-180 fmol of cGMP.Because unstimulated glands contain 10-20 pmol (approxi-mately 10-5 M) of cAMP, it is important to verify the specificityof the assay for cGMP in the presence of a large excess of cAMP.The addition of a 400-fold excess of cAMP to either standardcGMP or to pineal extract increased the apparent level of cGMPby 15 fmol (Table 1), which is within the range of variationbetween individual glands. In other experiments, a 1000-foldexcess of cAMP did not increase the error. In both standardmixtures and tissue extract, 95% of the assayed material was

sensitive to hydrolysis by phosphodiesterase (Table 1).A further assessment of the validity of this assay for cGMP

was made by comparing the apparent levels of cGMP with andwithout purification from the pineal extracts. Cyclic GMP inpineal extracts was purified on a Dowex AG 1X-10 formatecolumn as described above. Recovery after purification variedbetween 75 and 80%. There was no difference between theapparent amounts of cGMP in purified and unpurified tissueextracts (Table 1).

These data demonstrate that the acetylation radioimmu-noassay for pineal cGMP is highly specific and that purificationof pineal extracts prior to assay is unnecessary.

RESULTSTime Course of cGMP Response to I-Norepinephrine. 1-

Norepinephrine and other adrenergic agonists, such as 1-iso-proterenol, provoke a rapid increase in rat pineal cAMP levels,which reaches a maximum after 10-20 min (15, 25). 1-Nor-epinephrine can also stimulate increases in cGMP in cerebellum(11) and ductus deferens (10). Therefore, the effects of cate-cholamines on cGMP in cultured pineals were examined. Theaddition of 10-5M l-norepinephrine produced an 8- to 10-foldincrease in cGMP within 10 min (Fig. 1). Thereafter, a rela-tively stable level, three to four times control, was maintainedfor at least 2 hr. A 60 min incubation was routinely used in laterexperiments. The basal level of pineal cGMP fell from 40-500fmol per gland immediately after decapitation to a stable levelof 150-200 fmol per gland (150 nmol/kg of tissue wet weight)in vitro after 15 min of preincubation. These levels are similar

100

50-j

z

° 20.

0 2

w

10 2 10UjDz

U2o o a~ k

INCUBATION (min)

FIG. 1. In vitro time course of pineal cGMP and cAMP afterexposure to l-norepinephrine. Pineal glands were incubated in organculture for various times as indicated. After 15 mmn preincubation,10 juM l-norepinephrine (1-NE) was added to the medium (0 time).Control incubations were also run in the absence of added drug. In-cubations were terminated at specified times by removal of glandsfrom culture and immediate homogenization in 500 1u1 of 1.5 M per-chiloric acid. Cyclic GMP (a-A) and cyclic AMP (0-0) were as-*sayed as described in Materials and Methods. Values shown are themeans aSEM of triplicate determinations in six or more glands.

Biochemistry: O'Dea and Zatz

3400 Biochemistry: O'Dea and Zatz

-j

wU 0.5.

0.1

-0j

0.1 I<

[AGONIST] mol / liter

FIG. 2. Dose-response of l-norepinephrine (0-0) and l-iso-proterenol (@-@) on pineal cGMP. After 15 min preincubation inorgan culture, pineal glands were exposed to varying concentrationsof drugs as indicated. Sixty minutes after exposure to either agonist,pineals were removed from culture, homogenized, and assayed forcGMP as described in Materials and Methods. Values shown are themeans ±SEM of triplicate determinations in six or more glands.* Signifies greater than control with P < 0.001, by Student's t-test.** Signifies greater than 10-7 1-isoproterenol with P < 0.001.

to those reported in rat or mouse cerebellum (26). cAMP dis-plays a similar time course after l-norepinephrine, although thelevels of this cyclic nucleotide tend to fall slowly after 20 minincubation (Fig. 1). Control levels of both cyclic nucleotidesremained stable after the period of preincubation.Dose Response of Pineal.cGMP to l-Norepinephrine and

I-Isoproterenol. To further characterize the catecholamine-stimulated elevation of cGMP in rat pineal glands, the effectsof both a specific fl-agonist, I-isoproterenol, and a mixed a- and,B-agonist, I-norepinephrine, were tested at various concen-trations (Fig. 2). I-Isoproterenol was 100 times more potent thanl-norepinephrine in elevating pineal cGMP. However, themaximal response to l-norepinephrine was consistently greater,

by approximately 2-fold, than the response to l-isoproterenol.This finding raised the possibility that the maximal cGMP re-sponse consisted of more than one component. a-adrenergicreceptors have been implicated in the cGMP response to cate-cholamines (10, 11). In addition, other data suggest a possiblerole of fl-adrenergic receptors in the regulation of cGMP (27).It has been observed that injection of l-isoproterenol (5 mg/kgof body weight, subcutaneously) increases rat pineal cGMP 2-to 3-fold within 15 min after administration (data not shown).The multiple components of the cGMP response in the rat pi-neal might also derive, however, from changes occurring indifferent cellular entities, e.g., pinealocytes and other cell types,or presynaptic and postsynaptic sites.

Effects of Adrenergic Agonists and Antagonists. To de-termine the roles of a- and ,B-adrenergic receptors in the ele-vation of pineal cGMP by catecholamines, we studied the ef-fects of various specific agonists and antagonists (Table 2).Exposure of incubated glands to l-norepinephrine produceda 5-fold increase in cGMP. The d-stereoisomer of norepi-nephrine produced only a 3fold elevation. Thus, approximatelyhalf of the response to l-norepinephrine is stereospecific andhalf is nonspecific. l-Isoproterenol produced a 3-fold elevation,which was also produced by d-isoproterenol. Thus, the cGMPresponse to l-isoproterenol is not stereospecific. Although highconcentrations of propranolol blocked the effect of l-norepi-nephrine, no inhibition was observed at lower doses of eitherd- or l-propranolol. dl-Sotalol, another specific ,B-adrenergicantagonist (28), failed to reduce the response to l-norepineph-rine. None of the ,B-antagonists tested affected the response tol-isoproterenol. Therefore, the response to l-isoproterenol doesnot reflect a specific ,B-adrenergic-mediated component.The specific a-blockers, phenoxybenzamine and phentol-

amine, reduced the elevation of cGMP by l-norepinephrine toa level equivalent to that produced by the noncatecholamineagonists, histamine and serotonin (Table 2). Lower doses ofphenoxybenzamine (100 nM) also inhibited the response to 1-norepinephrine. These data suggest that a stereospecific a-adrenergic-like component contributes to the response. How-ever, the a-adrenergic agonists phenylephrine and methox-amine failed to elevate pineal cGMP, even to the level achievedby the nonspecific agonists. Therefore, the stereospecific re-sponse of cGMP to l-norepinephrine cannot be attributed to theclassical postsynaptic a-adrenergic receptor.

Effects of Cholinergic Agents and Melatonin. Cholinergicagonists (4-9), have been shown to increase-cGMP in other

Table 2. Effects of agonists and antagonists on pineal cGMP

Agonists cGMP Antagonists cGMP(mol/liter) (fmol/pineal) (mol/liter) (fmol/pineal)

Control 139 ± 14 (23) l-Norepinephrine (10-5)l-Norepinephrine (10-5) 733 ± 61 (44) + dl-Propranolol (2 x 10-5) 147 ± 42 (6)d-Norepinephrine (10-5) 378 ± 38 (6) + I-Propranolol (10-7) 869 ± 162 (12)I-Isoproterenol (10-7) 440 ± 24 (23) + d-Propranolol (10-7) 613 ± 112 (11)d-Isoproterenol (10-7) 452 ± 74 (6) + dl-Sotalol (10-6) 1020 ± 274 (6)

l-Isoproterenol (10-7)Phenylephrine (10-5) 112 ± 7 (6) + I-Propranolol (10-7) 400 ± 18 (12)Methoxamine (10-4) 184 ± 18 (6) + d-Propranolol (10-7) 483 ± 22 (12)Histamine (10-4) 361 ± 15 (6) + dl-Sotalol (10-6) 559 ± 33 (6)

I-Norepinephrine (10-5)Serotonin (10-4) 319 ± 70 (12) + Phenoxybenzamine (10-5) 395 + 76 (6)Dopamine (10-5) 244 ± 30 (6) + Phentolamine (2 x 10-5) 397 i 70 (6)

Pineal glands were homogenized and cGMP assayed as described after a 60 min exposure to the various agonists listed above. Adrenergicantagonists were present in the culture media during the 15 min preincubation prior to the addition of agonist. Values shown above are the*means +SEM for triplicate determinations in the number of separate experiments indicated in parentheses.

Proc. Nati. Acad. Sci. USA 73 (1976)

Proc. Nati. Acad. Sci. USA 73 (1976) 3401

2.4

-

-i

0C-J20.8

CONTROL/-NE I-ISO KCI OUABAIN

0.01 mM 0.1 MM 80mM 0.1 mM

FIG. 3. Effect of denervation on the pineal cGMP response tocatecholamines, potassium, and ouabain. Bilateral superior cervicalganglionectomized (a) or sham-operated (1) animals were surgicallyprepared 1 month prior to use. Pineal glands were removed and placedinto organ culture. After 15 min preincubation, the various agentswere added to each group and incubations continued for an additional60 min. Individual pineals were homogenized and assayed for cGMPas described in Materials and Methods. Each value shown representsthe mean +SEM of triplicate determinations on six or more glands.1-NE, 1-norepinephrine; I-Iso, 1-isoproterenol. Signifies greater thandenervated with P < 0.01, by Student's t-test. ** Signifies greaterthan denervated with P < 0.001. +.Signifies greater than denervatedcontrol with P < 0.001.

tissues. Despite the lack of cholinergic input to the rat pineal(29), the effects of acetylcholine and carbamlycholine wereexamined. Neither compound (at 10 AM) increased pinealcGMP. Cholinergic agonists also failed to increase cGMP inmouse brain (30). Melatonin, which raises cGMP levels in pe-ripheral blood monocytes (31), also failed to affect cGMP levelsin rat pineal.

Effects of Ouabain, Potassium, and Calcium. Calcium-dependent increases in cGMP have been shown in response toI-norepinephrine, ouabain, and high concentrations of potas-sium in brain tissue (30). Incubation of intact rat pineals with1 mM ouabain or 80 mM KC1 produced marked increases incGMP levels (Fig. 3). Although both agents increase cAMP inbrain slices (30, 32), neither compound affected pineal cAMP(ref. 33; data not shown). There was negligible response ofcGMP to l-norepinephrine, ouabain, or 80mM KCI in the ab-sence of calcium (data not shown). In contrast, the response ofpineal cAMP to l-norepinephrine was not blocked by the ab-sence of calcium. In addition, 10MM phentolamine completelyblocked the increase in cGMP observed after high potassium.Neither K+-free or Ca++-free medium altered the basal levelsof cGMP. Therefore, in contrast to pineal cAMP, the elevationof pineal cGMP by all agents examined is Ca++-dependent.

Effects of Denervation on Pineal cGMP. The possibilitythat the a-adrenergic-like and nonspecific components of thecGMP response to l-norepinephrine reside in discrete cellularcompartments was examined by comparing responses in intactand denervated glands. The rat pineal gland is innervated bypostganglionic sympathetic fibers originating in the superiorcervical ganglia (13). Bilateral ganglionectomy results in thedegeneration of the presynaptic nerve terminals and the de-velopment of a supersensitive response to ,6-adrenergic stimu-lation in terms of adenylate cyclase activation (17), cAMP ac-

cumulation (18), and protein kinase activity (M. Zatz and R. F.O'Dea, submitted for publication). In contrast, the accumula-tion of cGMP in response to l-norepinephrine is markedly re-duced after ganglionectomy (Fig. 3). Indeed, the response isquantitatively identical to the nonspecific effect of d-norepi-nephrine (Table 2) or of l-isoproterenol (Fig. 3). These dataindicate that the stereospecific, a-adrenergic-like componentof the noradrenergic stimulation of cGMP depends upon thepresence of intact nerve terminals. Furthermore, the responseof cGMP to high concentrations of potassium or to ouabain isabolished after ganglionectomy (Fig. 3). Thus, the pineal nerveterminals may be a site of cGMP accumulation.

DISCUSSIONCatecholamines elevate cGMP in cultured rat pineal glands.I-Norepinephrine has a similar effect in ductus deferens (10)and cerebellum (11). The accumulation of cGMP in these tissuesappears to be related to the stimulation of the a-adrenergicreceptor. In the pineal, examination of the dose-dependenteffects of l-norepinephrine and l-isoproterenol suggested thatboth a- and fl-adrenergic receptors might have a role in theregulation of cGMP. However, experiments with catecholaminestereoisomers and ,3-antagonists ruled against the participationof the stereospecific ,B-receptor. The effect of l-isoproterenol,like that of certain other compounds, appears to be nonspecific.In contrast to l-isoproterenol, l-norepinephrine-affected boththe nonspecific and another, stereosp-ecific, component. Thisstereospecific component is susceptibli to blockade by a-ad-renergic antagonists. Thus, as in certain other tissues, cGMP inthe pineal gland responds to catecholamines via an a-adren-ergic-like mechanism.The possibility that the nonspecific and stereospecific com-

ponents reflect pre- and postsynaptic compartments was ex-amined by comparing the increases in cGMP accumulation indenervated and sham-operated pineals. The nonspecificcomponent was unaffected by denervation. It appears that thiscomponent does not require intact nerve endings and, therefore,may be postsynaptic.

In contrast to cGMP, the accumulation of cAMP in responseto noradrenergic stimulation is greatly enhanced by denervation(18). Thus, postsynaptic adenylate cyclase, which is- closelylinked to the fl-adrenergic receptor (21), becomes supersensitiveafter denervation (17). In addition, the cAMP response to 1-norepinephrine is enhanced after prolonged exposure to light(18), whereas alteration in lighting had no effect on either basalor stimulated cGMP levels. Furthermore, 1 mM dibutyrylcGMP had no effect on N-acetyltransferase induction in tro,under several assay conditions (R. F. O'Dea and M. Zatz, un-published observation). Therefore, the mechanisms whichenhance the sensitivity of the postsynaptic ,B-adrenergic re-ceptor do not participate in the regulation of cGMP, nor doescGMP appear to influence the function of this receptor.The stereospecific component of the response of cGMP to

l-norepinephrine was abolished by denervation, as was theresponse to high potassium and ouabain. Thus, the stereospe-cific, a-adrenergic-like component in the generation of cGMPdepends upon the presence of intact nerve endings. A presyn-aptic role for cAMP also appears likely in view of the recentreport that it can activate tyrosine hydroxylase in synaptomalfractions prepared from rat brain (34).High potassium and ouabain were as effective as l-norepi-

nephrine in elevating cGMP. Although both agents producedepolarization of nervous tissue, each does so by a differentmechanism. High levels of K+ depolarize by decreasing thedifference between intracellular and extracellular K+ con-

4.2~~~ ~~I.*AnAMIk6I V

Biochemistry: O'Dea and Zatz

3402 Biochemistry: O'Dea and Zatz

centrations (35) whereas ouabain inhibits (Na+- K+)adenosinetriphosphatase, thereby preventing extrusion of Na+ from cells(36). Since it has been reported that l-norepinephrine hyper-polarizes pinealocyte membranes, whereas potassium andouabain reverse this effect (33), l-norepinephrine and the"depolarizing" agents may elevate cGMP via dissimilarmechanisms. However, the observation that each agent requiresextracellular calcium and is inhibited by a-adrenergic antag-onists, suggests that they may stimulate the generation of cGMPthrough a common mechanism. It is tempting to speculate thatthe presynaptic generation and action of cGMP might be re-lated functionally to the presynaptic a-adrenergic receptorwhich modulates neurotransmitter release (37).The response of cGMP to l-norepinephrine in the pineal

gland depends on extracellular calcium and consists of a ster-eospecific, presynaptic, a-adrenergic-like component and asecond component which appears to be nonspecific, postsy-naptic, and unrelated to either adrenergic receptor. The datapresented here may serve to clarify some of the complexitiesobserved in cGMP regulation in other nervous tissue (11) andelsewhere (1). In conclusion, this report provides evidence fora catecholamine-sensitive mechanism for the generation ofcGMP, which is dependent upon intact nerve terminals.

We are grateful to Dr. J. Axelrod for stimulating discussions on thiswork and for helpful criticism of the manuscript. Both authors areResearch Associates in the Pharmacology-Toxicology Program of theNational Institute of General Medical Sciences.

1. Goldberg, N. D., O'Dea, R. F. & Haddox, M. K. (1973) Adv.Cyclic Nucleotide Res. 3, 156-213.

2. Robison, G. A., Butcher, R. W. & Sutherland, E. W. (1971) CyclicAMP (Academic Press, New York).

3. Goldberg, N. D., Haddox, M. K., Dunham, E., Lopez, C. &Hadden, J. W. (1974) in the Cold Spring Harbor Symposium onthe Regulation of Proliferation in Animal Cells, eds. Clarkson,B. & Baserga, R. (Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.), pp. 609-625.

4. George, W. J., Polson, J. B., O'Toole, A. G. & Goldberg, N. D.(1970) Proc. Natl. Acad. Sci. USA 66,398-403.

5. Goldberg, N. D., Haddox, M. K., Nicol, S. E., Glass, D. B., San-ford, C. H., Kuehl, F. A. & Estensen, R. (1975) Adv. Cyclic Nu-cleotide Res. 5, 307-330.

6. Ferrendelli, J. A., Steiner, A. L., McDougal, D. B. & Kipnis, D.M. (1970) Biochem. Biophys. Res. Commun. 41,1061-1067.

7. Kuo,. J. F., Lee, T. P., Reyes, P. L., Walton, K. G., Donnelly, T.E. & Greengard, P. (1972) J. Biol. Chem. 247, 16-22.

8. Lee, T. P., Kuo, J. F. & Greengard, P. (1972) Proc. Natl. Acad.Sci. USA 69,3287-3291.

9. Kebabian, J. W., Bloom, F. E., Steiner, A. L. & Greengard, P.(1975) Science 190, 157-159.

10. Schultz, G., Hardman, J. G., Schultz, K., Davis, J. W. & Suther-land, E. W. (1973) Proc. Natl. Acad. Sci. USA 70, 1721-1725.

11. Ferrendelli, J. A., Kinscherf, D. A. & Chang, M. M. (1975) BrainRes. 84, 63-73.

12. Axelrod, J. (1974) Science 184, 1341-1348.13. Kappers, J. A. (1960) Z. Zellforsch. Mikrosk. Anat. 52, 163-

215.14. Klein, D. C., Berg, R. & Weller, J. C. (1970) Science 168,979-

980.15. Deguchi, T. (1972) Mol. Pharmacol. 9, 184-190.16. Axelrod, J., Shein, H. M. & Wurtman, R. J. (1969) Proc. Natl.

Acad. Sci. USA 62,544-549.17. Weiss, B. & Strada, S. J. (1972) Adv. Cyclic Nucleotide Res. 1,

357-374.18. Deguchi, T. & Axelrod, J. (1973) Mol. Pharmacol. 9,612-618.19. Deguchi, T. & Axelrod, J. (1972) Proc. Natl. Acad. Sci. USA 69,

2208-2211.20. O'Dea, R. F., Bodley, J. W., Lin, L., Haddox, M. K. & Goldberg,

N. D. (1974) in Methods in Enzymology, eds. Hardman, J. G.& O'Malley, B. W. (Academic Press, New York), Vol. 38C, pp.85-90.

21. Kebabian, J. W., Zatz, M., Romero, J. & Axelrod, J. (1975) Proc.Natl. Acad. Sci. USA 72,3735-3739.

22. Steiner, A. L., Parker, C. W. & Kipnis, D. M. (1972) J. Biol. Chem.247, 1106-1113.

23. Harper, J. & Brooker, G. (1975) J. Cyclic Nucleotide Res. 1,207-218.

24. Gilman, A. G. (1970) Proc. Natl. Acad. Sci. USA 67,305-312.25. Romero, J. & Axelrod, J. (1975) Proc. Natl. Acad. Sci. USA 72,

1661-1665.26. Kinscherf, D. A., Chang, M. M., Rubin, E. H., Schneider, D. R.

& Ferrendelli, J. A. (1976) J. Neurochem. 26,527-530.27. Durham, J. P., Baserga, R. & Butcher, F. R. (1974) Biochim.

Biophys. Acta 372, 196-217.28. Mukherjee, C., Caron, M. G., Coverstone, M. & Lefkowitz, R.

J. (1975) J. Biol. Chem. 250,4869-4876.29. Machado, A. B. M. & Lemos, J. P. J. (1971) J. Neuro-Visc. Relat.

32, 104-111.30. Ferrendelli, J. A., Kinscherf, D. A. & Chang, M. M. (1973) Mol.

Pharmacol. 9, 445-454.31. Sandler, J. A., Clyman, R. I., Manganiello, V. C. & Vaughn, M.

(1975) J. Clin. Invest. 55, 431-435.32. Shimizu, H., Takenoshita, M., Huang, M. & Daly, J. W. (1973)

J. Neurochem. 20,91-95.33. Parfitt, A., Weller, J. L., Klein, D. C., Sakai, K. K. & Marks, B.

H. (1975) Mol. Pharmacol. 11, 241-255.34. Harris, J. E., Baldessarini, R. J., Morgenrath, V. H. & Roth, R. J.

(1975) Proc. Natl. Acad. Sci. USA 72,789-793.35. Woodbury, J. W. (1965) in Physiology and Biophysics, eds. Ruch,

T. C. & Patton, H. D. (Saunders, Philadelphia, Pa.), pp. 26-58.36. Albers, R. W. (1972) in Basic Neurochemistry, eds. Albers, R. W.,

Siegel, G. J., Katzman, R. & Agranoff, B. W. (Little, Brown andCo., Boston), pp. 5-40.

37. Langer, S. Z. (1974) Biochem. Pharmacol. 23, 1793-1800.

Proc. Natl. Acad. Sci. USA 73 (1976)