Potentiation of nicotinic transmission in the rat superior cervical sympathetic ganglion: effects of cyclic GMP and nitric oxide generators

Brain Research, 573 (1992) 139-146 ~) 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$05.00

139

BRES 17457

Potentiation of nicotinic transmission in the rat superior cervical sympathetic ganglion: effects of cyclic GMP and nitric oxide generators Clark A. Briggs Neuroscience Research, Abbott Laboratories, Abbott Park, IL 60064 (U.S.A.) (Accepted 8 October 1991) Key words: Sympathetic ganglion; Guanosine cyclic monophosphate; Adenosine cyclic monophosphate; Neural transmission; Endothelium-derived relaxant factor; Nitric oxide; Synaptic potentiation

The efficacy of nicotinic transmission in the rat superior cervical ganglion in vitro (24-26°C) was estimated by extracellular recording of the postganglionic compound action potential response to stimulation of the preganglionic nerve at a slow rate (one shock every 60 s). Atropine (2/~M) was included to block muscarinic transmission, and hexamethonium (200-250/~M) was used to produce a submaximal response sensitive to potentiation and inhibition of nicotinic transmission. Upon exposure to 1-100 ~tM 8-bromo-guanosine 3',5'-cyclic monophosphate (8-Br-cGMP), nicotinic transmission was potentiated by 6 + 1% (n = 4) to 89 + 5% (n = 5) in a dose-dependent manner. 8-Bromo-adenosine 3',5'-cyclic monophosphate (8-Br-cAMP, 10-100/~M) also potentiated nicotinic transmission (3.8 + 0.3% (n = 3) to 43 + 4% (n = 3)). However, 8-Br-cGMP was at least 2-fold more effective than 8-Br-cAMP. Sodium nitroprusside (0.1 ~M to 1 mM) and sodium azide (0.1-100/~M) were used to stimulate the formation of endogenous cGMP52. Nicotinic transmission was potentiated by these substances also. The response was increased by 3.4 + 0.7% (n = 4) to 32 + 2% (n = 5) upon exposure to 0.1-100 ~M sodium nitroprusside, and by 5.5 + 0.9% (n = 3) to 18 + 4% (n = 4) upon exposure to 0.1-100/zM sodium azide. Ferricyanide ion (10-100/~M) appeared to be ineffective, as would be expected if the effect of nitroprusside was due to the nitric oxide rather than the cyanide or ferric moieties. A small increase in the response was observed when the extracellular concentration of K+ was increased by about 300/tM (100/zM potassium ferricyanide or 300/~M KC1). These data are consistent with the hypothesis that cGMP may mediate a potentiation of nicotinic synaptic transmission and/or increased neuronal excitability in the sympathetic ganglion, and that this system could be activated by local production of a substance similar to endothelium-derived relaxant factor. INTRODUCTION While it has been k n o w n for many years that guanosine 3',5'-cyclic m o n o p h o s p h a t e (cGMP) is synthesized in the central and peripheral nervous systems and that the formation of c G M P increases with stimulation, the roles and mechanisms of c G M P in regulating n e u r o n a l activity are not yet clearly defined. Interest in this area has increased recently since the finding that activation of the N-methyl-D-aspartate ( N M D A ) class of glutamate receptors in the brain may stimulate the formation of c G M P through a system rather similar to that of the endothelium-derived relaxant factor ( E D R F ) or nitric oxide (NO) system in the vasculature 2°'22'23'4°'47. However, the activation of c G M P formation is not necessarily limited to the activation of N M D A receptors. Presumably, other processes which cause a sufficient increase in the intracellular concentration of free Ca 2÷, through influx or intracellular release, could activate the Ca2+-depen dent synthesis of E D R F / N O . Additionally, c G M P synthesis can be stimulated through a receptor-mediated ac-

tivation of particulate guanylate cyclase 16'24'45. In the superior cervical ganglion (SCG), cGMP formation can be activated by synaptic activity elicited by electrical stimulation of the preganglionic nerve for 1 min or less 2'13'19'52-54'58 as well as by application of muscarinic agonists 15'19'27'28'54'57. Furthermore, this c G M P is localized in the sympathetic neurons 2'15'27 and possibly also in preganglionic nerve terminals L43'44,52. In the brain, guanylate cyclase, c G M P and c G M P - d e p e n d e n t protein kinase also have been localized in neurons3-5'41' 55. It seems likely that c G M P is important in the regulation of neuronal function in the peripheral and central nervous systems. O n e proposed role for c G M P is as an intracellular mediator for the muscarinic activation of a slow excitatory synaptic potential (slow-EPSP) in the rabbit SCG 25' 26,29,33. However, the ionic and molecular mechanisms have not been completely identified, and there has been some discussion about how closely cGMP mimics the muscarinic action in m a m m a l i a n and amphibian sympathetic ganglia 18'26,59. Additionally, it seems likely that

Correspondence: C.A. Briggs, Neuroscience Research, Abbott Laboratories, Abbott Park, IL 60064, U.S.A.

140 there are other roles for cGMP in sympathetic ganglia. Muscarinic transmission accounts for only a portion of the cGMP response to stimulation of the preganglionic nerve l'x3A9,S2-54. The formation of cGMP in the SCG is also increased by atrial natriuretic factor 16'24 which stimulates particulate guanylate cyclase, and by nitroprusside and azide 1'2'43'44'52 which stimulate soluble guanylate cyclase 4°'56. Atrial natriuretic factor has been found to inhibit M-channel currents in the frog sympathetic ganglion but not the rat SCG42; it is not known whether this action is mediated by cGMP. The electrophysiological effects of nitroprusside and azide do not appear to have been investigated in the sympathetic ganglion, apart from the use of millimolar concentrations of azide as a metabolic inhibitor 3°'37. Most studies of cGMP in sympathetic ganglia have used dibutyryl-cGMP. This analogue will depolarize postganglionic sympathetic neurons 17'18,21'eS'a6, an action that would be expected to affect nicotinic synaptic transmission. However, in the frog sympathetic ganglion, dibutyryl-cGMP was found to have no effect on low-frequency nicotinic transmission while it appeared to inhibit the formation of post-tetanic potentiation 48. In the rat SCG, we also found that monobutyryl- and dibutyrylcGMP (1 mM) did not alter low-frequency nicotinic transmission, but 8-Br-cGMP (1 mM) did cause a longlasting potentiation of the same response 12. Nevertheless, we questioned the specificity of the 8-Br-cGMP effect because of the discrepant results with the different cGMP analogues compared with the uniform ability of all of cAMP analogues (monobutyryl-, dibutyryl-and 8-Br) and forskolin to cause a potentiation of nicotinic transmission. To test the possibility that 8-Br-cGMP may have acted non-specifically by stimulating a cAMP-dependent system, the response to low concentrations of 8-Br-cGMP was investigated and the apparent potency of 8-Br-cGMP was compared to that of 8-Br-cAMP. Additionally, nitroprusside and azide were used to stimulate the formation of endogenous cGMP. The results argue against the suggestion that the potentiation caused by 8-Br-cGMP was a non-specific effect. Instead, the data are consistent with the hypotheses that cGMP can mediate a potentiation of nicotinic synaptic transmission and/or neuronal excitability, and that sufficient cGMP could be produced through activation of soluble guanylate cyclase in the sympathetic ganglion. A preliminary account of these studies has been published as an abstract 7. MATERIALS AND METHODS

Physiological saline solutions Ganglia were maintained in vitro at ambient temperature (2426°C) and superfused (1 ml/min) with modified Locke's solution

continuously gassed with 95% 02-5% CO 2. The modified Lockes solution contained 10 mM dextrose and 10 BM choline chloride in addition to the following salts (in mM): NaC1 136, KCI 5.6, CaCI 2 2.2, MgCI 2 1.2, NaHePO 4 1.2 and NaHCO 3 20.

Tissue preparation Superior cervical ganglia from male Sprague-Dawley rats (170250 g) were isolated and prepared for in vitro extracellular recording essentially as described previously9. Briefly, rats were anesthetized with CO 2 (20% in O2), decapitated, and the ganglia removed from the neck and immediately superfused with oxygenated modified Locke's solution (1 ml/min). The ganglia were then desheathed and placed in a second chamber with superfusion of the modified Locke's solution. One ganglion was mounted in bipolar suction electrodes for cxtracellular stimulation of the preganglionic cervical sympathetic nerve and extracellular recording of the postganglionic action potential response from the internal carotid nerve.

Experimental protocols The ganglia were continuously stimulated with single supramaximal preganglionic shocks (isolated 0.5 ms rectangular pulse) at a slow rate, one per 60 s (0.017 Hz). Each shock elicited a single postganglionic compound action potential which was recorded through a high-impedance differential amplifier and digitized (0.1 Hz and 3 kHz low and high filters; ~<5/~V resolution). The amplitude of each response was determined on-line using a custom program, and the results were stored on magnetic media for further analysis. The computer was programmed to display the waveform of each response and to mark the baseline, peak amplitude, beginning of the response, and end of the response as determined by the on-line analysis. The accuracy and precision of the analysis was monitored on one computer display, while the computed response amplitudes were plotted as a function of time on another display. Additionally, computed peak amplitudes were compared to the response displayed on an independent analog oscilloscope. After determining the maximal response and its stability, hexamethonium (200-250/xM) and atropine (2 #M) were added to the superfusion solution and the ganglion was allowed to stabilize under these conditions (at least 90 min) before the addition of any test drug (8-Br-cGMP, nitroprusside, etc.). The purpose of the atropine was to block muscarinic transmission, and the purpose of the hexamethonium was to partially inhibit nicotinic transmission so that the postganglionic response was sensitive to both inhibition and potentiation of nicotinic transmission while the preganglionic nerve was stimulated supramaximally for better stability9. In the present experiments, the postganglionic response amplitude was inhibited by 47 +_ 1% (mean + S.E.M., n = 35) 35 min after the addition of hexamethonium (200 to 250 ~M) and atropine (2/~M). The ganglion was exposed to a test drug only after the control postganglionic responses were stable and predictable for at least 30 min. The superfusion was switched temporarily to the drug-containing solution while stimulation of the preganglionic nerve remained continuous (0.017 Hz). Each drug was applied for 30 min to allow diffusion of the drug through the ganglion. In previous experiments, small molecules such as calcium, acetylcholine and hexamethonium approached near-equilibrium after about 15-20 min. In the present experiments, the initial exponential time constant by which hexamethonium inhibited nicotinic transmission was found to be 5.1 _+ 0.4 min (n = 18). This would suggest that the average extracellular concentration of hexamethonium in the ganglion would be about 95% of the bath concentration after a 15-rain exposure and 99.7% after a 30-min exposure, with simplifying assumptions regarding the action of hexamethonium and the ganglionic inputoutput function.

Data analysis and presentation The postganglionic compound action potential response amplitudes were plotted as a function of time, and a least-squares regression line was fitted to the control responses and projected to predict the control amplitude during and after exposure to the test

141 1.o - I E ~,

1.iii~' 21.0 , , i , - , i~-T , , I , ,mI

1.0 i!! 1.2 ~ ' ' ~.~ I I M~L~B~_jB I I ' ' Ir _~IM_I ' ' I ' ' I I~

~

~L , I i L I i i I L t I I , L I L -DO -30 0 30 60 90 120 ~E

0'4

0

30

60

0

z~ 0.4

i i I , , i , , i , , i , i i , i

O

z 0 0.0 w -60 -30

0

30

60

90

120

60

30

0

30

60

J

I

0.8

~ 0.2

1.2

I

0.6

10 pM B-Dr-cAMP t ii i ii ii ii , i i i

-60 -30 2,2

~

'i!

90

i 1

10 100 [CYCLIC NUCLEOTIDE], /zM

120

TIME, rninufes Fig. 1. Potentiation of ganglionic transmission by 8-BR-cGMP and 8-Br-cAMP. The amplitude of each postganglionic compound action potential response to supramaximal stimulation of the preganglionic nerve (0.017 Hz) is plotted as a function of time. Data shown are subsequent to the addition of 200-250 HM hexamethonium to partially inhibit nicotinic synaptic transmission and 2/~M atropine to block muscarinic transmission. Responses were increased by a 30-min exposure to 8-Br-cGMP (3 /zM, upper left panel; 10/~M, lower left panel), or to 8-Dr-cAMP (10 #M, upper right panel; 30/~M, lower right panel). The time of drug application is indicated by a horizontal bar in each panel. Note that the ordinate minima are not 0. Data in each panel are from different ganglia. Similar results were obtained in other experiments (see Fig. 2).

drug 12. The change in response amplitude caused by the drug was computed according to the equation I t = (Vt - V¢)/Vc where It is the increment or fractional change in the response at time t after beginning the drug application, Vt is the response amplitude at that time, and Vc is the projected control response amplitude at that time. Data averaged among experiments are presented as mean + S.E.M., with 'n' representing the number of experiments (preparations). Materials

Atropine sulfate, 8-bromoadenosine 3',5'-cyclic monophosphate (8-Br-cAMP, sodium salt), 8-bromoguanosine 3',5'-cyclic monophosphate (8-Br-cGMP, sodium salt), choline chloride, hexamethonium chloride, sodium azide (NAN3), and sodium nitroprusside (Na2Fe(CN)DNO) were purchased from Sigma Chemical Company (St. Louis, MO). Dextrose was purchased from EM Science (Cherry Hill, NJ). Potassium ferricyanide (K3Fe(CN)6) was purchased from JT Baker Chemical Company (Phillipsburg, NJ). Stock solutions of the experimental compounds were made fresh daily in distilled water, kept on ice, protected from light, and diluted at least 1000-fold into the modified Locke's superfusion solution within 2 min before application.

RESULTS

Cyclic nucleotide analogues 8-Br-cGMP elicited a potentiation of transmission through the superior cervical ganglion (SCG) at all concentrations tested. When applied for 30 min, micromolar concentrations of 8-Br-cGMP elicited a clear and re-

Fig. 2. Dose-response relationships for the potentiations caused by 8-Br-cGMP and 8-Dr-cAMP. Data are from experiments similar to those shown in Fig. 1. The potentiation caused by the cyclic nucleotide was measured for each response as a fractional change relative to a linear projection of the control response amplitudes (It, see Methods). This increase was averaged over 5-min periods (5 responses) surrounding 15, 30, 45, 60 and 90 min after the start of cyclic nucleotide application. The increase was maximal at about 45 min (145) for both cyclic nucleotides, and is plotted here. Each point represents the mean +S.E.M. for 4 experiments (1 ~M 8-BrcGMP) or 3 experiments (all others). For some points, the standard error bars are within the symbol.

producible increase in the evoked postganglionic response (Fig. 1). Indeed, concentrations as low as 0.1 HM 8-Br-cGMP elicited a slight but clear increase in two experiments (not shown). 8-Br-cAMP also elicited a potentiation of nicotinic transmission (Fig. 1). However, 8-Br-cGMP was more effective than 8-Dr-cAMP at similar concentrations. In 4 experiments, each SCG was exposed for 30 min to 10 HM 8-Br-cGMP and to 10 HM 8-Br-cAMP separately. The potentiation caused by 8-Br-cGMP was 3.1 + 0.3 fold larger than the potentiation caused by the same concentration of 8-Dr-cAMP in the same ganglia. The dose-response relationships for 8-Br-cGMP and 8-Br-cAMP (Fig. 2) were constructed in separate ganglia because of the possibility that one cyclic nucleotide could affect the response to another cyclic nucleotide 29'33. Each ganglion was exposed to either 8-Br-cGMP at various concentrations or to 8-Br-cAMP at various concentrations, but not to both cyclic nucleotides. Full dose-response curves were not constructed because at concentrations of 8-Br-cGMP above 100 HM the apparent potentiation could have been limited by attaining the maximum postganglionic response, rather than by attaining a maximum in the potentiation itself. Nevertheless, the data shown in Fig. 2 indicate that 3-100 HM 8-BrcGMP was at least 2-fold more effective than a similar concentration of 8-Br-cAMP in causing a potentiation of nicotinic transmission through the SCG. Upon the completion of these experiments, the re-

142

ta.I

0o°L...............t --

o.2

0,1 J rl

0 0

-

'

I

-30

'

'

I

0

I

I

I

30

'

'

I

60

'

L~

0.5

0.4

w

-

I

0.2

'

I

L

I

-

0

NITROPRUSSIDE

_

•

AZIDE

i

I

'

I

T-

FERRICYANIDE

-

(_9

. . . . . . . . . . . . . . . . .

~

-

0.3

0.1

1 ~M NITROPRUS$10[ '

-60

"---~-

o

0 '

I

90

'

0.O

120

-60

-30

0

,30

60

90

120

Z 0

0.2

I-'-

(/')

1.2

0

1.2

Z

0

1.0

1.0

I:1. (/1

0.s

O.B

0.0

I,I

r~

100 M,M AZIOE

o.8 -60

0.1

t",," LI-

_9",

,

0.t

1

0

30

60

90

120

TIME,

-60

, ,. . ,. .

10

....

,,

100

,

, 1000

CONCENTRATION, ~M

0.6 -30

,

-30

0

30

60

90

120

rninutes

Fig. 3. Potentiation of ganglionic transmission by sodium nitroprusside and sodium azide. Data were obtained and are shown in a manner similar to Fig. 1. The results for 1 and 100 MM nitroprusside (top) are from one ganglion, while the results for 1 MM and 100 azide (bottom) are from a second ganglion. Note that the ordinate scaling and origins of the two top panels differ from those of the two bottom panels.

maining 8 - B r - c A M P and 8-Br-cGMP were analyzed chemically by tH and ]3C nuclear magnetic resonance, F A B mass spectroscopy, and combustion analysis (Dr. James Kerwin, d a t a not shown). Both compounds were found to be intact and t>95% pure. Thus, the lesser potency of 8 - B r - c A M P c o m p a r e d to 8 - B r - c G M P did not a p p e a r to be due to a difference in the integrity of the materials.

Stimulation of cyclic GMP synthesis Sodium nitroprusside and sodium azide were used to stimulate the synthesis of endogenous cGMP, presumably through the activation of soluble guanylate cyclase. The formation of endogenous c G M P in the rat superior cervical ganglion can be stimulated by 6-22 fold by 1-100 ttM sodium nitroprusside, and 6-12 fold by 10 tiM to 1 m M sodium azide 2'43'44. As shown in Fig. 3 both nitroprusside and azide also elicited a potentiation of nicotinic transmission. F u r t h e r m o r e , each of these comounds affected nicotinic transmission through the S C G at concentrations within the range found to stimulate c G M P formation in the SCG. Fig. 4 shows the d o s e - r e s p o n s e curves for the potentiations elicited by nitroprusside and azide. Each comp o u n d elicited small but clear potentiations at concentrations as low as 0.1 MM. A t concentrations <~1 MM, azide and nitroprusside a p p e a r e d to be approximately equi-effective. A t concentrations >1 # M , nitroprusside a p p e a r e d to be m o r e effective than azide. H o w e v e r , the effects of azide were variable, with ranges of 0.0960.218 at 1 ,uM (n = 5), 0.085-0.222 at 10 MM (n = 5), and 0.118-0.267 at 100 # M (n = 4). In stimulating

Fig. 4. Dose-response relationships for sodium nitroprusside and sodium azide, with potassium ferricyanide and potassium chloride controls. Data were computed and are plotted in a manner similar to that of Fig. 3 except that the fractional increase at 30 min (/3o) rather than at 45 min (145) is shown. 130 was maximal for azide at all concentrations, and for nitroprusside at concentrations <_100 pM. For 1 mM nitroprusside, I45 was 0.47 _+ 0.04 while/3o was 0.37 +_ 0.04 (n = 4). For 100 ~M nitroprusside, 145 was 0.34 _+ 0.03 and 130 was 0.32 + 0.02 (n = 5); for 100/~M azide, these values were 20 _+ 4 and 18 + 4 (n = 4), respectively. Each point shows the mean -+S.E.M. for 3-5 experiments except for 1 and 10 pM nitroprusside (n = 7 and 10, respectively).

c G M P formation in the SCG, nitroprusside has been stated to be about 3-fold more effective than azide at concentrations > 1 ~tM 43. As a control for the assumption that the effect of sodium nitroprusside (Na2Fe(CN)sNO) was due to its ability to generate nitric oxide, 3 ganglia were exposed to potassium ferricyanide (K3Fe(CN)6) and elevated potassium chloride (KC1) as well as to sodium nitroprusside. A s shown in Fig. 4, potassium ferricyanide had very little effect c o m p a r e d to nitroprusside. A slight potentiation was detected with 100 MM potassium ferricyanide. H o w e v e r , this may have been due to the slightly elevated concentration of K ÷, because 300 ttM potassium chloride p r o d u c e d a similar, small potentiation (300 # M KC1 and 100 MM K3Fe(CN)6 would increase [K ÷] by the same amount, assuming both salts are equally dissociated). Thus, the results were consistent with the idea that the potentiation caused by nitroprusside was due to its ability to generate nitric oxide. DISCUSSION

In the superior cervical ganglion (SCG), the content of c G M P is increased several fold, in a CaZ+-dependent manner, by tetanic stimulation of the preganglionic nerve 13'19'52-54'58. F u r t h e r m o r e , immunohistochemical studies indicate that this c G M P is preferentially localized in postganglionic neurons as o p p o s e d to non-neuronal cells 2'15'27. Additionally, it has been suggested that a portion of the c G M P response may be localized in the

143 preganglionic nerve terminals, based upon studies with denervated ganglia 1'43'44. cGMP has been reported to stimulate protein phosphorylation with greater efficacy than cAMP in the rabbit SCG, although the proteins phosphorylated differed depending upon whether cGMP was applied to homogenates or the intact ganglion 49. While it seems likely that cGMP is important in neuronal function in the SCG, the role(s) of cGMP in the regulation of synaptic transmission and neuronal excitability are not entirely clear. In previous experiments, a potentiation of nicotinic transmission was observed following exposure of the rat SCG to 1 mM 8-Br-cGMP 12. However, at that time it was not clear whether this represented a cGMP-mediated process. Another cGMP analog, 1 mM dibutyrylcGMP, did n o t cause a potentiation of nicotinic transmission. This discrepancy between cGMP analogues contrasted with the uniform ability of similar cAMP analogues and forskolin to cause a potentiation of nicotinic transmission. In the present experiments, 8-Br-cGMP was found to be effective in micromolar concentrations, and to be several-fold more potent than 8-Br-cAMP in potentiating nicotinic transmission. This is the opposite of what one would expect if 8-Br-cGMP acted non-specifically as an analogue of cAMP, and suggests that the potentiation caused by 8-Br-cGMP was not due to a non-specific activation of the cAMP system. However, the results do not exclude the possibility that cGMP may act in concert with c A M E for example by inhibiting the hydrolysis of cAMP 33'35 or by acting separately upon a common intracellular process. It is not clear why dibutyryl-cGMP does not appear to cause a potentiation of nicotinic transmission while 8-Br-cGMP does. Several assumptions are often made in comparing the effects of cyclic nucleotide analogues applied to intact tissue: (1) that the analogues are chemically intact; (2) that the analogues diffuse into and out from the tissue equally well; (3) that the analogue acts intracellularly as a cyclic nucleotide mimic, not for example non-specifically upon an extracellular receptor; (4) that the analogues are equally permeant across the cell membranes; (5) that the catabolism of the analogues is similar; (6) that the analogues are pharmacologically similar with respect to their activation of signal transduction processes. The first assumption was addressed by chemical analysis of the 8-Br-cAMP and 8-Br-cGMP used in the present experiments. The lesser potency of 8-Br-cAMP compared to 8-Br-cGMP could not be explained by degradation of the lyophilized material. Additionally, the rate at which the induced potentiations appeared and reversed would suggest that 8-Br-cAMP diffused into

and out from the SCG as well as did 8-Br-cGMP. However, we know of no specific information regarding the movement and metabolism of these cyclic nucleotide analogues in this tissue. Additional biochemical and pharmacological experiments are needed to address the relative efficacy, potency and specificity of the various cGMP analogues in the SCG. Nitroprusside and azide provide a means of increasing the intracellular content of c G M P 1'2'5'43'44'52'56, seemingly with better selectivity than preganglionic nerve stimulation which would also activate cAMP synthesis and phosphatidylinositol hydrolysis 2'9'13'14'36'53'59'6°. Both nitroprusside and azide, like 8-Br-cGMP, elicited a potentiation of nicotinic transmission. Furthermore, both nitroprusside and azide were effective at concentrations similar to or less than those known to stimulate cGMP formation in this tissue 1'43'44'52. The effect of nitroprusside appeared to be due to its ability to generate nitric oxide, because ferricyanide ion did not cause a potentiation at similar concentrations. These results are consistent with the hypothesis that cGMP can mediate a potentiation of nicotinic transmission, and suggest that sufficient endogenous cGMP can be formed by stimulation of soluble guanylate cyclase (e.g. by nitric oxide). This assumes that the actions of nitroprusside and azide were mediated by cGMP. The assumption seems reasonable in view of the cGMP response to these agents in the same tissue, but further experiments with phosphodiesterase inhibitors and guanylate cyclase inhibitors would be needed to confirm the assumption. The mechanism by which cGMP potentiates nicotinic transmission is not entirely clear. While the extracellular approach used in this study provides a measure of the efficacy of nicotinic transmission in the SCG, it would not distinguish modulations specific to nicotinic synaptic transmission (e.g. acetylcholine release or nicotinic receptor sensitivity) from more generalized changes in neuronal excitability. In the rabbit SCG, it has been proposed that cGMP mediates a muscarinic slow excitatory postsynaptic potential (slow-EPSP) 25'26'29'33 (however, see refs. 18,59). This slow-EPSP could effect a potentiation of nicotinic transmission as observed in the present experiments 59'62. If so, there remains an apparent discrepancy in the effect of dibutyryl-cGMP among species. While dibutyryl-cGMP has been found to elicit a slowEPSP-like response in the rabbit SCG 25'26 (but see ref. 18), dibutyryl-cGMP has not been found to potentiate nicotinic transmission in the rat SCG 12 or the frog sympathetic ganglion 48. While 8-Br-cGMP does elicit a potentiation in the rat SCG, it remains to be determined whether 8-Br-cGMP mimics the slow-EPSP the mammalian SCG. Perhaps dibutyryl-cGMP differs pharmacolog-

144 ically in its efficacy or metabolism in sympathetic ganglia from different species. H o w e v e r , it also should be noted that c G M P probably also mediates some n o n - m u s carinic responses in sympathetic ganglia (see Introduction), and a slow-EPSP-like response is only one of several possible explanations for the potentiation observed in the present experiments. It has been further p r o p o s e d in the rabbit S C G that c A M P 33'38 and/or calmodulin 39 mediates a long-lasting potentiation of the slow-EPSP. In the present experiments with the rat SCG, 10/zM 8 - B r - c A M P and 10 ~ M 8-Br-cGMP each p r o d u c e d a submaximal potentiation of nicotinic transmission, but there was no a p p a r e n t synergistic effect when the two were applied simultaneously. However, cGMP, applied several minutes after cAMP, has been found to inhibit the ability of c A M P to potentiate the c G M P - m e d i a t e d slow-EPSP 29'33 (see also interactions in frog sympathetic ganglia31'32). We have not examined the t i m e - d e p e n d e n c y of 8 - B r - c A M P and 8-Brc G M P co-application, nor have we a t t e m p t e d to determine whether c A M P analogues interact with butyryl analogues of cGMP. In the hippocampus, it has been suggested that nitric oxide or another substance similar to E D R F may be involved in the formation of long-term potentiation (LTP) a3'61, and supportive evidence has a p p e a r e d during the preparation of this r e p o r t 6. Additionally, there is some recent evidence to suggest that an E D R F - l i k e system is involved in the formation of long-term depression in the cerebellum 46. In the rat SCG, brief tetanic stim-

REFERENCES 1 Ando, M., Nanba, T., Okuya, M. and Nagata, Y., Effect of axotomy on the cyclic GMP increase induced by preganglionic stimulation and high extracellular K ÷ concentration in superior cervical sympathetic ganglion of the rat, Brain Research, 267 (1983) 313-321. 2 Ariano, M.A., Briggs, C.A. and McAfee, D.A., Cellular localization of cyclic nucleotide changes in rat superior cervical ganglion, Cell. Mol. Neurobiol., 2 (1982) 143-156. 3 Ariano, M.A., Lewicki, J.A., Brandwein, H.J. and Murad, F., Immunohistochemical localization of guanylate cyclase within neurons of rat brain, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 1316-1320. 4 Ariano, M.A. and Matus, A.I., Ultrastructural localization of cyclic GMP and cyclic AMP in rat striatum, J. Cell Biol., 91 (1981) 287-292. 5 Berkelmans, H.S., Schipper, J., Hudson, L., Steinbusch, H.W.M. and De Vente, J.J., cGMP immunocytochemistry in aorta, kidney, retina and brain tissues of the rat after perfusion with nitroprusside, Histochemistry, 93 (1989) 143-148. 6 Bohme, G.A., Bon, C., Stutzmann, J.M., Doble, A. and Blanchard, J.C., Possible involvement of nitric oxide in long-term potentiation, Eur. J. Pharmacol., 199 (1991) 379-381. 7 Briggs, C.A., Cyclic GMP potentiation of nicotinic transmission in the rat superior cervical ganglion (SCG), Soc. Neurosci. Abstr., 17 (1991) p. 384. 8 Briggs, C.A., Brown, T.H. and McAfee, D.A., Neurophysiol-

ulation of the preganglionic nerve (5-20 Hz for 5-20 s) has been found to cause a long-term potentiation of nicotinic synaptic transmission and acetylcholine release s'l~ ~3. Tetanic stimulation of the preganglionic nerve (10-20 Hz for 30-60 s) also stimulates c G M P formation severalfold in the rat SCG 2'13'19'52-54'58, and both LTP and the c G M P response are Ca2+-dependent. H o w e v e r , other second messenger systems also respond to similar tetanic stimulation 2'9'13'53'6°, and the signal transduction processes responsible for nicotinic LTP are not yet known. Interestingly, in the present experiments higher concentrations (->10/~M) of nitroprusside, azide, and 8-Brc G M P a p p e a r e d to elicit a small long-lasting potentiation in addition to the potentiation that decayed upon washout of the drug (see Figs. 1 and 3). These results are consistent with the hypothesis that c G M P may mediate or contribute to the formation of LTP in the SCG. However, the long-lasting effect was not observed in all experiments, and lower concentrations ( < 1 0 / ~ M ) of nitroprusside, azide and 8 - B r - c G M P did not a p p e a r to elicit a long-lasting potentiation. The effects of higher concentrations of nitroprusside, azide, and 8-Br-cGMP are intriguing with respect to LTP, but further study is required to determine if this represents a physiological process underlying LTP.

Acknowledgements. We thank Dr. James Kerwin for the chemical analysis, and Drs. John Garthwaite, Ferid Murad, and Michael Williams for valuable suggestions.

ogy and pharmacology of long-term potentiation in the rat sympathetic ganglion, J. Physiol., 359 (1985) 503-521. 9 Briggs, C.A., Horwitz, J., McAfee, D.A., Tsymbalov, S. and Perlman, R.L., Effects of neuronal activity on inositol phospholipid metabolism in the rat autonomic nervous system, J. Neurochem., 44 (1985) 731-739. 10 Briggs, C.A. and McAfee, D.A., Long-term potentiation at nicotinic synapses in the rat superior cervical ganglion, J. Physiol., 404 (1988) 129-144. 11 Briggs, C.A. and McAfee, D.A. and McCaman, R.E., Longterm potentiation of synaptic acetylcholine release in the superior cervical ganglion of the rat, J. Physiol., 363 (1985) 181-190. 12 Briggs, C.A. and McAfee, D.A. and McCaman, R.E., Longterm regulation of synaptic acetylcholine release and nicotinic transmission: the role of cyclic AMP, Br. J. Pharmacol., 93 (1988) 399-411. 13 Briggs, C.A., Whiting, G.J., Ariano, M.A. and McAfee, D.A., Cyclic nucleotide metabolism in the sympathetic ganglion, Cell. Mol. Neurobiol., 2 (1982) 129-141. 14 Burt, D.R. and Larrabee, M.G., Phosphatidylinositol and other lipids in a mammalian sympathetic ganglion: effects of neuronal activity on incorporation of labelled inositol, phosphate, glycerol and acetate, J. Neurochem., 27 (1976) 753-763. 15 De Vente, J., Garssen, J., Tilders, F.J.H., Steinbusch, H.W.M. and Schipper, J., Single cell quantitative immunocytochemistry of cyclic GMP in the superior cervical ganglion of the rat, Brain Research, 411 (1987) 120-128. 16 Debinski, W., Kuchel, O., Buu, N.T., Tremblay, J. and Hamet,

145

17 18 19

20 21 22 23

24

25 26 27

28

29 30 31 32 33

34

35 36 37

P., Atrial natriuretic factor constitutes an intrinsic functional unit within superior cervical ganglia of the rat, Neuroscience, 36 (1990) 21-26. Dun, N.J., Kaibara, K. and Karczmar, A.G., Direct postsynaptic membrane effect of dibutyryl cyclic GMP on mammalian sympathetic neurons, Neuropharmacology, 16 (1977) 715-717. Dun, N.J., Kaibara, K. and Karczmar, A.G., Muscarinic and cGMP induced membrane potential changes: differences in electrogenic mechanisms, Brain Research, 150 (1978) 658-661. Frey, E.A. and McIsaac, R.J., A comparison of cyclic guanosine 3':5'-monophosphate and muscarinic excitatory responses in the superior cervical ganglion of the rat, J. PharmacoL Exp. Ther., 218 (1981) 115-121. Furchgott, R.E and Vanhoutte, P.M., Endothelium-derived relaxing and contracting factors, FASEB J., 3 (1989) 2007-2018. Gallagher, J.E and Shinnick-Gallagher, P., Cyclic nucleotides injected intracellularly into rat superior cervical ganglion cells, Science, 198 (1977) 851-852. Garthwaite, J., Glutamate, nitric oxide and cell-cell signalling in the nervous system, Trends Neurosei., 14 (1991) 60-67. Garthwaite, J., Charles, S.L. and Chess-Williams, R., Endothelium-derived relaxing factor release on activation of NMDA receptors suggests role as intercellular messenger in the brain, Nature, 336 (1988) 385-388. Gutkind, J.S., Kurihara, M., Castren, E. and Saavedra, J.M., Atrial natriuric peptide receptors in sympathetic ganglia: biochemical response and alterations in genetically hypertensive rats, Biochem. Biophys. Res. Commun., 149 (1987) 65-72. Hashiguchi, T., Ushiyama, N.S., Kobayashi, H. and Libet, B., Does cyclic GMP mediate the slow excitatory synaptic potential in sympathetic ganglia? Nature, 271 (1978) 267-268. Hashiguchi, T., Kobayashi, H., Tosaka, T. and Libet, B., Two muscarinic depolarizing mechanisms in mammalian sympathetic neurons, Brain Research, 242 (1982) 378-382. Kebabian, J.W., Blood, EE., Steiner, A.L. and Greengard, P., Neurotransmitters increase cyclic nucleotides in postganglionic neurons: immunocytochemical demonstration, Science, 190 (1975) 157-159. Kebabian, J.W., Steiner, A.L. and Greengard, P., Muscarinic cholinergic regulation of cyclic guanosine 3,5-monophosphate in autonomic ganglia: possible role in synaptic transmission, J. Pharmacol. Exp. Ther., 193 (1975) 474-488. Kobayashi, H., Roles of cyclic nucleotides in the synaptic transmission in sympathetic ganglia of rabbits, Comp. Biochem. Physiol. [C], 72 (1982) 197-202. Kobayashi, H. and Libet, B., Generation of slow postsynaptic potentials without increases in ionic conductance, Proc. Natl. Acad. Sci. U.S.A., 60 (1968) 1304-1311. Kuba, K. and Kumamoto, E., Long-term potentiation of transmitter release induced by adrenaline in bull-frog sympathetic ganglia, J. Physiol., 374 (1986) 515-530. Kumamoto, E. and Kuba, K., Mechanisms regulating the adrenaline-induced long-term potentiation in bullfrog sympathetic ganglia, Pfl•gers Arch., 408 (1987) 573-577. Libet, B., Mediation of nonclassical postsynaptic responses by cyclic nucleotides. In M. Avoli, T.A. Reader, R.W. Dykes and P. Gloor (Eds.) Neurotransmitters and Cortical Function. From Molecules to Mind, Plenum, New York, 1988, pp. 453-469. Maurice, D.H. and Haslam, R.J., Molecular basis of the synergistic inhibition of platelet function by nitrovasodilators and activators of adenylate cyclase: inhibition of cyclic AMP breakdown by cyclic GMP, MoL PharmacoL, 37 (1990) 671-681. Maurice, D.H. and Haslam, R.J., Nitroprusside enhances isoprenaline-induced increases in cAMP in rat aortic smooth muscle, Eur. J. Pharmacol., 191 (1990) 471-475. McAfee, D.A., Schorderet, M. and Greengard, P., Adenosine 3',5'-monophosphate in nervous tissue: increase associated with synaptic transmission, Science, 171 (1971) 1156-1158. McIsaac, R.J., Post-tetanic enhancement of stimulus-induced muscarinic afterdischarge in the rat superior cervical ganglion,

J. Pharmacol. Exp. Ther., 207 (1978) 72-82. 38 Mochida, S., Kobayashi, H. and Libet, B., Stimulation of adenylate cyclase in relation to dopamine-induced long-term enhancement (LTE) of muscarinic depolarization in the rabbit superior cervical ganglion, J. Neurosci., 7 (1987) 311-318. 39 Mochida, S. and Libet, B., Postsynaptic long-term enhancement (LTE) by dopamine may be mediated by Ca 2+ and calmodulin, Brain Research, 513 (1990) 144-148. 40 Murad, F., Ishii, K., Forstermann, U., Gorsky, L., Kerwin, J.E, Pollock, J. and Heller, M., EDRF is an intracellular 2nd messenger and autacoid to regulate cyclic GMP synthesis in many cells, Biol. Med. Signal Transduct., 24 (1990) 441-448. 41 Nakane, M., Ichikawa, M. and Degeuchi, T., Light and electron microscopic demonstration of guanylate cyclase in rat brain, Brain Research, 273 (1983) 9-15. 42 Pant, K.K. and Smith, P.A., Atrial natriuretic factor suppresses M-current in frog but not rat sympathetic neurones, Neurosci. Lett., 100 (1989) 243-248. 43 Quenzer, L.E, Patterson, B.A. and Voile, R.L., The cyclic nucleotide content of the rat superior cervical ganglion, J. Pharmacol. Exp. Ther., 215 (1980) 297-303. 44 Quenzer, L.F., Patterson, B.A. and Volle, R.L., K+-induced accumulation of guanosine 3',5'-monophosphate in sympathetic ganglia, J. Neurochem., 34 (1980) 1782-1784. 45 Schulz, S., Chinkers, M. and Gathers, D.L., The guanylate cyclase/receptor family of proteins, FASEB J., 3 (1989) 20262035. 46 Shibuki, K. and Okada, D., Endogenous nitric oxide release required for long-term synaptic depression in the cerebellum, Nature, 349 (1991) 326-328. 47 Snyder, S.H. and Bredt, D.S., Nitric oxide as a neuronal messenger, Trends Pharmac. Sci., 12 (1991) 125-128. 48 Suria, A., Cyclic GMP modulates the intensity of post-tetanic potentiation in bullfrog sympathetic ganglia, Neuropharmacology, 15 (1976) 11-16. 49 Takahashi, S.Y., Mochida, S. and Kobayashi, H., Cyclic GMPdependent protein kinase and phosphorylation of the endogenous substrate proteins in the rabbit superior cervical ganglion, J. Neurochem., 51 (1988) 1300-1307. 50 Volle, R.L. and Patterson, B., cAMP in guinea-pig superior cervical ganglia during preganglionic nerve stimulation, Experientia, 39 (1983) 1345-1346. 51 Voile, R.L. and Patterson, B.A., Regulation of cyclic AMP accumulation in a rat sympathetic ganglion: effects of vasoactive intestinal polypeptide, J. Neurochem., 39 (1982) 1195-1197. 52 Voile, R.L. and Quenzer, L.F., Regulation of cyclic GMP levels in nerve tissue, Fed. Proc., 42 (1983) 3099-3102. 53 Voile, R.L., Quenzer, L.F. and Patterson, B.A., The regulation of cyclic nucleotides in a sympathetic ganglion, J. Auton. " Nerv. Syst., 6 (1982) 65-72. 54 Voile, R.L., Quenzer, L.F., Patterson, B.A., Alkadhi, K.A. and Henderson, E.G., Cyclic guanosine 3':5'-monophosphate accumulation and 45Ca-uptake by rat superior cervical ganglia during preganglionic stimulation, J. Pharmacol. Exp. Ther., 219 (1981) 338-343. 55 Walaas, S.I., Girault, J.A. and Greengard, P., Localization of cyclic GMP-dependent protein kinase in rat basal ganglia neurons, J. Mol. Neurosci., 1 (1989) 243-250. 56 Waldman, S.A. and Murad, F., Cyclic GMP synthesis and function, Pharmacol. Rev., 39 (1987) 163-196. 57 Wamsley, J.K., West, J.R., Black Jr., A.C. and Williams, T.H., Muscarinic cholinergic and preganglionic physiological stimulation increase cyclic GMP levels in guinea-pig superior cervical ganglia, J. Neurochem., 32 (1979) 1033-1035. 58 Weight, EE, Petzold, G. and Greengard, P., Guanosine Y,5'monophosphate in sympathetic ganglia: increase associated with synaptic transmission, Science, 186 (1974) 942-944. 59 Weight, EE, Schulman, J.A., Smith, P.A. and Busis, N.A., Long-lasting synaptic potentials and the modulation of synaptic transmission, Fed. Proc., 38 (1979) 2084-2094.

146 60 White, G.L. and Larrabee, M.G., Phosphoinositides and other phospholipids in sympathetic ganglia and nerve trunks of rats. Effects of neuronal activity and inositol analogs [6- and 7-hexachlorocyclohexane (lindane)] on [32p]-labelling, synaptic transmission and axonal conduction, J. Neurochem., 20 (1973) 783-798. 61 Williams, J.H., Errington, M.L., Lynch, M.A. and Bliss, T.V.P.,

Arachidonic acid induces a long-term activity-dependent enhancement of synaptic transmission in the hippocampus, Nature, 341/342 (1989) 739-742. 62 Yarosh, C.A,, Acosta, C.G. and Ashe, J.H., Modification of nicotinic ganglionic transmission by muscarinic slow postsynaptic potentials in the in vitro rabbit superior cervical ganglion, Synapse, 2 (1988) 174-182.