002X-39OX,X3’050609-06803.00!0

Copyright 0 19X3 Pergamon Press Ltd

EFFECT OF ADENOSINE, ADENOSINE TRIPHOSPHATE, ADENOSINE DEAMINASE,

DIPYRIDAMOLE AND AMINOPHYLLINE ON ACETYLCHOLINE RELEASE FROM

ELECTRICALLY-STIMULATED BRAIN SLICES

FEIXITA PII~A~A’. T~~~ANA ANTONLLI ‘. L. LAMBERT&, L. BEAN’ and G. PEPEU’

‘Department of Pharmacology. University of Florence, Viale Morgagni 65. 50134 Florence and ‘University of Ferrara. Via Fossato di Mortara 64. 44100 Ferrarn. Italy

Summary-The effect of adenosine on release of acetylcholine (ACh) was investigated in slices of rat cortex perfused with Krebs solution, at rest and during electrical stimulation at frequencies between 0.2 and 20 Hz. Electrical stimulation brought about a linear increase in release of ACh. Adenosine, in concentrations ranging from 1 to 100 PM, reduced in a dose-dependent manner the release of ACh and was more active on the stimulated than on the resting release. However, the fractional reduction by adenosine of stimulated release of ACh did not vary with increasing stimulation rate. Adenosine triphos- phate was less active than adenosine in reducing release of ACh. The inhibitory effect of adenosine was antagonized by aminophylline (0.5 mM) and did not occur when the stimulated release of ACh was enhanced by blocking muscarinic autoreceptors with atropine (15 nM). Aminophylline (0.1 and 0.5 mM) itself exerted a biphasic effect on release of ACh. increasing it at rest and during stimulation at low frequencies, and decreasing it at higher stimulation rates. The manipulation of endogenous adenosine concentrations by adding adenosine deaminase or diphyridamole. an inhibitor of adenosine uptake, had little effect on release of ACh. Dipyridamole. (4 uM). only significantly decreased release of ACh at the 20 Hz stimulation rate.

According to Stone (198 1) one of the actions of aden-

osine which has been most frequently and clearly de- scribed is its ability to depress the evoked release of transmitters from both cholinergic and adrenergic neurones. However, information on the effect of aden- osine on cholinergic neurones in the brain is limited. Vizi and Knoll (1976) investigated the inhibitory effect of adenosine and ATP on the release of endogenous ACh from brain slices. Harms. Wardeh and Mulder (1979), and Corrieri. Barberis and Gayet (1981) dem- onstrated that adenosine decreased the release of labelled ACh from striatal slices and neo-cortical synaptosomes respectively. preloaded with radioactive choline. In these experiments the output of ACh was stimulated by ouabain (Vizi and Knoll, 1976) or by potassium depolarization.

Much evidence for a stimulated release of aden- osine derivatives from the synaptic terminals in the mammalian central nervous system has been provided (see ref. in Phillis and Wu, 1981: Lee, Schubert, Grib- koff. Sherman and Lynch. 1982). As the firing fre- quency of the neurone increased, the extracellular adenosine concentration would be expected to rise and could be considered to be a means of restricting otherwise excessive transmitter release (Stone, I98 I).

Key words: adenosine. acetylcholine release, aminophyl- line.

Address correspondence to: Professor G. Pepeu. Depart- ment of Pharmacology. University of Florence. Viale Mor- gagni 65, 50134 Florence. Italy.

In order to test this hypothesis the possibility that changes in adenosine concentrations in the extracellu- lar fluid, obtained by adding adenosine and by mani- pulating its endogenous levels, would affect the release of ACh from brain slices, stimulated electri-

cally at different frequencies. was investigated. En- dogenous adenosine levels were modified by adding adenosine deaminase (Daly, Padgett, Creveling, Can- tacuzine and Kirk, I98 I) or by blocking uptake of adenosine by dipyridamole (Meunier and Morel. 1978). The effect of aminophylline. an adenosine an- tagonist (Sawynok and Jhamandas, 1976; Perkins and Stone, 1980), and the possible relationship between adenosine and the muscarinic presynaptic receptors, regulating the output of ACh from nerve terminals (Molenaar and Polak. 1980) were also investigated.

A preliminary communication of part of the results was presented at a meeting of the British Pharmaco- logical Society (Pedata and Pepeu. 1981).

YIETHODS

The experiments were carried out on male Wistar rats of 15&200 g body weight. The animals were decapitated, the skull opened and the right and left parietal cor:ices were rapidly removed and plunged into cold Krebs solution with the following compo- sition (mM): NaCl 118.5, KC1 4.7, CaCl, 2.5, MgSO, 1.2. KH,P04 1.2. glucose 10, NaHCO, 25, choline

609

0.02. The cortical slices were prepared and stimulated according to the method of Beani. Bianchi. Giaco- mclli and Tamberi (1978).

Briefly. the cortical samples. submerged in Krebs solution, \vcrc cut into slices 400 /[rn thick by means of ;I microtome for fresh tissue. The slices were kept tioating for 30 min in Krebs solution. bubbled with 05” 0 wd 5”(, C02. and for another 30 min in the 2 ‘ prc;cncc of physostigmine sulphatc (3.8 /tM). They 11 crc then transferred to Perspex superfusion

chambers of 0.9 ml vol. and superfused with gassed Krebs solution. containing physostigmine sulphate. at the rate of 0.5 ml:min at 37 C. After equilibration for 20 min. the first sample was collected during 5 min rest, Then the slices v,crc stimulated at frequencies from 0.2 to 20 Hz Nith rectangular pulses of alternat- ing polarity. with ;I current strength of 3OmA,cm’ and ;I pulse duration of 5 msec. Usually two or three cycles of stimulation at the same or at different fre- quencies were carried out. The cycles were separated by ;I 20 min interval. Within each cycle. 5 min stimu- lation periods were followed by IO min rest. Samples from each stimulation period and the following rest period were collected in order to allow the uashout of all ACh released by the electrical stimulation. The extra release caused by electrical stimulation aas esti- mated by subtracting the ACh release expected during I5 min rest (calculated on the basis of the first sample) from the whole amount found in the I5 min perfusate sample.

Drugs were added to the perfusion fluid at the beginning of the 20 min interval between two stimu- lation cycles.

The ACh content of the superfusate samples was

quantified on the isolated guinea-pig ileum perfused with Tyrode solution containing cyproheptadine (3 nM). morphine (3 {IM) and preincubated for 60 min with tetrodotoxin (0.03 mM) according to the pro- cedure described by Beani rf ~1. (1978).

The identity of the active substance in the samples with ACh was routinely checked by adding atropine or by alkaline hydrolysis of the samples. In order to prevent drugs, added to the Krebs solution perfusing

the slices. from interfering with the bioassay, standard solutions of ACh. containing the same concentration of the drugs. \vcrc used. When the effect of adenosine \\as studied in the presence of atropine. the ACh con- tent of the superfusate was bioassayed on the dorsal muscle of the leech according to the method of Mur- naghan (1958). The results obtained with the two bio- assays gave differences uithin lo”,,.

The ACh release was expressed as ng,‘g of wet tis- sue, min of superfusion k SE.

Dr1rg.s

Freshly prepared solutions of the following drugs were used: adenosine (Calbiochem), ntropine sulphate and physostigmine sulphate (BDH). ATP. acetylcho-

line, aminophylline and adcnosinc deummasc (Sigma). morphine sulphate (Carlo Erba). tctrodotoxin (Blo- chemia). Cyproheptadine \+as :I gift from Merck. Sharp and Dome and dipyridnmole from Bochringcr. Ingelheim.

RESL’LTS

Figure I shous the dose effect relationship bet\\een increasing concentrations of adcnosinc in the supcr- fusate and ACh release from cortical sliccb. both at rest and during stimulation at 5 Hz. The basal relcasc at rest was X.66 _t 0.59 and increased appro\imatel> II-fold during stimulation. Adenosinc \\IIS less active on the release at rest than during stimulation. Fat instance. adenosine at the concentration of I /tM had no detectable effect on the former \\hile It caucetl

about a 35”,~ decrease on the latter.

The intermediate concentration of 30 HIM of adcn-

osine. causing approxim:itely a N”,, decrease in AC‘h release at 5 H7, \V;IS selected in order to inicstigatc the influence of diRerent stimulation frequencies on the inhibitory effect of adenosins on output of AC‘h From the results she\\ n in Figure 2 it can bc seen th;lt the increase in stimulation frcclucnq LI;I~ asociated

with a linear increase in rcleasc of AC‘h (r = O.Y)x: P < 0.01 ). In the presence of adenosine the Increase in release of AC‘h was still linear (I’ = 0.05: I’ < 0.05~ but markedly reduced b) I 5C’,I at rest and b! 5i”,, at I II/ stimulation frequency. Although at higher frequcnclcs the difference from the control level of rclcasc W;IS not

proportionally larger. the amount of ACh released which LV;~S inhibited by the addition of adenasine W:IS

17 ng,‘g,min at I Hr and 71 ng g min at IO H7 stimu-

lation frequency. An equimolar concentration of ATP ;IISO inhibited

release of ACh but its effect was smaller than that ot :idenosine. The increase in ACh relcasc at ditTcrsn1

Adenosine and cortical ACh rrleasc hl I

40

20

I , I 2 5 IO 20 HZ

Fig. 2. Relcasc of ACh from cortical slices stimulated elec- trvzall) at different frequencies in normal Krebs solution (*-o), low calcium (CaCI, 1.2 mM) (V-0). aden- osine 30k1M (0-Z) and ATP 3OktM (n-n). Ekh point is the mean k SE (vertical bars) of at least 7 experi- ments for normal Krebs and at least 5 experiments in all other conditions. Difference from normal Krebs solution

statistically signilicant with *P < 0.01 : **P < 0.05.

stimulation frequencies in the presence of ATP was linear (r = 0.99: P < 0.01). There was no difference from control release at I Hz and reductions of 20 and 3X”,, at 5 and 10 Hz. respectively.

For comparison with the effect of adenosine, the concentration of Ca” in the superfusing Krebs sol- ution was halved and ACh release at various frequen- cies was studied. Under these conditions the release of ACh was significantly smaller than with normal Krcbs solution. A linear relationship (r = 0.99;

rlrlfl

*

T f

AD AD AD Al A’M A+M

Fig. 3. Antagonism between aminophylline (AM) 0.5 mM and adenosine (AD) 30 /tM on release of ACh from cortical slices at rest and stimulated at 5 Hz. Each column is the mean + SE (vertical bars) of at lcast 4 experiments. *The difference is statistically significant from control and

AD + AM aith P < 0.01.

01 rnM

Fig. 4. Effect of aminophqlline on release of ACh from cortical slices at rest nnd stimulated at different frequen- cies. Each column is the mean of 4 eupcriments: vertical bars: standard error of the mean. *Statistically significant

percentage differences with P < 0.001.

P < 0.01) between the increase in stimulation fre- quencies and release of ACh was found and it was parallel to that obtained in the presence of adenosine.

Figure 3 shows that the decrease in release of ACh caused by adenosine (30 /LM) at the 5 Hz stimulation frequency was completely prevented by the addition of aminophylline (0.5 mM) to the superfusing Krebs

solution. Besides antagonizing the depression of ACh release

induced by adenosine. aminophylline exerted a direct biphasic effect. As illustrated in Figure 4. 0.5 mM aminophylline brought about a significant increase in release of ACh at rest and during stimulation at 0.2 Hz. Conversely, there was a marked decrease in

the release at higher stimulation frequencies. Amino- phylline. 0.1 mM. caused a small. but statistically sig- nificant. increase in release of ACh at rest. it had no effect at the 0.2. I and 2 Hz stimulation frequencies and brought about a marked decrease in ACh release at 5 Hz. This effect was observed also if the stimu- lation was carried out at the beginning of the per- fusion with aminophylline.

As shown in Table I. atropine (I5 mM) brought about ;I large increase in release of ACh from the stimulated brain slices. This effect is attributed to a blockade of presynaptic muscarinic nutoreceptors modulating release of ACh from cholincrgic nerve endings (Hadhazy and Szerb. 1977). The addition of adenosine. at concentrations of 30 and lOO/tM &as not followed by the expected decrease in release of ACh.

In the following cxpt‘rimrnts an attemp! was made to reduce the synaptic concentration of adenosine by adding adenosine deaminase. Conversely, the block- ade of adenosine uptake by dipyridamole (4 /iM) was expected to increase adenosine concentrations at the receptor level.

The results are shown in Table 2. in which only two of the four stimulation frequencies tested are shown. They demonstrate that no significant differences in release of ACh from cortical slices. incubated in nor- mal Krebs. could be detected between the first and second stimulation cycles. The addition of adenosine deaminase (lO~g,mI) to the perfusing Krebs solution between the two cycles did not affect the release of ACh. On the other hand, in the presence of dipyrida- mole (4 PM) release of ACh in the second stimulation cycle was significantly smaller than in the first at 20 Hz. In two experiments (not showni dip~rid~lnlole

(lO$@) brought about a X0”,, increase in the output of ACh at rest. and had no effect at i and 5 Hz stimu- lation frequencies.

Several reviews (Burnstock, 1980; Fredholm and

Hedquist, 1980; Stone. 1981; Phillis and Wu, 19x1) list the large number OF papers demonstrating that adenosine and other adenine nucieotides decrease

nctlrotraIlsmitter relcasc from peripherd and central

neuronal endings through a presynaptic mechanism In the present study it was shown that both aden-

osine and ATP also inhibited release of ACh from cortical slices, electrically stimulated at different fre- quencies. Both nucleotides were more effective in de- creasing release of ACh from stimulated slices than from slices at rest. However. over the range of stimu- lation frequencies from 1 to 10 Hz the inhibitory effect of 30 ;fM adenosine on ACh release resulted in outputs of ACh consistently between 40 and 50”,, of the control. Under the experimental conditions used adenosine was almost twice as active in reducing release of i\Ch than in experiments (Harms er d..

1979) in which labelled ACh output was stimulated by potassium depolarization. On the other hand, Vizi and Knoll (1976) found that. in the isolated guinea- pig ileum, adenine nucleotides are more effective in inhibiting the t\sitch response evoked hy stimul~~ti~)l~ at 0. I Hz. than at higher stimulatian frequencies.

Adenosine tripbosph~lte was less uffectire than adenosine. p~~rti~L]l~~rly at lotncr frequencies. iti inhi- biting release of AC’h from cortical slices. I!nlike

adenosine, its efQctiveness MUS enhanced by mcrcas- inp the st~rnl~i~~ti~~n rate. The antagonism of adenosine by aminophylline confirms that specific purinergic receptors are involved in modulating reieasc of ACh from the cerebral cortex. Moody and Burnstock (1982) have demonstrated the presence of P, purinn-

Table 1. Effect of ndenosine deaminase (10 pg:ml) and dipyridamole (5 biMI on release of ACh from cortical slices at rest and stimulated electrically

Stimulation frequency (Hz\

Conditions Cycles 0 1 ‘0

(ACh release ng:g:min & SE) I..

Dipyridamole isi

* Significantly different from 1 cycle \vith 1’ < 0.01. The drugs were added to ihe Krebs solution Ii min before the second almu-

lation cvcle. The number of experiments are in parenthesis.

Adenosine and cortical ACh release hl 3

ceptors in cholinergic nerve terminals in the gumea- pig ileum.

It seems likely that a reduction of uptake of Ca” + is involved in mediating the presynaptic action of aden- osine on cholinergic nerve terminals (cf. Burnstock. IWO). In this regard. a remarkable parallel was found between the effect of adenosine on ACh release from the stimulated slices and that of Krebs solution con- taining one half of the Ca’+ concentration.

It was found that adenosine had no effect when the release of ACh was stimulated by atropine. According to Nordstrom. Westlind, Hedlund, Unden and Bartfai (1981) there is the possibility that the principal ion involved in the action of the presynaptic muscarinic

receptor is Ca’ ’ and that atropinc increases the

rcleasc of ACh by facilitating Ca’+ influx in cholin- ergic nerve terminals. This could prevent the reduc- tion of the depolarization Ca” entry induced by

adenosine, thus suppressing its effect.

cortical slices is strongly inhibited bq exogenous

adenosine. though this effect does not occur if the muscarinic autoreceptors are blocked by atropine.

The experiments do not clarify. however. v,hcther cn- dogenous ndenosine plays a role in modulating AC‘h release from cholinergic nerve endings in the cerebral cortex. as demonstrated by Gustafsson. Fredholm and Hedquist (19x1) in the guinea-pig ileum. Furthermore. the effects of dipyridamole and aminophyllinc on ttlc

release of ACh deserve further reswrch.

The attempt to manipulate endogenous adenosine levels as a way of defining its physiological role in modulating ACh release met with limited success. The

addition to the Krebs solution of adenosine de- aminase, in a concentration found active in brain slices by Daly ef ul. (1981), did not increase resting and stimulated release of ACh. It must be mentioned.

however. that according to Willemat and Paton (198 1). as well as Fredholm, Jonzon. Lindgren and LindstrBm (1982). the uptake of adenosine is a more important inactivation pathway than is enzymatic deamination.

Beani. L.. Bianchi. C.. Giacomelll. A. and T;~mher~. F. (1978). Noradrenaline Inhibition of acctylcholine release from guinea-pig brain. Etrr. J. P/~cvmc~. 48: 17Y lY3.

Burnstock. G. (1980). Purinerglc modulation of cholinerglc transmission. &I. Phurmu~. 1 I: 15 IX.

Corrieri. A. G.. Barberis. C. and Gayet. J. (19X1). High affinity choline uptake and acetylcholine release by guinea-pig neocortex synaptosomes: inhibItIon bl adenosine derivatives. Biochrrn. Phurr~uc~. 30: 3731 2734.

Daly, J. W.. Padgett, W.. Creveling. C. R.. Cantacurene. D. and Kirk, K. L. (1981). Cyclic AMP-generating slyterns: regional differences in action by adrenergic receptor< 111

rat brain. J. .Vruro.~i. I: 49-59. Fredholm, B. B. (1980). Are methylxanthme effects due to

antagonism of endogenour adenosine’? T,er~rl\ P/~trr,,~oc,. Sci. 1: 129-132.

Adenosine uptake was therefore blocked with dipyridamole. The concentration of this drug used in those experiments was greater than that which doubles the effectiveness of adenosine in increasing accumulation of [‘HIcyclic AMP in ‘hippocampal slices (Fredholm rt al.. 1982) and which causes a 50”,, inhibition of uptake of adenosine by cholinergic synaptosomes (Meunier and Morel, 1978). Neverthe- less. a decrease in ACh output in the presence of dipyridnmole was only found at the 20 Hz stimulation frequency, Further investigations are needed to vcrifl whether this effect of dipyridamole was realI\

mediated through adenosine.

Fredholm, B. B. and Hedquist. P. (1980). Modulation of neurotransmission by purine nucleotides and nucleo- sides. Biochr~n Phmuac. 29: 1635-I 643.

Fredholm. B. B.. Jonzon. B.. Lindgren. E. and Lindstr8m. K. (1982). Adenosine receptors mediating cyclic AMP production in the rat hippocampus. J. .2’cww/1cw. 39: 165Sl75.

Gustafsson. L.. Fredholm. B. B. and Hedquist. P. (IYXI). Theophyllme Interferes Lbith the modulatory role of cn- dogenous adcnosine on cholincrgic neurotr~~nsmissisn 111 guinea-pig ileum. Act0 II/I~X~O/. bwid. 1 II : 269 2X0.

Hadharq. P. and Srerb. J. C. (1977). The effect of cholln- crgic drugs on [zH]acetylcholine release from sl~ccs 01 rat hippocampua. striatum and cortex. Ur[ri~l Rc,\. 123: 31 I- 3X!.

Aminophylline enhanced the release of ACh from the cortical slices at rest and at the low stimulation frequency. This finding could be interpreted as due to a blockade of adenosine receptors. removing an inhi- bition mediated by endogenous adenosine. However. it is difficult to reconcile this interpretation with the lack of effect of dipyridamole on the release of ACh at rest and at low frequencies. At higher frequencies, aminophylline brought about a marked decrease in ACh output. Since, at the concentrations used in these experiments, methylxanthines possess a number of other properties, such as phosphodiesterase inhibition and catecholamine release (Fredholm. 1980). it is difi- cult to interpret this unexpected finding.

Harms. H. H.. Wardeh. G. and Muldcr. A H. (lY7Y). Effects of adenosmr on depolariratlon-IridLlccd rcleasc 01 various radlolabelled neurotr~lnsmlttrrs from slices 01 rat corpus strlatum. Vc1r~o/,/lc1~,~1~rc,0/(,[/~. 18: 577 5X0.

Let. K.. Schubert. P.. GribkoH: V.. Sherman. B. and Lynch. G. (IYX2). .A combined “I~I r~ro 11, tirro” atrid! 01 the prcsynaptlc rzleaic of adenohinc dcri\atl\c\ ln the hippo- campus. J. Xcwoc/~m. 38: X0 X3.

Mcunier. F. M. and Morell. N. (197X). Adenoslnc uptake by chollnergic synaptosomes from torpedo electric organ. J. .Vcwo~ho~~ _ 11: 835~XSl.

Molenaar. P. C. and Polak. R. L. (1980). InhibItion of ;tce- tylcholme release b) activation of acetylcholinr recep- tors. In: Procjrrs.~ irl Plrtrr~r~trco/o~~~. Vol. 3 3. pp. 39-33. Gustav Fisher-Verlag. Stuttgart.

Moody. C. J. and Burnstock, G. (19X3). Evidence for the presence of P,-purinoceptors on cholinergic nerve ter- minals in the guinea-pig ileum. Eur. J. Plrcr~n~trc. 77: t 9.

Murhaghan. M. F. (195X). The morphinized-eserinifed leech muscle for the assay of acetylcholine. \ </twC’. Lwtl. 182: 317.

In conclusion, from these experiments it appears Nordstrom. A.. Westlind. A.. Hedlund. B.. Unden. A. and that the release of ACh from electrically-stimulated Bartfai. T. (1981). On the ionic mechanism of pres!naptlc

muscarinic receptors in rut hippocampus. In: Clwliurrqrc~ !\~whutlis/tl.\ (Pepeu. G. and Ladinskq. H.. Eds). pp. 579-586. Plenum Press, New York.

Pedata. F. and Pepeu. G. (19X1). Effect of adenosme on acetylcholine output from electrically stimulated brain SI~CXS. Br. J. P/WUW. 74: 764P 765P.

Perkins. M. K. and Stone. T. W. (19X0). Amlnophylline and theophylline derlvativcs as antagonists of neuronal de- pression by adenosine: a microiontophoretic study. Arc/u irlt. P/~trrm~c~od~u. 7‘her. 246: 205 -2 14.

Phillis. J. W. and Wu. P. H. (1981). The role of adenosme and its nuclcotides in central synaptic transmission. Proq. .\‘ewohiol. 16: I X7-239.

Sawynok. J. and Jhamandas. K. H. (1976). InhIbitIon 01. acetylcholine release from cholincrgic ncrws bk aden- osine. adenine nucleotides and morphine: antagonism h! theophylline. J. Piwww. t’\p. Tlrvr. 197: 37’) 390.

Stow. T. W. (19X1 I. Physiological roles for ndeno\lnc and adenosine 5’.triphosphate in the nervous system. .~vw~~- ,SUC,XC 6: 52.1 555.

VIZI. E. S. and Knoll. J. (1976). The Inhibitory efIect 01 adenosme and related nucleotldes on the release of uce- tylcholine. h’ewwx irw~ 1 : 39 1-39X.

Willemat. J and Paton. D. M. (19X1). Mctabollsm and presynaptlc inhibitory ellects of adenosine in rat LX\ defcrens. J. Aufcm. Phtrrmc~ I : 2 I7 224