Apparent enhancement of cholinergic transmission in rabbit bronchi via adenosine A2 receptors

European Journal of Pharmacoloyv, 120 (1986) 179-185 Elsevier

179

A P P A R E N T E N H A N C E M E N T OF CHOLINERGIC T R A N S M I S S I O N IN RABBIT B R O N C H I VIA A D E N O S I N E A2 R E C E P T O R S LARS E. GUSTAFSSON 1.2., N. PETERWIKLUNDI2 and BOCEDERQVIST1 Department of Physiology J. Karolinska Institutet and the Swedish National Institute of Eneironmental Medicine :. S-10401 Stockhohn, Sweden

Received 27 June 1985, revised MS received 2 October 1985, accepted 29 October 1985

L.E. GUSTAFSSON, N.P. WIKLUND and B. CEDERQVIST, Apparent enhancement of cholinergic tran.rrnission in rabbit bronchi via adenosine A_, receptors, European J. Pharmacol. 120 (1986) 179-185. Adenosine and its derivatives enhanced the contractile responses to transmural nerve stimulation in rabbit isolated bronchial smooth muscle. 5'-N-Ethylcarboxamideadenosine (NECA) was the most potent adenosine analogue studied. Enhancement of contractile responses by NECA was competitively antagonized by 8-p-sulfophenyltheophylline. Guanethidine, mepyramine, capsaicin or eicosatetraynoic acid did not antagonize the enhancement elicited by adenosine or NECA. NECA did not enhance the contractile responses to exogenously applied acetylcholme or contractile responses elicited after administration of tetrodotoxin. We suggest that adenosine, via an action at A 2 receptors, enhances contractile responses to nerve stimulation in rabbit bronchial muscle. Methylxanthines are competitive antagonists at these extracellular receptors. The enhancement probably involves a sodium-dependent mechanism hut not adrenergic mechanisms or release of histamine, substance P or arachidonate metabolites. The enhancement indicates increased cholinergic transmitter release or action, but release of a secondary spasmogenic or decreased release of an inhibitor mediator cannot be excluded. The results may indicate a role for adenosine in asthma. 8-p-Sulphophenyltheophylline Airway smooth muscle

Acetylcholine

Adenosine

1. Introduction Bronchial smooth muscle plays an i m p o r t a n t role in the increased airway resistance in obstructive lung diseases. M e t h y l x a n t h i n e s such as theophylline are k n o w n to relax respiratory smooth muscle ( F r e d h o l m et al., 1979; Jeppson et al., 1982; Karlsson a n d Persson, 1981) a n d have long been used in the treatment of obstructive lung diseases (Persson, 1980). Several m e c h a n i s m s have been proposed to explain the effects of theophylline on respiratory smooth muscle (Rail, 1980). Of these mechanisms, two are of special interest: theophylline may either act classically, intracellularly * To whom all correspondence should be addressed: Dept of Physiology, Karolinska Inslitutet, Box 60400, S-10401 Stockholm, Sweden. 0014-2999/86/$03.50 ~"~1986 Elsevier Science Publishers B.V.

Modulation

Neurotransmission

to inhibit phosphodiesterases and increase the level of cyclic nucleotides, or it may act extracellularly to antagonize the effect of purines at adenosine receptors. The latter action, which is the subject of this paper, would require that e n d o g e n o u s adenosine stimulate the airway smooth muscle, and such an effect is difficult to reconcile with the traditional smooth muscle relaxing effect of adenosine (cf. Baer et al., 1983). A d e n o s i n e has been reported to relax guinea-pig tracheal preparations with a high resting tension, and at times to be able to contract preparations with a low resting tension ( F r e d h o l m et al., 1979). A d e n o s i n e was also reported to have no effect on cholinergic contractile responses in this tissue (Grundstr/3m et al., 1981). However, a d e n o s i n e has been shown to potentiate contractile responses to cholinergic nerve stimulation in the rabbit stomach and bronchial smooth

180

muscle, and in guinea-pig ileum (Gustafsson, 1981; 1983; Gustafsson et al., 1985a,b). Moreover, adenosine has been shown to cause bronchoconstriction in asthmatic but not in normal human subjects (Cushley et al., 1983). We therefore further investigated the capacity of adenosine to enhance responses to cholinergic nerve stimulation in rabbit bronchi, with the aim to characterize the adenosine receptor type and to explore the ability of xanthines to antagonize the adenosine effect. Ligand binding studies have shown that adenosine actions are mediated via at least two extracellular receptor types: A 1 and A~ (cf. Daly, 1982). The subclasses of adenosine receptors may be distinguished by the relative agonist potencies of certain adenosine analogues (cf. Daly, 1982). At the adenosine A~ receptor, 5'-N-ethylcarboxamideadenosine (NECA) is more potent than L-N% phenylisopropyladenosine (L-PIA), whereas L-PIA is equi- or more potent than NECA at the A~ receptor. L-P1A is 50-100 times more potent than its stereoisomer D-PIA at the A t receptor. The difference in potency between L-PIA and D-PIA at the A~ receptor is at most 5-fold (Bruns et al., 1980). Both receptor subtypes are blocked by methylxanthines, although no receptor selective antagonist is available at present. We have used 8-p-sulfophenyltheophylline as an adenosine receptor antagonist, since it has recently been identified as a competitive adenosine antagonist without direct smooth muscle inhibitory action (Gustafsson, 1984). We now report that adenosine can act at an extracellular A : receptor to enhance cholinergic responses in rabbit bronchi, and that this receptor is blocked by xanthines in a competitive fashion.

2. Materials and methods

2.1. General procedure Thirty one New Zealand White rabbits of either sex were stunned and bled. The lungs and heart were removed and placed on ice. Bronchial segments of approximately the 3rd to 6th generation of division were dissected from the lungs then cut in spirals and suspended vertically in 6 ml organ

baths containing Tyrode solution (composition; concentrations in mM: Na + 149, K + 4.8, Ca 2+ 2.5, Mg 2+ 0.5, C1 147, HCO~ 11.9, H2PO 4 0.4 and glucose 5.5) kept at 37°C and continuously aerated with 5% CO 2 in O~. Contractile responses were measured isotonitally at a resting tension of 2.5 mN by means of a Harvard Apparatus smooth muscle transducer and a Grass Polygraph or a Honeywell Elektronik 184 ordinate writer. Transmural stimulation was applied through a pair of 10 mm long electrodes in parallel with the tissue, placed 10 mm apart and opposing each other in the bath walls, and connected to either a Grass $88 or a Somedic AB stimulator. Specific nerve stimulation was elicited by monophasic pulses at supramaximal voltage, 5-10 Hz, 0.2-1 ms pulse duration, 50-100 pulses at 1-2 rain intervals. Direct smooth muscle stimulation was performed at a pulse duration of 5 ms in the presence of tetrodotoxin 3 x 10 7 M (further details in Results). 2.2. Drugs Acetylcholine chloride, adenosine, atropine sulphate, capsaicin, carbamylcholine chloride (carbachol), eserine sulphate, mepyramine, tetrodotoxin and theophylline were purchased from Sigma Co. (St. Louis, USA). Other drugs used were eicosatetraynoic acid (F. Hoffman-La Roche & Co, Basel, Switzerland), guanethidine sulphate (Ciba-Geigy AG, Basel, Switzerland), L-P1A (LN%phenylisopropyladenosine) and its ( + )-isomer D-PIA (Boehringer Mannheim G m b H , Mannheim, West Germany) and N E C A (5'-N-ethylcarboxamideadenosine; Byk-Gulden Lomberg Chem. Fabr., Konstanz, West Germany). 8-pSulfophenyltheophylline was synthetized according to Daly et al. (1985) and was pure as analyzed by HPLC, using as reference a sample kindly provided by Dr. R.F. Bruns (Warner-Lambert Co, Ann Arbor, MI, USA). 2.3. Statistics Experimental data were expressed as mean values +S.E.M. Statistical significance was tested according to Student's t-test for unpaired variates.

181

Schild plot analysis was performed as previously described (Gustafsson, 1984). Dose ratios were obtained by comparing the concentrations of N E C A necessary to cause 20% enhancement of contractile responses.

3. Results Transmural stimulation of the rabbit bronchial spiral at 3-10 Hz, ~< 1 ms pulse duration and 10-100 pulses at 1-2 min intervals, elicited contractile responses that were blocked by atropine (3 X 10-7-10 6 M) or tetrodotoxin (3 x 10 7 M), and enhanced by eserine ( 4 x 10 7 M). These contractile responses were thus likely to be due to transmural stimulation of postganglionic cholineric nerves. W h e n the pulse duration was increased to 3 ms or more, atropine- and tetrodotoxin-resistant contractile responses were obtained. The latter responses were then regarded as due to direct smooth muscle stimulation. Adenosine (10--%10 4 M) dose dependently and reversibly enhanced the contractile response to transmural nerve stimulation in the rabbit bronchial circular muscle (figs. 1 and 2). The adenosine receptor type responsible for the enhancing effect of adenosine was determined by using adenosine analogues with a receptor selective action. The stable adenosine analogues N E C A (10 ~-10 -6 M) and L-PIA (10 6 M) enhanced contractile responses to transmural nerve stimulation in the bronchial muscle. N E C A was more

5 rain

THEO10uM

t 200

°

t 200

°

GUANETH 4tJM

t 200

°

t

°

200 ,uM ADO

Fig. 1. Contractile responses of rabbit isolated bronchial spiral in response to transmural stimulation at 10 Hz, 1 ms, 50 pulses at 1 rain intervals. Note enhancing effect of adenosine and its antagonism by theophylline, and the incapacity of guanethidine to modify the adenosine effect. Wash shown by dots.

• ADENOSINE J

200Z 0 t-j mO if)iT :71.Oz15[ o.,.O

rrm wO F-(D <

QNECA O L-PIA n= 3 - 1 0

T

T ~J' //

~/41

/

//'

10(

Z O O -LOG CONC.,M

Fig. 2. Concentration-dependent enhancement of contractile responses to transmural nerve stimulation (3-10 Hz, 0.3-1.0 ms. 30-100 pulses) in rabbit bronchial spirals as induced by adenosine (filled triangles), NECA (filled circles) and L-PIA (open circles). Symbols represent means, vertical bars indicate S.E.M. and n = number of experiments.

potent than L-PIA (P < 0.001) and adenosine (P < 0.01) (figs. 2 and 3a), whereas the stereoisomer D - P I A (10 s-10 ~' M) did not modify the contractile responses to transmural nerve stimulation (fig. 3a). L-PIA is known to inhibit contractile cholinergic responses in smooth muscle via a strong prejunctional inhibitory action (Paton, 1981; Gustafsson et al., 1985b). We studied the possibility that L-PIA was equipotent to or even more potent than N E C A in stimulating the enhancing receptors in bronchi but that the enhancing effect by L-PIA was masked by a simultaneously occurring strong prejunctional inhibitory effect. For this experiment the rabbit bronchial spirals were pretreated with L-PIA 10 ~' M just before the application of N E C A . The enhancing effect by N E C A ( 1 0 7 M, 3 × 1 0 - 7 M) was not obliterated by L-PIA (fig. 4). The adenosine receptor antagonist and phosphodiesterase inhibitor theophylline (10 5-10 4 M) reversibly antagonized the enhancing effect of adenosine on contractile responses to transmural nerve stimulation in the bronchial spirals (fig. 1). However, smooth muscle relaxation occurred simultaneously (fig. 1), making it difficult to compare the efficacy of antagonism at different doses and a Schild plot could not be obtained. We

182 10min

A)

1 0 8 1 ~ 71{36

10`8

L-PIA

l ( J 3x107

10810710 6 (M)

10~6

D PIA

NECA

10rain I I PSOTIOaM

NECA

NECA

NECA

10 mm

NECA I()6MD,

1

c/

t

ACH10-7

"I

ACHIO 7

" I

ACH 1() 7

"

T"

ACH 1(57 (M)

Fig. 3. Contractile responses to transmural nerve stimulation (10 Hz, 1 ms, 10 pulses at 1 min intervals in panels A and B, and 100 pulses at 2 min intervals in panel C) in rabbit bronchial circular smooth muscle. In the experiment in panel C contractions were also evoked by repeated application of acetylcholine (ACH), 10 "7 M. Wash shown by dots. (A) Enhancing effects by L-PIA and NECA but lack of effect by D-PIA(10 s-10 ~ M). (B) Enhancing effect by NECA (10 ~' M) and its antagonism by 8-p-sulfophenyltheophylline (PSoT) (10 4 M). (C) Enhancement of contractile responses to nerve stimulation during NECA 10 {' M and incapacity of NECA to enhance the contractile response to acetylcholine.

I

10 m l n

I

l

t

t

10-7 3 x l o 7 NECA

I

L-PIA

therefore used the analogue 8-p-sulfophenyltheophylline as ad en o si n e receptor antagonist. 8-pS u l f o p h e n y l t h e o p h y l l i n e dose d e p e n d e n t l y and reversibly antagonized the e n h a n c i n g effect of N E C A (fig. 3b). Th e Schild plot thus obtained showed a correlation coefficient of 0.95, a pA 2 corresponding to 2.7 x 10 ~' M of antagonist, and a slope of 0.97, indicating c o m p e t i t i v e a n t a g o n i s m (fig. 5). Th e possible i n v o l v e m e n t of histamine in the e n h a n c i n g effect by adenosine was studied by using m e p y r a m i n e 3 × 10 ~ M. M e p y r a m i n e at this conc e n t r a l i o n did not affect the contractile responses to transmural nerve stimulation and did not antagonize the e n h a n c i n g effect by N E C A (10 ~' M), but totally abolished a p r o f o u n d contractile effect of exogenously applied histamine (3 × 10 7 M, ,1 = 4). We also studied whether the e n h a n c i n g effect by ad en o si n e was due to m o d u l a t i o n of adrenergic mechanisms. The e n h a n c i n g effect elicited by ad en o si n e and its analogues was not an t ag o n i zed by 20 min p r e i n c u b a t i o n of rabbit bronchial spirals with g u a n e t h i d i n e 4 × 10 -~ M (n = 3) (fig. 1). G u a n e t h i d i n e at this c o n c e n t r a t i o n has been shown to c o m p l e t e l y block adrenergic transmission in e.g. guinea-pig ileum (Gustafsson et al., 1980), but it neither modified the contractile repsonses to transmural nerve stimulation nor blocked the e n h a n c e m e n t by adenosine in the rabbit bronchi. A possible i n v o l v e m e n t of substance P or a r a c h i d o n i c acid metabolites was studied by preinc u b a t i n g the p r e p a r a t i o n s with capsaicin 3 × 10 s M (n = 3) or eicosatetraynoic acid ( E T A ) 3 X 10 5

106M

t

10 7 3x107 NECA

10 7 3x10 7 NECA

C ONC M

Fig. 4. Contractile responses to transmural nerve stimulation (3 Hz, 1 ms, 30 pulses at 2 rain intervals) in a rabbit bronchial spiral. Note the enhancing effect by NECA (10 v M, 3×10 7 M), and that L-PIA 10 6 M did not obliterate the enhancing effect by NECA. Wash shown by dots.

183 /

20

¸

<

U uJ z I0i

@

:< . . . . . . . . . . . . . .

/

- L O G M. ANTAGONIST

Fig. 5. Schild plot for 8-p-sulfophenyhheophylline (PSeT)

antagonism of NECA-induced enhancement of contractile responses to transmural nerve stimulation (3-10 Hz, 0.3-1.0 ms, 30-100 pulses) in rabbit bronchial spirals, Curved broken lines indicate the 95% confidence limits for the regression line; n=9.

M (n = 6) for 20 min. Capsaicin at this concentration could be expected to release substance P and block neurotransmission at unmyelinated sensory neurons (Bj6rkroth, 1983) but it modified neither the contractile responses to transmural nerve stimulation nor the enhancing effect of NECA 10 6 M in these experiments. ETA at the concentration used is known to inhibit arachidonic acid metabolism and thus can profoundly inhibit arachidonate-dependent responses in certain smooth muscles (Gustafsson et al., 1980). ETA 3 × 10-s M showed only a weak inhibitory action in the bronchi on the contractile response to transmural nerve stimulation (remaining at 86 + 3.4% of control, n = 6) and did not antagonize the enhancing effect by N E C A (10 6 M). Interference with inhibitory neurotransmission seems uniikely as explanation for the adenosine effect since relaxation in response to transmural nerve stimulation could not be evoked in rabbit bronchi, despite several trials with guanethidine, atropine, BaC12 and histamine application• These same treatments very effectively revealed neuronally induced relaxations in a number of other smooth muscle preparations (data not shown). The effect of N E C A on contractions elicited by

exogenously applied acetylcholine and carbachol was studied to determine whether the enhancing effect by adenosine and its analogues was due to a postjunctional enhancement of released acetylcholine. NECA 10 -6 M did not enhance acetylcholine (10 7 M)-induced contractions (fig. 3c). The mean contractile response elicited by carbachol at 3 × 10- v M was 62_+ 9.6% (n = 9) of the contractile response to transmural nerve stimulation (10 Hz, 1.0 ms and 100 pulses at 2 rain intervals). After N E C A (10 6 M) the contractile response to carbachol (3 × 10 v M) remained at 61 _+ 10.3% (n = 9) of the contractile response to nerve stimulation in untreated preparations. Thus, NECA did not enhance the contractions elicited by exogenously applied carbachol. After application of tetrodotoxin, 3 × 10 ? M, the smooth muscle of the bronchial spiral was activated via direct muscle stimulation (10 Hz, 5 ms pulse duration and 100 pulses at 2 rain intervals). The contractile responses to electrical stimulation were not enhanced by NECA (10 s-10 ~' M) (n = 5).

4. Discussion The present observations demonstrate that adenosine could enhance the contractile responses elicited by transmural nerve stimulation of postganglionic cholinergic nerves in rabbit bronchial smooth muscle, This effect was most likely exerted via an action at A 2 receptors since NECA was more potent than L-PIA and adenosine. The effect is in all probability extracellular, since NECA and L-PIA are not agonists at the intracellular 'P-site' for adenosine action (cf. Daly, 1982). The possibility that a strong enhancing effect by L-PIA was masked by a simultaneously occurring prejunctional inhibition of contractile responses was investigated by studying the effect elicited by NECA after preincubation of the preparation with L-PIA at a 10-fold higher concentration. The enhancing effect elicited by NECA was however not obliterated. Thus, the potency difference of NECA and L-PIA in enhancing contractile responses is real, and not due to an interfering prejunctional action by L-PIA. The enhancement by adenosine could not be

184 ascribed directly to postjunctional potentiation of the effect of released acetylcholine, since contractions elicited by exogenously administered acetylcholine or carbachol were not enhanced by NECA. Since tetrodotoxin blocked the enhancement, the effect of the adenosine derivatives is dependent on fast sodium channels, Thus, adenosine might act prejunctionally either to increase cholinergic transmitter release or to increase the release of some other unknown agent and then would act postjunctionally. A third possibility, that fast sodium channels are involved in the postjunctional response to nerve stimulation but not to application of exogenous cholinomimetics cannot be excluded. As a fourth possibility, adenosine might, as mentioned above, inhibit the release of an unknown inhibitory mediator. However, since we could not demonstrate inhibitory transmission in the bronchi, in spite of following several different approaches, this mechanism seems less likely at the moment. It has been shown that antigen can induce adenosine release from rat lung and in human asthmatic subjects (Fredholm, 1981; Mann et al., 1983), and that adenosine can potentiate the antigen-induced release of histamine from human lung mast cells via an action at A 2 receptors (Hughes et al., 1984). It has also been shown that theophylline can antagonize the adenosine-induced histamine release in rat mast cells (Sydbom and Fredholm, 1982), However, mepyramine did not antagonize the enhancing effect by NECA in our experiments and thus adenosine-induced histamine release is not likely to explain the enhancing effect of adenosine and its analogues. The enhancing effect of adenosine and N E C A was not antagonized by pretreatment with guanethidine, making it unlikely that the enhancing effect of adenosine is exerted via interference with adrenergic mechanisms. Furthermore, since capsaicin or ETA did not antagonize the enhancing effect by NECA, the release of substance P or of arachidonic acid metabolites is not likely to be responsible for the enhancing effect by adenosine in the rabbit bronchi. The exact mechanism for adenosine enhancement can thus not yet be identified, although there is obviously pre- or postjunctional interference with

cholinergic neurotransmission, via a sodiumdependent mechanism. Theophylline and 8-p-sulfophenyltheophylline antagonized the enhancement of contractile responses elicited by the purines. The antagonism obtained with the sulfo compound is especially interesting since 8-p-sulfophenyltheophylline has been shown to be a competitive adenosine receptor antagonist without direct smooth muscle relaxing effect (Gustafsson, 1984). The Schild plot for 8-psulfophenyltheophylline versus NECA showed a slope not different from unity indicating that 8-psulfophenyltheophylline is a competitive receptor antagonist of the enhancing effects of N E C A in rabbit bronchial smooth muscle. The pA 2 value suggests that it is as potent an antagonist at the A 2 receptor in the rabbit bronchi as it is at the prejunctional A~ receptor in guinea-pig ileum (Wiklund and Gustafsson, 1985). The lack of direct smooth muscle relaxing effect by 8-psulfophenyltheophylline is likely due to its poor penetration of cell membranes because of its permanent charge in aqueous solution. The 8sulfophenylxanthines should also be poor inhibitors of phosphodiesterases (Daly et al., 1985). Thus, our findings suggest that theophylline can antagonize the enhancing effect elicited by adenosine and its analogues without acting as a phosphodiesterase inhibitor. This is interesting since it has been proposed that phosphodiesterase inhibition is not an obligatory mechanism of action for methylxanthines in asthma (cf. Cushley et al., 1984). Since phosphodiesterase inhibition also causes smooth muscle relaxation, a combination of adenosine antagonism and phosphodiesterase inhibition might thus be valuable characteristics in a bronchodilating compound. In conclusion, this study showed that adenosine could enhance the contractile responses to transmural nerve stimulation in rabbit bronchial smooth muscle via an action at extracellular adenosine A , receptors. It also shows that purine action at this receptor is blocked by methyl xanthines in a competitive fashion. These findings may be of great physiological interest since antagonism of endogenous adenosine may be a useful action by xanthines in asthma (Cushley et al., 1983; Gustafsson, 1983: Holgate et al., 1984).

185

Acknowledgements Supported by The Swedish Society of Medicine, the Swedish National Institute of Environmental Medicine and the Swedish Medical Research Council (project 14X-4342).

References Baer, H.P., M.J. Muller and R. Vriend, 1983, Adenosine receptors in smooth muscle, in: Physiology and Pharmacology of Adenosine Derivatives, eds. J.W. Daly, Y. Kuroda, J.W. Phillis, H. Shimizu and M. Ui (Raven Press, New York) p. 77. Bj6rkroth, U., 1983, Inhibition of smooth muscle contractions induced by capsaicin and electrical transmural stimulation by a substance P antagonist, Acta Physiol. Scand. 118, Suppl. 515, 11. Bruns, R.F., J.W. Daly and S.H. Snyder, 1980, Adenosine receptors in brain m e m b r a n e s : Binding of N 6cyclohexyl[3Hladenosine and 1,3-diethyl-8-[3H]phenyl xanthines, Proc. Natl. Acad. Sci. U.S.A. 77, 5547. Cushley, M.J., A.E. Tattersfield and S.T. Holgate, 1983, Inhaled adenosine and guanosine on airway resistance in normal and asthmatic subjects, Br. J. Clin. Pharmacol. 15, 161. Cushley, M.J.. A.E. Tattersfield and S.T. Holgate, 1984, Adenosine-induced bronchoconstriction in asthma, Am. Rev. Respir. Dis. 129, 380. Daly, J.W., 1982, Adenosine receptors: Targets for future drugs, J. Med. Chem. 25, 197. Daly, J.W., W. Padgett, M.T. Shamim, P. Butts-Lamb and J. Waters, 1985, 1,3-Dialkyl-8-(p-sulfophenyl)xanthines: Potent water-soluble antagonists for A 1 and Ae-adenosine receptors, J. Med. Chem. 28, 487. Fredholm, B.B., 1981, Release of adenosine from rat lung by antigen and compound 48/80, Acta Physiol. Scand. 111, 507. Fredholm, B.B., K, Brodin and K. Strandberg, 1979, On the mechanism of relaxation of tracheal muscle by theophylline and other cyclic nucleotide phosphodiesterase inhibitors, Acta Pharmacol. Toxicol. 45, 336. Grundstr6m, N., R,G.G. Andersson and J.E.S. Wikberg, 1981, Investigations on possible presynaptic effect of adenosine and noradrenaline on cholinergic neurotransmission in guinea pig trachea, Acta Pharmacol. Toxicol. 49, 158. Gustafsson, L.E., 1981, Influence of adenosine on responses to vagal nerve stimulation in the anaesthetized rabbit, Acta Physiol. Scand. 111,263. Gustafsson, L.E., 1983, Endogenous purines as modulators of autonomic cholinergic neurotransmission, in: Regulatory Function of Adenosine, eds. R.M. Berne, T.W. Rail and R. Rubio (Martinus Nijhoff, The Hague) p. 514.

Gustafsson, L.E., 1984, Adenosine antagonism and related effects of theophylline derivatives in guinea pig ileum, Acta Physiol. Scand. 122, 191. Gustafsson, L.E., P. Hedqvist and G. Lundgren, 1980, Pre- and postjunctional effects of prostaglandin E 2, prostaglandin synthetase inhibitors and atropine on cholinergic neurotransmission in guinea pig ileum and bovine iris, Acta Physiol. Scand. 110, 401. Gustafsson, L.E., N.P. Wiklund and B. Cederqvist, 1985a, Evidence for adenosine as a smooth muscle stimulating agent in bronchi and the gastrointestinal tract, and its antagonism by xanthines, in: Antiasthmatic Xanthines and Adenosine, eds. C.G.A. Persson and K.-E. Andersson (Elsevier Press) (in press). Gustafsson, L.E., N,P. Wiklund, J. Lundin and P. Hedqvist, 1985b, Characterization of pre- and postjunctional adenosine receptors in guinea-pig ileum, Acta Physiol. Scand. 123, 195. Holgate, S.T., J.S. Mann and M,J. Cushley, 1984, Adenosine as a bronchoconstrictor mediator in asthma and its antagonism by methylxanthines, J. Allergy Clin. lmmunol. 74, 302. Hughes, P.J., S.T, Holgate and M.T. Church, 1984, Adenosine inhibits and potentiates IgE-dependent histamine release from human lung mast cells by an A2-purinoceptor mediated mechanism, Biochem. Pharmacol. 33, 3847. Jeppson, A.-B., U. Johansson and B. Waldeck, 1982, Dissociation between the effects of some xanthine derivatives on the tracheal smooth muscle and on skeletal muscle, Acta Pharmacol. Toxicol. 51, 115. Karlsson, J.-A. and C.G.A. Persson, 1981, Influence of tracheal contraction on relaxant effects in vitro of theophylline and isoprenaline, Br. J. Pharmacol. 74, 73. Mann, J.S,, A.G. Renwick and S.T. Holgate, 1983, Antigen bronchial provocation causes an increase in plasma adenosine levels in asthma, Clin. Sci. 65, 22P. Paton, D.M., 1981, Structure-activity relations for presynaptic inhibition of noradrenergic and cholinergic transmission by adenosine: Evidence for action on A~ receptors, J. Auton. Pharmacol. 1, 287. Persson, C.G.A., 1980, Some pharmacological aspects on xanthines in asthma, J. Respir. Dis. Suppl. 109, 61, 7. Rail, T.W., 1980, The xanthines, in: The Pharmacological Basis of Therapeutics, eds. A. Goodman Gilman, L.S. Goodman and A. Gilman (Macmillan, New York) p. 592. Sydbom, A. and B.B. Fredholm, 1982, On the mechanism by which theophylline inhibits histamine release from rat mast cells, Acta Physiol. Scand. 114, 243. Wiklund, N.P. and L.E. Gustafsson, 125, Pre- and postjunctional modulation of cholinergic neuroeffector transmission by adenine nucleotides. Experiments with agonist and antagonist, Acta Physiol. Scand. 125 (in press).