Theophylline antagonizes some effects of purines in the intestine but not those of intramural inhibitory nerve stimulation

A2A and A3 receptors mediate the adenosine-induced relaxation in spontaneously active possum duodenum in vitro

1. The aim of this study was to define the P1 purinergic receptors that regulate spontaneous or adenosine-induced duodenal motor activity. 2. Spontaneous contractile activity was recorded isometrically from possum longitudinal duodenal muscle strips. Adenosine (0.5 micro M-1 mM) was administered noncumulatively and repeated after pretreatment with a P1 antagonist or tetrodotoxin (TTX, 1 micro M), (n=4-7). Antagonists used were: A(1), 8-cyclopentyl-1,3-dipropylxanthine (DPCPX, 10 nM); A(2A), 8-(3-chlorostyryl)caffeine (CSC, 10 micro M); A(2B), 3-isobutyl-8-pyrrolidinoxanthine (IPDX, 10 micro M); A(3), 9-chloro-2-(2-furanyl)-5-[(phenylacetyl) amino][1,2,4]-triazolo(1,5-c)quinazoline (MRS1220, 10 micro M). Changes in activity are expressed as percentage of baseline. Statistical analysis utilised nonparametric tests. 3. Adenosine (n=34) induced a long-lasting, concentration-dependent decrease in activity by 55.6+/-3.2% area under curve (AUC), 47.3+/-4.0% contraction amplitude, 31.6+/-3.6% basal tension and 10.4+/-1.7% contraction frequency (all P<0.001). The adenosine-induced decrease in contraction amplitude was blocked by CSC (P<0.01) or inhibited by MRS1220 (P<0.03) pretreatment, but not modified by TTX, DPCPX or IPDX pretreatment. 4. Adenosine antagonists modified spontaneous contractile activity. Pretreatment with DPCPX or CSC increased basal tension, whereas IPDX or MRS1220 pretreatment decreased contractile activity. 5. In conclusion, exogenous adenosine reduced duodenal longitudinal motor activity via A(2A) and A(3) receptors. Our findings suggest that endogenous purines may modulate spontaneous duodenal motor activity.

Characterization and tissue location of the neural adenosine receptor in the rat ileum

1. The aim of the present investigation was to characterize and determine the tissue location of the adenosine receptors present in the rat ileum using a method that detects drug action on the cholinergic nerves innervating the longitudinal and circular muscles. 2. The non-selective adenosine agonist, NECA (10 and 100 nM) caused significant concentration-related reductions in the circular muscle responses to transmural stimulation over the frequency range of 2.5-40 Hz, but did not affect the responses of the longitudinal muscle, nor did it reduce the muscle responses of the guinea-pig ileum. 3. The affinity order of antagonists at inhibiting the effect of NECA on the circular muscle was: CPDPX>8-PT>DMPX with apparent pA2 values of 9.31, 7.54 and 5.63 respectively. CPDPX (10-100 nM) caused parallel displacements of the concentration-effect curves to CPA with a pKb value of 9.15 and Schild slope of 1.03. 4. The agonists previously tested in the rat jejunum peristaltic reflex preparation were also shown to inhibit responses of the rat ileum in the following decreasing order of potency: CPA>NECA>2-CADO>R-PIA>S-PIA>>PAA. In addition, CHA and CCPA were also potent agonists. NECA (100 nM) and CPA (32 nM) did not inhibit carbachol (1 microM)-induced tone of tissues pre-treated with TTX (1 microM). 5. In conclusion, the rat ileum contains inhibitory A1 adenosine receptors situated on cholinergic nerve endings innervating the circular muscle.

Characterization and tissue location of the neural adenosine receptor in the rat ileum

1. The aim of the present investigation was to characterize and determine the tissue location of the adenosine receptors present in the rat ileum using a method that detects drug action on the cholinergic nerves innervating the longitudinal and circular muscles. 2. The non-selective adenosine agonist, NECA (10 and 100 nM) caused significant concentration-related reductions in the circular muscle responses to transmural stimulation over the frequency range of 2.5-40 Hz, but did not affect the responses of the longitudinal muscle, nor did it reduce the muscle responses of the guinea-pig ileum. 3. The affinity order of antagonists at inhibiting the effect of NECA on the circular muscle was: CPDPX>8-PT>DMPX with apparent pA2 values of 9.31, 7.54 and 5.63 respectively. CPDPX (10-100 nM) caused parallel displacements of the concentration-effect curves to CPA with a pKb value of 9.15 and Schild slope of 1.03. 4. The agonists previously tested in the rat jejunum peristaltic reflex preparation were also shown to inhibit responses of the rat ileum in the following decreasing order of potency: CPA>NECA>2-CADO>R-PIA>S-PIA>>PAA. In addition, CHA and CCPA were also potent agonists. NECA (100 nM) and CPA (32 nM) did not inhibit carbachol (1 microM)-induced tone of tissues pre-treated with TTX (1 microM). 5. In conclusion, the rat ileum contains inhibitory A1 adenosine receptors situated on cholinergic nerve endings innervating the circular muscle.

Prejunctional autoinhibition of purinergic transmission in circular muscle of guinea-pig ileum; a mechanism distinct from P1-purinoceptor activation

The role of prejunctional purinoceptors (P1-subtype) in the control of ATP-release from inhibitory motoneurons was investigated electrophysiologically, by studying fast purinergic inhibitory junction potentials (IJPs) in guinea-pig ileal circular muscle. Pressure ejections of adenosine and ATP (but not of alpha,beta-methylene ATP) onto circular muscle depressed the amplitude of fast IJPs, indicating the presence of prejunctional P1-purinoceptors. An adenosine (A1/2)-receptor antagonist, theophylline (10(-8)-10(-4) M), increased the amplitude of fast IJPs in a dose-related manner (EC50 = 17.5 microM), suggesting the existence of a basal 'adenosine tone' that regulated ATP-release from ileal motoneurons. However, three methylxanthine derivatives, caffeine (10(-8)-10(-4) M), 3-isobutyl-1-methylxanthine (IBMX; 10(-8)-10(-4) M) and the potent A1-receptor antagonist 1,3-dipropyl-8-(2-amino-4-chlorophenyl)-xanthine (DPCPX; 10(-8)-10(-4) M), failed to potentiate fast IJPs and placed in doubt the existence of this inhibitory adenosine tone. Caffeine and IBMX, but not DCPCX, hyperpolarised ileal circular muscle in a dose-related manner and reduced IJP-amplitude; DPCPX did not alter the amplitude of IJPs. The non-specific inhibitor of phosphodiesterases, Ro-20-1724 (5 x 10(-7)-5 x 10(-5) M), increased the amplitude of fast IJPs, mimicking the actions of theophylline. To this extent, facilitation of inhibitory transmission appeared to involve phosphodiesterase inhibition and modification of intra-axonal cAMP levels and phosphorylation of intra-axonal protein kinases. The phenomenon of IJP rundown, presumed to be a manifestation of prejunctional autoinhibition, was studied using theophylline and DPCPX as A1-receptor antagonists.(ABSTRACT TRUNCATED AT 250 WORDS)

The relaxing response of the isolated rat duodenum to nicotine

1. The relaxing responses of the atropine-treated isolated rat duodenum to 12 microM nicotine (N) and transmural electrical stimulation (S) have been compared. 2. The effects of both N and S were unaffected by 3-30 microM guanethidine or previous reserpinization (1 mg/kg, i.p. during 5 days). 3. The effects of N and S were completely blocked by 0.5-2.0 mM procaine, 0.1 mM butacaine or 0.3-1.2 microM tetrodotoxin. 4. The relaxing response to N but not that to S was blocked by 0.3 mM hexamethonium or 3 microM mecamylamine and absent in the nicotine-desensitized preparation. 5. The effects of both N and S were unaffected in the adenosine-5'-triphosphate (ATP)-desensitized preparation. 6. It is suggested that the nonadrenergic and noncholinergic (NANC) relaxation of the rat duodenum in response to N and S is also nonpurinergic.

Role of adenine compounds in autonomic neurotransmission

Clearly adenine compounds exert numerous effects throughout the autonomic nervous system. The responses of various peripheral tissues to purines are summarized in Table 2. The evidence supporting a possible excitatory neurotransmitter function for ATP is very good in the vas deferens and good in both the bladder detrusor and certain blood vessels. ATP may also be an excitatory neurotransmitter in the colon, hepatocytes and frog atrium. These responses appear to be mediated by P2x-purinoceptors. There is good evidence supporting a role for ATP as an inhibitory neurotransmitter in the taenia coli and duodenum, and some support in the anal sphincter and possibly the rabbit portal vein; these responses appear to be mediated by P2y-purinoceptors. There is good evidence against ATP being an inhibitory neurotransmitter in the stomach fundic muscle and ileum. ATP (or more likely its metabolite adenosine) may act as an inhibitory neurotransmitter by interacting with postsynaptic P1-purinoceptors in cultured sympathetic neurones and also in the parasympathetic vesicle ganglion of the cat. It seems likely that ATP released from heart, platelets or vascular endothelium could be an endogenous relaxant of blood vessels through its actions on the endothelium. Although the addition of exogenous adenosine affects many tissues, evidence supporting modulatory functions for endogenous extracellular adenosine has only been clearly demonstrated in the ileum, gallbladder, vas deferens, fallopian tubes, kidney, blood vessels, carotid sinus, heart and adipose tissue. Both ATP and adenosine, released during periods of hypoxia or ischemia, could exert negative inotropic, chronotropic and dromotropic actions in the heart. In many cases, the potential sources of extracellular purines have not been established. This is particularly important when attempting to establish a neurotransmitter function for ATP in a tissue. For instance, the one outstanding piece of evidence required to confirm that ATP is an excitatory neurotransmitter released from sympathetic nerves in blood vessels is the unequivocal demonstration that it is, in fact, released from the sympathetic nerves when they are stimulated. To date, only the release of radiolabeled metabolites of ATP, possibly from post- rather than presynaptic sites, has been detected. Studies of the release of ATP are complicated by its rapid degradation extracellularly by ecto-ATPase. Unfortunately, there are no specific inhibitors of ecto-ATPase available at present, but one hopes that a suitable inhibitor will be developed shortly.(ABSTRACT TRUNCATED AT 400 WORDS)

Neural inhibition of gastrointestinal smooth muscle motility in guinea-pigs

The neural inhibition of the gastrointestinal (GI) smooth muscle motility was studied by means of electrical transmural nerve stimulation (ETNS) in guinea-pigs. In untreated muscle strips, ETNS induced four types of response, consisting of three basic components, i.e., contraction and relaxation during ETNS, and after-contraction. Following atropinization, all the responses changed to only one type, i.e., relaxation during ETNS followed by after-contraction. The relaxation of the muscle strips induced by ETNS after atropinization was not significantly reduced under superimposing guanethidine treatment in any site of the GI tract. The maximal relaxation of the muscle strips induced by ETNS after atropine and guanethidine treatments was not uniform throughout the GI tract. The relaxation of the gastric body and colon was greater than that of the jejunum and ileum. The extent of the relaxation was significantly different even in the colon. Theophylline and phentolamine did not reduce the ETNS-induced relaxation following atropine and guanethidine treatments in the distal colon and taenia coli. The findings suggest the following: the non-adrenergic inhibitory (NAI) nerve is the main postganglionic nerve to inhibit the GI smooth muscle motility; and the role of adrenergic nerve is of little importance in respect to postganglionic inhibition. The innervation of the NAI nerve seems not to be uniform throughout the GI tract. Theophylline and phentolamine are not the specific antagonists of the NAI nerve.

[3H]adenosine binding to bovine coronary arteries and myocardium

The characteristics of [3H]adenosine binding to bovine myocardial and coronary tissue were investigated using the microsomal fraction obtained by fractionated centrifugation. Binding experiments consisting of equilibrium, kinetic and competition studies were performed employing a filtration assay. The coronary microsomal fraction yielded two classes of binding sites with different affinities and binding capacities. The approximate binding parameters were: KD = 1.7 x 10(-7) M, Bmax = 1.2 pmol/mg protein, and KD = 2.6 x 10(-6) M, Bmax = 6.6 pmol/mg protein, respectively. Only one single class of binding sites was found in the myocardium as indicated by a linear Scatchard plot (KD = 3.7 x 10(-7) M, Bmax = 11.7 pmol/mg protein). In both tissues competition experiments performed with a number of adenosine analogues displayed similar specificity, i.e. high competing potency of adenosine nucleotides, whereas, 2-chloroadenosine, L-PIA, 2'-deoxyadenosine and theophylline showed little or no affinity. The results presented in this paper do not provide evidence for the predominance of typical Ra sites according to the definition of Londos et al. (Proc. Natl. Acad. Sci. U.S.A. 77, 2551, 1980) in bovine coronary arterial tissue or myocardium.

Adenosine and ATP effects on isolated guinea pig gallbladder

Effects of adenosine, ATP and several derivatives of adenosine were measured in isolated strips of guinea pig gallbladder. Adenosine caused relaxations which were antagonized by theophylline and potentiated by an inhibitor of adenosine uptake, 6-(1-hydroxy-5-nitrobenzylthio)-guanosine (HNBTG). Among several adenosine derivatives, 2-chloroadenosine and 5'-N-ethylcarboxy-midoadenosine were similarly effective while 1-N6-phenylisopropyladenosine was only a weak relaxant. None of the derivatives caused maximal relaxations at 100 microM, and thus absolute potencies could not be determined. ATP caused predominantly contractile effects, with relaxations sometimes being evident at high concentrations. Indomethacin abolished contractile effects of ATP, suggesting prostaglandin involvement, and only relaxations were evident in its presence. Adenosine deaminase abolished the effects of adenosine and partly reduced the relaxant effects of ATP in the presence of indomethacin. In view of the low potency of adenosine and ATP, physiological roles for these compounds in gallbladder motility are not readily evident.

Dilazep: an inhibitor of adenosine uptake with intrinsic calcium antagonistic properties

Concentrations of dilazep which were ineffective in altering the muscular tone of the guinea-pig taenia caeci (0.03, 0.3 microM) or the phasic mechanical activity of the rabbit proximal ileum (0.03 microM) markedly potentiated the inhibitory action of adenosine on both these parameters. Dilazep, 0.3 microM or greater, dose-dependently inhibited the mechanical activity of the proximal ileum. This inhibitory action was probably mediated by more than one mechanism, as shown by the fact that theophylline (50, 100 microM) antagonized the effect at lower dilazep concentrations (up to 3 microM) leaving essentially unchanged the response to higher concentrations (6, 10 microM). Similarly, the responses to low doses of dilazep were reduced after desensitization of the organ to adenosine, whilst the responses to higher doses were unaffected by this procedure. In a Ca2+-free, high-K+ medium, dilazep (1-10 microM) caused a parallel shift to the right of the Ca2+-induced contractions of the guinea-pig taenia caeci. Adenosine showed only slight Ca2+-antagonistic properties within the mM range of concentrations. These findings suggest that, at the higher concentration tested, dilazep exhibits Ca2+-antagonistic properties unrelated to its adenosine-mediated mode of action.

Purine receptors in the guinea-pig internal anal sphincter

1 In the isolated internal anal sphincter of the guinea-pig, adenosine 5'-triphosphate (ATP) and adenosine induced a concentration-dependent and tetrodotoxin-insensitive relaxation. 2 Pretreatment with theophylline (25-50 microM) had no significant effect on the concentration-response curves obtained with either purine compound. 3 Reactive blue 2 (25-100 microM) shifted the curve to ATP to the right in a dose-dependent fashion leaving that to adenosine unaltered. The antagonism appeared to be non-competitive. 4 Neither reactive blue 2 nor purine receptor occupation by ATP or adenosine altered the electrically-induced non-adrenergic, non-cholinergic inhibitory response. 5 The actions of ATP and adenosine in the guinea-pig internal anal sphincter appear to be mediated by separate receptors. These receptors are not involved in the nerve-mediated relaxation.

Neither a purine nor VIP is the mediator of inhibitory nerves of opossum oesophageal smooth muscle

Effects of stimulation of intramural nerves in the circular smooth muscle layer of the body of the oesophagus of the opossum (Didelphis marsupialis) were studied, simultaneously measuring the membrane potential of muscle cells using the sucrose-gap technique and contractions of the muscle. Electrical field stimulation of the preparation, superfused with Krebs solution at 27 degrees C, induced a transient hyperpolarization of the smooth muscle cell membrane (i.j.p.) followed by a transient depolarization on which muscle action potentials were often superimposed. The muscle did not develop active tension spontaneously; it therefore did not relax during the i.j.p., but often contracted during the 'off' depolarization. The i.j.p. and the responses following it were characterized as mediated by intramural, non-adrenergic, non-cholinergic (n.a.n.c.) nerves. The i.j.p. amplitude was reduced by raising the external K concentration. When Cl was replaced by isethionate, a small hyperpolarization of the smooth muscle cell membrane ensued along with a comparable small reduction of the i.j.p. amplitude. The 'off' activity following the i.j.p. disappeared completely in Cl-free medium. Apamin (10(-7)-10(-5) M) did not influence this preparation nor the i.j.p. Adenosine and its related adenine nucleotides in concentrations up to 10(-3) M hardly affected the preparation. Prolonged superfusion with adenosine and 6-chloroadenosine revealed a gradually increasing attenuation of the i.j.p. Exogenously applied vasoactive intestinal polypeptide (VIP) induced rhythmic depolarizations of the smooth muscle cell membrane and spontaneous contractions, which were insensitive to neurotoxins but absent in Cl-free media. Field stimulation in the presence of VIP caused an i.j.p. which transiently interrupted the VIP-induced contractile responses. It is concluded that the inhibitory mediator of the intramural n.a.n.c. nerves present in the circular smooth muscle layer of the opossum oesophagus is neither a purine nor VIP. The i.j.p. may result from a selective increase in K permeability of the smooth muscle cell membrane, the 'off' depolarization may involve an increase in the Cl ion conductance. The suggestion is made that the release of neurotransmitter from intramural n.a.n.c. nerves is modulated presynaptically via P1-purinoceptors and that VIP is a likely candidate for an excitatory transmitter, released simultaneously with the inhibitory transmitter.

Purine receptors in the rat anococcygeus muscle

Adenosine triphosphate (ATP) caused contraction of the resting isolated rat anococcygeus muscle. Non-phosphorylated purines did not cause contraction of the resting muscle but did so in muscles in which the tone was raised by carbachol or guanethidine. Adenosine, (-)N6-phenylisopropyladenosine (PIA) and 5'-N-ethyl-carboxamide adenosine (NECA) were approximately equipotent, and these responses were not prevented by theophylline, quinidine, 2,2'-pyridylisatogen tosylate, phentolamine, methysergide, dipyridamole, hexobendine or indomethacin. The contractions became smaller as muscle tone progressively declined, and it is suggested that this effect may explain the apparent blockade of ATP responses by indomethacin reported previously. Adenosine, 2-chloroadenosine, ATP, PIA and NECA inhibited contractile responses of the anococcygeus to field stimulation of the excitatory adrenergic innervation. This inhibitory action was blocked by theophylline, and as PIA was easily the most potent purine tested, it may involve activation of an A1/Ri receptor. It is also argued, however, that the A/R scheme of classification may be inappropriate for the description of responses of intact tissues. As response to noradrenaline were not changed by the purines, the inhibitory effect on stimulation-evoked contractions is probably mediated at a presynaptic site. None of the purines tested had any effect on the neurally mediated inhibition of the anococcygeus which is seen when intrinsic tone is raised and the excitatory adrenergic nerves are blocked.

Mode of action of ATP on propulsive activity in rabbit colon

ATP induced a concentration-dependent reduction of the velocity of propulsion in isolated segments of rabbit colon as assessed by the aboral displacement of an intraluminal rubber balloon, and delayed the onset of the propulsive wave. ATP depressed both the reflex contraction of the circular coat above the distended balloon and the response of the circular muscle to transmural (cholinergic) stimulation. On the contrary, ATP (up to 200 muM), while causing relaxation of the circular muscle, had no effect on either the muscular contractile response induced by carbachol and histamine or the non-adrenergic inhibitory responses elicited by electrical stimulation and by radial distension of the gut wall. Within the concentration range used (10-200 muM), ATP concentration-depression curves for propulsion and transmural excitatory stimulation were shifted to the right in the presence of theophylline (10 muM). Theophylline, however, had no influence on either the direct inhibitory action of ATP on circular smooth muscle or the non-adrenergic relaxation in response to electrical stimulation. These data are consistent with the concept that at least two populations of purinergic receptors are present in intestinal tissue. Those populations located presynaptically, unlike those located postsynaptically, are blocked by theophylline. Since the contractile machinery does not appear to be affected by ATP concentrations up to 200 muM, the mechanism by which ATP impairs propulsive activity is probably dependent on activation of presynaptic purinergic receptors located on the nervous pathways subserving the wave of contraction, without having any appreciable influence on descending inhibition.

Effect of arylazido aminopropionyl ATP (ANAPP3), on ATP responses of isolated guinea pig smooth muscle

Arylazido aminopropionyl ATP (ANAPP3), a photoaffinity analogue of adenosine 5'-triphosphate, photoactivated with visible light (+hv), specifically and irreversibly antagonized ATP contractions of the guinea pig vas deferens. ANAPP3 (30 microM) antagonized responses to exogenously added ATP in untreated, and in tissues pretreated with indomethacin (2.9 microM) and 6-(2-hydroxy-5-nitrobenzyl)-thio guanosine (10 microM). It was of interest to see if this pharmacological antagonist of ATP could be used to assess the validity of the purinergic nerve hypothesis by allowing a differentiation between, or proof of the identity of, responses to ATP and the non-adrenergic inhibitory transmitter in guinea pig stomach fundus. After photoactivation (+hv) in the organ bath and subsequent washout, ANAPP3 (30 and 100 microM) failed to antagonize relaxant responses to ATP (1.0 - 1000 microM) in fundic strips. In addition ANAPP3 failed to antagonize ATP-induced inhibition of the twitch response in electrically stimulated guinea pig ileum longitudinal muscle strips. We conclude that ANAPP3 does not antagonize all actions of ATP, which may limit its usefulness in assessing the above hypothesis. Results with this compound suggest that ATP excitatory receptors may differ from those mediating relaxation and other ATP actions.

Attempt to antagonized the stimulatory effect or ATP on insulin secretion

An attempt was made to antagonize the stimulatory effect of ATP on insulin secretion from the isolated perfused rat pancreas. The insulin secretory effect of ATP does not seem to be mediated by cholinergic or beta-adrenergic receptors since neither atropine 0.3 micro mol/1 nor propranolol 1 micro mol/1 had any antagonistic action on insulin secretion induced by ATP (16.5 micro mol/1). Theophylline (50 micro mol/1) did not antagonize the insulin secretion evoked by ATP (16.5 micro mol/1). Apamin (10 nmol/1) was also without antagonistic action. 2-2'Pyridylisatogen tosylate (5 micro mol/1) had not effect on insulin secretion induced by glucose )8.33 mmol/1) or acetylcholine (0.5 micro mol/1 and 0.1 micro mol/1) but inhibited the insulin secretory effect of ATP (16.5 micro mol/1). Thus, the antagonism of 2-2'pyridylisatogen for ATP seems selective. We conclude that a purinoceptor of the P2-type is likely to be present on the B-cell of the rat pancreas.

Theophylline interferes with the modulatory role of endogenous adenosine on cholinergic neurotransmission in guinea pig ileum

The ability of theophylline and other phosphodiesterase inhibitors to alter contractile responses to cholinergic nerve stimulation was investigated in isolated longitudinal muscle of the guinea pig ileum. Theophylline in low concentrations (10-100 microM), having no or little effect on measured phosphodiesterase activity, antagonized inhibitory effects of exogenous adenosine. In higher concentrations (0.1-10 mM), shown to be effective in inhibiting phosphodiesterase, theophylline as well as a "pure" cAMP phosphodiesterase inhibitor, ZK 62, 711, inhibited contractile responses. Dipyridamole and dilazep, inhibitors of adenosine inactivation, and also selective inhibitors of cAMP and cGMP phosphodiesterase, respectively, were found to enhance effects of exogenous adenosine and to cause a marked leftward shift to adenosine threshold dose. When dipyridamole and dilazep by themselves had inhibitory effects these could be antagonized by theophylline, suggesting an action through increased levels of endogenous adenosine. As a further indication of endogenous adenosine modulating neurotransmission low concentrations of theophylline enhanced responses to transmural stimulation. Endogenous purine concentrations in tissues and bath media were measured by HPLC. Because of tissue and microbial adenosine inactivation direct estimates of extracellular adenosine concentration could not be obtained. However, adenosine levels increased during transmural stimulation, and during inhibition of adenosine inactivation were sufficient, even in the bath medium, to interfere with the cholinergic neurotransmission.

Relaxation of guinea-pig fundic strip by adenosine, adenosine triphosphate and electrical stimulation: lack of antagonism by theophylline or ATP treatment

1 Theophylline relaxed isolated strips of guinea-pig stomach fundus in a dose-dependent manner; above 50 to 100 microM responses showed no fade for up to 90 min. 2 Relaxant responses to adenosine, adenosine triphosphate (ATP), noradrenaline, and to electric field stimulation of non-adrenergic inhibitory nerves were not affected in a significant manner in the presence of 50 microM theophylline. 3 In tissues which showed complete fade of initial responses in the continued presence of 50 microM ATP, the effects of stimulation of non-adrenergic inhibitory nerves remained unaltered, suggesting that the ATP receptor has no function in non-adrenergic inhibitory transmission in this tissue. 4 These findings are opposite to those of Okwuasaba, Hamilton & Cook (1977), who claimed that 50 microM theophylline almost fully inhibited relaxation induced by adenosine ATP and nerve stimulation and that ATP-induced fade also abolished sensitivity to inhibitory nerve stimulation.