Potentiation by dilazep on the negative inotropic effect of adenosine on guinea-pig atria.

1 Dilazep, a coronary dilator, has been reported to potentiate the negative inotropic and negative chronotropic responses of guinea-pig atria to adenosine. Studies were made on the mechanism of the potentiating action of dilazep with special reference to the degradation and uptake of adenosine. 2 The negative inotropic actions of adenosine and adenine nucleotides, such as ATP, ADP, AMP and cyclic AMP, on guinea-pig atria were selectively and dose-dependently augmented by dilazep at concentrations insufficient to produce any effect alone (0.01 to 1 microM). 3 Incubation of atrial tissue with 8.8 nM adenosine, containing 0.1 microCi of [3H]-adenosine, resulted in accumulation of [3H]-adenosine in the tissue; dilazep (0.01 to 1 microM) inhibited this accumulation. 4 Adenosine (10 microM to 10 mM) was degraded to inosine and hypoxanthine during incubation with atrial tissue; dilazep (0.1 to 10 microM) retarded the disappearance of adenosine and the formation of inosine and hypoxanthine. 5 These results suggest that dilazep potentiates the negative inotropic effect of adenosine on guinea-pig atria by preventing both its accumulation by atrial tissue and degradation by deaminase.     

Endothelin modulation of neuroeffector transmission in rat and guinea pig vas deferens

The effects of endothelin-1 (human, porcine) on contractions induced by transmural nerve stimulation, exogenous ATP or noradrenaline, and on the release of [3H]noradrenaline were studied in guinea pig and rat vas deferens. Endothelin enhanced nerve-induced contractile responses, increased basal muscle tone and increased the contractile response to exogenous ATP in both guinea pig and rat vas deferens. Endothelin did not affect the contractile responses to exogenous noradrenaline. The calcium channel blocker felodipine antagonized the stimulating effects of endothelin in the rat vas deferens, whereas blockade of lipoxygenase and cyclooxygenase pathways by a combination of BW 755C and indomethacin was without effect. In rat vas deferens preparations preincubated with [3H]noradrenaline, endothelin inhibited the 3H overflow induced by transmural stimulation, although the contractile responses were enhanced by endothelin. Pretreatment with forskolin or felodipine did not abolish the endothelin inhibition of radiotracer overflow. In conclusion, endothelin can modulate adrenergic and purinergic neuroeffector transmission in both guinea pig and rat vas deferens via inhibitory prejunctional and stimulant postjunctional mechanisms. The stimulant postjunctional effect seemed to predominate in our experiments.     

Release of endogenous ATP from the vasa deferentia of the rat and guinea-pig by the indirect sympathomimetic tyramine

1 Adenosine 5′triphosphate (ATP) as well as [³H]-noradrenaline ([³H]-NA) is released by perfusion of the vas deferens with the indirect sympathomimetic tyramine (100 μM); this result is consistent with the concept of sympathetic cotransmission. 2 While tyramine produced a strong contraction in the vas deferens of the rat, it had little mechanical action in the guinea-pig vas deferens. This appears to be largely because tyramine induces considerably lower levels of release of both ATP and NA from the guinea-pig vas deferens compared to that of the rat. Furthermore, NA released by tyramine appears to release ATP from a secondary pool in the rat vans deferens, but not that of the guinea-pig, since prazosin reduced the tryamine-induced release of ATP in the rat vas deferens. 3 α, β -Methylene ATP (α, β -meATP) increased both the spontaneous release of ATP and the tyramine-evoked efflux of ATP and [³H]-NA. The basal and tyramine-induced efflux of [³H]-NA was also enhanced by the α₁-adrenoceptor antagonist, prazosin, suggesting that prejunctional α₁-adrenoceptors may modulate neurotransmitter release.     

Effects of noradrenaline,the calcium ionophore A23187, forskolin,sodium nitroprusside and glibenclamide on the degradation of extracellular adenosine 5′-triphosphate by the rat isolated vas deferens

1 The effects of noradrenaline (NA), the calcium ionophore A23187, forskolin, sodium nitroprusside (SNP) and the K⁺-channel blocker glibenclamide on the degradation by ectonucleotidases of extracellular adenosine 5′-triphosphate (ATP) were studied in the rat vas deferens. 2 ATP (100 μm ) was rapidly broken down by the rat vas deferens with a half-life of 5.83 ± 0.40 min via adenosine 5′-diphosphate (ADP) and adenosine 5′-monophosphate (AMP), with the final degradation product being inosine and with little adenosine being detected in the samples. 3 Preincubation for 1 h with NA (10 μm ), A23187 (10 μm ), or glibenclamide (100 μm ) had no significant effect on the breakdown of ATP or the production of metabolites. However, both forskolin (10 μm ) and SNP (1 μm ) significantly increased the concentrations of AMP detected with time. In the case of SNP (1 μm ) there was also a significant reduction in the rate of production of inosine, while in the case of forskolin (10 μm ) there was a significant increase in the rate of removal of ATP. 4 These results suggest that preincubation with SNP may inhibit 5′-nucleotidase and so reduce the metabolism of AMP, while preincubation with forskolin may increase the activity of the ectonucleotidases responsible for production of AMP from ATP.     

Suppression by adenine compounds of histamine-induced contraction in isolated guinea pig tracheal rings

The effects of purine compounds on the contraction of isolated guinea pig tracheal pigs induced by histamine were studied in vitro. Adenosine, at 0.1-1 mM concentrations, suppressed the contraction induced by histamine in the isolated guinea pig tracheal rings in a concentration-dependent manner. ATP, AMP and adenine were also effective in blocking tracheal constriction induced by histamine. The other purine compounds such as inosine, guanosine, guanine and hypoxanthine had no effect on tracheal smooth muscle. Coformycin, which is a competitive inhibitor of adenosine deaminase, did not potentiate adenosine action. Dipyridamole, which blocks adenosine uptake, did not impair the suppression caused by adenosine. These results suggest that adenosine action on tracheal smooth muscle cells may be mediated by direct action on cell surface receptors. In addition, it seems that adenosine is an important modulator of tracheal smooth muscle cell function.     

Analgesic,receptor binding,and peripheral opioid activities of synthetic dermorphin-dynorphin hybrid peptides

The effects of substituting the enkephalin moiety of dynorphin with the dermorphin sequence were studied on the receptor preference, analgesic, and peripheral opioid potencies by using synthetic dermorphin-dynorphin hybrid peptides as the probe. Replacement of the enkephalin moiety of dynorphin with the dermorphin or dermorphin^1–5 sequence caused a remarkable increase in analgesic potency, and a 3-6 fold increase in potency of binding against [³H]-dihydromorphine. The potency of receptor binding against [³H]-EKC was also increased by incorporation of the whole dermorphin sequence into the dynorphin molecule. In the presence of NaCl (100mM), the effect of enhancing binding against [³H]-EKC due to dermorphin substitution disappeared, suggesting the contribution of opioid μreceptor. Peripheral opioid activities assayed by various smooth muscle preparations showed that dermorphin incorporation caused a decrease in the potency of inhibition of the contractions of the guinea pig ileum and the rabbit vas deferens, no change in potency on the mouse vas deferens, and a marked increase in the inhibition of the rat vas deferens. Among the peripheral opioid activities only that assayed with the rat vas deferens appears to correlate approximately with the analgesic and the receptor binding activities. Judging from the relative potencies obtained from all assays, it is evident that the N-terminal dermorphin moiety, but not the C-terminal dynorphin fragment, dominates the opioid activity and receptor preference of the hybrid peptide.     

Effects of hexavalent chromium on the adenylate pool of hamster fibroblasts

The mechanism through which Cr(VI) reduces the intracellular levels of ATP in hamster fibroblasts (BHK 21 line) were investigated. Leakage of pool components during treatment of cells prelabelled with [³H]adenosine was examined. Adenylate pool composition was analyzed by thin-layer chromatography (TLC) in cells incubated with [³H]adenosine immediately after treatment with Cr(VI), and the endogenous concentrations of different nucleotides were determined by high-performance liquid chromatography (HPLC). Interference of Cr(VI) with different steps of ATP biosynthesis could be assessed.Treated cells shown increased loss of radioactive hypoxanthine and inosine in the treatment medium, but no nucleotides were detected outside the cells. A marked unbalance of the endogenous adenylate pool was produced by chromium, consisting in a strong decrease of ATP, accompanied by a very pronounced increase of ADP and AMP. Treatment with Cr(VI) caused an altered distribution of label among the different adenylate pool components, especially an accumulation of radioactivity in the ADP and AMP fractions. Also other effects on the soluble pool were detected by HPLC analysis, namely the reduction of NAD content and variations in th guanylate pool which closely paralleled the ones observed in the adenylate pool.Energy charge values, calculates on the basis of HPLC and TLC data, allowed to detect a specific inhibition of ADP phosphorylation to ATP. An interference with the early steps of adenylate synthesis could be inferred from the enchanced amounts of hypoxanthine and ionsine in the treated cultures.     

Potentiation of adenosine triphosphate-induced contractile responses of the guinea-pig isolated vas deferens by adenosine monophosphate and adenosine 5''-monophosphorothioate.

The effects of incubating the guinea-pig isolated vas deferens in the presence of adenine nucleotides (adenosine triphosphate, ATP; adenosine diphosphate, ADP; and adenosine monophosphate, AMP), or in the presence of their phosphorothioate analogues (adenosine 5'-O-(3-thiotriphosphate), ATP gamma S; adenosine 5'-O-(2-thiodiphosphate), ADP beta S; and adenosine 5'-monophosphorothioate, AMP alpha S), on contractile responses to ATP were compared. After challenge with a low (1 microM) or high (300 microM) concentration of ATP to obtain control responses, one vas deferens of a pair was incubated for 5 min with one of the adenine nucleotides, while the contralateral preparation was incubated with the corresponding phosphorothioate analogue. At the conclusion of the incubation the preparations were challenged again with ATP. Incubation with AMP or AMP alpha S resulted in a transient potentiation of responses to 1 microM and 300 microM ATP. The potentiation following incubation with AMP alpha S was larger than that produced by AMP. After incubation with ADP, ADP beta S, ATP and ATP gamma S, responses to 1 microM ATP were decreased, while those to 300 microM ATP were unaffected. Thus, incubation with AMP and AMP alpha S results in potentiation, rather than inhibition, of ATP-induced responses. On the other hand, 5'-diphosphate, 5'-triphosphate, 5'-O-(2-thiodiphosphate) and 5'-O-(3-thiotriphosphate) moieties on adenosine have no effect or cause autoinhibition. These results indicate that AMP exerts a potentiating effect on reactivity to exogenous ATP. AMP arising from the enzymatic degradation of ATP might modulate the level of response to ATP released endogenously as a cotransmitter.     

Comparison of binding of [3H]-methionine-enkephalin, [3H]-naltrexone and [3H]-dihydromorphine in the mouse vas deferens and the myenteric plexus and brain of the ginea pig.

[3H]-Methionine-enkephalin, [3H]-naltrexone and [3H]-dihydromorphine are specifically bound in homogenates of not only the brain and myenteric plexus of the guinea pig but also the vas deferens of the mouse. Brain has a ratio of methionine-enkephalin to dihydromorphine binding in favour of methionine-enkephalin binding and the myenteric plexus a ratio in favour of dihydromorphine binding, with the mouse vas deferens being intermediate.     

Cyclic AMP in freshly prepared thymocyte suspensions, Evidence for stimulation by endogenous adenosine.

Thymocyte suspensions were prepared from guinea pig thymus by gentle mincing and repeated washings. The mincing resulted in a ten-fold increase of the cyclic AMP level, most of the cyclic AMP being intracellular. The level remained high during the washing procedures. During incubation at 37° the level of intracellular cyclic AMP gradually fell by a temperature dependent process.The cyclic AMP content was increased by non-methylxanthine phosphodiesterase inhibitors (papaverine, Ro 20-1724, ZK 62.711) but decreased in the presence of theophylline or isobutylmethylxanthine. Adenosine and phenylisopropyl adenosine acutely increased cyclic AMP by a theophylline inhibited mechanism. An increase was also obtained by inhibitors of adenosine uptake (dipyridamol, dilazep) and of adenosine deaminase (EHNA).The results were compatible with a release of adenosine during the isolation of thymocytes, and with an adenosine-induced increase of intracellular cyclic AMP. This interpretation was supported by in vivo labelling of purine nucleotides with [³H]adenine. A marked rapid fall of purine nucleotides was found during the preparation of the thymocytes and at the same time adenosine metabolites appeared in the medium.     

Activation of presynaptic A1‐receptors by endogenous adenosine inhibits acetylcholine release in the guinea‐pig ileum

1 It is well established that presynaptic adenosine A₁‐receptor activation inhibits acetylcholine (ACh) release in the guinea‐pig ileum. The present study extends this observation and examines a possible role for endogenous adenosine in modulating cholinergic nerve function. 2 The actions of the adenosine uptake blocker, dipyridamole, the adenosine deaminase inhibitor, erythro‐9(2‐hydroxy‐3‐nonyl)adenine (EHNA) and the A₁‐receptor antagonist, 1,3‐dipropyl‐8‐cyclopentylxanthine (DPCPX) were examined on electrically evoked neurogenic, cholinergic twitch contractions of the guinea‐pig ileum. Some additional studies measuring [³H]‐ACh release were also performed. 3 Adenosine and the selective A₁‐receptor agonist, 2‐chloroadenosine (2‐CA), inhibited electrically evoked contractions and, in the case of 2‐CA, [³H]‐ACh release. The actions were antagonized by DPCPX. At low concentrations, dipyridamole and EHNA enhanced the effect of adenosine causing a leftward shift of the concentration–response curve. In contrast, inhibition induced by 2‐CA was unaffected by either dipyridamole or EHNA. 4 When applied alone at higher concentrations, EHNA and dipyridamole produced a concentration‐dependent suppression of cholinergic neurotransmission. In both cases, the effect could be reversed by DPCPX. At the same concentration, DPCPX alone produced a small but consistent increase in twitch height and [³H]‐ACh release. 5 The data confirm the existence of inhibitory presynaptic adenosine A₁‐receptors modulating cholinergic nerve function in the guinea‐pig ileum and suggests that these receptors can be activated by endogenous adenosine released either as adenosine itself or as an ATP metabolite.     

Effect of the intraperitoneal administration of salmon-calcitonin on the “in vitro” actions of opioid agonists

• 1.1. The interaction of intraperitoneal administration of salmon-calcitonin with opioids was studied. The study was carried out using guinea pig ileum (μ and κ-opioid receptors), rabbit vas deferens (κ-opioid receptors) and mouse vas deferens (δ-opioid receptors), and selective μ, δ and κ agonists were used in the pertinent tissues.
• 2.2. The treatment with salmon-calcitonin increased, in a dose-dependent manner, the effect of U-50,488H in guinea pig ileum and rabbit vas deferens and the effects of [d-Pen², d-Pen⁵]enkephalin in mouse vas deferens.
• 3.3. The treatment with analgesic doses of salmon-calcitonin enhances the in vitro effects of κ- and δ-opioid agonists. The increase of the effectiveness of the opioid agonists may be one of the mechanisms involved on the analgesia induced by salmon-calcitonin.

     

Evidence that the inhibition of ATP release from sympathetic nerves by adenosine is a physiological mechanism.

1. Perfusion with the P1-purinoceptor agonist adenosine (1-500 microM) greatly reduced the stimulation-induced release of ATP and the initial contractile phase of the response of the guinea pig vas deferens to field stimulation. 2. The inhibitory effects of adenosine (100 microM) were readily antagonised by the P1-purinoceptor antagonist, 8-phenyltheophylline (10 microM). 3. Dipyridamole (10 microM), inhibited the stimulation-evoked release of ATP from the guinea pig vas deferens and reduced the initial component of contraction. 4. These results support the view that adenosine, resulting from ectoenzymatic breakdown of ATP released as a cotransmitter from sympathetic nerve terminals, acts as a physiological prejunctional regulator of transmitter release.     

2-Deoxycoformycin inhibition of intracellular phosphorylation of adenosine in Novikoff rat hepatoma cells

Treatment of cultured wild type and azaguanine-resistant Novikoff rat hepatoma cells with 0.1 to 100 μM 2-deoxycoformycin resulted in an inhibition of more than 50 per cent of the incorporation of 100 μM [8-³H]adenosine into intracellular ATP and nucleic acids. In wild type cells, part of the effect resulted from an inhibition of adenosine deamination of deoxycoformycin, with a consequent decrease of incorporation of radioactivity from adenosine via inosine → hypoxanthine → IMP. This pathway, however, was blocked in azaguanine-resistant cells because of the lack of hypoxanthine/guanine phosphoribosyltransferase. The inhibition of adenosine incorporation by deoxycoformycin in these cells was not mediated at the level of adenosine transport or phosphorylation of AMP. We conclude, therefore, that the intracellular phosphorylation of adenosine is impaired in deoxycoformycin-treated cells. There was a close correlation between inhibition of adenosine deamination and adenosine incorporation, both with respect to effective concentrations of deoxycoformycin and to irreversibility of the inhibition. In addition, intracellular concentrations of adenosine above 1–5 μM were found to strongly inhibit the phosphorylation of adenosine in sity, reflecting substrate inhibition of adenosine kinase. The results indicate that the inhibition of adenosine phosphorylation in deoxycoformycin-treated cells was caused by the accumulation of free adenosine in these cells due to adenosine deaminase inhibition.     

Biotransformation of 9-beta-D-arabinofurano-syladenine by rat and human erythrocytes

Studies were performed to test whether 9-β-d-arabinofuranosyladenine (ara-A) would accumulate in erythrocytes as a result of phosphorylation to the nucleotide level. When [³H]ara-A was incubated with whole blood from rat, monkey or man for 2 hr at 37°, the drug rapidly equilibrated between erythrocytes and plasma but did not concentrate in the cells. Incubation of [³H]ara-A with rat and human erythrocyte lysates for 2 hr followed by Chromatographie analysis showed that 2–5 per cent of ara-A was converted to nucleotides. In contrast, 10–35 per cent of [¹⁴C]adenosine was converted to adenine nucleotides under the same conditions. Incubation of [³H]ara-A with human erythrocyte lysates for 18 hr resulted in a conversion of approx. 40 per cent of the labeled drug to nucleotides. Additional chromatography revealed, however, that the nucleotide fraction contained almost no arabinosyl nucleotides. Rather, 90 per cent of the label in the nucleotide fraction was identified as IMP. These results indicate that only a minor amount, if any, of ara-A was phosphorylated by erythrocyte enzymes to yield arabinosyl nucleotides. An alternative pathway converted much of the labeled drug to ribosyl nucleotides via the deamination of ara-A to ara-hypoxanthine, cleavage to hypoxanthine and conversion of the free hypoxanthine to IMP.     

Effects of reserpine and serotonin on adenine nucleotide metabolism and glucose oxidation of washed rabbit platelets

Reserpine inhibits the transfer of ATP from the metabolically active pool into the releasable pool (amine storage organelles) of washed rabbit platelets. Since reserpine is known to interfere with oxidative phosphorylation of isolated mitochondria, possible effects of reserpine on the energy metabolism of intact platelets were examined with the aim of determining whether or not such effects could explain the reduced transfer of metabolic ATP into the releasable pool of ATP. Reserpine caused a small increase in [¹⁴C]hypoxanthine and [¹⁴C]inosine accumulation in a suspension of washed rabbit platelets prelabeled with [¹⁴C]adenosine or [¹⁴C]adenine. Reserpine also caused increased oxidation of [¹⁴C]glucose. Serotonin mimicked these effects of reserpine, and imipramine inhibited the serotonin-induced as well as the reserpine-induced increases in [¹⁴C]hypoxanthine and [¹⁴C]inosine formation and glucose oxidation. However, imipramine did not prevent the reserpine-induced inhibition of [¹⁴C]ATP transfer from the metabolically active pool into the releasable pool. It is concluded, that the effects of reserpine on energy metabolism are at least partially attributable to the increased demand for energy for the uptake of serotonin, liberated from the platelets during incubation with reserpine, and not to uncoupling of oxidative phosphorylation. Furthermore, the small decrease of the metabolically active ATP pool upon incubation of platelets with reserpine cannot explain the reduced transfer of ATP from the metabolic pool into the amine storage granule pool.     

Post-receptor pathway of the ATP-induced relaxation in smooth muscle of the mouse vas deferens.

1. The post-receptor pathway of the ATP relaxant effect in K(+)-precontracted vas deferens smooth muscle (VD) was examined. 2. The relaxation to ATP was not antagonized either by 10 microM methylene blue, a cyclic GMP inhibitor, by 10 microM indomethacin, an inhibitor of prostaglandin synthesis or by 100 microM NG-nitro-L-arginine, an inhibitor of NO production. 3. The Rp-diastereomer of adenosine 3':5'-cyclic monophosphorothioate (Rp-cAMPS) 200 microM, a competitive inhibitor of cyclic AMP significantly diminished the relaxant response to ATP. 4. Isoprenaline 10 microM, a beta-adrenoceptor agonist, produced a sustained relaxation, inhibited by Rp-cAMPS, without a significant change in [Ca2+]i, thereby mimicking the ATP-induced relaxant effect. 5. The level of the phosphorylated myosin light chain in the precontracted VD was significantly lowered by 1000 microM ATP. 6. ATP (1000 microM) and isoprenaline (10 microM) produced the same increase (+ 50%) of [cyclic AMP] when applied to a resting VD. 7. The effect of simultaneous increases of [Ca2+]i and of [cyclic AMP] produced by externally applied ATP are discussed. 8. These results suggest that ATP-induced relaxation in K(+)-precontracted VD is mediated by the activation of adenylyl cyclase.     

Pharmacokinetics of inhibition of adenosine deaminase by erythro-9-(2-hydroxy-3-nonyl)adenine in CBA mice

The pharrnacokinetics oferythro-9-(2-hydroxy-3-nonyl)adenine (EHNA) inhibition of adenosine deaminase (ADA) was measured in vivo in CBA mice. The in vivo assay utilized injection of 10–100 nmoles [2-³H]adenosine and measurement of blood ³H₂O 20 min later. A single oral dose of EHNA (50 mg/kg) totally inhibited ADA for 4 hr and caused a large increase in conversion of [2-³H]adenosine to [2-³H]ATP. EHNA (3 mg/kg) decreased deamination by 50% for 2–6 hr, depending on the dose of adenosine used. Mice dosed with EHNA (100 mg/kg) once daily for 7 days showed the same ADA recovery rate as mice dosed only once. High single oral doses of EHNA had no effect on blood ATP and GTP pools.     

Existence and pharmacological properties of dopamine D4 receptors in guinea pig vas deferens.

This study was carried out to identify subtypes of dopamine D2-like receptors in guinea pig isolated vas deferens. Dopamine had no effect on the muscle tone in the presence of prazosin, an alpha1-adrenoceptor antagonist. However, contractile responses to adenosine triphosphate (ATP), noradrenaline and acetylcholine were potentiated in a concentration dependent manner by dopamine in the presence of prazosin. This potentiation was not inhibited by raclopride, an antagonist for dopamine D2 and D3 receptors. However, the potentiation of ATP- and noradrenaline-induced contraction was inhibited by clozapine and 8-methyl-6-(4-methyl-1-piperazinyl)-11H-pyrido[2,3b][1,4]benzodiazepine (JL-18), dopamine D4 receptor antagonists. Further, the potentiation of noradrenaline- and acetylcholine-induced contraction was also inhibited by spiperone, an antagonist for dopamine D2, D3 and D4 receptors. These results suggest that the dopamine D4 receptor is located on the postsynaptic site of guinea pig vas deferens and that activation of the dopamine D4 receptor enhances contractile responses to agonists without affecting muscle tone.     

Correlation between adenine nucleotide-induced cyclic AMP elevation and extracellular adenosine formation in NG108-15 cells

We previously demonstrated that extracellular adenine nucleotides induced cyclic AMP elevation in NG108-15 cells. This response was resistant to adenosine deaminase (ADA) and the ecto-5'-nucleotidase (CD73) inhibitor alpha,beta-methylene ADP (alpha,beta-MeADP), but was inhibited by both P1- and P2-receptor antagonists. In the present study, we investigated the relationship between adenine nucleotide-induced cyclic AMP elevation and extracellular adenosine formation. ATP, AMP and beta,gamma-methylene ATP (beta,gamma-MeATP) were time-dependently metabolized to adenosine in NG108-15 cells. Adenosine formations from ATP, AMP and beta,gamma-MeATP were not affected by alpha,beta-MeADP, but suppressed by the P2-receptor antagonist pyridoxalphosphate-6-azophenyl-2',4'-disulphonic acid (PPADS). A close correlation between extracellular adenosine formation and cyclic AMP increasing effects were obtained with several adenine nucleotide agonists in NG108-15 cells as well as their parent cell line C6Bu-1 and N18TG-2 cells, all of which possess functional adenosine A2 receptors. When NG108-15 cells were incubated with [3H]ATP or [3H]AMP in the presence of ADA, [3H]adenosine was found to distribute dominantly on the cell surface. NG108-15 cells expressed mRNA for the ecto-ATPase and nucleotide pyrophosphatase, but not for CD73. These results suggest that local adenosine formation by an ecto-enzyme distinct from CD73 is involved in adenine nucleotide-induced cyclic AMP formation in NG108-15 cells.