Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Endogenous adenosine modulates the metabolic activity and function of a wide variety of excitable as well as nonexcitable cells. The effects of adenosine are mediated by at least four receptors coupled to adenylyl cyclase through G proteins. Activation of A₁ and A₃ receptors by adenosine leads to inhibition of the enzyme, whereas A_2A and A_2B receptors are positively linked to it (Fredholm et al., 1994). Among these four subtypes, A₁ and A_2A receptors can be activated under basal physiological conditions due to the relatively high affinity for the endogenous ligand (Fredholm, 1995).

PC12 rat pheochromocytoma cells have a number of neuronal characteristics, and they undergo terminal neuronal differentiation in response to cyclic AMP (cAMP)- and cAMP-elevating agents (Roth et al., 1991; Vossler et al., 1997). Therefore, this cell line is often used as an experimental model to study signal transduction mechanisms involved in a variety of processes occurring in neuronal cells. Some of these, such as catecholamine synthesis and neurotransmitter release, are known to undergo physiological control by adenosine through cAMP-dependent pathways (McMahon and Sabban, 1992; Oda et al., 1995; Chae and Kim, 1997; Ono et al., 1998).

PC12 cells posses A₁, A_2A, and A_2B receptors (Arslan et al., 1997, 1999), but A₁ receptors are not functionally relevant (Noronha-Blob et al., 1986). Some pieces of evidence indicate that in these cells, endogenous adenosine exerts a tonic control on adenylyl cyclase activity. In fact, both inactivation of endogenous adenosine by means of adenosine deaminase (ADA) and blockade of adenosine receptors with theophylline lead to a decrease in cAMP level in unstimulated PC12 cells, whereas an opposite effect has been observed after inhibition of adenosine uptake by dipyridamole (Roskoski and Roskoski, 1989).

The physiological control operated by adenosine can be further stressed on stimulation of adenylyl cyclase with forskolin, a diterpene that acts through a receptor-independent mechanism, by directly interacting with the catalytic site of the enzyme (Dessauer et al., 1997). The cAMP-elevating effect of forskolin is sensitive to inhibition by ADA (Kim et al., 1993; Rabin et al., 1993; Florio et al., 1999). In accordance with these findings, adenosine receptor desensitization has been shown to reduce the adenylyl cyclase response to subsequent stimulation with adenosine receptor agonists as well as with forskolin in PC12 cell membranes (Chern et al., 1993, 1995). Moreover, in intact cells, an increase in forskolin-induced cAMP accumulation has also been reported after a 4-day exposure to 5'-N-ethylcarboxamido adenosine (NECA; Rabin et al., 1993), an hydrolysis-resistant adenosine derivative. These observations prompted us to better define the dependence of the cAMP response to forskolin on the occupancy of purinergic receptors by endogenous adenosine in PC12 cells and to identify the specific receptor or receptors involved.

The direct precursor of adenosine, AMP, is dephosphorylated by 5'-nucleotidase, whose membrane-bound form is an ectoenzyme present in essentially all nervous tissues, including PC12 cells (Zimmermann and Braun, 1996). Because in many cell systems AMP mimics a variety of adenosine receptor-mediated effects, and in PC12 cells it increases cAMP level in an adenosine-like manner (Yakushi et al., 1996), part of the present study was intended to evaluate the dependence of AMP action on 5'-nucleotidase activity and the possible interaction of the mononucleotide with forskolin.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. [2,8-³H]cADP (specific activity, 27 Ci/mmol) was obtained from DuPont-New England Nuclear (Bad Homburg, Germany). Nonlabeled cAMP, adenosine, AMP, ,-methyleneadenosine 5'-diphosphate (AOPCP), and ADA type VIII (175 U/mg, 25 mg/ml) were obtained from Sigma Chemical Co. (St. Louis, MO). Forskolin, 2-[p-(2-carbonylethyl)phenylethylamino]-5'-N-ethylcarboxamidoadenosine (CGS 21680), NECA, 8-(p-sulfophenyl)theophylline (8-SPT), 3,7-dimethyl-1-propargylxanthine (DMPX), xanthine amine congener (XAC), and 8-cyclopentyl-1,3-dipropylxanthine (DPCPX) were purchased from Research Biochemicals International (Natick, MA). N-[2-(Dimethylamino)ethyl]-N-methyl-4-(2,3,6,7-tetrahydro-2,6dioxo-1,3-dipropyl-1H-purin-8-yl)benzenesulfonamide (PD 115199) was obtained from Warner-Lambert/Parke-Davis Pharmaceutical Research Division (Ann Arbor, MI). Forskolin was dissolved in ethanol, kept at 20°C as stock solution, and diluted with culture medium, pH 7.4, immediately before use. Stock solutions (10 mM) of XAC, DPCPX, or PD 115199, dissolved in dimethyl sulfoxide, and of NECA, made in 50% ethanol, were diluted with culture medium on the day of the experiment. Final vehicle concentrations in incubation wells were always less than 0.5% (v/v). All the other reagents were dissolved in distilled water.

Cell Culture. PC12 cells were cultured in RPMI-1640 (GIBCO BRL, Paisley, Scotland) supplemented with 10% heat-inactivated horse serum (Sigma Chemical Co.), 5% fetal calf serum (GIBCO BRL), 100 U/ml penicillin, and 100 µg/ml streptomycin (Sigma Chemical Co.) at 37°C in a humidified 95% air/5% CO₂ atmosphere. The cells were routinely subcultured once weekly. For determination of cAMP production, the cells were seeded at a cell density of 0.3 × 10⁶ cells in 24-well plates 48 h before the experiments.

cAMP Accumulation. Cells were incubated for 30 min at 37°C in 1 ml/well HEPES-buffered culture medium, pH 7.4, containing 1% horse serum-fetal calf serum and, where appropriate, adenosine receptor antagonists. After removal of the preincubation medium, reactions were started by adding 1 ml of the same medium containing test agents or vehicle. After the appropriate time, incubation was terminated by the removal of medium and the addition of 0.25 ml of ice-cold 0.1 N hydrochloric acid to each well. The cells were sonicated, and the supernatants, neutralized by the addition of 0.25 ml of ice-cold 0.1 M Tris, were collected in Eppendorf tubes, centrifuged, and stored at 20°C until assayed.

Measurement of cAMP. cAMP content was determined by displacing [³H]cAMP binding to a bovine adrenal extract as described by Nordstedt and Fredholm (1990), with slight modifications (Florio et al., 1999). Samples or unlabeled cAMP for standard curve were added to a 96-well MultiScreen-FB microtiter plate (Millipore, MA) in a final volume of 250 µl of 100 mM Tris-HCl, pH 7.4, containing 250 mM NaCl, 10 mM EDTA, 0.5 pmol [³H]cAMP, and binding protein. After 150 min at 4°C, the incubation was stopped by vacuum filtration using a MultiScreen vacuum manifold, and the filters were washed twice with 200-µl aliquots of ice-cold Tris-HCl. After the addition of SuperMix liquid scintillation cocktail (25 µl/well), the plate was counted in a MicroBeta Trilux liquid scintillation counter (Wallac, Turku, Finland).

Calculations and Statistical Analysis. The computer program SigmaPlot (Jandel Scientific, Erkrath, Germany) was used to generate IC₅₀ parameters for antagonist concentration-response curves. Data were analyzed using an ANOVA followed by a post hoc multiple-comparison Newman-Keuls test, as indicated in the text. Results from kinetic studies were compared using an ANOVA followed by the post hoc multiple-comparison Bonferroni's test.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Effect of Various Adenosine Receptor Antagonists on Forskolin-Induced Intracellular cAMP Accumulation. A 15-min exposure to 10 µM forskolin caused a large increase in intracellular cAMP levels in PC12 cells (from 1.52 ± 0.13 to 1452 ± 136 pmol/10⁶ cells). In the presence of several adenosine receptor antagonists, forskolin-dependent cAMP accumulation was inhibited in a concentration-dependent manner with the following order of potency: PD 115199 > XAC DPCPX > 8-SPT = DMPX (Fig. 1). The calculated IC₅₀ values were 2.76 ± 1.16 nM, 17.4 ± 1.08 nM, 443 ± 1.03 nM, 2.00 ± 1.01 µM, and 2.25 ± 1.05 µM, respectively. Maximally effective antagonist concentrations did not fully suppress the response to forskolin, and the residual increase over basal values ranged from 39 ± 10 to 63 ± 5 pmol cAMP/10⁶ cells. Moreover, a maximally effective 8-SPT concentration (25 µM) did not alter basal levels of the cyclic nucleotide (1.60 ± 0.15 pmol/10⁶ cells).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Effect of increasing concentrations of adenosine antagonists on intracellular cAMP levels in cells exposed to 10 µM forskolin for 15 min. Cells were incubated for 30 min in the presence of the antagonists before the addition of forskolin. Each point is the mean ± S.E.M. of at least two separate experiments, each performed in triplicate. , PD 115199; , XAC; , DPCPX; , 8-SPT; , DMPX.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Effect of increasing concentrations of forskolin on intracellular cAMP levels in the presence () or in the absence (bars) of 2.5 nM PD 115199. PC12 cells were incubated for 30 min in the presence of the adenosine receptor antagonist before the addition of forskolin. Stimulation was carried out for 15 min. Values are mean ± S.E.M. of two separate experiments performed in triplicate.

Time Course of Intracellular cAMP Accumulation Induced by Forskolin in Absence and Presence of PD 115199. To further investigate the characteristics of the inhibitory effect of adenosine antagonists on forskolin activity, PC12 cells were incubated with 10 µM forskolin in the absence and the presence of 2.5 nM PD 115199, and intracellular cAMP was measured at different incubation times. The inhibitory effect of the antagonist was already detectable after 5 min of incubation. In the presence of PD 115199, cAMP progressively increased until 30 min of stimulation, with no significant changes occurring thereafter. At the end of the incubation (120 min), when a decline in cAMP levels occurred in cells treated with 10 µM forskolin, the inhibitory effect of the antagonist was no longer significant (Fig. 3).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Time course of intracellular cAMP accumulation in PC12 cells exposed to 10 µM forskolin (), 1 µM forskolin (), or 10 µM forskolin in the continuous presence of 2.5 nM PD 115199 (). Before the addition of drugs, cells were incubated for 30 min in prewarmed medium. Values are mean ± S.E.M. of three separate experiments performed in triplicate. The curve for 10 µM forskolin was significantly different from the curve for 1 µM forskolin (P < .01 for all time points, with the exception of the 120-min time point, where P < .05), as well as from the curve for 10 µM forskolin plus 2.5 nM PD 115199 (P < .01 for all time points, with the exception of the 120-min time point, not significant). No significant differences were found between the curve for 1 µM forskolin and the curve for 10 µM forskolin plus PD 115199, with the exception of the 15-min time point, where P < .05 (ANOVA followed by Bonferroni's test).

Effect of Adenosine Receptor Agonists on Forskolin-Dependent cAMP Accumulation. The above results strongly suggest a role for endogenous adenosine in enhancing the ability of forskolin to stimulate adenylyl cyclase activity. To verify if exogenously applied adenosine could accordingly facilitate forskolin action, the effect of increasing concentrations of forskolin was investigated in the absence and in the presence of adenosine at 0.1 µM, a concentration below its EC₅₀ value of 513 ± 61 nM (Florio et al., 1999). In this set of experiments, 0.1 µM adenosine-induced intracellular cAMP accumulation was equal to 114 ± 20 pmol/10⁶ cells. As shown in Fig. 4, adenosine markedly potentiated the effect of forskolin during 30 min of incubation, causing a parallel shift of the concentration-response curve for forskolin to the left. Adenosine potentiated the response to 1 and 3 µM forskolin to a comparable extent, enhancing intracellular cAMP levels by 5.3 ± 0.3- and 5.4 ± 0.8-fold, respectively, whereas 30 µM forskolin-induced cAMP levels were augmented by 1.8 ± 0.2-fold. At 100 µM forskolin, a concentration that was shown previously to be maximally effective (Florio et al., 1999) the potentiation (1.7 ± 0.1-fold increase) was still significant.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Effect of the presence of 0.1 µM adenosine on intracellular cAMP in PC12 exposed to increasing concentrations of forskolin. Cells were exposed to forskolin alone () or to forskolin plus adenosine () for 30 min. Adenosine (0.1 µM)-dependent cAMP synthesis was equal to 114 ± 20 pmol/106 cells. Data are mean ± S.E.M. of four separate experiments performed in triplicate. Results were compared using ANOVA followed by the Newman-Keuls range test; the curve for forskolin plus adenosine was significantly different from the forskolin curve (P < .01 for all forskolin concentrations, with the exception of the lowest concentration, where P < .05).

View larger version (33K): [in this window] [in a new window]   Fig. 5.   Effect of increasing concentrations of the adenosine receptor agonists NECA (black columns) and CGS 21680 (gray columns) on intracellular cAMP levels. Each point is the mean ± S.E.M. of two separate experiments performed in duplicate.

Effect of Exogenous AMP and Inosine on Forskolin-Dependent cAMP Accumulation. The possible interaction between forskolin and the adenosine precursor AMP was also evaluated. AMP was less potent than adenosine in promoting cAMP accumulation, with a calculated EC₅₀ value equal to 49 ± 13 µM (not shown). AMP, at concentrations both below (10 µM) and above (100 µM) its EC₅₀ value, mimicked the ability of adenosine to potentiate the effect of forskolin (Fig. 6). The response to 1 µM forskolin was increased by 4.8 ± 0.5- and 6.7 ± 0.7-fold in the presence of 10 and 100 µM AMP, respectively, whereas the same concentrations of the nucleotide augmented the effect of 10 µM forskolin by 1.1 ± 0.1- and 1.2 ± 0.2-fold, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Effect of 10 or 100 µM AMP on intracellular cAMP in PC12 cells exposed to 1 and 10 µM forskolin. Cells were exposed to forskolin alone (open columns) or to forskolin plus AMP (open plus filled columns) for 30 min. AMP (10 and 100 µM)-dependent intracellular cAMP synthesis was equal to 24 ± 5 and 120 ± 11 pmol/106 cells, respectively. Data are mean ± S.E.M. of 6 to 12 separate determinations assayed in duplicate. Results were compared using ANOVA followed by the Newman-Keuls range test, and the results for 1 and 10 µM forskolin plus 10 and 100 µM AMP were found to be significantly different from those obtained with forskolin alone (**P < .01).

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Effect of 10 µM inosine on intracellular cAMP in PC12 cells exposed to 1 and 10 µM forskolin. Cells were exposed to forskolin alone (open columns) or to forskolin plus inosine (INO, open plus filled columns) for 30 min. Data are mean ± S.E.M. of three to nine separate determinations assayed in duplicate. Results were compared using ANOVA followed by the Newman-Keuls range test. No significant difference was found between 1 and 10 µM forskolin plus 100 µM inosine with respect to forskolin alone.

Effect of ADA and of 8-SPT on AMP- and Adenosine-Dependent Intracellular Accumulation of cAMP. ADA, the enzyme that promotes the deamination of adenosine into inosine; 8-SPT; and AOPCP, an inhibitor of ecto-5'-nucleotidase activity, were used to evaluate whether the ability of exogenous AMP to increase intracellular cAMP was due to interaction of the nucleotide with adenosine receptors per se or after conversion to adenosine or to some other unrelated mechanism. cAMP accumulation caused by 100 µM AMP was reduced after a 30-min incubation together with 1 U/ml ADA (78%) or with 25 µM 8-SPT (54%) (Fig. 8). At these concentrations, ADA nearly suppressed the response to 1 and 10 µM exogenous adenosine, whereas 8-SPT decreased the response to 1 and 10 µM exogenous adenosine by 96 and 48%, respectively (Fig. 9). The effect of 100 µM AMP was also sensitive to inhibition by AOPCP (100 µM), which reduced cAMP accumulation by 54% (Fig. 8).

View larger version (14K): [in this window] [in a new window]   Fig. 8.   Effect of 1 U/ml ADA, 25 µM 8-SPT, and 100 µM AOPCP on intracellular cAMP accumulation in PC12 cells exposed to 100 µM AMP. Cells were exposed to AMP alone or plus test agents for 30 min. Basal values (3.34 ± 0.53 pmol/106 cells) were not significantly affected by the presence of ADA (2.80 ± 0.58 pmol/106 cells) or 8-SPT (3.52 ± 0.37). Data are mean ± S.E.M. of 5 to 12 separate determinations assayed in duplicate. Results were compared using ANOVA followed by the Newman-Keuls range test, and the results for the three different treatments (AMP plus ADA, 8-SPT, or AOPCP) were found to be significantly different from those with AMP alone (**P < .01).

View larger version (14K): [in this window] [in a new window]   Fig. 9.   Effect of 1 U/ml ADA and 25 µM 8-SPT on intracellular cAMP accumulation in PC12 cells exposed to 1 or 10 µM adenosine (ADO). Cells were exposed to adenosine alone or to adenosine plus test agents for 30 min. Data are mean ± S.E.M. of six separate determinations assayed in duplicate. Results were compared using ANOVA followed by the Newman-Keuls range test, and the results for 1 and 10 µM adenosine plus ADA or 8-SPT were found significantly different from those with adenosine alone (**P < .01).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The present results demonstrate that the ability of forskolin to stimulate cAMP production in PC12 cells is strictly modulated by the occupancy of adenosine receptors by endogenous adenosine. In fact, several adenosine antagonists dramatically impaired activation of adenylyl cyclase by 10 µM forskolin. The inhibition was concentration-dependent with the following order of potency: PD 115199 > XAC DPCPX > 8-SPT = DMPX. The antagonist potencies are in good agreement with the IC₅₀ values reported in several studies for antagonism at adenosine A₂ receptors (Daly et al., 1986; Bruns et al., 1987; Jarvis et al., 1989; Hide et al., 1992; Kirk and Richardson, 1995).

PC12 cells predominantly express adenosine A_2A receptors but also possess a small proportion of A_2B receptors (Van der Ploeg et al., 1996), and both binding sites are positively coupled to adenylyl cyclase. On the basis of the potency of two antagonists that have high affinity at the A_2A receptor, PD 115199 and XAC (Bruns et al., 1987; Jacobson et al., 1987), it appears that the receptor subtype predominantly involved in modulating forskolin activity is the A_2A subtype. This would be in keeping with the relatively high affinity of this receptor for the endogenous ligand, allowing its activation under basal physiological conditions (Fredholm, 1995).

Because it is highly improbable that forskolin and adenosine antagonists may share a common binding site, the remaining alternative is that adenosine antagonists impair forskolin activity by acting in a noncompetitive manner (i.e., by blocking the chain of events that leads to the production of cAMP in response to the diterpene). To better clarify the type of antagonism involved, the effect of the reversible and relatively selective A_2A receptor antagonist PD 115199 (Bruns et al., 1987) was investigated in the presence of increasing concentrations of forskolin. The results show that the inhibition of the response to 10 µM forskolin caused by PD 115199 could be surmounted by increasing the concentration of the stimulant but that the reversibility of PD 115199 antagonism was only partial because a maximally effective forskolin concentration did not fully restore the cAMP response. These results mirror previous findings showing that the nonselective P₁ receptor antagonist 8-SPT at a concentration that was maximally effective in inhibiting 10 µM forskolin-dependent cAMP accumulation caused a downward shift in the concentration-response curve for forskolin (Florio et al., 1999).

Inclusion of the antagonist PD 115199 in the incubation medium reduced the early burst of cAMP accumulation evoked by 10 µM forskolin, thus allowing it to maintain its levels unchanged during the second hour of incubation. A similar profile was found in time course experiments performed using a 10-fold lower concentration of forskolin. Thus, the presence of the antagonist mimics a condition in which a low concentration (1 µM) of forskolin is used, possibly indicating that the recruitment of A_2A receptors for the amplification of forskolin response is directly proportional to forskolin concentration.

Time course studies also demonstrated that the ability of forskolin to stimulate adenylyl cyclase activity was not only prevented but also reversed by PD 115199. Thus, a prolonged increase in intracellular cAMP levels caused by forskolin does not seem to impair the function of the purinergic receptor system, in accordance with previous reports indicating that adenosine A_2A receptor desensitization, which occurs rapidly in PC12 cells (Mundell and Kelly, 1998), is a homologous process (Palmer et al., 1994). However, when PD 115199 was added together with forskolin at the beginning of the incubation, after 120 min its inhibitory effect was no longer significant. A 2-h preincubation of this compound with the culture medium alone or with culture medium preincubated in cell-plated wells did not reduce its ability to inhibit the response to a subsequent challenge of untreated cells with forskolin, excluding a significant chemical or metabolic instability of PD 115199. Thus, PC12 cells appear to undergo time-dependent "adaptation" to the antagonist. The cellular mechanisms involved in this phenomenon remain to be identified.

In recent years, much experimental work has been done to better understand the molecular processes that lead to activation of adenylyl cyclase by forskolin (McHugh Sutkowski et al., 1994; Juska and de Foresta, 1995; Dessauer et al., 1997; Tesmer et al., 1997; Yan et al., 1997). Several lines of evidence now indicate the presence of at least one binding site for the drug that resides on the catalyst, whose affinity for forskolin is enhanced by almost 500-fold in the presence of the G protein-derived G_s subunit (Dessauer et al., 1997). The recently demonstrated atypical interaction between adenosine receptors and G proteins may explain the dramatic effect of adenosine antagonists on forskolin-dependent cAMP accumulation observed in PC12 cells. Compared with other G protein-coupled receptors, adenosine A₁ and A₂ receptors are tightly coupled to G proteins, with the ternary complex stabilized in a high-affinity conformation (Freissmuth et al., 1991; Nanoff et al., 1991, 1995; Nanoff and Stiles, 1993; Nanoff and Freissmuth, 1997). Displacement of endogenous adenosine from the receptor by A₂ receptor antagonists is likely to unhinge the equilibrium between endogenous adenosine and its binding site, thus facilitating the uncoupling of the ternary complex. This would finally result in a reduction of forskolin affinity toward its binding site, the complex G_s subunit/catalyst. Interestingly, it has been reported that adenylyl cyclase type VI, which in PC12 cells mediates adenosine-dependent cAMP synthesis (Chern et al., 1995), is not detectable by photolabeling with iodinated forskolin derivatives in PC12 cell membranes unless G_s subunits are present (McHugh Sutkowski et al., 1994).

The presence of adenosine antagonists would not affect the binding of forskolin on the catalyst, which is independent from the presence of G_s. In agreement with this hypothesis is the finding that the ability of forskolin to stimulate adenylyl cyclase was not totally dependent on adenosine receptor stimulation. In fact, even at maximally effective antagonist concentrations, a residual stimulatory response to forskolin was always found, resulting in a 25-fold increase of cAMP over basal levels.

The prominent role of adenosine in facilitating adenylyl cyclase activation by forskolin was further stressed by the finding that exogenous adenosine strongly potentiated cAMP synthesis and that the relative enhancing effect was more evident at low forskolin concentrations, which are apparently more largely dependent on occupancy of adenosine receptors. In fact, the response to 1 µM forskolin was increased 5-fold by 0.1 µM adenosine, whereas that of 10 µM forskolin was only doubled. The parallel leftward shift of the concentration-response curve for forskolin caused by adenosine indicates an increased affinity of the diterpene for its binding site on adenylyl cyclase.

As reported previously (Chern et al., 1993), the nonselective adenosine receptor agonist NECA was equieffective to the selective adenosine receptor agonist CGS 21680 in increasing cAMP levels. The effect of forskolin was potentiated by NECA and, to an even larger extent, by CGS 21680. These results straighten the conclusion that the enhancing effect is predominantly mediated by adenosine receptor of the A_2A subtype.

Another possibility worthy of investigation was that, besides adenosine, other endogenous purines might contribute in facilitating forskolin response. The present data indicate that, like adenosine, its precursor, AMP, causes intracellular accumulation of the cyclic nucleotide and significantly enhances forskolin activity when added exogenously. In contrast, inosine, the deamination product of adenosine, neither alters basal cAMP nor affects the response evoked by forskolin. The ability of 100 µM AMP to stimulate intracellular cAMP accumulation in the absence of forskolin was reduced by 50% by the inhibitor of ecto-5'-nucleotidase, AOPCP, and by the adenosine receptor antagonist 8-SPT, suggesting that conversion of AMP to adenosine by ecto-5'-nucleotidase and the interaction of the nucleoside with its receptors are at least in part responsible for the enhancing effect of AMP. In this respect, Roskoski and Roskoski (1989) reported that conversion of 100 µM [¹⁴C]AMP into [¹⁴C]adenosine by PC12 is relatively slow, with approximately 35% of the nucleotide being lost within 20 min of incubation. This may explain the lower potency of AMP in stimulating cAMP accumulation compared with adenosine or NECA. The stimulatory effect of 100 µM AMP was markedly inhibited by 1 U/ml ADA, and 5 U/ml concentration of the enzyme was previously shown to completely abolish 100 µM AMP-induced accumulation of cAMP in these cells (Yakushi et al., 1996). Thus, adenosine fully or at least in part accounts for the observed effects of AMP. It should be stressed that the cAMP response to forskolin in PC12 cells is inhibited by ADA but is not significantly affected by AOPCP (Florio et al., 1999). Thus, even though these cells are capable of releasing AMP (Braumann et al., 1986), which can undergo extracellular degradation to adenosine via ecto-nucleotidase, the latter pathway would not contribute significantly to the adenosine pool involved in the amplification of forskolin response.

In conclusion, the present study demonstrates that occupancy of adenosine A_2A receptors by adenosine receptor ligands strictly modulates the ability of forskolin to stimulate adenylyl cyclase and that endogenous adenosine has a determinant role in facilitating forskolin-dependent cAMP accumulation in PC12 cells.

Although no putative endogenous ligand for the forskolin-binding site or sites has been so far identified, the dependence of forskolin action on the neuroprotective metabolite adenosine opens new perspectives in the pharmacological control of neuronal function.