Inhibition of brain adenylate cyclase by A1 adenosine receptors: pharmacological characteristics and locations

Adenosine modulates the dopaminergic function in the nigro-striatal system by interacting with striatal dopamine dependent adenylate cyclase

Behavioral and pharmacological evidences suggest that dopaminergic mechanisms in striatum might be counteracted by adenosine or potentiated by its pharmacological antagonists methylxanthines. To test whether adenosine modulation of the dopaminergic function could be, at least in part, due to an interaction at the level of the adenylate cyclase complex, we studied the effects of the adenosine analog R-Phenyl-isopropil-adenosine (R-PIA) on basal and dopamine-sensitive adenylate cyclase in rat striatum. R-PIA, which interacts with both adenosine A1-inhibitory and A2-stimulatory receptors, dose-dependently inhibited the stimulation induced by dopamine, and seemed to utilize the same pool of enzyme linked to dopaminergic D1 receptors. Two experimental approaches leading to supersensitivity of striatal dopaminergic receptors, (i.e., 6-hydroxy-dopamine injection in substantia nigra and reserpine administration) also induced upregulation of adenosine-dependent adenylate cyclase in striatum, and altered R-PIA modulation of dopamine-sensitive adenylate cyclase. Conversely, after subchronic treatment with neuroleptics such as haloperidol or sulpiride, upregulation of 3H-Spiroperidol binding in striatum was not associated with changes of R-PIA dependent adenylate cyclase in this area. It is concluded that adenosine might modulate post-synaptic responses to dopamine via adenosine receptors which functionally interact with dopaminergic D1 receptors in striatum.

Adenosine receptors in post-mortem human cerebral cortex and the effect of carbamazepine

Representative A1 site- and A2 site-selective ligands for adenosine receptors were found to bind to a single class of non-co-operative sites in synaptic membranes prepared from pieces of normal human temporal lobe cerebral cortex obtained at autopsies performed within 24 h post-mortem. The known adenosine antagonists theophylline and 1-isobutyl-3-methylxanthine also bound to these sites and could completely displace either ligand. Computer analysis combining data from 21 separate experiments which used membranes prepared from seven different autopsy cases gave the following kinetic parameters for the site: equilibrium dissociation constants (KD) for L-PIA, 4.1 +/- 0.9 nmol/l; KD for NECA, 26 +/- 4 nmol/l; KD for theophylline, 28 +/- 5 mumol/l; receptor number (Bmax), 510 +/- 77 fmol/mg protein. Carbamazepine could also displace either radioligand from the sites, although solubility limits for this drug were reached at 50% receptor occupancy. The KD for carbamazepine was 138 +/- 18 mumol/l. Other anticonvulsants tested were generally ineffective at their therapeutic concentrations, although phenytoin showed a small amount of ligand binding inhibition. The results are in line with earlier studies in rodent tissue preparations, and suggest a possible purinergic component in the anticonvulsant activity of carbamazepine in a man.

Calcium-dependent currents in cultured rat dorsal root ganglion neurones are inhibited by an adenosine analogue

Dorsal root ganglion neurones from 2-day-old rats were grown in dissociated culture. The effect of the adenosine analogue 2-chloroadenosine (2-CA) was investigated on action potential duration and on Ca2+ current (ICa) activation. 2-CA (0.5 microM) shortened both control action potentials and those prolonged by tetraethylammonium (TEA), Ba2+, or intracellular Cs+. This effect was prevented by two adenosine antagonists isobutylmethylxanthine (IBMX, 1-2 mM) and 8-phenyltheophylline (8-PT, 2.5 microM). The inward current, ICa, recorded using the 'whole-cell' patch-clamp technique in medium containing 2.5 microM-tetrodotoxin, 25 mM-TEA and 2.5 mM-Ba2+ was reduced by 2-CA (0.05 microM). The activation of ICa was decreased, but its reversal potential was unchanged. The effect of 2-CA was antagonized by IBMX (1 mM) and 8-PT (1 microM). 2-CA also reduced the large inward tail currents which occurred at the termination of the depolarizing voltage step command in a proportion of neurones. Brief application of 2-CA (0.05 microM) did not affect the inward holding current required to maintain the cells at -80 mV. In the presence of TTX (2.5 microM) and Ca2+ (5 mM), 2-CA decreased the activation of outward K+ currents caused by 5 s depolarizing voltage commands from -80 mV or -40 mV. The GABAB agonist (-)-baclofen (50-100 microM) also shortened the action potential duration and reduced ICa. 8-PT (1 microM) did not prevent the effect of baclofen on ICa. It is concluded that in cultured rat dorsal root ganglion neurones 2-CA reduces ICa activation by a direct effect on an A1 adenosine receptor.

1,3-Dipropyl-8-cyclopentylxanthine (DPCPX) inhibition of [3H]N-ethylcarboxamidoadenosine (NECA) binding allows the visualization of putative non-A1 adenosine receptors

The binding of the adenosine receptor agonists, [3H]N-ethylcarboxamidoadenosine (NECA) and [3H]cyclohexyladenosine (CHA) to membrane preparations and to cryostat sections of the rat brain was examined. The xanthine derivative, 1,3-dipropyl-8-cyclopentylxanthine (DPCPX) was ca. 500-fold more effective at A1 than at A2 sites. [3H]CHA binding to A1 adenosine receptors was virtually eliminated by the inclusion of DPCPX (50 nM), while [3H]NECA binding was only partially inhibited. The pattern of DPCPX-insensitive [3H]NECA binding sites was strikingly different from that of A1 receptors and is believed to represent an association with A2 type adenosine receptors and perhaps another or several, previously undescribed non-A1 sites.

A1 and A2 adenosine receptors regulate adenylate cyclase in cultured human lung fibroblasts

Adenosine stimulates and inhibits adenylate cyclase activity and cAMP levels in WI-38 and VA13 fibroblasts. The inhibitory effects appear to be mediated by both A1 receptors and the P-site. Results supporting these conclusions are as follows: Adenosine by itself increased cAMP accumulation in these cells. PGE1-stimulated cAMP accumulation was inhibited by adenosine in a concentration-dependent fashion. IAP treatment blocked adenosine inhibition of cAMP accumulation and adenylate cyclase activity and enhanced adenosine stimulation of cAMP accumulation in VA13 cells. Theophylline and MIX attenuated adenosine inhibition of cAMP accumulation. Adenosine analogs with substitutions in the purine ring inhibited PGE1-stimulated cAMP accumulation and adenylate cyclase activity. PGE1-stimulated cAMP accumulation was inhibited by the P-site agonist 2'5'-dideoxyadenosine, but this inhibition was not attenuated by MIX or IAP treatment. These data support the idea that adenosine may inhibit cAMP accumulation in VA13 or WI-38 cells by acting at an A1 receptor of the P-site. The decrease in cAMP accumulation mediated by the A1 receptor appeared to be due at least in part to an Ni-mediated inhibition of adenylate cyclase.

Effect of adenosine receptor agonists and other compounds on cyclic AMP accumulation in forskolin-treated hippocampal slices

The effect of adenosine analogues and some putative neurotransmitters have been studied on cyclic AMP accumulation in rat hippocampal slices treated with the adenylate cyclase activator forskolin. The effects of PGE2 and histamine were potentiated by forskolin (0.1 microM). Isoprenaline and NECA had essentially additive effects with 0.1 microM forskolin and serotonin (above 10(-4) M) inhibited forskolin-stimulated cyclic AMP accumulation. The A1-adenosine receptor selective adenosine analogue R-PIA inhibited forskolin stimulated cyclic AMP accumulation in low doses and stimulated in high. NECA, adenosine and 2-chloroadenosine uniformly stimulated cyclic AMP accumulation. 2',5'-dideoxyadenosine inhibited, but only at high concentrations. Both the stimulatory and the inhibitory effects of R-PIA were antagonized by 8-phenyltheophylline (10 microM). Enprofylline (100 microM) selectively inhibited the stimulatory effect. In the presence of enprofylline both 2-chloroadenosine showed an inhibitory effect on cyclic AMP accumulation. It is concluded that the forskolin-treated rat hippocampal slice is a useful preparation to study both stimulatory and inhibitory effects of transmitters and modulators on adenylate cyclase. The results also show that the rat hippocampus has both A1-receptors that are linked to inhibition of cyclic AMP accumulation and A2-receptors that are linked to stimulation. Furthermore, enprofylline is shown to selectively antagonize the stimulatory response, revealing inhibitory effects of compounds such as 2-chloroadenosine and adenosine.

Adenosine agonists reduce voltage-dependent calcium conductance of mouse sensory neurones in cell culture

Adenosine and several of its analogues produced a concentration-dependent shortening of calcium-dependent action potential (c.a.p.) duration of mouse dorsal root ganglion (d.r.g.) neurones in dissociated cell culture. The following rank order of potency was obtained: N6-(L-phenylisopropyl)adenosine greater than N6-(D-phenylisopropyl)adenosine greater than N6-cyclohexyladenosine greater than 2-chloroadenosine much greater than 1-methylisoguanosine greater than adenosine. Effects of adenosine agonists on c.a.p. duration were blocked by methylxanthine adenosine antagonists. Adenosine monophosphate (AMP) and cyclic AMP shortened c.a.p.s in d.r.g. neurones, while ATP also depolarized cells. Voltage-clamp analysis revealed that the effect arose from reduction of a voltage-dependent calcium conductance. Adenosine agonists reduced depolarization-evoked inward currents but did not alter membrane conductance following blockade of calcium channels by cadmium. Additionally, adenosine reduced the instantaneous current-voltage slope (chord conductance) during step commands that produced maximal activation of voltage-dependent calcium conductance. If effects of adenosine on neuronal somata and synaptic terminals are similar, adenosine agonists may inhibit neurotransmitter release in the central nervous system by inhibiting a voltage-dependent calcium conductance. Since effects of adenosine agonists did not correspond with their relative potencies as modulators of adenylate cyclase activity or inhibitors of neurotransmitter release in peripheral tissues, a novel adenosine receptor may be involved in regulation of this conductance.

Barbiturates are selective antagonists at A1 adenosine receptors

Barbiturates in pharmacologically relevant concentrations inhibit binding of (R)-N6-phenylisopropyl[3H]adenosine ([3H]PIA) to solubilized A1 adenosine receptors in a concentration-dependent, stereospecific, and competitive manner. Ki values are similar to those obtained for membrane-bound receptors and are 31 microM for (+/-)-5-(1,3-dimethyl)-5-ethylbarbituric acid [(+/-)-DMBB] and 89 microM for (+/-)-pentobarbital. Kinetic experiments demonstrate that barbiturates compete directly for the binding site of the receptor. The inhibition of rat striatal adenylate cyclase by unlabelled (R)-N6-phenylisopropyladenosine [(R)-PIA] is antagonized by barbiturates in the same concentrations that inhibit radioligand binding. The stimulation of adenylate cyclase via A2 adenosine receptors in membranes from N1E 115 neuroblastoma cells is antagonized only by 10-30 times higher concentrations of barbiturates. It is concluded that barbiturates are selective antagonists at the A1 receptor subtype. In analogy to the excitatory effects of methylxanthines it is suggested that A1 adenosine receptor antagonism may convey excitatory properties to barbiturates.

Development of the A1 adenosine receptors in the visual cortex of cats, dark-reared and normally reared

The ontogeny of the distribution of the binding sites for [3H]chlorohydroxyladenosine, an A1 adenosine receptor-specific ligand, was visualized autoradiographically within coronal sections of the visual cortical areas of developing cats. In adults, the A1 adenosine receptors were found in all lamina except for lamina IV, and in particularly high concentration within laminas I-III. In brains of kittens 2 months old and younger who were within the critical period for the development of visual neural function, the receptor distribution was less defined and sparser, except that in contrast to adults, it was found in relatively high concentration within lamina VI. Animals dark-reared from birth, so that the critical period was postponed, exhibited an ontogenetic pattern identical to that of the normally reared animals. These results indicate that, at least with respect to ocular dominance determination, A1 adenosine receptors are probably not involved in determining the state of plasticity that is seen during the critical period.

Adenosine uptake sites in rat brain: identification using [3H]nitrobenzylthioinosine and co-localization with adenosine deaminase

The binding characteristics of [3H]nitrobenzylthioinosine ([3H]NBI) to rat brain membrane preparations was examined, and the autoradiographic distribution of this ligand in brain sections was compared with the immunohistochemical localization of adenosine deaminase (ADA). It was found that [3H]NBI labels sites for which adenosine has far higher affinity than do other nucleosides, that these sites are heterogeneously distributed and that there is an exact correspondence between areas containing [3H]NBI sites and ADA-immunoreactive neurons. Our results indicate that [3H]NBI and ADA are potential markers for revealing anatomical sites at which actions of adenosine may be expressed.

Heterogeneous distribution of adenosine transport sites labelled by [3H]nitrobenzylthioinosine in rat brain: an autoradiographic and membrane binding study

A highly heterogeneous distribution of [3H]nitrobenzylthioinosine [( 3H]NBI) binding sites was observed using both autoradiographic and membrane binding methodology. Of the 24 brain regions examined in the radio-ligand binding studies, the highest levels of [3H]NBI sites were found in the thalamus, followed by midbrain, superior colliculus, olfactory cortex and hypothalamus. The thalamus contained over 5 times more sites than cerebellum which exhibited the lowest [3H]NBI binding levels. The results obtained from autoradiographic analysis agreed well with quantitative measurements and revealed that subnuclei of thalamus and hypothalamus as well as specific layers of the superior colliculus contained particularly high concentrations of [3H]NBI sites. When the [3H]NBI autoradiograms were compared with the distribution of adenosine deaminase in brain it was found that brain regions richest in neural elements immunoreactive for adenosine deaminase contained the greatest numbers of [3H]NBI sites. In contrast, a poor correlation was found between the distribution of [3H]NBI binding and adenosine receptors labelled with [3H]cyclohexyladenosine. The co-localization of [3H]NBI binding and adenosine deaminase in brain indicates the existence of neural systems having a high capacity to take up and metabolize adenosine.

Effects of several 5'-carboxamide derivatives of adenosine on adenosine receptors of human platelets and rat fat cells

The effects of several 5'-carboxamide derivatives of adenosine on stimulatory (Ra) adenosine receptors of human platelets and inhibitory (Ri) adenosine receptors of rat fat cells have been compared. 5'-N-Cyclopropylcarboxamidoadenosine (CPCA) and 5'-N-ethylcarboxamidoadenosine (NECA) most potently inhibited ADP-induced aggregation of human platelets as shown by IC50-values of 0.24 and 0.34 mumol/l. 5'-N-Methylcarboxamidoadenosine (MECA; IC50 0.81 mumol/l) and 5'-N-carboxamidoadenosine (NCA; IC50 2.1 mumol/l) were less potent, whereas adenosine, 2-chloroadenosine and (-)N6-phenylisopropyladenosine [(-)PIA] exhibit IC50-values of about 1.5 mumol/l. Nearly the same rank order of potency was obtained for stimulation of adenylate cyclase activity of platelet membranes and for inhibition of [3H]NECA binding to human platelets. In order to examine the effects of the carboxamide analogues on Ri adenosine receptors of rat fat cells inhibition of lipolysis and adenylate cyclase were studied. (-)PIA was the most potent inhibitor of lipolysis as shown by an IC50 of 0.5 nmol/l followed by CPCA (IC50 1.1 nmol/l) and NECA (IC50 1.3 nmol/l), whereas MECA (IC50 17.9 nmol/l) and NCA (IC50 20.1 nmol/l) were much less potent than NECA in inhibiting lipolysis. Similar results were obtained for inhibition of adenylate cyclase activity of fat cell membranes and for competition with [3H]PIA binding to fat cell membranes. The relative potencies of the adenosine analogues at both receptor subclasses were calculated from the ratio of the IC50-values for inhibition of platelet aggregation and of lipolysis. (-)PIA showed the highest selectivity for Ri receptors as indicated by a 2,900-fold lower IC50 for the antilipolytic than for the anti-aggregatory effect.(ABSTRACT TRUNCATED AT 250 WORDS)

Interaction of barbiturates with adenosine receptors in rat brain

The effects of barbiturates on radioligand binding to inhibitory Ri adenosine receptors of rat brain membranes were investigated. Binding of the adenosine receptor agonist (-)N6-phenylisopropyl[3H]adenosine and the antagonist 1,3-diethyl-8-[3H]phenylxanthine was inhibited by several barbiturates. This inhibition was concentration-dependent and occurred in the range of pharmacologically effective concentrations. Pentobarbital was the most potent of the barbiturates tested with a Ki of 92 mumol/l. The (+)isomers of hexobarbital and mephobarbital were more potent than the respective (-)isomers. Barbituric acid itself did not displace either radioligand in concentrations up to 1 mmol/l. The inhibitory effect of pentobarbital was reversed by a single wash of membranes preincubated with the barbiturate. The presence of pentobarbital caused a decrease of the affinity of the receptor for the antagonist radioligand but did not alter the number of binding sites, suggesting a competitive antagonism. The effects of pentobarbital on radioligand binding to the receptor were not changed by the presence of picrotoxinin nor by the absence of chloride ions. This indicates that they are not mediated via the picrotoxinin binding site. The barbiturates could not be classified as either agonists or antagonists at the Ri adenosine receptor. The presence of GTP did not influence the inhibition of radioligand binding by pentobarbital; this is also observed for antagonists, whereas the affinity of agonists is markedly reduced by GTP. Binding of antagonists to the receptor is enthalpy-driven; the interaction of pentobarbital with the receptor was entropy-driven and the same was true for agonists.(ABSTRACT TRUNCATED AT 250 WORDS)

Differential location of adenosine A1 and A2 receptors in striatum

Stimulation of either adenosine A1 or A2 receptors results in a decrease or an increase in the adenylate cyclase activity, respectively. With various concentrations of the adenosine agonist N6-phenylisopropyladenosine, both responses on cyclase are observed in a crude membrane preparation from rat striatum. The A2 receptors appeared to be associated primarily with intrinsic striatal neurons, as a kainic acid injection into the striatum resulted in about a 70% loss of receptor responsiveness to cyclase. The A1 receptors were more diffuse, being associated with intrinsic neurons and with cortical-striatal nerve terminals. From studies with 6-hydroxydopamine, we were unable to associate either receptor with dopaminergic nerve terminals within the striatum.