Storage, release, uptake, and inactivation of purines

Mechanisms of apoptosis induced by purine nucleosides in astrocytes

Astrocytes release adenine-based and guanine-based purines under physiological and, particularly, pathological conditions. Thus, the aim of this study was to determine if adenosine induced apoptosis in cultured rat astrocytes. Further, if guanosine, which increases the extracellular concentration of adenosine, also induced apoptosis determined using the TUNEL and Annexin V assays. Adenosine induced apoptosis in a concentration-dependent manner up to 100 microM. Inosine, hypoxanthine, guanine, and guanosine did not. Guanosine or adenosine (100 microM) added to the culture medium was metabolized, with 35% or 15%, respectively, remaining after 2-3 h. Guanosine evoked the extracellular accumulation of adenosine, and particularly of adenine-based nucleotides. Cotreatment with EHNA and guanosine increased the extracellular accumulation of adenosine and induced apoptosis. Inhibition of the nucleoside transporters using NBTI (100 microM) or propentophylline (100 microM) significantly decreased but did not abolish the apoptosis induced by guanosine + EHNA or adenosine + EHNA, respectively. Apoptosis produced by either guanosine + EHNA or adenosine + EHNA was unaffected by A(1) or A(2) adenosine receptor antagonists, but was significantly reduced by MRS 1523, a selective A(3) adenosine receptor antagonist. Adenosine + EHNA, not guanosine + EHNA, significantly increased the intracellular concentration of S-adenosyl-L-homocysteine (SAH) and greatly reduced the ratio of S-adenosyl-L-methioine to SAH, which is associated with apoptosis. These data demonstrate that adenosine mediates apoptosis of astrocytes both, via activation of A(3) adenosine receptors and by modulating SAH hydrolase activity. Guanosine induces apoptosis by accumulating extracellular adenosine, which then acts solely via A(3) adenosine receptors.

In vivo effects of adenosine A(2) receptor agonist and antagonist on neuronal and astrocytic intermediary metabolism studied with ex vivo (13)C MR spectroscopy

The effect of adenosine A(2) receptor agonist 2-[p-(2-carboxyethyl)phenylethylamino]-5'-ethylcarboxamidoadenosine (CGS 21680) and antagonist 3,7-dimethyl-1-propargylxanthine (DMPX) on [1-(13)C]glucose and [1,2-(13)C]acetate metabolism was studied in rats by (13)C magnetic resonance (MR) spectroscopy and HPLC. In the cortex a significant reduction was observed in the amounts of [2-(13)C]GABA and [3-(13)C]aspartate from [1-(13)C]glucose in CGS 21680. In the subcortex the concentration of labelled [4-(13)C]glutamate was increased in both treatment groups. The amounts of [2 + 3-(13)C]succinate and [3-(13)C]lactate were increased in the CGS 21680 group compared to control, and the DMPX group showed an increase in the total amount of [6-(13)C]N-acetyl aspartate compared to control in the subcortex. Astrocyte metabolism was only affected in the cortex as shown by a decrease in the pyruvate carboxylase/pyruvate dehydrogenase ratio in glutamate and glutamine in the treatment groups. Labelling from [1,2-(13)C]acetate was not much affected by CGS 21680 or DMPX. However, the amount of [1,2-(13)C]acetate in cortex and subcortex was reduced in the DMPX group. In the cortex a reduction in the labelling of [3-(13)C]GABA in the DMPX group compared to control and an increase in the total amount of taurine in both treatment groups was detected. The present study shows that A(2) receptor agonist and antagonist have similar effects; however, in cortex GABAergic neurones and astrocytes were affected in contrast to subcortex, where glutamatergic neurones showed the greatest changes.

In vivo effects of adenosine A1 receptor agonist and antagonist on neuronal and astrocytic intermediary metabolism studied with ex vivo 13C NMR spectroscopy

Adenosine is a neuromodulator, and it has been suggested that cerebral acetate metabolism induces adenosine formation. In the present study the effects that acetate has on cerebral intermediary metabolism, compared with those of glucose, were studied using the adenosine A1 receptor agonist 2-chloro-N6-cyclopentyladenosine (CCPA) and antagonist 8-cyclopentyl-1,3-dipropylxanthine (DPCPX). Fasted rats received an intravenous injection of CCPA, DPCPX, or vehicle. Fifteen minutes later either [1,2-13C]acetate or [1-13C]glucose was given intraperitoneally; after another 30 min the rats were decapitated. Cortical extracts were analyzed with 13C NMR spectroscopy and HPLC analysis. DPCPX affected neuronal and astrocytic metabolism. De novo synthesis of GABA from neuronal and astrocytic precursors was significantly reduced. De novo syntheses of glutamate and aspartate were at control levels, but their degradation was significantly elevated. In glutamine the anaplerotic activity and the amount of label in the position representing the second turn in the tricarboxylic acid cycle were significantly increased, suggesting elevated metabolic activity in astrocytes. CCPA did not influence GABA, aspartate, or glutamine synthesis. In glutamate the contribution from the astrocytic anaplerotic pathway was significantly decreased. In the present study the findings in the [1,2-13C]acetate and [1-13C]glucose control, CCPA, and DPCPX groups were complementary, and no adenosine A1 agonist effects arising from cerebral acetate metabolism were detected.

Hypoxia and neuronal function under in vitro conditions

Neurons in the mammalian CNS are highly sensitive to the availability of oxygen. Hypoxia can alter neuronal function and can lead to neuronal injury or death. The underlying changes in the membrane properties of single neurons have been studied in vitro in slice preparations obtained from various brain areas. Hypoxic changes of membrane potential and input resistance correspond to a decrease in ATP concentration and an increase in internal Ca2+ concentration. Functional modifications consisting of substantial membrane depolarization and failure of synaptic transmission can be observed within a few minutes following onset of hypoxia. The hypoxic depolarization accompanied by a hyperexcitability is a trigger signal for induction of neuronal cell death and is mediated mainly by activation of glutamate receptors. The mechanisms of the hypoxic hyperpolarization are more complex. Two types of potassium channels contribute to the hyperpolarization, the Ca(2+)- and the ATP-activated potassium channel. A number of neurotransmitters and neuromodulators is involved in the preservation of normal cell function during hypoxia. Therefore, hypoxia-induced cellular changes are unlikely to have a single, discrete pathway. The complexity of cellular changes implies that several strategies may be useful for neuroprotection and a successful intervention may be dependent upon drug action at more than one target site.

Increase in myocardial interstitial adenosine and net lactate production in brain-dead pigs: an in vivo microdialysis study

BACKGROUND

Brain death-related cardiovascular dysfunction has been documented; however, its mechanisms remain poorly understood. We investigated changes in myocardial function and metabolism in brain-dead and control pigs.

METHODS

Heart rate, systolic (SAP) and mean (MAP) arterial pressure, left ventricular (LV) dP/dtmax, rate-pressure product, cardiac output (CO), left anterior descending coronary artery blood flow, lactate metabolism, and interstitial myocardial purine metabolite concentrations, monitored by cardiac microdialysis, were studied. A volume expansion protocol was performed at the end of the study.

RESULTS

After brain death, a transient increase in heart rate (from 90 [67-120] to 158 [120-200] beats/min) (median, with range in brackets), MAP (82 [74-103] to 117 [85-142] mmHg), LV dP/dtmax (1750 [1100-2100] to 5150 [4000-62,000] mmHg x sec(-1), rate-pressure product (9100 [7700-9700] beats mmHg/min to 22,750 [20,000-26,000] beats mmHg/min), CO (2.2 [2.0-4.0] to 3.3 [3.0-6.0] L/min), and a limited increase in left anterior descending coronary artery blood flow (40 [30-60] to 72 [50-85] ml/min) were observed. Net myocardial lactate production occurred (27 [4-40] to -22 [-28, -11] mg/L, P<0.05) and persisted for 2 hr. A 6-7-fold increase in adenosine dialysate concentration was observed after brain death induction (2.9 [1.0-5.8] to 15.8 [7.0-50.7] micromol/L), followed by a slow decline. Volume expansion significantly increased MAP, CO, and LV dP/dtmax in control animals, but decreased LV dP/dtmax and slightly increased CO in brain-dead animals. A significant increase in adenosine concentration was observed in both groups, with higher levels (P<0.05) in brain-dead animals.

CONCLUSIONS

Brain death increased oxygen demand in the presence of a limited increase in coronary blood flow, resulting in net myocardial lactate production and increased interstitial adenosine concentration consistent with an imbalance between myocardial oxygen demand and supply. This may have contributed to the early impairment of cardiac function in brain-dead animals revealed by rapid volume infusion.

An ecto-ATPase and an ecto-ATP diphosphohydrolase are expressed in rat brain

Extracellular nucleotides acting as signaling molecules are inactivated by hydrolysis catalyzed by ecto-nucleotidases. ATP is sequentially degraded via ADP and AMP to adenosine. Enzymes that can be involved in the extracellular hydrolysis chain are ecto-ATP diphosphohydrolase (ecto-apyrase), ecto-ATPase, ecto-ADPase and 5'-nucleotidase. Mammalian ecto-ATP diphosphohydrolase is a member of a family of apyrases sharing four "apyrase conserved regions" that presumably participate in the formation of the catalytic site. We report the presence of ecto-ATP diphosphohydrolase in rat brain and the primary structure of a new mammalian member of the apyrase family. Expression in CHO cells shows that it represents an ecto-ATPase. As revealed by Northern analysis of rat tissues, the ecto-ATPase is co-expressed with ecto-ATP diphosphohydrolase in heart, kidney, spleen, thymus, lung, skeletal muscle and brain. Signals for both ecto-nucleotidases are very weak in liver. mRNAs for both proteins are present in PC12 cells, suggesting that the two nucleotidases may be co-expressed in the same neural cell. Using computer-aided sequence analysis, primary structure and membrane topography are compared with those of other members of the apyrase family.

Effect of adenosine and some of its structural analogues on the conductance of NMDA receptor channels in a subset of rat neostriatal neurones

1. In order to investigate the modulatory effects of adenosine on excitatory amino acid projections onto striatal medium spiny neurons, whole-cell patch clamp experiments were carried out in rat brain slices. The effects of various agonists for P1 (adenosine) and P2 (ATP) purinoceptors and their antagonists were investigated. The A2A receptor agonist 2-p-(2-carboxyethyl)phenythylamino-5'-N-ethylcarboxamidoadenosine (CGS 21680; 0.1 microM), the A1 receptor agonist 2-chloro-N6-cyclcopentyladenosine (CCPA; 10 microM) and the non-selective P1 purinoceptor antagonist 8-(p sulphophenyl)-theophylline (8-SPT; 100 microM) did not alter the resting membrane potential, the threshold current necessary to elicit an action potential, the amplitude of spikes, their rise time, the amplitude of the afterhyperpolarization (AHP) and the time to peak of the AHP. 2. N-methyl-D-aspartate (NMDA; 1-1000 microM) caused a concentration-dependent inward current which was larger in the absence than in the presence of Mg2+ (1.3 mM). In a subset of striatal neurones, the current response to NMDA (10 microM) and the accompanying increase in conductance were both inhibited by CGS 21680 (0.01-1 microM). The effect of CGS 21680 (0.1 microM) persisted in the presence of tetrodotoxin (0.5 microM) or in a Ca(2+)-free medium, under conditions when synaptically mediated influences may be negligible. 3. The A3 receptor agonist N6-2-(4-aminophenyl)ethyladenosine (APNEA; 0.1-10 microM) also diminished the effect of NMDA (10 microM), while the A1 receptor agonists CCPA (0.1-10 microM) and (2S)-N6-[2-endonorbornyl] adenosine [S(-)-ENBA; 10 microM] as well as the endogenous, non-selective P1 purinoceptor agonist adenosine (100 microM) were inactive. The endogenous non-selective P2 purinoceptor agonist ATP (1000 microM) also failed to alter the current response to NMDA (10 microM). Adenosine (100 microM), but not ATP (1000 microM) became inhibitory after blockade of nucleoside uptake by S[4-nitrobenzyl)-6-thioguanosine (NBTG; 30 microM). 4. 8-(p-Sulphophenyl)-theophylline (8-SPT; 100 microM), as well as the A2A receptor antagonist 8-(3-chlorostyryl) caffeine (CSC; 1 microM) and the A1 receptor antagonist 8-cyclopentyl-1,3-dipropylxanthine (DPCPX) at 0.03, but not 0.003 microM abolished the inhibitory action of CGS 21,680 (0.1 microM). None of these compounds altered the effect of NMDA (10 microM) by itself. DPCPX (0.03 microM) prevented the inhibition of APNEA (10 microM). 5. There was no effect of CGS 21,680 (0.1 microM), when guanosine 5'-O-(3-thiodiphosphate (GDP-beta-S; 300 microM) was included in the pipette solution in order to block G protein-mediated reactions. 6. In conclusion, adenosine receptors, probably of the A2A-subtype, inhibit the conductance of NMDA receptor channels in a subset of medium spiny neurones of the rat striatum by a transduction mechanism which involves a G protein.

Extracellular metabolism of nucleotides in the nervous system

1. A variety of surface-located enzymes are involved in the metabolism of extracellular nucleotides. The biochemical properties of some of these are briefly discussed. 2. The molecular identity of ecto-diadenosine polyphosphate hydrolase has not yet been revealed. On neural cells the enzyme catalyses the hydrolysis of ApnA to Apn-1 and AMP. 3. The molecular structure of ATP-diphosphohydrolase has recently been identified. The enzyme occurs in essentially all tissues where it catalyses the extracellular hydrolysis of ATP and ADP with the formation of AMP. 4. Ecto-5'-nucleotidase is a GPI-anchored glycoprotein and catalyses the formation of AMP to adenosine. In the adult brain, and as revealed by immunocytochemistry, the enzyme is mainly associated with astrocytes. It is associated with developing nerve cells and cultured neural cells. In vitro its inhibition or suppression of its synthesis result in the inhibition of neurite formation and long-time survival of neural cells. Continued extracellular hydrolysis of AMP and formation of adenosine thus appear to be essential for neural differentiation and survival.

Biochemistry, localization and functional roles of ecto-nucleotidases in the nervous system

Nucleotides such as ATP, ADP, UTP or the diadenosine polyphosphates and possibly even NAD+ are extracellular signaling substances in the brain and in other tissues. Enzymes located on the cell surface catalyze the hydrolysis of these compounds and thus limit their spatio-temporal activity. As a final hydrolysis product they generate the nucleoside and phosphate. The paper discusses the biochemical properties, cellular localization and functional properties of surface-located enzymes that hydrolyse nucleotides released from nervous tissue. This is preceded by a brief discussion of nucleotide receptors, cellular storage and mechanisms of nucleotide release. In nervous tissue nucleoside 5'-triphosphates are hydrolysed by ecto-ATP-diphosphohydrolase and possibly in addition also by ecto-nucleoside triphosphatase and ecto-nucleoside diphosphatase. The molecular identity of the ATP-diphosphohydrolase has now been revealed. The hydrolysis of nucleoside 5'-monophosphates is catalysed by 5'-nucleotidase whose biochemical properties and molecular structure have been studied in detail. Little is known about the molecular properties of the diadenosine polyphosphatases. Surface located enzymes for the extracellular hydrolysis of NAD+ and also ecto-protein kinases are discussed briefly. The cellular localization of the ecto-nucleotidases is only partly defined. Whereas in adult mammalian brain activity for hydrolysis of ATP and ADP may be associated with nerve cells or glial cells 5'-nucleotidase appears to have a preferential glial allocation in the adult mammal. The extracellular hydrolysis of the nucleotides is of functional importance not only during synaptic transmission where it functions in signal elimination. It plays a crucial role also for the survival and differentiation of neural cells in vitro and presumably during neuronal development in vivo.

The source of brain adenosine outflow during ischemia and electrical stimulation

Adenosine outflow and adenosine and adenine nucleotide content of hippocampal slices were evaluated under two different experimental conditions: ischemia-like conditions and electrical stimulation (10 Hz). Five minutes of ischemia-like conditions brought about an 8-fold increase in adenosine outflow in the following 5 min during reperfusion, and a 2-fold increase in adenosine content, a 43% decrease in ATP, a 72% increase in AMP and a 30% decrease in energy charge (E.C.) at the end of the ischemic period. After 10 min of reperfusion ATP, AMP and E.C. returned to control values, while the adenosine content was further increased. Five minutes of electrical stimulation brought about an 8-fold increase in adenosine outflow that peaked 5 min after the end of stimulation, a 4-fold increase in adenosine content and an 18% decrease in tissue E.C. at the end of stimulation. After 10 min of rest conditions the adenosine content and E.C. returned to basal values. The origin of extracellular adenosine from S-adenosylhomocysteine (SAH) was examined under the two different experimental conditions. The SAH hydrolase inhibitor, adenosine-2,3-dialdehyde (10 microM), does not significantly modify the adenosine outflow evoked by electrical stimulation or ischemia-like conditions. This finding excludes a significant contribution by the transmethylation pathway to adenosine extracellular accumulation evoked by an electrical or ischemic stimulus, and confirms that the most likely source of adenosine is from AMP dephosphorylation.

Inhibition of protein synthesis induced by adenine nucleotides requires their metabolism into adenosine

Adenine nucleotides and adenosine inhibit the incorporation of radiolabelled leucine into proteins of isolated hepatocytes. Impairment occurred with nucleotides which can be converted into 9-beta-D-ribofuranosyladenine (adenosine) but was not observed after treatment with adenine or AMPCPP (the alpha, beta-methylene analogue of ATP). Metabolism into adenosine was further suggested by the increase in cellular ATP levels following treatment of hepatocytes with ATP, adenosine or AMPPCP (the beta, gamma-methylene ATP analogue) while AMPCPP was without any significant effect. The inhibition of protein synthesis caused by adenosine was not due to a lytic effect nor to a general disturbance in hepatic functions and was reversed when the cells were washed and transferred to a nucleoside-free medium. This impairment, however, was not coupled to the activation of adenylate cyclase, as preincubation of hepatocytes with P1 purinoceptor antagonists failed to prevent protein synthesis inhibition. In contrast, L-homocysteine enhanced the inhibitory effect of adenosine on the incorporation of radiolabelled leucine into proteins. Our results thus suggest that the inhibition of protein synthesis caused by adenine nucleotides requires their conversion into adenosine. They also indicate that the inhibitory effect of adenosine does not involve a receptor-mediated effect but may be related to an increase in S-adenosylhomocysteine content and a subsequent low level of macromolecule methylation.

Inhibition of extracellular ATP degradation in endothelial cells

The plasma membrane ATPase on the human umbilical vein endothelial cell line (ECV304) was demonstrated to be an ecto-enzyme. Hydrolysis of ATP was measured by monitoring the appearance of inorganic phosphorus. Hydrolysis of extracellular ATP was insensitive to oligomycin, vanadate, ouabain and N-ethylmaleimide, compounds that inhibit the intracellular ion pumping ATPases. Beta-Glycerophosphate (1-10 mM) or p-nitrophenyl phosphate (1-10 mM) did not inhibit hydrolysis of ATP, ruling out the involvement of non-specific phosphatases. Enzyme activity in buffer that had previously been incubated with cells was < 7%, showing that the enzyme activity measured did not result from release of intracellular enzymes. Consistent with this, the cell preparations used were estimated to be > 95% intact as judged by release of cytosolic enzyme lactate dehydrogenase. The enzyme activity was Ca2-/Mg2- dependent. Gramicidin S (20 microM), suramin (100-300 microM), chlorpromazine (250 microM), trifluoperazine (50-250 microM), and thioridazine (100 microM) inhibited the hydrolysis of ATP (3 mM) by 45-80%. The percentage inhibition produced by these substances was not altered in the presence of a concentration of alpha, beta-methylene ADP (10 microM) which inhibited hydrolysis of AMP (3 mM) by 90%, suggesting that these compounds inhibit ecto-ATPase and/or ecto-ADPase. Measurements of absolute amounts of ATP released from various tissues, including the heart, have been hindered because ATP is rapidly and sequentially hydrolysed to adenosine. Identification of compounds that inhibit ATP degradation would prove to be useful to overcome this problem and would lead to the development of invaluable pharmacological tools in many other aspects of purine research.

Role of cyclic AMP in stimulation-evoked release of 3H-noradrenaline from rabbit isolated ear artery

The possible involvement of cyclic AMP in the stimulation-evoked 3H-overflow from rabbit isolated ear artery preloaded with 3H-noradrenaline was studied. Cyclic AMP (10(-5)-3 x 10(-4)M), 8-bromo-cyclic AMP (3 x 10(-4)M) and adenosine (10(-5)-3 x 10(-4)M) enhanced stimulation-evoked 3H-overflow. Dibutyryl-cyclic AMP (10(-5)-3 x 10(-4)M) had no effect. Theophylline (3 x 10(-5)M) and the phosphodiesterase inhibitor AH 21-132 (3 x 10(-5)M) did not alter the enhancement of 3H-overflow caused by either cyclic AMP or adenosine. Forskolin (3 x 10(-6)M) and the phosphodiesterase inhibitors ICI 63 197 (10(-4)M) and AH 21-132 (3 x 10(-6)-3 x 10(-5)M) increased stimulation-evoked 3H-overflow. Forskolin (10(-6)M) enhanced the effect of ICI 63 197 (3 x 10(-5)M) but it did not alter the effect of AH 21-132. It is concluded that cyclic AMP is involved in the stimulation-evoked release of noradrenaline from postganglionic sympathetic nerves in the rabbit ear artery.

Effects of cyclopiazonic acid on contractility and ecto-ATPase activity in guinea-pig urinary bladder and vas deferens

1. Cyclopiazonic acid (CPA), an inhibitor of sarcoplasmic ATPase, was tested on guinea-pig urinary bladder and vas deferens for its ability: (1) to modify contractile responses to electrical field stimulation (EFS), exogenous ATP, alpha,beta-methylene ATP (alpha,beta-MeATP), carbachol, noradrenaline (NA), histamine, and KCl; (2) to affect ecto-ATPase activity; (3) to modify the release of ATP evoked by EFS. 2. In the urinary bladder, CPA (10 microM) potentiated contractile responses to EFS, exogenous ATP (100 microM), alpha,beta-meATP (1 microM), carbachol (0.5 microM), histamine (30 microM) and KCl (30 mM). In the vas deferens, CPA (10 microM) potentiated responses to EFS, ATP, alpha,beta-meATP, NA (100 microM) and KCl. CPA at a concentration of 1 microM had no effect on ATP-induced relaxation of carbachol-precontracted guinea-pig taenia coli, and at a concentration of 10 microM it markedly increased spontaneous contractile activity of taenia. 3. Ecto-ATPase was estimated to have Vmax and Km values of 0.98 nmol Pi 30 min-1 mg-1 wet tissue and 881 microM ATP in the urinary bladder, and 0.75 nmol Pi 30 min-1 mg-1 wet tissue and 914 microM ATP in the vas deferens, respectively. CPA at a concentration of 10 microM significantly inhibited ecto-ATPase activity by 18% in the urinary bladder and by 24% in the vas deferens. 4. In the guinea-pig vas deferens, CPA significantly potentiated ATP release evoked by EFS from 2.2 +/- 0.8 (6) pmol ATP min-1 g-1 wet tissue to 35.2 +/- 4.8 (6) pmol ATP min-1 g-1 wet tissue (P < 0.01). 5. In conclusion, the potentiation of contractile responses of the guinea-pig urinary bladder and vas deferens by CPA has a non-specific character. CPA inhibited ecto-ATPase activity and increased ATP release, but these effects do not appear to contribute to the potentiation of Pu-purinoceptor-mediated responses since the contractile actions of all the agonists studied were potentiated to the same extent.

Stimulation of reactive astrogliosis in vivo by extracellular adenosine diphosphate or an adenosine A2 receptor agonist

Adenosine and its nucleotides adenosine triphosphate (ATP) and adenosine diphosphate (ADP) stimulate the proliferation of brain astrocytes in vitro and augment the effects of other growth factors. Following brain injury, hypoxia, or around solid tumors with necrotic centers, such as glioblastoma multiformes, high concentrations of adenine nucleotides and adenosine are released into the extracellular space; extracellular adenosine concentrations can rise 30-100-fold to a concentration in excess of 100 microM. Increased concentrations of extracellular adenosine and adenine nucleotides may contribute to reactive astrocytic proliferation following brain injury. To test this hypothesis, adenosine, an adenosine analog 5'-(N-cyclopropyl)-carboxamidoadenosine (CPCA), or ADP was micro-injected into rat cortex. The number of glial fibrillary acidic protein-immunopositive cells was compared between the treated and contralateral saline-injected hemispheres. Within 48 hr, astrocyte density around the CPCA (100 microM) infusion site was almost double that around the control saline infusion site. In hemispheres into which CPCA was infused, there was an increase in astrocytes in the subpial region along fiber tracts and around blood vessels, characteristic of Scherer's secondary structures found in association with malignant astrocytic brain tumors. The increased astrogliosis elicited by CPCA was abolished by coinfusion of the adenosine A2 receptor antagonist 1,3-dipropyl-7-methylxanthine (DPMX). While microinjection of adenosine (1 mM) failed to stimulate astrogliosis, microinjection of ADP (500 microM) also resulted in a significant reactive astrogliosis and accumulation of astrocytes similar to Scherer's secondary structures.(ABSTRACT TRUNCATED AT 250 WORDS)