Rat brain astrocyte and microglia cultures express different members of ATP-binding-cassette (ABC) proteins. RT-PCR analysis showed that astrocytes are equipped with P-glycoprotein (mdr1a, mdr1b), multidrug resistance-associated-protein (mrp1, mrp4, mrp5) and cystic fibrosis transmembrane conductance regulator (CFTR). No transcripts for mrp5 and CFTR were detected in microglia. The ABC protein functional activities are shown by the following results: (i) cyclosporin A (50 microM), verapamil (50 microM), probenecid (1 mM) or sulfinpyrazone (2 mM) enhanced [3H]vincristine accumulation; (ii) cyclosporin A or verapamil but not probenecid or sulfinpyrazone enhanced [3H]digoxin accumulation; (iii) glibenclamide (100 microM) inhibited 36Cl efflux from astrocytes. ATP release from glial cells was inhibited by the pretreatment with ABC protein inhibitors indicating that ABC proteins are involved in nucleotide efflux from glial cells which represent the main source of cerebral extracellular purines.

Guanosine has many trophic effects in the CNS, including the stimulation of neurotrophic factor synthesis and release by astrocytes, which protect neurons against excitotoxic death. Therefore, we questioned whether guanosine protected astrocytes against apoptosis induced by staurosporine. We evaluated apoptosis in cultured rat brain astrocytes, following exposure (3 h) to 100 nM staurosporine by acridine orange staining or by oligonucleosome, or caspase-3 ELISA assays. Staurosporine promoted apoptosis rapidly, reaching its maximal effect (approximately 10-fold over basal apoptotic values) in 18-24 h after its administration to astrocytes. Guanosine, added to the culture medium for 4 h, starting from 1 h prior to staurosporine, reduced the proportion of apoptotic cells in a concentration-dependent manner. The IC50 value for the inhibitory effect of guanosine is 7.5 x 10(-5) M. The protective effect of guanosine was not affected by inhibiting the nucleoside transporters by propentophylline, or by the selective antagonists of the adenosine A1 or A2 receptors (DPCPX or DMPX), or by an antagonist of the P2X and P2Y purine receptors (suramin). In contrast, pretreatment of astrocytes with pertussis toxin, which uncouples Gi-proteins from their receptors, abolished the antiapoptotic effect of guanosine. The protective effect of guanosine was also reduced by pretreatment of astrocytes with inhibitors of the phosphoinositide 3-kinase (PI3K; LY294002, 30 microM) or the MAPK pathway (PD98059, 10 microM). Addition of guanosine caused a rapid phosphorylation of Akt/PKB, and glycogen synthase kinase-3beta (GSK-3beta) and induced an upregulation of Bcl-2 mRNA and protein expression. These data demonstrate that guanosine protects astrocytes against staurosporine-induced apoptosis by activating multiple pathways, and these are mediated by a Gi-protein-coupled putative guanosine receptor.

1 Extracellular guanosine has diverse e ects on many cellular components of the central nervous system, some of which may be related to its uptake into cells and others to its ability to release adenine-based purines from cells. Yet other e ects of extracellular guanosine are compatible with an action on G-protein linked cell membrane receptors. 4 This site was speci®c for guanosine, and the order of potency in displacing 50 nM [ 3 H]-guanosine was: guanosine=6-thio-guanosine4inosine46-thio-guanine4guanine. Other naturally occurring purines, such as adenosine, hypoxanthine, xanthine ca eine, theophylline, GDP, GMP and ATP were unable to signi®cantly displace the radiolabelled guanosine. Thus, this binding site is distinct from the well-characterized receptors for adenosine and purines. 5 The addition of GTP produced a small concentration-dependent decrease in guanosine binding, suggesting this guanosine binding site was linked to a G-protein.

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.

Like adenine-based purines, extracellular nonadenine-based purines have a multitude of trophic effects on the growth, differentiation, and survival of target cells. The nonadenine-based purines, which include guanosine, inosine, and GTP, apparently exert their trophic effects by interacting with both intercellular targets as well as those on the cell surface. Specifically, guanosine and inosine target the protein kinase N-kinase, in promoting remarkable nerve process extension, even in long tracts of the central nervous system after injury. In contrast, GTP may exert its effects via a cell surface receptor coupled to the release of calcium from internal stores. In other cases trophic effects may be mediated by the enhancement of release of adenine-based purines by guanosine. Additionally, evidence is presented for the existence of a high-affinity binding site for guanosine with receptor-like characteristics on the plasma membranes of astrocytes and brain tissue. This site may be G-protein-coupled and exert its effects through activation of the MAP kinase cascade. One effect apparently mediated through this mechanism is the production and release by astrocytes of trophic protein growth factors such as NGF and TGFβ. These have substantial neuroprotective effects. Additionally, this pathway is apparently involved in modulating the expression of P2Y 1 and P2Y 2 receptors in response to extracellular guanosine. Extracellular nonadenine-based purines can interact with other growth factors, but these interactions are not always synergistic. For example, combinations of guanosine and FGF are antagonistic and reduce the growth of microvascular cells in vitro. Some of the properties of the nonadenine-based purines likely derive from their unique intracellular metabolism in which conversion of guanine to xanthine is the final catabolic step. This step is catalyzed by guanase, the activity of which varies markedly in different brain regions, raising the possibility that guanine or guanosine are involved in neurotransmission. Together these data suggest several potentially useful pharmacological approaches involving nonadenine-based purines to modulate trophic effects in the central nervous system. Drug Dev. Res. 52:303-315, 2001.

Inflammation is widely recognized as contributing to the pathology of acute and chronic neurodegenerative conditions. Microglial cells are pathologic sensors in the brain and activated microglia have been viewed as detrimental. Leukotriene, including cysteinylleukotrienes (CysLTs) are suggested to be involved in brain inftammation and neurological diseases andATP, by its receptors is a candidate for microglia activation.A23187 (l0f1M) stimulated microglia to co-release CysLTs and [3H]adenine based purines ([3H]ABPs), mainly ATP. The biosynthetic production of CysLTs was abolished by 10f1MMK-886, an inhibitor of 5-lipoxygenase-activating protein activity. RT-PCR analysis showed that microglia expressed both CysLT I / CysLT 2 receptors, P 2Y1 ATPreceptors and several members of the ATP binding cassette (ABC) transporters including MRP1, MRP4 and Pgpo The increase in [Ca 2+]i elicited by LTD4 (0.1 f1M) and 2MeSATP (l00J.1M), agonists for CysLT-and P 2Y1 -receptors, was abolished by the respective antagonists, BAYu9773 (0.5 f1M) and suramin (50 f1M). The stimulation of both receptor subtypes, induced a concomitant increase in the release of both [3H]ABPs and CysLTs that was blocked by the antagonists and significantly reduced by a cocktail ofABC transporter inhibitors, BAPTAIAM (intracellular Ca2+ chelator) and staurosporine (0.1 f1M, PKC blocker). P2Y antagonist was unable to antagonise the effects ofLTD 4 and BAYu9773 did not reduce the effects of2MeSATP. These data suggest that: i) the efffux of purines and cysteinyl-leukotrienes is specifically and independently controlled by the two receptor types, ii) calcium, PKC and the ABC transporter system can reasonably be considered common mechanisms underlying the release ofABPs and CysLTs from microglia. The blockade ofP 2Y1 orCysLT/CysLT 2 receptors by specific antagonists that abolished the raise in [Ca"]! and drastically reduced the concomitant efffux of both compounds, as well as the effects ofBAPTA and staurosporine support this hypothesis. In conclusion, the data of the present study suggest a cross talk between the purine and leukotriene systems in a possible autocrine/paracrine control ofthe microglia-mediated initiation and progression of an inftammatory response.Proinflammatory products of 5-lipoxygenase (5-LO) pathway, including cysteinylleukotrienes (LTC4, LTD4, LTE4) are suggested to be involved in brain inflammation and neurological diseases. In aging brain and in Alzheimer's disease an increase in the activity of 5-LO has been shown (41, 57). The inhibition of the enzyme or 5-LO activating protein (FLAP) reduced the microglia-mediated toxicity towards neuronal cells, whereas the toxicity was enhanced by the cysteinyl leukotriene LTD4 (35).

No abstract

* Equally contributed to this paperAstrocytes have been recognized as important elements in controlling inflammatory as well as immune processes in the central nervous system (CNS). Recently, glial cells have been shown to produce cysteinyl leukotrienes (CysLTs) which are known lipid mediators of inflammation and whose extracellular concentrations rise under different pathological conditions in the brain. In the same conditions also extracellular concentrations of ATP dramatically increase reaching levels able to activate P2X 7 ionotropic receptors for which an emerging role in neuroinflammation and neurodegeneration has been claimed. RT-PCR analysis showed that primary cultures of rat brain astrocytes express P2X 7 receptors. Application of the selective P2X7 agonist benzoyl-benzolyATP(BzATP) markedly increased [Ca"], which was mediated by a calcium influx from the extracellular milieu. The P2X 7 antagonist, oATP,suppressed the BzATP-induced calcium increase. Consistent with the evidence that increased calcium levels activate the leukotriene biosynthetic pathway, challenge of astrocytes with either the calcium ionophore A23187 or BzATP significantly increased CysLT production and the cell pre-treatment with EGTA abolished these effects. Again the P2X7 antagonist prevented the BzATP-mediated CysLT efflux, whereas the astrocyte pretreatment with MK-571, a CysLT I receptor antagonist, was ineffective. The astrocyte pre-treatment with a cocktail of inhibitors of ATP binding cassette (ABC) proteins reduced the BzATP-mediated CysLT production confirming that ABC transporters are involved in the release of CysLTs. The astrocyte P2X 7-evoked rise of CysLT efflux was abolished in the presence of MK-886, an inhibitor of 5-lipoxygenase activating protein (FLAP) whose expression, along with that of 5'-lipoxygenase (5-LO) was reported by Northern Blot analysis. The stimulation ofP2X 7 induced an up-regulation of FLAP mRNA that was reduced by the antagonist oATP. These data suggest that in rat brain cultured astrocytes P2X 7 ATP receptors may participate in the control of CysLT release thus further supporting a role for extracellular ATP as an integral component of the inflammatory brain response.Inflammation is involved in several neurological diseases including brain injury, cerebral ischemia, multiple sclerosis,Alzheimer's and Parkinson's diseases (31,47,35). Neuroinflammation is characterized by the activation of both microglia and astrocytes (45). Glial cells, including astrocytes, are reported to produce