Martine Tencé

Sphingosine-1-phosphate induces proliferation of astrocytes: regulation by intracellular signalling cascades

Sphingosine‐1‐phosphate (S1P) is a potent lysophospholipid mediator mostly released by activated platelets. It is involved in several functions in peripheral tissues, but its effects in the central nervous system are poorly documented. Therefore, we have examined the effects of S1P on the proliferation of striatal astrocytes from the mouse embryo. These cells have been found to express mRNAs for the S1P receptors, Edg‐1 and Edg‐3. S1P stimulated thymidine incorporation and induced activation of extracellular signal‐regulated kinases (Erks). Both effects were prevented by U0126, an Erk kinase inhibitor. The S1P‐evoked activation of Erk1 was totally blocked in astrocytes pretreated with a combination of either phorbol ester (24 h) and LY294002, or phorbol ester (24 h) and pertussis toxin (PTX). Each individual treatment only partially inhibited Erk1 activation. This suggests that several separate mechanisms mediate this process, one involving protein kinase C and another involving Gi/Go proteins and phosphatidylinositol 3‐kinase. In contrast, the stimulatory effect of S1P on astrocyte proliferation was totally blocked by either PTX or LY294002, but not by a downregulation of protein kinase C. S1P dramatically inhibited the evoked production of cyclic AMP, a response that was impaired by PTX. Finally, S1P stimulated the production of inositol phosphates and increased intracellular calcium by mobilization from thapsigargin‐sensitive stores. These latter effects were mainly insensitive to PTX. Probably, Gi/Go protein activation and phosphoinositide hydrolysis are early events that regulate the activation of Erks by S1P. Altogether, these observations show that astrocytes are targets for S1P. Their proliferation in response to S1P could have physiopathological consequences at sites of brain lesions and alterations of the blood–brain barrier.

Endothelin-evoked Release of Arachidonic Acid from Mouse Astrocytes in Primary Culture

In striatal astrocytes, receptors for the vasoactive peptide endothelin (ET) are associated with several intracellular signalling pathways: ET‐1 increases the breakdown of phosphoinositides, induces a sustained influx of Ca2+ and inhibits the isoproterenol‐induced formation of cAMP (Marin et al., J. Neurochem., 56, 1270–1275, 1991). In the present study, it will be shown that ET‐1 and ET‐3 markedly stimulate the release of arachidonic acid (AA) from cultured astrocytes from the mouse striatum (EC50= 3 and 7 nM for ET‐1 and ET‐3, respectively), mesencephalon and cerebral cortex. The ET‐1‐evoked release of AA probably resulted from the activation of a phospholipase A2, since it required extracellular Ca2+ and was prevented by mepacrine but not by RHC 80267, an inhibitor of diacylglycerol lipase. The ET‐1‐induced release of AA was shown to be partially mediated by a guanine nucleotide‐binding protein sensitive to pertussis toxin but not to cholera toxin. A cAMP‐dependent process is not involved since the ET‐1‐evoked release of AA was not affected when cells were incubated with either isoproterenol or 8‐bromo‐cAMP. The ET‐1‐evoked release of AA could be mimicked by the co‐application of a calcium ionophore and a protein kinase C activator. However, staurosporine, a potent inhibitor of protein kinase C, which blocked the release of AA induced by the combined application of ionomycin and phorbol 12‐myristate 12‐acetate (PMA), was without effect on the ET‐1‐evoked response, indicating that protein kinase C is not directly involved in the ET‐1‐induced release of AA. Furthermore, the responses induced by ET‐1 and by PMA were found to be additive. These results suggest that (1) ET‐1 receptors are coupled to the release of AA by a mechanism independent of both protein kinase C activation and the adenylate cyclase pathway, possibly via the activation of phospholipase A2, (2) different mechanisms (or different phospholipase A2 subtypes) are involved in the control of AA release in astrocytes.

Synergistic Effects of Acetylcholine and Glutamate on the Release of Arachidonic Acid from Cultured Striatal Neurons

Abstract: The activation of muscarinic and NMDA receptors by carbachol and NMDA, respectively, stimulated the release of [3H]arachidonic acid ([3H]AA) from cultured striatal neurons. Striking synergistic effects were observed when both agonists were coapplied. This synergistic response was suppressed by atropine or (5R, 10S)‐(+)‐5‐methyl‐10,11‐dihydro‐5H‐dibenzo[a,d]cyclohepten‐5,10‐imine hydrogen maleate and inhibited by magnesium. It was markedly reduced in the absence of external calcium and suppressed by mepacrine. NMDA strongly elevated the intracellular calcium concentration ([Ca2+]i), but carbachol was ineffective. Ionomycin, α‐amino‐3‐hydroxy‐5‐methylisoxazole‐4‐propionate, or potassium depolarization, which increased [Ca2+]i but was ineffective on [3H]AA release, also potentiated the carbachol response. Sphingosine and Ro 31‐8220 suppressed the responses evoked by carbachol, NMDA, or both agonists. However, no synergistic responses could be observed when phorbol 12‐myristate 13‐acetate was associated with either carbachol or NMDA. Together, these results suggest that both the massive influx of calcium induced by NMDA and the coupling of muscarinic receptors with a putative phospholipase A2 are required for the strong synergistic effects of carbachol and NMDA on [3H]AA release. Synergistic effects were also observed with acetylcholine and glutamate in the presence of magnesium, further revealing the physiological relevance of this process.

Pleiotropic effects of lysophosphatidic acid on striatal astrocytes.

Lysophosphatidic acid (LPA) is a potent lipid mediator that is likely involved in diverse functions in the brain. Several recent studies have suggested that astrocytes are important target cells for LPA. In the present study, we have identified the signal transduction pathways activated following LPA stimulation in mouse striatal astrocytes in primary culture. In cells prelabeled with myo‐[3H]inositol, LPA stimulated the formation of [3H]inositol phosphates (EC50 = 0.7 μM). This effect was reproduced neither by other lysophospholipids nor by phosphatidic acid. Astrocyte pretreatment with pertussis toxin partially abolished this LPA response indicating the involvement of a Gi/Go protein. In [3H]adenine‐prelabeled cells, LPA strongly inhibited the formation of [3H]cyclic AMP induced by forskolin (EC50 = 0.3 μM) and by isoproterenol and PACAP‐38. These inhibitory effects were strongly reduced by pertussis toxin treatment. Although with a lesser potency (EC50 = 5 μM), LPA also stimulated the release of [3H]arachidonic acid from [3H]arachidonic acid‐prelabeled astrocytes. This latter effect was totally inhibited by mepacrine, did not involve a pertussis toxin‐sensitive G protein, and was highly dependent on external calcium. LPA also stimulated the activity of both extracellular signal‐regulated kinases (Erk) Erk1 and Erk2 by a mechanism involving a Gi/Go protein. Surprisingly, in contrast to that observed in fibroblasts, LPA was totally ineffective in stimulating DNA synthesis. These results provide additional evidence in favor of an important physiological role of LPA in the astrocytic functions. GLIA 28:25–33, 1999.

Synergistic Regulation of Cytosolic Ca2+ Concentration by Adenosine and α1-Adrenergic Agonists in Mouse Striatal Astrocytes

Adenosine has a broad array of actions on neurons but astrocytes also possess adenosine receptors. We have previously shown that adenosine, by acting on astrocytes in the striatum, can modulate neuronal responses mediated by receptors coupled to phospholipase C through an astrocyto‐neuronal interaction. In addition, adenosine was found to potentiate the α1‐adrenergic production of inositol phosphates in astrocytes. The mechanism involved in this potentiation was further investigated by examining the effects of adenosine and α1‐adrenergic receptor agonists on cytosolic Ca2+ in cultured striatal astrocytes from the embryonic mouse in primary culture. When used alone, methoxamine, a selective agonist of α‐adrenergic receptors or 2‐chloroadenosine, a stable analogue of adenosine, induced a transitory increase in cytosolic Ca2+, but their combined addition led to a sustained increase in cytosolic Ca2+, which seems to be due to a Ca2+ influx, because it was not observed in the absence of external Ca2+. Voltage independent Ca2+ channels contribute to this process and different blockers of voltage‐operated calcium channels, such as dihydropyridines, phenylalkylamines, La3+ or Co2+ were ineffective in suppressing the sustained cytosolic Ca2+ elevation. Three observations suggest the implication of arachidonic acid in the observed potentiation: (i) arachidonic acid induced a sustained elevation of cytosolic Ca2+ similar to that evoked by the coapplication of methoxamine and 2‐chloroadenosine; (ii) the addition of arachidonic acid during the calcic plateau produced by the combined application of the agonists did not increase further cytosolic Ca2+ levels; (iii) in the presence of methoxamine, 2‐chloroadenosine induced a release of arachidonic acid. The stimulation of phospholipase C and the resulting activation of protein kinase C induced by methoxamine seem to be required for the potentiating effect of 2‐chloroadenosine on cytosolic Ca2+. In fact, the direct activation of protein kinase C by an exogenous diacylglycerol analogue mimicked the effect of methoxamine because, in this condition, 2‐chloroadenosine alone evoked a sustained elevation of cytosolic Ca2+. Therefore, methoxamine, through the successive activation of phospholipase C and protein kinase C, could allow a lipase, probably phospholipase A2, to be stimulated by 2‐chloroadenosine. Arachidonic acid has already been shown to trigger the opening of K+ channels and the formation of inositol phosphates in other cell types. Therefore, in striatal astrocytes, 2‐chloroadenosine, through an arachidonic acid‐mediated hyperpolarization, could increase the Ca2+ driving force and thus improve Ca2+ influx through inositol phosphate‐gated channels. This hypothesis is further supported by the suppressing effect of a 50 mM KCI‐induced depolarization on the long lasting elevation of cytosolic Ca2+ seen in the combined presence of 2‐chloroadenosine and methoxamine.

2-Chloroadenosine potentiates the alpha 1-adrenergic activation of phospholipase C through a mechanism involving arachidonic acid and glutamate in striatal astrocytes

In cultured striatal astrocytes, 2-chloroadenosine, an adenosine analog resistant to adenosine deaminase, although inactive alone, markedly potentiated the activation of phospholipase C induced by methoxamine, an alpha 1-adrenergic agonist. This effect was suppressed by antagonists of either A1 adenosine or alpha 1-adrenergic receptors. An influx of calcium and two distinct G-proteins are involved in this phenomenon since the potentiating effect of 2-chloradenosine was suppressed in the absence of external calcium or when cells were pretreated with pertussis toxin. In addition, arachidonic acid is likely involved in this potentiating effect. This was shown first by examining the effects of inhibitors of phospholipase A2 or arachidonic metabolism, then by examining the action of arachidonic acid on the production of inositol phosphates in either the presence or absence of methoxamine, and finally by measuring the release of arachidonic acid. The sequential activation of phospholipase C and of protein kinase C is required for the 2-chloroadenosine-induced activation of phospholipase A2 since 2-chloroadenosine markedly stimulated phospholipase C activity in the absence of methoxamine when protein kinase C was activated by a diacylglycerol analog. Finally, the enhancing effect of 2- chloroadenosine on the methoxamine-evoked response seems to result from an inhibition of glutamate reuptake into astrocytes by arachidonic acid. Indeed, the potentiating effect of 2-chloroadenosine was suppressed when external glutamate was removed enzymatically and mimicked by either selective inhibitors of the glutamate reuptake process or direct application of glutamate.

Functional Coupling of the NK1 Tachykinin Receptor to Phospholipase D in Chinese Hamster Ovary Cells and Astrocytoma Cells

Abstract: In [3H]myristic acid‐prelabeled Chinese hamster ovary cells stably expressing the rat NK1 tachykinin receptor, the selective NK1 agonist [Pro9]substance P ([Pro9]SP) time and concentration dependently stimulated the formation of [3H]phosphatidylethanol in the presence of ethanol. This [Pro9]SP‐induced activation of phospholipase D (PLD) was blocked by NK1 receptor antagonists and poorly or not mimicked by NK2 and NK3 agonists, respectively. In confirmation of previous observations, [Pro9]SP also stimulated the hydrolysis of phosphoinositides, the release of arachidonic acid, and the formation of cyclic AMP (cAMP). All these [Pro9]SP‐evoked responses could be mimicked by aluminum fluoride, but they remained unaffected in cells pretreated with pertussis toxin, suggesting that a Gi/Go protein is not involved in these different signaling pathways. The activation of PLD by [Pro9]SP was sensitive to external calcium and required an active protein kinase C because the inhibition of this kinase (Ro 31‐8220) or its down‐regulation (long‐term treatment with a phorbol ester) abolished the response. In contrast, a cAMP‐dependent process was not involved in the activation of PLD because the [Pro9]SP‐evoked response was neither affected by Rp‐8‐bromoadenosine 3′,5′‐cyclic monophosphorothioate nor mimicked by cAMP‐generating compounds (cholera toxin or forskolin) or by 8‐bromo‐cyclic AMP. A functional coupling of NK1 receptors to PLD was also demonstrated in the human astrocytoma cell line U 373 MG stimulated by SP or [Pro9]SP. These results suggest that PLD activation could be an additional signaling pathway involved in the mechanism of action of SP in target cells expressing NK1 receptors.

Increased interaction of connexin43 with zonula occludens-1 during inhibition of gap junctions by G protein-coupled receptor agonists.

Astrocytes are extensively coupled through gap junctions (GJs) that are composed of channels mostly constituted by connexin43 (Cx43). This astroglial gap junctional intercellular communication (GJIC) allows propagation of ions and signaling molecules critical for neuronal activity and survival. It is drastically inhibited by a short-term exposure to endothelin-1 (ET-1) or to sphingosine-1-phosphate (S1P), both compounds being inflammatory mediators acting through activation of GTP-binding protein-coupled receptors (GPCRs). Previously, we have identified the GTPases G(i/o) and Rho as key actors in the process of S1P-induced inhibition. Here, we asked whether similar mechanisms underlied the effects of ET-1 and S1P by investigating changes in the phosphorylation status of Cx43 and in the molecular associations of Cx43 with zonula occludens (ZO) proteins and occludin. We showed that the inhibitory effect of ET-1 on GJIC was entirely dependent on the activation of G(i/o) but not on Rho and Rho-associated kinase. Both ET-1 and S1P induced dephosphorylation of Cx43 located at GJs through a process mediated by G(i/o) and calcineurin. Thanks to co-immunoprecipitation approaches, we found that a population of Cx43 (likely junctional Cx43) was associated to ZO-1-ZO-2-occludin multiprotein complexes and that acute treatments of astrocytes with ET-1 or S1P induced a G(i/o)-dependent increase in the amount of Cx43 linked to these complexes. As a whole, this study identifies a new mechanism of GJIC regulation in which two GPCR agonists dynamically alter interactions of Cx43 with its molecular partners.

Endothelin stimulates phospholipase D in striatal astrocytes.

Abstract: In primary cultures of mouse striatal astrocytes prelabeled with [3H]myristic acid, endothelin (ET)‐1 induced a time‐dependent formation of [3H]phosphatidic acid and [3H]diacylglycerol. In the presence of ethanol, a production of [3H]phosphatidylethanol was observed, indicating the activation of a phospholipase D (PLD). ET‐1 and ET‐3 were equipotent in stimulating PLD activity (EC50 = 2–5 nM). Pretreatment of the cells with pertussis toxin partially abolished the effect of ET‐1, indicating the involvement of a Gi/Go protein. Inhibition of protein kinase C by Ro 31‐8220 or down‐regulation of the kinase by a long‐time treatment with phorbol 12‐myristate 13‐acetate (PMA) totally abolished the ET‐1‐induced stimulation of PLD. In contrast, a cyclic AMP‐dependent process is not involved in the activation of PLD, because the ET‐1‐evoked formation of [3H]phosphatidylethanol was not affected when cells were coincubated with either isoproterenol, 8‐bromo‐cyclic AMP, or forskolin. Acute treatment with PMA also stimulated PLD through a protein kinase C‐dependent process. However, the ET‐1 and PMA responses were additive. Furthermore, the ET‐1‐evoked response, contrary to that of PMA, totally depended on the presence of extracellular calcium. These results suggest that at least two distinct mechanisms are involved in the control of PLD activity in striatal astrocytes. Finally, ET‐1, ET‐3, and PMA also stimulated PLD in astrocytes from the mesencephalon, the cerebral cortex, and the hippocampus.

Albumin stimulates monocyte chemotactic protein-1 expression in rat embryonic mixed brain cells

Albumin, a blood protein absent from the adult brain in physiological situations, can be brought into contact with brain cells during development or, in adult, following breakdown of the blood–brain barrier occurring as a result of local inflammation. In the present study, we show that ovalbumin and albumin induce the release of monocyte chemotactic protein 1 (MCP‐1/CCL2) from rat embryonic mixed brain cells. A short‐term exposure to ovalbumin during the cell dissociation procedure is sufficient to generate MCP‐1 mRNA. A comparable effect is observed when the cells are incubated for 4 hr with ovalbumin or rat albumin, while MCP‐1 messengers are barely detectable following bovine albumin exposure. The amount of MCP‐1 protein measured in 4 hr‐supernatants of albumin‐treated cells followed the same albumin‐inducing pattern as that of MCP‐1 mRNA, while all albumins tested induced MCP‐1 protein after a 17 hr‐incubation period. The albumin‐induced MCP‐1 production is significantly inhibited in calphostin C‐treated cells, suggesting the implication of a protein kinase C‐dependent signaling pathway. This MCP‐1‐inducing activity is maintained after a lipid extraction procedure but abolished by proteinase K or trypsin treatments of albumin. The MCP‐1 secretion following albumin contact with nervous cells could thus interfere, by chemotactic gradient formation, with the brain infiltration program of blood‐derived cells during development or brain injury.